* Re: IPv6 multipath routes
From: Ulrich Weber @ 2012-09-11 12:57 UTC (permalink / raw)
To: Vincent Bernat; +Cc: netdev
In-Reply-To: <1346952649-5716-1-git-send-email-bernat@luffy.cx>
Hi Vincent,
there is no multipath support for IPv6 in the kernel.
RTA_MULTIPATH attributes are not parsed for IPv6.
Cheers
Ulrich
On 09/06/2012 07:30 PM, Vincent Bernat wrote:
> Hi!
>
> It appears that "ip -6 route add" expects IPv4 addresses with nexthop directives:
>
> $ ip -6 route add to 2a01:c9c0:a1:982::/64 proto bird \
> nexthop via fe80::ea39:35ff:febd:f9e dev bai2.2008 weight 1 \
> nexthop via fe80::ea39:35ff:febd:fd6 dev bai1.2009 weight 1
> Error: an IP address is expected rather than "fe80::ea39:35ff:febd:f9e"
>
> The following patch fix this problem. However, it does not work. I now have:
> RTNETLINK answers: No such device
>
> I have little knowledge of netlink so it is likely that my patch is
> buggy. However, I have found this problem by trying to debug what
> appears to be a valid netlink message refused by the kernel with the
> same error. Therefore, I suspect that there is also a bug in the kernel.
>
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
* [PATCH V3 7/8] ipvs: API change to avoid rescan of IPv6 exthdr
From: Jesper Dangaard Brouer @ 2012-09-11 12:38 UTC (permalink / raw)
To: Hans Schillstrom, Hans Schillstrom, netdev, Patrick McHardy,
Pablo Neira Ayuso, lvs-devel, Julian Anastasov
Cc: Jesper Dangaard Brouer, Thomas Graf, Wensong Zhang,
netfilter-devel, Simon Horman
In-Reply-To: <20120911123531.4305.40304.stgit@dragon>
Reduce the number of times we scan/skip the IPv6 exthdrs.
This patch contains a lot of API changes. This is done, to avoid
repeating the scan of finding the IPv6 headers, via ipv6_find_hdr(),
which is called by ip_vs_fill_iph_skb().
Finding the IPv6 headers is done as early as possible, and passed on
as a pointer "struct ip_vs_iphdr *" to the affected functions.
This patch reduce/removes 19 calls to ip_vs_fill_iph_skb().
Notice, I have choosen, not to change the API of function
pointer "(*schedule)" (in struct ip_vs_scheduler) as it can be
used by external schedulers, via {un,}register_ip_vs_scheduler.
Only 4 out of 10 schedulers use info from ip_vs_iphdr*, and when
they do, they are only interested in iph->{s,d}addr.
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
---
V3:
- This patch is introduced in V3, to allow easier integration of
Patricks idea and my RFC patch of caching exthdr info in skb->cb[]
include/net/ip_vs.h | 81 +++++++++++-----------
net/netfilter/ipvs/ip_vs_conn.c | 15 ++--
net/netfilter/ipvs/ip_vs_core.c | 116 ++++++++++++++-----------------
net/netfilter/ipvs/ip_vs_proto_ah_esp.c | 9 +-
net/netfilter/ipvs/ip_vs_proto_sctp.c | 42 ++++-------
net/netfilter/ipvs/ip_vs_proto_tcp.c | 40 ++++-------
net/netfilter/ipvs/ip_vs_proto_udp.c | 41 ++++-------
net/netfilter/ipvs/ip_vs_xmit.c | 61 +++++++---------
net/netfilter/xt_ipvs.c | 2 -
9 files changed, 177 insertions(+), 230 deletions(-)
diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h
index 98806b6..a681ad6 100644
--- a/include/net/ip_vs.h
+++ b/include/net/ip_vs.h
@@ -487,27 +487,26 @@ struct ip_vs_protocol {
int (*conn_schedule)(int af, struct sk_buff *skb,
struct ip_vs_proto_data *pd,
- int *verdict, struct ip_vs_conn **cpp);
+ int *verdict, struct ip_vs_conn **cpp,
+ struct ip_vs_iphdr *iph);
struct ip_vs_conn *
(*conn_in_get)(int af,
const struct sk_buff *skb,
const struct ip_vs_iphdr *iph,
- unsigned int proto_off,
int inverse);
struct ip_vs_conn *
(*conn_out_get)(int af,
const struct sk_buff *skb,
const struct ip_vs_iphdr *iph,
- unsigned int proto_off,
int inverse);
- int (*snat_handler)(struct sk_buff *skb,
- struct ip_vs_protocol *pp, struct ip_vs_conn *cp);
+ int (*snat_handler)(struct sk_buff *skb, struct ip_vs_protocol *pp,
+ struct ip_vs_conn *cp, struct ip_vs_iphdr *iph);
- int (*dnat_handler)(struct sk_buff *skb,
- struct ip_vs_protocol *pp, struct ip_vs_conn *cp);
+ int (*dnat_handler)(struct sk_buff *skb, struct ip_vs_protocol *pp,
+ struct ip_vs_conn *cp, struct ip_vs_iphdr *iph);
int (*csum_check)(int af, struct sk_buff *skb,
struct ip_vs_protocol *pp);
@@ -607,7 +606,7 @@ struct ip_vs_conn {
NF_ACCEPT can be returned when destination is local.
*/
int (*packet_xmit)(struct sk_buff *skb, struct ip_vs_conn *cp,
- struct ip_vs_protocol *pp);
+ struct ip_vs_protocol *pp, struct ip_vs_iphdr *iph);
/* Note: we can group the following members into a structure,
in order to save more space, and the following members are
@@ -858,13 +857,11 @@ struct ip_vs_app {
struct ip_vs_conn *
(*conn_in_get)(const struct sk_buff *skb, struct ip_vs_app *app,
- const struct iphdr *iph, unsigned int proto_off,
- int inverse);
+ const struct iphdr *iph, int inverse);
struct ip_vs_conn *
(*conn_out_get)(const struct sk_buff *skb, struct ip_vs_app *app,
- const struct iphdr *iph, unsigned int proto_off,
- int inverse);
+ const struct iphdr *iph, int inverse);
int (*state_transition)(struct ip_vs_conn *cp, int direction,
const struct sk_buff *skb,
@@ -1163,14 +1160,12 @@ struct ip_vs_conn *ip_vs_ct_in_get(const struct ip_vs_conn_param *p);
struct ip_vs_conn * ip_vs_conn_in_get_proto(int af, const struct sk_buff *skb,
const struct ip_vs_iphdr *iph,
- unsigned int proto_off,
int inverse);
struct ip_vs_conn *ip_vs_conn_out_get(const struct ip_vs_conn_param *p);
struct ip_vs_conn * ip_vs_conn_out_get_proto(int af, const struct sk_buff *skb,
const struct ip_vs_iphdr *iph,
- unsigned int proto_off,
int inverse);
/* put back the conn without restarting its timer */
@@ -1343,9 +1338,10 @@ extern struct ip_vs_scheduler *ip_vs_scheduler_get(const char *sched_name);
extern void ip_vs_scheduler_put(struct ip_vs_scheduler *scheduler);
extern struct ip_vs_conn *
ip_vs_schedule(struct ip_vs_service *svc, struct sk_buff *skb,
- struct ip_vs_proto_data *pd, int *ignored);
+ struct ip_vs_proto_data *pd, int *ignored,
+ struct ip_vs_iphdr *iph);
extern int ip_vs_leave(struct ip_vs_service *svc, struct sk_buff *skb,
- struct ip_vs_proto_data *pd);
+ struct ip_vs_proto_data *pd, struct ip_vs_iphdr *iph);
extern void ip_vs_scheduler_err(struct ip_vs_service *svc, const char *msg);
@@ -1404,33 +1400,38 @@ extern void ip_vs_read_estimator(struct ip_vs_stats_user *dst,
/*
* Various IPVS packet transmitters (from ip_vs_xmit.c)
*/
-extern int ip_vs_null_xmit
-(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp);
-extern int ip_vs_bypass_xmit
-(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp);
-extern int ip_vs_nat_xmit
-(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp);
-extern int ip_vs_tunnel_xmit
-(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp);
-extern int ip_vs_dr_xmit
-(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp);
-extern int ip_vs_icmp_xmit
-(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp,
- int offset, unsigned int hooknum);
+extern int ip_vs_null_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
+ struct ip_vs_protocol *pp, struct ip_vs_iphdr *iph);
+extern int ip_vs_bypass_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
+ struct ip_vs_protocol *pp,
+ struct ip_vs_iphdr *iph);
+extern int ip_vs_nat_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
+ struct ip_vs_protocol *pp, struct ip_vs_iphdr *iph);
+extern int ip_vs_tunnel_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
+ struct ip_vs_protocol *pp,
+ struct ip_vs_iphdr *iph);
+extern int ip_vs_dr_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
+ struct ip_vs_protocol *pp, struct ip_vs_iphdr *iph);
+extern int ip_vs_icmp_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
+ struct ip_vs_protocol *pp, int offset,
+ unsigned int hooknum, struct ip_vs_iphdr *iph);
extern void ip_vs_dst_reset(struct ip_vs_dest *dest);
#ifdef CONFIG_IP_VS_IPV6
-extern int ip_vs_bypass_xmit_v6
-(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp);
-extern int ip_vs_nat_xmit_v6
-(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp);
-extern int ip_vs_tunnel_xmit_v6
-(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp);
-extern int ip_vs_dr_xmit_v6
-(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp);
-extern int ip_vs_icmp_xmit_v6
-(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp,
- int offset, unsigned int hooknum);
+extern int ip_vs_bypass_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
+ struct ip_vs_protocol *pp,
+ struct ip_vs_iphdr *iph);
+extern int ip_vs_nat_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
+ struct ip_vs_protocol *pp,
+ struct ip_vs_iphdr *iph);
+extern int ip_vs_tunnel_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
+ struct ip_vs_protocol *pp,
+ struct ip_vs_iphdr *iph);
+extern int ip_vs_dr_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
+ struct ip_vs_protocol *pp, struct ip_vs_iphdr *iph);
+extern int ip_vs_icmp_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
+ struct ip_vs_protocol *pp, int offset,
+ unsigned int hooknum, struct ip_vs_iphdr *iph);
#endif
#ifdef CONFIG_SYSCTL
diff --git a/net/netfilter/ipvs/ip_vs_conn.c b/net/netfilter/ipvs/ip_vs_conn.c
index d6c1c26..30e764a 100644
--- a/net/netfilter/ipvs/ip_vs_conn.c
+++ b/net/netfilter/ipvs/ip_vs_conn.c
@@ -308,13 +308,12 @@ struct ip_vs_conn *ip_vs_conn_in_get(const struct ip_vs_conn_param *p)
static int
ip_vs_conn_fill_param_proto(int af, const struct sk_buff *skb,
const struct ip_vs_iphdr *iph,
- unsigned int proto_off, int inverse,
- struct ip_vs_conn_param *p)
+ int inverse, struct ip_vs_conn_param *p)
{
__be16 _ports[2], *pptr;
struct net *net = skb_net(skb);
- pptr = frag_safe_skb_hp(skb, proto_off, sizeof(_ports), _ports, iph);
+ pptr = frag_safe_skb_hp(skb, iph->len, sizeof(_ports), _ports, iph);
if (pptr == NULL)
return 1;
@@ -329,12 +328,11 @@ ip_vs_conn_fill_param_proto(int af, const struct sk_buff *skb,
struct ip_vs_conn *
ip_vs_conn_in_get_proto(int af, const struct sk_buff *skb,
- const struct ip_vs_iphdr *iph,
- unsigned int proto_off, int inverse)
+ const struct ip_vs_iphdr *iph, int inverse)
{
struct ip_vs_conn_param p;
- if (ip_vs_conn_fill_param_proto(af, skb, iph, proto_off, inverse, &p))
+ if (ip_vs_conn_fill_param_proto(af, skb, iph, inverse, &p))
return NULL;
return ip_vs_conn_in_get(&p);
@@ -432,12 +430,11 @@ struct ip_vs_conn *ip_vs_conn_out_get(const struct ip_vs_conn_param *p)
struct ip_vs_conn *
ip_vs_conn_out_get_proto(int af, const struct sk_buff *skb,
- const struct ip_vs_iphdr *iph,
- unsigned int proto_off, int inverse)
+ const struct ip_vs_iphdr *iph, int inverse)
{
struct ip_vs_conn_param p;
- if (ip_vs_conn_fill_param_proto(af, skb, iph, proto_off, inverse, &p))
+ if (ip_vs_conn_fill_param_proto(af, skb, iph, inverse, &p))
return NULL;
return ip_vs_conn_out_get(&p);
diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c
index 4b9995e..f72d64f 100644
--- a/net/netfilter/ipvs/ip_vs_core.c
+++ b/net/netfilter/ipvs/ip_vs_core.c
@@ -222,11 +222,10 @@ ip_vs_conn_fill_param_persist(const struct ip_vs_service *svc,
*/
static struct ip_vs_conn *
ip_vs_sched_persist(struct ip_vs_service *svc,
- struct sk_buff *skb,
- __be16 src_port, __be16 dst_port, int *ignored)
+ struct sk_buff *skb, __be16 src_port, __be16 dst_port,
+ int *ignored, struct ip_vs_iphdr *iph)
{
struct ip_vs_conn *cp = NULL;
- struct ip_vs_iphdr iph;
struct ip_vs_dest *dest;
struct ip_vs_conn *ct;
__be16 dport = 0; /* destination port to forward */
@@ -236,20 +235,18 @@ ip_vs_sched_persist(struct ip_vs_service *svc,
union nf_inet_addr snet; /* source network of the client,
after masking */
- ip_vs_fill_iph_skb(svc->af, skb, &iph);
-
/* Mask saddr with the netmask to adjust template granularity */
#ifdef CONFIG_IP_VS_IPV6
if (svc->af == AF_INET6)
- ipv6_addr_prefix(&snet.in6, &iph.saddr.in6, svc->netmask);
+ ipv6_addr_prefix(&snet.in6, &iph->saddr.in6, svc->netmask);
else
#endif
- snet.ip = iph.saddr.ip & svc->netmask;
+ snet.ip = iph->saddr.ip & svc->netmask;
IP_VS_DBG_BUF(6, "p-schedule: src %s:%u dest %s:%u "
"mnet %s\n",
- IP_VS_DBG_ADDR(svc->af, &iph.saddr), ntohs(src_port),
- IP_VS_DBG_ADDR(svc->af, &iph.daddr), ntohs(dst_port),
+ IP_VS_DBG_ADDR(svc->af, &iph->saddr), ntohs(src_port),
+ IP_VS_DBG_ADDR(svc->af, &iph->daddr), ntohs(dst_port),
IP_VS_DBG_ADDR(svc->af, &snet));
/*
@@ -266,8 +263,8 @@ ip_vs_sched_persist(struct ip_vs_service *svc,
* is created for other persistent services.
*/
{
- int protocol = iph.protocol;
- const union nf_inet_addr *vaddr = &iph.daddr;
+ int protocol = iph->protocol;
+ const union nf_inet_addr *vaddr = &iph->daddr;
__be16 vport = 0;
if (dst_port == svc->port) {
@@ -342,14 +339,14 @@ ip_vs_sched_persist(struct ip_vs_service *svc,
dport = dest->port;
flags = (svc->flags & IP_VS_SVC_F_ONEPACKET
- && iph.protocol == IPPROTO_UDP)?
+ && iph->protocol == IPPROTO_UDP) ?
IP_VS_CONN_F_ONE_PACKET : 0;
/*
* Create a new connection according to the template
*/
- ip_vs_conn_fill_param(svc->net, svc->af, iph.protocol, &iph.saddr,
- src_port, &iph.daddr, dst_port, ¶m);
+ ip_vs_conn_fill_param(svc->net, svc->af, iph->protocol, &iph->saddr,
+ src_port, &iph->daddr, dst_port, ¶m);
cp = ip_vs_conn_new(¶m, &dest->addr, dport, flags, dest, skb->mark);
if (cp == NULL) {
@@ -392,22 +389,20 @@ ip_vs_sched_persist(struct ip_vs_service *svc,
*/
struct ip_vs_conn *
ip_vs_schedule(struct ip_vs_service *svc, struct sk_buff *skb,
- struct ip_vs_proto_data *pd, int *ignored)
+ struct ip_vs_proto_data *pd, int *ignored,
+ struct ip_vs_iphdr *iph)
{
struct ip_vs_protocol *pp = pd->pp;
struct ip_vs_conn *cp = NULL;
- struct ip_vs_iphdr iph;
struct ip_vs_dest *dest;
__be16 _ports[2], *pptr;
unsigned int flags;
*ignored = 1;
-
/*
* IPv6 frags, only the first hit here.
*/
- ip_vs_fill_iph_skb(svc->af, skb, &iph);
- pptr = frag_safe_skb_hp(skb, iph.len, sizeof(_ports), _ports, &iph);
+ pptr = frag_safe_skb_hp(skb, iph->len, sizeof(_ports), _ports, iph);
if (pptr == NULL)
return NULL;
@@ -427,7 +422,7 @@ ip_vs_schedule(struct ip_vs_service *svc, struct sk_buff *skb,
* Do not schedule replies from local real server.
*/
if ((!skb->dev || skb->dev->flags & IFF_LOOPBACK) &&
- (cp = pp->conn_in_get(svc->af, skb, &iph, iph.len, 1))) {
+ (cp = pp->conn_in_get(svc->af, skb, iph, 1))) {
IP_VS_DBG_PKT(12, svc->af, pp, skb, 0,
"Not scheduling reply for existing connection");
__ip_vs_conn_put(cp);
@@ -438,7 +433,8 @@ ip_vs_schedule(struct ip_vs_service *svc, struct sk_buff *skb,
* Persistent service
*/
if (svc->flags & IP_VS_SVC_F_PERSISTENT)
- return ip_vs_sched_persist(svc, skb, pptr[0], pptr[1], ignored);
+ return ip_vs_sched_persist(svc, skb, pptr[0], pptr[1], ignored,
+ iph);
*ignored = 0;
@@ -460,7 +456,7 @@ ip_vs_schedule(struct ip_vs_service *svc, struct sk_buff *skb,
}
flags = (svc->flags & IP_VS_SVC_F_ONEPACKET
- && iph.protocol == IPPROTO_UDP)?
+ && iph->protocol == IPPROTO_UDP) ?
IP_VS_CONN_F_ONE_PACKET : 0;
/*
@@ -469,9 +465,9 @@ ip_vs_schedule(struct ip_vs_service *svc, struct sk_buff *skb,
{
struct ip_vs_conn_param p;
- ip_vs_conn_fill_param(svc->net, svc->af, iph.protocol,
- &iph.saddr, pptr[0], &iph.daddr, pptr[1],
- &p);
+ ip_vs_conn_fill_param(svc->net, svc->af, iph->protocol,
+ &iph->saddr, pptr[0], &iph->daddr,
+ pptr[1], &p);
cp = ip_vs_conn_new(&p, &dest->addr,
dest->port ? dest->port : pptr[1],
flags, dest, skb->mark);
@@ -500,18 +496,16 @@ ip_vs_schedule(struct ip_vs_service *svc, struct sk_buff *skb,
* no destination is available for a new connection.
*/
int ip_vs_leave(struct ip_vs_service *svc, struct sk_buff *skb,
- struct ip_vs_proto_data *pd)
+ struct ip_vs_proto_data *pd, struct ip_vs_iphdr *iph)
{
__be16 _ports[2], *pptr;
- struct ip_vs_iphdr iph;
#ifdef CONFIG_SYSCTL
struct net *net;
struct netns_ipvs *ipvs;
int unicast;
#endif
- ip_vs_fill_iph_skb(svc->af, skb, &iph);
- pptr = frag_safe_skb_hp(skb, iph.len, sizeof(_ports), _ports, &iph);
+ pptr = frag_safe_skb_hp(skb, iph->len, sizeof(_ports), _ports, iph);
if (pptr == NULL) {
ip_vs_service_put(svc);
return NF_DROP;
@@ -522,10 +516,10 @@ int ip_vs_leave(struct ip_vs_service *svc, struct sk_buff *skb,
#ifdef CONFIG_IP_VS_IPV6
if (svc->af == AF_INET6)
- unicast = ipv6_addr_type(&iph.daddr.in6) & IPV6_ADDR_UNICAST;
+ unicast = ipv6_addr_type(&iph->daddr.in6) & IPV6_ADDR_UNICAST;
else
#endif
- unicast = (inet_addr_type(net, iph.daddr.ip) == RTN_UNICAST);
+ unicast = (inet_addr_type(net, iph->daddr.ip) == RTN_UNICAST);
/* if it is fwmark-based service, the cache_bypass sysctl is up
and the destination is a non-local unicast, then create
@@ -535,7 +529,7 @@ int ip_vs_leave(struct ip_vs_service *svc, struct sk_buff *skb,
int ret;
struct ip_vs_conn *cp;
unsigned int flags = (svc->flags & IP_VS_SVC_F_ONEPACKET &&
- iph.protocol == IPPROTO_UDP)?
+ iph->protocol == IPPROTO_UDP) ?
IP_VS_CONN_F_ONE_PACKET : 0;
union nf_inet_addr daddr = { .all = { 0, 0, 0, 0 } };
@@ -545,9 +539,9 @@ int ip_vs_leave(struct ip_vs_service *svc, struct sk_buff *skb,
IP_VS_DBG(6, "%s(): create a cache_bypass entry\n", __func__);
{
struct ip_vs_conn_param p;
- ip_vs_conn_fill_param(svc->net, svc->af, iph.protocol,
- &iph.saddr, pptr[0],
- &iph.daddr, pptr[1], &p);
+ ip_vs_conn_fill_param(svc->net, svc->af, iph->protocol,
+ &iph->saddr, pptr[0],
+ &iph->daddr, pptr[1], &p);
cp = ip_vs_conn_new(&p, &daddr, 0,
IP_VS_CONN_F_BYPASS | flags,
NULL, skb->mark);
@@ -562,7 +556,7 @@ int ip_vs_leave(struct ip_vs_service *svc, struct sk_buff *skb,
ip_vs_set_state(cp, IP_VS_DIR_INPUT, skb, pd);
/* transmit the first SYN packet */
- ret = cp->packet_xmit(skb, cp, pd->pp);
+ ret = cp->packet_xmit(skb, cp, pd->pp, iph);
/* do not touch skb anymore */
atomic_inc(&cp->in_pkts);
@@ -911,7 +905,7 @@ static int ip_vs_out_icmp(struct sk_buff *skb, int *related,
ip_vs_fill_ip4hdr(cih, &ciph);
ciph.len += offset;
/* The embedded headers contain source and dest in reverse order */
- cp = pp->conn_out_get(AF_INET, skb, &ciph, offset, 1);
+ cp = pp->conn_out_get(AF_INET, skb, &ciph, 1);
if (!cp)
return NF_ACCEPT;
@@ -922,7 +916,7 @@ static int ip_vs_out_icmp(struct sk_buff *skb, int *related,
#ifdef CONFIG_IP_VS_IPV6
static int ip_vs_out_icmp_v6(struct sk_buff *skb, int *related,
- unsigned int hooknum)
+ unsigned int hooknum, struct ip_vs_iphdr *ipvsh)
{
struct icmp6hdr _icmph, *ic;
struct ipv6hdr _ip6, *ip6; /* The ip header contained
@@ -931,10 +925,6 @@ static int ip_vs_out_icmp_v6(struct sk_buff *skb, int *related,
struct ip_vs_protocol *pp;
union nf_inet_addr snet;
- struct ip_vs_iphdr ipvsh_stack;
- struct ip_vs_iphdr *ipvsh = &ipvsh_stack;
- ip_vs_fill_iph_skb(AF_INET6, skb, ipvsh);
-
*related = 1;
ic = frag_safe_skb_hp(skb, ipvsh->len, sizeof(_icmph), &_icmph, ipvsh);
if (ic == NULL)
@@ -977,7 +967,7 @@ static int ip_vs_out_icmp_v6(struct sk_buff *skb, int *related,
ipvsh->daddr.in6 = ip6->daddr;
/* The embedded headers contain source and dest in reverse order */
- cp = pp->conn_out_get(AF_INET6, skb, ipvsh, ipvsh->len, 1);
+ cp = pp->conn_out_get(AF_INET6, skb, ipvsh, 1);
if (!cp)
return NF_ACCEPT;
@@ -1016,17 +1006,17 @@ static inline int is_tcp_reset(const struct sk_buff *skb, int nh_len)
*/
static unsigned int
handle_response(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd,
- struct ip_vs_conn *cp, int ihl)
+ struct ip_vs_conn *cp, struct ip_vs_iphdr *iph)
{
struct ip_vs_protocol *pp = pd->pp;
IP_VS_DBG_PKT(11, af, pp, skb, 0, "Outgoing packet");
- if (!skb_make_writable(skb, ihl))
+ if (!skb_make_writable(skb, iph->len))
goto drop;
/* mangle the packet */
- if (pp->snat_handler && !pp->snat_handler(skb, pp, cp))
+ if (pp->snat_handler && !pp->snat_handler(skb, pp, cp, iph))
goto drop;
#ifdef CONFIG_IP_VS_IPV6
@@ -1125,7 +1115,7 @@ ip_vs_out(unsigned int hooknum, struct sk_buff *skb, int af)
if (unlikely(iph.protocol == IPPROTO_ICMPV6)) {
int related;
int verdict = ip_vs_out_icmp_v6(skb, &related,
- hooknum);
+ hooknum, &iph);
if (related)
return verdict;
@@ -1160,10 +1150,10 @@ ip_vs_out(unsigned int hooknum, struct sk_buff *skb, int af)
/*
* Check if the packet belongs to an existing entry
*/
- cp = pp->conn_out_get(af, skb, &iph, iph.len, 0);
+ cp = pp->conn_out_get(af, skb, &iph, 0);
if (likely(cp))
- return handle_response(af, skb, pd, cp, iph.len);
+ return handle_response(af, skb, pd, cp, &iph);
if (sysctl_nat_icmp_send(net) &&
(pp->protocol == IPPROTO_TCP ||
pp->protocol == IPPROTO_UDP ||
@@ -1375,7 +1365,7 @@ ip_vs_in_icmp(struct sk_buff *skb, int *related, unsigned int hooknum)
/* The embedded headers contain source and dest in reverse order.
* For IPIP this is error for request, not for reply.
*/
- cp = pp->conn_in_get(AF_INET, skb, &ciph, offset, ipip ? 0 : 1);
+ cp = pp->conn_in_get(AF_INET, skb, &ciph, ipip ? 0 : 1);
if (!cp)
return NF_ACCEPT;
@@ -1444,7 +1434,7 @@ ignore_ipip:
ip_vs_in_stats(cp, skb);
if (IPPROTO_TCP == cih->protocol || IPPROTO_UDP == cih->protocol)
offset += 2 * sizeof(__u16);
- verdict = ip_vs_icmp_xmit(skb, cp, pp, offset, hooknum);
+ verdict = ip_vs_icmp_xmit(skb, cp, pp, offset, hooknum, &ciph);
out:
__ip_vs_conn_put(cp);
@@ -1453,8 +1443,8 @@ out:
}
#ifdef CONFIG_IP_VS_IPV6
-static int
-ip_vs_in_icmp_v6(struct sk_buff *skb, int *related, unsigned int hooknum)
+static int ip_vs_in_icmp_v6(struct sk_buff *skb, int *related,
+ unsigned int hooknum, struct ip_vs_iphdr *iph)
{
struct net *net = NULL;
struct ipv6hdr _ip6h, *ip6h;
@@ -1465,10 +1455,6 @@ ip_vs_in_icmp_v6(struct sk_buff *skb, int *related, unsigned int hooknum)
struct ip_vs_proto_data *pd;
unsigned int offs_ciph, verdict;
- struct ip_vs_iphdr iph_stack;
- struct ip_vs_iphdr *iph = &iph_stack;
- ip_vs_fill_iph_skb(AF_INET6, skb, iph);
-
*related = 1;
ic = frag_safe_skb_hp(skb, iph->len, sizeof(_icmph), &_icmph, iph);
@@ -1522,7 +1508,7 @@ ip_vs_in_icmp_v6(struct sk_buff *skb, int *related, unsigned int hooknum)
/* The embedded headers contain source and dest in reverse order
* if not from localhost
*/
- cp = pp->conn_in_get(AF_INET6, skb, &ciph, ciph.len,
+ cp = pp->conn_in_get(AF_INET6, skb, &ciph,
(hooknum == NF_INET_LOCAL_OUT) ? 0 : 1);
if (!cp)
@@ -1540,7 +1526,7 @@ ip_vs_in_icmp_v6(struct sk_buff *skb, int *related, unsigned int hooknum)
IPPROTO_SCTP == ciph.protocol)
offs_ciph = ciph.len + (2 * sizeof(__u16));
- verdict = ip_vs_icmp_xmit_v6(skb, cp, pp, offs_ciph, hooknum);
+ verdict = ip_vs_icmp_xmit_v6(skb, cp, pp, offs_ciph, hooknum, &ciph);
__ip_vs_conn_put(cp);
@@ -1610,7 +1596,8 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af)
}
if (unlikely(iph.protocol == IPPROTO_ICMPV6)) {
int related;
- int verdict = ip_vs_in_icmp_v6(skb, &related, hooknum);
+ int verdict = ip_vs_in_icmp_v6(skb, &related, hooknum,
+ &iph);
if (related)
return verdict;
@@ -1633,8 +1620,7 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af)
/*
* Check if the packet belongs to an existing connection entry
*/
- cp = pp->conn_in_get(af, skb, &iph, iph.len, 0);
-
+ cp = pp->conn_in_get(af, skb, &iph, 0);
if (unlikely(!cp) && !iph.fragoffs) {
/* No (second) fragments need to enter here, as nf_defrag_ipv6
* replayed fragment zero will already have created the cp
@@ -1642,7 +1628,7 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af)
int v;
/* Schedule and create new connection entry into &cp */
- if (!pp->conn_schedule(af, skb, pd, &v, &cp))
+ if (!pp->conn_schedule(af, skb, pd, &v, &cp, &iph))
return v;
}
@@ -1680,7 +1666,7 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af)
ip_vs_in_stats(cp, skb);
ip_vs_set_state(cp, IP_VS_DIR_INPUT, skb, pd);
if (cp->packet_xmit)
- ret = cp->packet_xmit(skb, cp, pp);
+ ret = cp->packet_xmit(skb, cp, pp, &iph);
/* do not touch skb anymore */
else {
IP_VS_DBG_RL("warning: packet_xmit is null");
@@ -1854,7 +1840,7 @@ ip_vs_forward_icmp_v6(unsigned int hooknum, struct sk_buff *skb,
if (!net_ipvs(net)->enable)
return NF_ACCEPT;
- return ip_vs_in_icmp_v6(skb, &r, hooknum);
+ return ip_vs_in_icmp_v6(skb, &r, hooknum, &iphdr);
}
#endif
diff --git a/net/netfilter/ipvs/ip_vs_proto_ah_esp.c b/net/netfilter/ipvs/ip_vs_proto_ah_esp.c
index 5b8eb8b..5de3dd3 100644
--- a/net/netfilter/ipvs/ip_vs_proto_ah_esp.c
+++ b/net/netfilter/ipvs/ip_vs_proto_ah_esp.c
@@ -57,7 +57,7 @@ ah_esp_conn_fill_param_proto(struct net *net, int af,
static struct ip_vs_conn *
ah_esp_conn_in_get(int af, const struct sk_buff *skb,
- const struct ip_vs_iphdr *iph, unsigned int proto_off,
+ const struct ip_vs_iphdr *iph,
int inverse)
{
struct ip_vs_conn *cp;
@@ -85,9 +85,7 @@ ah_esp_conn_in_get(int af, const struct sk_buff *skb,
static struct ip_vs_conn *
ah_esp_conn_out_get(int af, const struct sk_buff *skb,
- const struct ip_vs_iphdr *iph,
- unsigned int proto_off,
- int inverse)
+ const struct ip_vs_iphdr *iph, int inverse)
{
struct ip_vs_conn *cp;
struct ip_vs_conn_param p;
@@ -110,7 +108,8 @@ ah_esp_conn_out_get(int af, const struct sk_buff *skb,
static int
ah_esp_conn_schedule(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd,
- int *verdict, struct ip_vs_conn **cpp)
+ int *verdict, struct ip_vs_conn **cpp,
+ struct ip_vs_iphdr *iph)
{
/*
* AH/ESP is only related traffic. Pass the packet to IP stack.
diff --git a/net/netfilter/ipvs/ip_vs_proto_sctp.c b/net/netfilter/ipvs/ip_vs_proto_sctp.c
index b903db6..746048b 100644
--- a/net/netfilter/ipvs/ip_vs_proto_sctp.c
+++ b/net/netfilter/ipvs/ip_vs_proto_sctp.c
@@ -10,28 +10,26 @@
static int
sctp_conn_schedule(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd,
- int *verdict, struct ip_vs_conn **cpp)
+ int *verdict, struct ip_vs_conn **cpp,
+ struct ip_vs_iphdr *iph)
{
struct net *net;
struct ip_vs_service *svc;
sctp_chunkhdr_t _schunkh, *sch;
sctp_sctphdr_t *sh, _sctph;
- struct ip_vs_iphdr iph;
- ip_vs_fill_iph_skb(af, skb, &iph);
-
- sh = skb_header_pointer(skb, iph.len, sizeof(_sctph), &_sctph);
+ sh = skb_header_pointer(skb, iph->len, sizeof(_sctph), &_sctph);
if (sh == NULL)
return 0;
- sch = skb_header_pointer(skb, iph.len + sizeof(sctp_sctphdr_t),
+ sch = skb_header_pointer(skb, iph->len + sizeof(sctp_sctphdr_t),
sizeof(_schunkh), &_schunkh);
if (sch == NULL)
return 0;
net = skb_net(skb);
if ((sch->type == SCTP_CID_INIT) &&
- (svc = ip_vs_service_get(net, af, skb->mark, iph.protocol,
- &iph.daddr, sh->dest))) {
+ (svc = ip_vs_service_get(net, af, skb->mark, iph->protocol,
+ &iph->daddr, sh->dest))) {
int ignored;
if (ip_vs_todrop(net_ipvs(net))) {
@@ -47,10 +45,10 @@ sctp_conn_schedule(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd,
* Let the virtual server select a real server for the
* incoming connection, and create a connection entry.
*/
- *cpp = ip_vs_schedule(svc, skb, pd, &ignored);
+ *cpp = ip_vs_schedule(svc, skb, pd, &ignored, iph);
if (!*cpp && ignored <= 0) {
if (!ignored)
- *verdict = ip_vs_leave(svc, skb, pd);
+ *verdict = ip_vs_leave(svc, skb, pd, iph);
else {
ip_vs_service_put(svc);
*verdict = NF_DROP;
@@ -64,20 +62,16 @@ sctp_conn_schedule(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd,
}
static int
-sctp_snat_handler(struct sk_buff *skb,
- struct ip_vs_protocol *pp, struct ip_vs_conn *cp)
+sctp_snat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp,
+ struct ip_vs_conn *cp, struct ip_vs_iphdr *iph)
{
sctp_sctphdr_t *sctph;
- unsigned int sctphoff;
+ unsigned int sctphoff = iph->len;
struct sk_buff *iter;
__be32 crc32;
- struct ip_vs_iphdr iph;
- ip_vs_fill_iph_skb(cp->af, skb, &iph);
- sctphoff = iph.len;
-
#ifdef CONFIG_IP_VS_IPV6
- if (cp->af == AF_INET6 && iph.fragoffs)
+ if (cp->af == AF_INET6 && iph->fragoffs)
return 1;
#endif
@@ -110,20 +104,16 @@ sctp_snat_handler(struct sk_buff *skb,
}
static int
-sctp_dnat_handler(struct sk_buff *skb,
- struct ip_vs_protocol *pp, struct ip_vs_conn *cp)
+sctp_dnat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp,
+ struct ip_vs_conn *cp, struct ip_vs_iphdr *iph)
{
sctp_sctphdr_t *sctph;
- unsigned int sctphoff;
+ unsigned int sctphoff = iph->len;
struct sk_buff *iter;
__be32 crc32;
- struct ip_vs_iphdr iph;
- ip_vs_fill_iph_skb(cp->af, skb, &iph);
- sctphoff = iph.len;
-
#ifdef CONFIG_IP_VS_IPV6
- if (cp->af == AF_INET6 && iph.fragoffs)
+ if (cp->af == AF_INET6 && iph->fragoffs)
return 1;
#endif
diff --git a/net/netfilter/ipvs/ip_vs_proto_tcp.c b/net/netfilter/ipvs/ip_vs_proto_tcp.c
index 8a96069..9af653a 100644
--- a/net/netfilter/ipvs/ip_vs_proto_tcp.c
+++ b/net/netfilter/ipvs/ip_vs_proto_tcp.c
@@ -33,16 +33,14 @@
static int
tcp_conn_schedule(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd,
- int *verdict, struct ip_vs_conn **cpp)
+ int *verdict, struct ip_vs_conn **cpp,
+ struct ip_vs_iphdr *iph)
{
struct net *net;
struct ip_vs_service *svc;
struct tcphdr _tcph, *th;
- struct ip_vs_iphdr iph;
- ip_vs_fill_iph_skb(af, skb, &iph);
-
- th = skb_header_pointer(skb, iph.len, sizeof(_tcph), &_tcph);
+ th = skb_header_pointer(skb, iph->len, sizeof(_tcph), &_tcph);
if (th == NULL) {
*verdict = NF_DROP;
return 0;
@@ -50,8 +48,8 @@ tcp_conn_schedule(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd,
net = skb_net(skb);
/* No !th->ack check to allow scheduling on SYN+ACK for Active FTP */
if (th->syn &&
- (svc = ip_vs_service_get(net, af, skb->mark, iph.protocol,
- &iph.daddr, th->dest))) {
+ (svc = ip_vs_service_get(net, af, skb->mark, iph->protocol,
+ &iph->daddr, th->dest))) {
int ignored;
if (ip_vs_todrop(net_ipvs(net))) {
@@ -68,10 +66,10 @@ tcp_conn_schedule(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd,
* Let the virtual server select a real server for the
* incoming connection, and create a connection entry.
*/
- *cpp = ip_vs_schedule(svc, skb, pd, &ignored);
+ *cpp = ip_vs_schedule(svc, skb, pd, &ignored, iph);
if (!*cpp && ignored <= 0) {
if (!ignored)
- *verdict = ip_vs_leave(svc, skb, pd);
+ *verdict = ip_vs_leave(svc, skb, pd, iph);
else {
ip_vs_service_put(svc);
*verdict = NF_DROP;
@@ -128,20 +126,16 @@ tcp_partial_csum_update(int af, struct tcphdr *tcph,
static int
-tcp_snat_handler(struct sk_buff *skb,
- struct ip_vs_protocol *pp, struct ip_vs_conn *cp)
+tcp_snat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp,
+ struct ip_vs_conn *cp, struct ip_vs_iphdr *iph)
{
struct tcphdr *tcph;
- unsigned int tcphoff;
+ unsigned int tcphoff = iph->len;
int oldlen;
int payload_csum = 0;
- struct ip_vs_iphdr iph;
- ip_vs_fill_iph_skb(cp->af, skb, &iph);
- tcphoff = iph.len;
-
#ifdef CONFIG_IP_VS_IPV6
- if (cp->af == AF_INET6 && iph.fragoffs)
+ if (cp->af == AF_INET6 && iph->fragoffs)
return 1;
#endif
oldlen = skb->len - tcphoff;
@@ -210,20 +204,16 @@ tcp_snat_handler(struct sk_buff *skb,
static int
-tcp_dnat_handler(struct sk_buff *skb,
- struct ip_vs_protocol *pp, struct ip_vs_conn *cp)
+tcp_dnat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp,
+ struct ip_vs_conn *cp, struct ip_vs_iphdr *iph)
{
struct tcphdr *tcph;
- unsigned int tcphoff;
+ unsigned int tcphoff = iph->len;
int oldlen;
int payload_csum = 0;
- struct ip_vs_iphdr iph;
- ip_vs_fill_iph_skb(cp->af, skb, &iph);
- tcphoff = iph.len;
-
#ifdef CONFIG_IP_VS_IPV6
- if (cp->af == AF_INET6 && iph.fragoffs)
+ if (cp->af == AF_INET6 && iph->fragoffs)
return 1;
#endif
oldlen = skb->len - tcphoff;
diff --git a/net/netfilter/ipvs/ip_vs_proto_udp.c b/net/netfilter/ipvs/ip_vs_proto_udp.c
index d6f4eee..503a842 100644
--- a/net/netfilter/ipvs/ip_vs_proto_udp.c
+++ b/net/netfilter/ipvs/ip_vs_proto_udp.c
@@ -30,23 +30,22 @@
static int
udp_conn_schedule(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd,
- int *verdict, struct ip_vs_conn **cpp)
+ int *verdict, struct ip_vs_conn **cpp,
+ struct ip_vs_iphdr *iph)
{
struct net *net;
struct ip_vs_service *svc;
struct udphdr _udph, *uh;
- struct ip_vs_iphdr iph;
- ip_vs_fill_iph_skb(af, skb, &iph);
-
- uh = skb_header_pointer(skb, iph.len, sizeof(_udph), &_udph);
+ /* IPv6 fragments, only first fragment will hit this */
+ uh = skb_header_pointer(skb, iph->len, sizeof(_udph), &_udph);
if (uh == NULL) {
*verdict = NF_DROP;
return 0;
}
net = skb_net(skb);
- svc = ip_vs_service_get(net, af, skb->mark, iph.protocol,
- &iph.daddr, uh->dest);
+ svc = ip_vs_service_get(net, af, skb->mark, iph->protocol,
+ &iph->daddr, uh->dest);
if (svc) {
int ignored;
@@ -64,10 +63,10 @@ udp_conn_schedule(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd,
* Let the virtual server select a real server for the
* incoming connection, and create a connection entry.
*/
- *cpp = ip_vs_schedule(svc, skb, pd, &ignored);
+ *cpp = ip_vs_schedule(svc, skb, pd, &ignored, iph);
if (!*cpp && ignored <= 0) {
if (!ignored)
- *verdict = ip_vs_leave(svc, skb, pd);
+ *verdict = ip_vs_leave(svc, skb, pd, iph);
else {
ip_vs_service_put(svc);
*verdict = NF_DROP;
@@ -125,20 +124,16 @@ udp_partial_csum_update(int af, struct udphdr *uhdr,
static int
-udp_snat_handler(struct sk_buff *skb,
- struct ip_vs_protocol *pp, struct ip_vs_conn *cp)
+udp_snat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp,
+ struct ip_vs_conn *cp, struct ip_vs_iphdr *iph)
{
struct udphdr *udph;
- unsigned int udphoff;
+ unsigned int udphoff = iph->len;
int oldlen;
int payload_csum = 0;
- struct ip_vs_iphdr iph;
- ip_vs_fill_iph_skb(cp->af, skb, &iph);
- udphoff = iph.len;
-
#ifdef CONFIG_IP_VS_IPV6
- if (cp->af == AF_INET6 && iph.fragoffs)
+ if (cp->af == AF_INET6 && iph->fragoffs)
return 1;
#endif
oldlen = skb->len - udphoff;
@@ -212,20 +207,16 @@ udp_snat_handler(struct sk_buff *skb,
static int
-udp_dnat_handler(struct sk_buff *skb,
- struct ip_vs_protocol *pp, struct ip_vs_conn *cp)
+udp_dnat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp,
+ struct ip_vs_conn *cp, struct ip_vs_iphdr *iph)
{
struct udphdr *udph;
- unsigned int udphoff;
+ unsigned int udphoff = iph->len;
int oldlen;
int payload_csum = 0;
- struct ip_vs_iphdr iph;
- ip_vs_fill_iph_skb(cp->af, skb, &iph);
- udphoff = iph.len;
-
#ifdef CONFIG_IP_VS_IPV6
- if (cp->af == AF_INET6 && iph.fragoffs)
+ if (cp->af == AF_INET6 && iph->fragoffs)
return 1;
#endif
oldlen = skb->len - udphoff;
diff --git a/net/netfilter/ipvs/ip_vs_xmit.c b/net/netfilter/ipvs/ip_vs_xmit.c
index e44933f..90122eb 100644
--- a/net/netfilter/ipvs/ip_vs_xmit.c
+++ b/net/netfilter/ipvs/ip_vs_xmit.c
@@ -424,7 +424,7 @@ do { \
*/
int
ip_vs_null_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
- struct ip_vs_protocol *pp)
+ struct ip_vs_protocol *pp, struct ip_vs_iphdr *ipvsh)
{
/* we do not touch skb and do not need pskb ptr */
IP_VS_XMIT(NFPROTO_IPV4, skb, cp, 1);
@@ -438,7 +438,7 @@ ip_vs_null_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
*/
int
ip_vs_bypass_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
- struct ip_vs_protocol *pp)
+ struct ip_vs_protocol *pp, struct ip_vs_iphdr *ipvsh)
{
struct rtable *rt; /* Route to the other host */
struct iphdr *iph = ip_hdr(skb);
@@ -493,17 +493,16 @@ ip_vs_bypass_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
#ifdef CONFIG_IP_VS_IPV6
int
ip_vs_bypass_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
- struct ip_vs_protocol *pp)
+ struct ip_vs_protocol *pp, struct ip_vs_iphdr *iph)
{
struct rt6_info *rt; /* Route to the other host */
- struct ip_vs_iphdr iph;
int mtu;
EnterFunction(10);
- ip_vs_fill_iph_skb(cp->af, skb, &iph);
- if (!(rt = __ip_vs_get_out_rt_v6(skb, NULL, &iph.daddr.in6, NULL, 0,
- IP_VS_RT_MODE_NON_LOCAL)))
+ rt = __ip_vs_get_out_rt_v6(skb, NULL, &iph->daddr.in6, NULL, 0,
+ IP_VS_RT_MODE_NON_LOCAL);
+ if (!rt)
goto tx_error_icmp;
/* MTU checking */
@@ -515,7 +514,7 @@ ip_vs_bypass_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
skb->dev = net->loopback_dev;
}
/* only send ICMP too big on first fragment */
- if (!iph.fragoffs)
+ if (!iph->fragoffs)
icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
dst_release(&rt->dst);
IP_VS_DBG_RL("%s(): frag needed\n", __func__);
@@ -559,7 +558,7 @@ ip_vs_bypass_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
*/
int
ip_vs_nat_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
- struct ip_vs_protocol *pp)
+ struct ip_vs_protocol *pp, struct ip_vs_iphdr *ipvsh)
{
struct rtable *rt; /* Route to the other host */
int mtu;
@@ -629,7 +628,7 @@ ip_vs_nat_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
goto tx_error_put;
/* mangle the packet */
- if (pp->dnat_handler && !pp->dnat_handler(skb, pp, cp))
+ if (pp->dnat_handler && !pp->dnat_handler(skb, pp, cp, ipvsh))
goto tx_error_put;
ip_hdr(skb)->daddr = cp->daddr.ip;
ip_send_check(ip_hdr(skb));
@@ -677,20 +676,18 @@ ip_vs_nat_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
#ifdef CONFIG_IP_VS_IPV6
int
ip_vs_nat_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
- struct ip_vs_protocol *pp)
+ struct ip_vs_protocol *pp, struct ip_vs_iphdr *iph)
{
struct rt6_info *rt; /* Route to the other host */
int mtu;
int local;
- struct ip_vs_iphdr iph;
EnterFunction(10);
- ip_vs_fill_iph_skb(cp->af, skb, &iph);
/* check if it is a connection of no-client-port */
- if (unlikely(cp->flags & IP_VS_CONN_F_NO_CPORT && !iph.fragoffs)) {
+ if (unlikely(cp->flags & IP_VS_CONN_F_NO_CPORT && !iph->fragoffs)) {
__be16 _pt, *p;
- p = skb_header_pointer(skb, iph.len, sizeof(_pt), &_pt);
+ p = skb_header_pointer(skb, iph->len, sizeof(_pt), &_pt);
if (p == NULL)
goto tx_error;
ip_vs_conn_fill_cport(cp, *p);
@@ -739,7 +736,7 @@ ip_vs_nat_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
skb->dev = net->loopback_dev;
}
/* only send ICMP too big on first fragment */
- if (!iph.fragoffs)
+ if (!iph->fragoffs)
icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
IP_VS_DBG_RL_PKT(0, AF_INET6, pp, skb, 0,
"ip_vs_nat_xmit_v6(): frag needed for");
@@ -754,7 +751,7 @@ ip_vs_nat_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
goto tx_error_put;
/* mangle the packet */
- if (pp->dnat_handler && !pp->dnat_handler(skb, pp, cp))
+ if (pp->dnat_handler && !pp->dnat_handler(skb, pp, cp, iph))
goto tx_error;
ipv6_hdr(skb)->daddr = cp->daddr.in6;
@@ -815,7 +812,7 @@ tx_error_put:
*/
int
ip_vs_tunnel_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
- struct ip_vs_protocol *pp)
+ struct ip_vs_protocol *pp, struct ip_vs_iphdr *ipvsh)
{
struct netns_ipvs *ipvs = net_ipvs(skb_net(skb));
struct rtable *rt; /* Route to the other host */
@@ -935,7 +932,7 @@ tx_error_put:
#ifdef CONFIG_IP_VS_IPV6
int
ip_vs_tunnel_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
- struct ip_vs_protocol *pp)
+ struct ip_vs_protocol *pp, struct ip_vs_iphdr *ipvsh)
{
struct rt6_info *rt; /* Route to the other host */
struct in6_addr saddr; /* Source for tunnel */
@@ -945,10 +942,8 @@ ip_vs_tunnel_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
unsigned int max_headroom; /* The extra header space needed */
int mtu;
int ret;
- struct ip_vs_iphdr ipvsh;
EnterFunction(10);
- ip_vs_fill_iph_skb(cp->af, skb, &ipvsh);
if (!(rt = __ip_vs_get_out_rt_v6(skb, cp->dest, &cp->daddr.in6,
&saddr, 1, (IP_VS_RT_MODE_LOCAL |
@@ -978,7 +973,7 @@ ip_vs_tunnel_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
skb->dev = net->loopback_dev;
}
/* only send ICMP too big on first fragment */
- if (!ipvsh.fragoffs)
+ if (!ipvsh->fragoffs)
icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
IP_VS_DBG_RL("%s(): frag needed\n", __func__);
goto tx_error_put;
@@ -1060,7 +1055,7 @@ tx_error_put:
*/
int
ip_vs_dr_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
- struct ip_vs_protocol *pp)
+ struct ip_vs_protocol *pp, struct ip_vs_iphdr *ipvsh)
{
struct rtable *rt; /* Route to the other host */
struct iphdr *iph = ip_hdr(skb);
@@ -1121,14 +1116,12 @@ ip_vs_dr_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
#ifdef CONFIG_IP_VS_IPV6
int
ip_vs_dr_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
- struct ip_vs_protocol *pp)
+ struct ip_vs_protocol *pp, struct ip_vs_iphdr *iph)
{
struct rt6_info *rt; /* Route to the other host */
int mtu;
- struct ip_vs_iphdr iph;
EnterFunction(10);
- ip_vs_fill_iph_skb(cp->af, skb, &iph);
if (!(rt = __ip_vs_get_out_rt_v6(skb, cp->dest, &cp->daddr.in6, NULL,
0, (IP_VS_RT_MODE_LOCAL |
@@ -1148,7 +1141,7 @@ ip_vs_dr_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
skb->dev = net->loopback_dev;
}
/* only send ICMP too big on first fragment */
- if (!iph.fragoffs)
+ if (!iph->fragoffs)
icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
dst_release(&rt->dst);
IP_VS_DBG_RL("%s(): frag needed\n", __func__);
@@ -1193,7 +1186,8 @@ tx_error:
*/
int
ip_vs_icmp_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
- struct ip_vs_protocol *pp, int offset, unsigned int hooknum)
+ struct ip_vs_protocol *pp, int offset, unsigned int hooknum,
+ struct ip_vs_iphdr *iph)
{
struct rtable *rt; /* Route to the other host */
int mtu;
@@ -1208,7 +1202,7 @@ ip_vs_icmp_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
translate address/port back */
if (IP_VS_FWD_METHOD(cp) != IP_VS_CONN_F_MASQ) {
if (cp->packet_xmit)
- rc = cp->packet_xmit(skb, cp, pp);
+ rc = cp->packet_xmit(skb, cp, pp, iph);
else
rc = NF_ACCEPT;
/* do not touch skb anymore */
@@ -1314,24 +1308,23 @@ ip_vs_icmp_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
#ifdef CONFIG_IP_VS_IPV6
int
ip_vs_icmp_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
- struct ip_vs_protocol *pp, int offset, unsigned int hooknum)
+ struct ip_vs_protocol *pp, int offset, unsigned int hooknum,
+ struct ip_vs_iphdr *iph)
{
struct rt6_info *rt; /* Route to the other host */
int mtu;
int rc;
int local;
int rt_mode;
- struct ip_vs_iphdr iph;
EnterFunction(10);
- ip_vs_fill_iph_skb(cp->af, skb, &iph);
/* The ICMP packet for VS/TUN, VS/DR and LOCALNODE will be
forwarded directly here, because there is no need to
translate address/port back */
if (IP_VS_FWD_METHOD(cp) != IP_VS_CONN_F_MASQ) {
if (cp->packet_xmit)
- rc = cp->packet_xmit(skb, cp, pp);
+ rc = cp->packet_xmit(skb, cp, pp, iph);
else
rc = NF_ACCEPT;
/* do not touch skb anymore */
@@ -1388,7 +1381,7 @@ ip_vs_icmp_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
skb->dev = net->loopback_dev;
}
/* only send ICMP too big on first fragment */
- if (!iph.fragoffs)
+ if (!iph->fragoffs)
icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
IP_VS_DBG_RL("%s(): frag needed\n", __func__);
goto tx_error_put;
diff --git a/net/netfilter/xt_ipvs.c b/net/netfilter/xt_ipvs.c
index 3f9b8cd..8d47c37 100644
--- a/net/netfilter/xt_ipvs.c
+++ b/net/netfilter/xt_ipvs.c
@@ -85,7 +85,7 @@ ipvs_mt(const struct sk_buff *skb, struct xt_action_param *par)
/*
* Check if the packet belongs to an existing entry
*/
- cp = pp->conn_out_get(family, skb, &iph, iph.len, 1 /* inverse */);
+ cp = pp->conn_out_get(family, skb, &iph, 1 /* inverse */);
if (unlikely(cp == NULL)) {
match = false;
goto out;
^ permalink raw reply related
* [PATCH V3 6/8] ipvs: Complete IPv6 fragment handling for IPVS
From: Jesper Dangaard Brouer @ 2012-09-11 12:38 UTC (permalink / raw)
To: Hans Schillstrom, Hans Schillstrom, netdev, Patrick McHardy,
Pablo Neira Ayuso, lvs-devel, Julian Anastasov
Cc: Jesper Dangaard Brouer, Thomas Graf, Wensong Zhang,
netfilter-devel, Simon Horman
In-Reply-To: <20120911123531.4305.40304.stgit@dragon>
IPVS now supports fragmented packets, with support from nf_conntrack_reasm.c
Based on patch from: Hans Schillstrom.
IPVS do like conntrack i.e. use the skb->nfct_reasm
(i.e. when all fragments is collected, nf_ct_frag6_output()
starts a "re-play" of all fragments into the interrupted
PREROUTING chain at prio -399 (NF_IP6_PRI_CONNTRACK_DEFRAG+1)
with nfct_reasm pointing to the assembled packet.)
Notice, module nf_defrag_ipv6 must be loaded for this to work.
Report unhandled fragments, and recommend user to load nf_defrag_ipv6.
To handle fw-mark for fragments. Add a new IPVS hook into prerouting
chain at prio -99 (NF_IP6_PRI_NAT_DST+1) to catch fragments, and copy
fw-mark info from the first packet with an upper layer header.
IPv6 fragment handling should be the last thing on the IPVS IPv6
missing support list.
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: Hans Schillstrom <hans@schillstrom.com>
---
V3:
- In case of no nf_defrag_ipv6, second fragments could create
strange conn entries, fix that.
- Report unhandled fragments, and recommend user to load nf_defrag_ipv6.
- Move ICMPv6 improvements to seperate patch
V2:
- In ip_vs_in_icmp_v6() hint that &ciph.len can be updated
- Add NULL pointer check in ip_vs_in_icmp_v6()
V1:
- Fixed refcnt bug since last.
include/net/ip_vs.h | 39 +++++++++++++
net/netfilter/ipvs/Kconfig | 6 +-
net/netfilter/ipvs/ip_vs_conn.c | 2 -
net/netfilter/ipvs/ip_vs_core.c | 117 ++++++++++++++++++++++++++++++++-------
net/netfilter/ipvs/ip_vs_xmit.c | 33 ++++++++---
5 files changed, 162 insertions(+), 35 deletions(-)
diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h
index 29265bf..98806b6 100644
--- a/include/net/ip_vs.h
+++ b/include/net/ip_vs.h
@@ -109,6 +109,7 @@ extern int ip_vs_conn_tab_size;
struct ip_vs_iphdr {
__u32 len; /* IPv4 simply where L4 starts
IPv6 where L4 Transport Header starts */
+ __u32 thoff_reasm; /* Transport Header Offset in nfct_reasm skb */
__u16 fragoffs; /* IPv6 fragment offset, 0 if first frag (or not frag)*/
__s16 protocol;
__s32 flags;
@@ -116,6 +117,35 @@ struct ip_vs_iphdr {
union nf_inet_addr daddr;
};
+/* Dependency to module: nf_defrag_ipv6 */
+#if defined(CONFIG_NF_DEFRAG_IPV6) || defined(CONFIG_NF_DEFRAG_IPV6_MODULE)
+static inline struct sk_buff *skb_nfct_reasm(const struct sk_buff *skb)
+{
+ return skb->nfct_reasm;
+}
+static inline void *frag_safe_skb_hp(const struct sk_buff *skb, int offset,
+ int len, void *buffer,
+ const struct ip_vs_iphdr *ipvsh)
+{
+ if (unlikely(ipvsh->fragoffs && skb_nfct_reasm(skb)))
+ return skb_header_pointer(skb_nfct_reasm(skb),
+ ipvsh->thoff_reasm, len, buffer);
+
+ return skb_header_pointer(skb, offset, len, buffer);
+}
+#else
+static inline struct sk_buff *skb_nfct_reasm(const struct sk_buff *skb)
+{
+ return NULL;
+}
+static inline void *frag_safe_skb_hp(const struct sk_buff *skb, int offset,
+ int len, void *buffer,
+ const struct ip_vs_iphdr *ipvsh)
+{
+ return skb_header_pointer(skb, offset, len, buffer);
+}
+#endif
+
static inline void
ip_vs_fill_ip4hdr(const void *nh, struct ip_vs_iphdr *iphdr)
{
@@ -141,12 +171,19 @@ ip_vs_fill_iph_skb(int af, const struct sk_buff *skb, struct ip_vs_iphdr *iphdr)
(struct ipv6hdr *)skb_network_header(skb);
iphdr->saddr.in6 = iph->saddr;
iphdr->daddr.in6 = iph->daddr;
- /* ipv6_find_hdr() updates len, flags */
+ /* ipv6_find_hdr() updates len, flags, thoff_reasm */
+ iphdr->thoff_reasm = 0;
iphdr->len = 0;
iphdr->flags = 0;
iphdr->protocol = ipv6_find_hdr(skb, &iphdr->len, -1,
&iphdr->fragoffs,
&iphdr->flags);
+ /* get proto from re-assembled packet and it's offset */
+ if (skb_nfct_reasm(skb))
+ iphdr->protocol = ipv6_find_hdr(skb_nfct_reasm(skb),
+ &iphdr->thoff_reasm,
+ -1, NULL, NULL);
+
} else
#endif
{
diff --git a/net/netfilter/ipvs/Kconfig b/net/netfilter/ipvs/Kconfig
index a97ae53..0c3b167 100644
--- a/net/netfilter/ipvs/Kconfig
+++ b/net/netfilter/ipvs/Kconfig
@@ -30,11 +30,9 @@ config IP_VS_IPV6
depends on IPV6 = y || IP_VS = IPV6
select IP6_NF_IPTABLES
---help---
- Add IPv6 support to IPVS. This is incomplete and might be dangerous.
+ Add IPv6 support to IPVS.
- See http://www.mindbasket.com/ipvs for more information.
-
- Say N if unsure.
+ Say Y if unsure.
config IP_VS_DEBUG
bool "IP virtual server debugging"
diff --git a/net/netfilter/ipvs/ip_vs_conn.c b/net/netfilter/ipvs/ip_vs_conn.c
index 1548df9..d6c1c26 100644
--- a/net/netfilter/ipvs/ip_vs_conn.c
+++ b/net/netfilter/ipvs/ip_vs_conn.c
@@ -314,7 +314,7 @@ ip_vs_conn_fill_param_proto(int af, const struct sk_buff *skb,
__be16 _ports[2], *pptr;
struct net *net = skb_net(skb);
- pptr = skb_header_pointer(skb, proto_off, sizeof(_ports), _ports);
+ pptr = frag_safe_skb_hp(skb, proto_off, sizeof(_ports), _ports, iph);
if (pptr == NULL)
return 1;
diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c
index caed44d..4b9995e 100644
--- a/net/netfilter/ipvs/ip_vs_core.c
+++ b/net/netfilter/ipvs/ip_vs_core.c
@@ -402,8 +402,12 @@ ip_vs_schedule(struct ip_vs_service *svc, struct sk_buff *skb,
unsigned int flags;
*ignored = 1;
+
+ /*
+ * IPv6 frags, only the first hit here.
+ */
ip_vs_fill_iph_skb(svc->af, skb, &iph);
- pptr = skb_header_pointer(skb, iph.len, sizeof(_ports), _ports);
+ pptr = frag_safe_skb_hp(skb, iph.len, sizeof(_ports), _ports, &iph);
if (pptr == NULL)
return NULL;
@@ -507,8 +511,7 @@ int ip_vs_leave(struct ip_vs_service *svc, struct sk_buff *skb,
#endif
ip_vs_fill_iph_skb(svc->af, skb, &iph);
-
- pptr = skb_header_pointer(skb, iph.len, sizeof(_ports), _ports);
+ pptr = frag_safe_skb_hp(skb, iph.len, sizeof(_ports), _ports, &iph);
if (pptr == NULL) {
ip_vs_service_put(svc);
return NF_DROP;
@@ -654,14 +657,6 @@ static inline int ip_vs_gather_frags(struct sk_buff *skb, u_int32_t user)
return err;
}
-#ifdef CONFIG_IP_VS_IPV6
-static inline int ip_vs_gather_frags_v6(struct sk_buff *skb, u_int32_t user)
-{
- /* TODO IPv6: Find out what to do here for IPv6 */
- return 0;
-}
-#endif
-
static int ip_vs_route_me_harder(int af, struct sk_buff *skb)
{
#ifdef CONFIG_IP_VS_IPV6
@@ -941,8 +936,7 @@ static int ip_vs_out_icmp_v6(struct sk_buff *skb, int *related,
ip_vs_fill_iph_skb(AF_INET6, skb, ipvsh);
*related = 1;
-
- ic = skb_header_pointer(skb, ipvsh->len, sizeof(_icmph), &_icmph);
+ ic = frag_safe_skb_hp(skb, ipvsh->len, sizeof(_icmph), &_icmph, ipvsh);
if (ic == NULL)
return NF_DROP;
@@ -961,6 +955,11 @@ static int ip_vs_out_icmp_v6(struct sk_buff *skb, int *related,
*related = 0;
return NF_ACCEPT;
}
+ /* Fragment header that is before ICMP header tells us that:
+ * it's not an error message since they can't be fragmented.
+ */
+ if (ipvsh->flags & IP6T_FH_F_FRAG)
+ return NF_DROP;
/* Now find the contained IP header */
ipvsh->len += sizeof(_icmph);
@@ -1117,6 +1116,12 @@ ip_vs_out(unsigned int hooknum, struct sk_buff *skb, int af)
ip_vs_fill_iph_skb(af, skb, &iph);
#ifdef CONFIG_IP_VS_IPV6
if (af == AF_INET6) {
+ if (!iph.fragoffs && skb_nfct_reasm(skb)) {
+ struct sk_buff *reasm = skb_nfct_reasm(skb);
+ /* Save fw mark for coming frags */
+ reasm->ipvs_property = 1;
+ reasm->mark = skb->mark;
+ }
if (unlikely(iph.protocol == IPPROTO_ICMPV6)) {
int related;
int verdict = ip_vs_out_icmp_v6(skb, &related,
@@ -1124,7 +1129,6 @@ ip_vs_out(unsigned int hooknum, struct sk_buff *skb, int af)
if (related)
return verdict;
- ip_vs_fill_iph_skb(af, skb, &iph);
}
} else
#endif
@@ -1134,7 +1138,6 @@ ip_vs_out(unsigned int hooknum, struct sk_buff *skb, int af)
if (related)
return verdict;
- ip_vs_fill_ip4hdr(skb_network_header(skb), &iph);
}
pd = ip_vs_proto_data_get(net, iph.protocol);
@@ -1167,8 +1170,8 @@ ip_vs_out(unsigned int hooknum, struct sk_buff *skb, int af)
pp->protocol == IPPROTO_SCTP)) {
__be16 _ports[2], *pptr;
- pptr = skb_header_pointer(skb, iph.len,
- sizeof(_ports), _ports);
+ pptr = frag_safe_skb_hp(skb, iph.len,
+ sizeof(_ports), _ports, &iph);
if (pptr == NULL)
return NF_ACCEPT; /* Not for me */
if (ip_vs_lookup_real_service(net, af, iph.protocol,
@@ -1468,7 +1471,7 @@ ip_vs_in_icmp_v6(struct sk_buff *skb, int *related, unsigned int hooknum)
*related = 1;
- ic = skb_header_pointer(skb, iph->len, sizeof(_icmph), &_icmph);
+ ic = frag_safe_skb_hp(skb, iph->len, sizeof(_icmph), &_icmph, iph);
if (ic == NULL)
return NF_DROP;
@@ -1487,6 +1490,11 @@ ip_vs_in_icmp_v6(struct sk_buff *skb, int *related, unsigned int hooknum)
*related = 0;
return NF_ACCEPT;
}
+ /* Fragment header that is before ICMP header tells us that:
+ * it's not an error message since they can't be fragmented.
+ */
+ if (iph->flags & IP6T_FH_F_FRAG)
+ return NF_DROP;
/* Now find the contained IP header */
ciph.len = iph->len + sizeof(_icmph);
@@ -1511,10 +1519,20 @@ ip_vs_in_icmp_v6(struct sk_buff *skb, int *related, unsigned int hooknum)
IP_VS_DBG_PKT(11, AF_INET6, pp, skb, offs_ciph,
"Checking incoming ICMPv6 for");
- /* The embedded headers contain source and dest in reverse order */
- cp = pp->conn_in_get(AF_INET6, skb, &ciph, ciph.len, 1);
+ /* The embedded headers contain source and dest in reverse order
+ * if not from localhost
+ */
+ cp = pp->conn_in_get(AF_INET6, skb, &ciph, ciph.len,
+ (hooknum == NF_INET_LOCAL_OUT) ? 0 : 1);
+
if (!cp)
return NF_ACCEPT;
+ /* VS/TUN, VS/DR and LOCALNODE just let it go */
+ if ((hooknum == NF_INET_LOCAL_OUT) &&
+ (IP_VS_FWD_METHOD(cp) != IP_VS_CONN_F_MASQ)) {
+ __ip_vs_conn_put(cp);
+ return NF_ACCEPT;
+ }
/* do the statistics and put it back */
ip_vs_in_stats(cp, skb);
@@ -1584,6 +1602,12 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af)
#ifdef CONFIG_IP_VS_IPV6
if (af == AF_INET6) {
+ if (!iph.fragoffs && skb_nfct_reasm(skb)) {
+ struct sk_buff *reasm = skb_nfct_reasm(skb);
+ /* Save fw mark for coming frags. */
+ reasm->ipvs_property = 1;
+ reasm->mark = skb->mark;
+ }
if (unlikely(iph.protocol == IPPROTO_ICMPV6)) {
int related;
int verdict = ip_vs_in_icmp_v6(skb, &related, hooknum);
@@ -1608,13 +1632,16 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af)
pp = pd->pp;
/*
* Check if the packet belongs to an existing connection entry
- * Only sched first IPv6 fragment.
*/
cp = pp->conn_in_get(af, skb, &iph, iph.len, 0);
if (unlikely(!cp) && !iph.fragoffs) {
+ /* No (second) fragments need to enter here, as nf_defrag_ipv6
+ * replayed fragment zero will already have created the cp
+ */
int v;
+ /* Schedule and create new connection entry into &cp */
if (!pp->conn_schedule(af, skb, pd, &v, &cp))
return v;
}
@@ -1623,6 +1650,14 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af)
/* sorry, all this trouble for a no-hit :) */
IP_VS_DBG_PKT(12, af, pp, skb, 0,
"ip_vs_in: packet continues traversal as normal");
+ if (iph.fragoffs && !skb_nfct_reasm(skb)) {
+ /* Fragment that couldn't be mapped to a conn entry
+ * and don't have any pointer to a reasm skb
+ * is missing module nf_defrag_ipv6
+ */
+ IP_VS_DBG_RL("Unhandled frag, load nf_defrag_ipv6\n");
+ IP_VS_DBG_PKT(7, af, pp, skb, 0, "unhandled fragment");
+ }
return NF_ACCEPT;
}
@@ -1707,6 +1742,38 @@ ip_vs_local_request4(unsigned int hooknum, struct sk_buff *skb,
#ifdef CONFIG_IP_VS_IPV6
/*
+ * AF_INET6 fragment handling
+ * Copy info from first fragment, to the rest of them.
+ */
+static unsigned int
+ip_vs_preroute_frag6(unsigned int hooknum, struct sk_buff *skb,
+ const struct net_device *in,
+ const struct net_device *out,
+ int (*okfn)(struct sk_buff *))
+{
+ struct sk_buff *reasm = skb_nfct_reasm(skb);
+ struct net *net;
+
+ /* Skip if not a "replay" from nf_ct_frag6_output or first fragment.
+ * ipvs_property is set when checking first fragment
+ * in ip_vs_in() and ip_vs_out().
+ */
+ if (reasm)
+ IP_VS_DBG(2, "Fragment recv prop:%d\n", reasm->ipvs_property);
+ if (!reasm || !reasm->ipvs_property)
+ return NF_ACCEPT;
+
+ net = skb_net(skb);
+ if (!net_ipvs(net)->enable)
+ return NF_ACCEPT;
+
+ /* Copy stored fw mark, saved in ip_vs_{in,out} */
+ skb->mark = reasm->mark;
+
+ return NF_ACCEPT;
+}
+
+/*
* AF_INET6 handler in NF_INET_LOCAL_IN chain
* Schedule and forward packets from remote clients
*/
@@ -1845,6 +1912,14 @@ static struct nf_hook_ops ip_vs_ops[] __read_mostly = {
.priority = 100,
},
#ifdef CONFIG_IP_VS_IPV6
+ /* After mangle & nat fetch 2:nd fragment and following */
+ {
+ .hook = ip_vs_preroute_frag6,
+ .owner = THIS_MODULE,
+ .pf = NFPROTO_IPV6,
+ .hooknum = NF_INET_PRE_ROUTING,
+ .priority = NF_IP6_PRI_NAT_DST + 1,
+ },
/* After packet filtering, change source only for VS/NAT */
{
.hook = ip_vs_reply6,
diff --git a/net/netfilter/ipvs/ip_vs_xmit.c b/net/netfilter/ipvs/ip_vs_xmit.c
index 428de75..e44933f 100644
--- a/net/netfilter/ipvs/ip_vs_xmit.c
+++ b/net/netfilter/ipvs/ip_vs_xmit.c
@@ -496,12 +496,13 @@ ip_vs_bypass_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
struct ip_vs_protocol *pp)
{
struct rt6_info *rt; /* Route to the other host */
- struct ipv6hdr *iph = ipv6_hdr(skb);
+ struct ip_vs_iphdr iph;
int mtu;
EnterFunction(10);
+ ip_vs_fill_iph_skb(cp->af, skb, &iph);
- if (!(rt = __ip_vs_get_out_rt_v6(skb, NULL, &iph->daddr, NULL, 0,
+ if (!(rt = __ip_vs_get_out_rt_v6(skb, NULL, &iph.daddr.in6, NULL, 0,
IP_VS_RT_MODE_NON_LOCAL)))
goto tx_error_icmp;
@@ -513,7 +514,9 @@ ip_vs_bypass_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
skb->dev = net->loopback_dev;
}
- icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
+ /* only send ICMP too big on first fragment */
+ if (!iph.fragoffs)
+ icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
dst_release(&rt->dst);
IP_VS_DBG_RL("%s(): frag needed\n", __func__);
goto tx_error;
@@ -685,7 +688,7 @@ ip_vs_nat_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
ip_vs_fill_iph_skb(cp->af, skb, &iph);
/* check if it is a connection of no-client-port */
- if (unlikely(cp->flags & IP_VS_CONN_F_NO_CPORT)) {
+ if (unlikely(cp->flags & IP_VS_CONN_F_NO_CPORT && !iph.fragoffs)) {
__be16 _pt, *p;
p = skb_header_pointer(skb, iph.len, sizeof(_pt), &_pt);
if (p == NULL)
@@ -735,7 +738,9 @@ ip_vs_nat_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
skb->dev = net->loopback_dev;
}
- icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
+ /* only send ICMP too big on first fragment */
+ if (!iph.fragoffs)
+ icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
IP_VS_DBG_RL_PKT(0, AF_INET6, pp, skb, 0,
"ip_vs_nat_xmit_v6(): frag needed for");
goto tx_error_put;
@@ -940,8 +945,10 @@ ip_vs_tunnel_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
unsigned int max_headroom; /* The extra header space needed */
int mtu;
int ret;
+ struct ip_vs_iphdr ipvsh;
EnterFunction(10);
+ ip_vs_fill_iph_skb(cp->af, skb, &ipvsh);
if (!(rt = __ip_vs_get_out_rt_v6(skb, cp->dest, &cp->daddr.in6,
&saddr, 1, (IP_VS_RT_MODE_LOCAL |
@@ -970,7 +977,9 @@ ip_vs_tunnel_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
skb->dev = net->loopback_dev;
}
- icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
+ /* only send ICMP too big on first fragment */
+ if (!ipvsh.fragoffs)
+ icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
IP_VS_DBG_RL("%s(): frag needed\n", __func__);
goto tx_error_put;
}
@@ -1116,8 +1125,10 @@ ip_vs_dr_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
{
struct rt6_info *rt; /* Route to the other host */
int mtu;
+ struct ip_vs_iphdr iph;
EnterFunction(10);
+ ip_vs_fill_iph_skb(cp->af, skb, &iph);
if (!(rt = __ip_vs_get_out_rt_v6(skb, cp->dest, &cp->daddr.in6, NULL,
0, (IP_VS_RT_MODE_LOCAL |
@@ -1136,7 +1147,9 @@ ip_vs_dr_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
skb->dev = net->loopback_dev;
}
- icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
+ /* only send ICMP too big on first fragment */
+ if (!iph.fragoffs)
+ icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
dst_release(&rt->dst);
IP_VS_DBG_RL("%s(): frag needed\n", __func__);
goto tx_error;
@@ -1308,8 +1321,10 @@ ip_vs_icmp_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
int rc;
int local;
int rt_mode;
+ struct ip_vs_iphdr iph;
EnterFunction(10);
+ ip_vs_fill_iph_skb(cp->af, skb, &iph);
/* The ICMP packet for VS/TUN, VS/DR and LOCALNODE will be
forwarded directly here, because there is no need to
@@ -1372,7 +1387,9 @@ ip_vs_icmp_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
skb->dev = net->loopback_dev;
}
- icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
+ /* only send ICMP too big on first fragment */
+ if (!iph.fragoffs)
+ icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
IP_VS_DBG_RL("%s(): frag needed\n", __func__);
goto tx_error_put;
}
^ permalink raw reply related
* [PATCH V3 5/8] ipvs: Fix faulty IPv6 extension header handling in IPVS
From: Jesper Dangaard Brouer @ 2012-09-11 12:37 UTC (permalink / raw)
To: Hans Schillstrom, Hans Schillstrom, netdev, Patrick McHardy,
Pablo Neira Ayuso, lvs-devel, Julian Anastasov
Cc: Jesper Dangaard Brouer, Thomas Graf, Wensong Zhang,
netfilter-devel, Simon Horman
In-Reply-To: <20120911123531.4305.40304.stgit@dragon>
IPv6 packets can contain extension headers, thus its wrong to assume
that the transport/upper-layer header, starts right after (struct
ipv6hdr) the IPv6 header. IPVS uses this false assumption, and will
write SNAT & DNAT modifications at a fixed pos which will corrupt the
message.
To fix this, proper header position must be found before modifying
packets. Introducing ip_vs_fill_iph_skb(), which uses ipv6_find_hdr()
to skip the exthdrs. It finds (1) the transport header offset, (2) the
protocol, and (3) detects if the packet is a fragment.
Note, that fragments in IPv6 is represented via an exthdr. Thus, this
is detected while skipping through the exthdrs.
This patch depends on commit 84018f55a:
"netfilter: ip6_tables: add flags parameter to ipv6_find_hdr()"
This also adds a dependency to ip6_tables.
Originally based on patch from: Hans Schillstrom
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
---
V3:
- No big API changes (these are postpone to later patch)
- Make ip_vs_in_icmp_v6() easier to review, avoid spliting related
changes across patches.
- Fix port write offset bug in ip_vs_in_icmp_v6()
- Move ip_vs_pe_sip.c changes to a seperate patch.
V2:
- Use macro IS_ENABLED(CONFIG_XXX)
- Re-add NULL pointer check in ip_vs_out_icmp_v6()
include/net/ip_vs.h | 72 +++++++++++-
net/netfilter/ipvs/Kconfig | 1
net/netfilter/ipvs/ip_vs_core.c | 191 +++++++++++++++------------------
net/netfilter/ipvs/ip_vs_dh.c | 2
net/netfilter/ipvs/ip_vs_lblc.c | 2
net/netfilter/ipvs/ip_vs_lblcr.c | 2
net/netfilter/ipvs/ip_vs_pe_sip.c | 2
net/netfilter/ipvs/ip_vs_proto_sctp.c | 22 ++--
net/netfilter/ipvs/ip_vs_proto_tcp.c | 22 ++--
net/netfilter/ipvs/ip_vs_proto_udp.c | 22 ++--
net/netfilter/ipvs/ip_vs_sh.c | 2
net/netfilter/ipvs/ip_vs_xmit.c | 5 +
net/netfilter/xt_ipvs.c | 2
13 files changed, 198 insertions(+), 149 deletions(-)
diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h
index c8b2bdb..29265bf 100644
--- a/include/net/ip_vs.h
+++ b/include/net/ip_vs.h
@@ -22,6 +22,9 @@
#include <linux/ip.h>
#include <linux/ipv6.h> /* for struct ipv6hdr */
#include <net/ipv6.h>
+#if IS_ENABLED(CONFIG_IPV6)
+#include <linux/netfilter_ipv6/ip6_tables.h>
+#endif
#if IS_ENABLED(CONFIG_NF_CONNTRACK)
#include <net/netfilter/nf_conntrack.h>
#endif
@@ -103,30 +106,79 @@ static inline struct net *seq_file_single_net(struct seq_file *seq)
/* Connections' size value needed by ip_vs_ctl.c */
extern int ip_vs_conn_tab_size;
-
struct ip_vs_iphdr {
- int len;
- __u8 protocol;
+ __u32 len; /* IPv4 simply where L4 starts
+ IPv6 where L4 Transport Header starts */
+ __u16 fragoffs; /* IPv6 fragment offset, 0 if first frag (or not frag)*/
+ __s16 protocol;
+ __s32 flags;
union nf_inet_addr saddr;
union nf_inet_addr daddr;
};
static inline void
-ip_vs_fill_iphdr(int af, const void *nh, struct ip_vs_iphdr *iphdr)
+ip_vs_fill_ip4hdr(const void *nh, struct ip_vs_iphdr *iphdr)
+{
+ const struct iphdr *iph = nh;
+
+ iphdr->len = iph->ihl * 4;
+ iphdr->fragoffs = 0;
+ iphdr->protocol = iph->protocol;
+ iphdr->saddr.ip = iph->saddr;
+ iphdr->daddr.ip = iph->daddr;
+}
+
+/* This function handles filling *ip_vs_iphdr, both for IPv4 and IPv6.
+ * IPv6 requires some extra work, as finding proper header position,
+ * depend on the IPv6 extension headers.
+ */
+static inline void
+ip_vs_fill_iph_skb(int af, const struct sk_buff *skb, struct ip_vs_iphdr *iphdr)
{
#ifdef CONFIG_IP_VS_IPV6
if (af == AF_INET6) {
- const struct ipv6hdr *iph = nh;
- iphdr->len = sizeof(struct ipv6hdr);
- iphdr->protocol = iph->nexthdr;
+ const struct ipv6hdr *iph =
+ (struct ipv6hdr *)skb_network_header(skb);
iphdr->saddr.in6 = iph->saddr;
iphdr->daddr.in6 = iph->daddr;
+ /* ipv6_find_hdr() updates len, flags */
+ iphdr->len = 0;
+ iphdr->flags = 0;
+ iphdr->protocol = ipv6_find_hdr(skb, &iphdr->len, -1,
+ &iphdr->fragoffs,
+ &iphdr->flags);
} else
#endif
{
- const struct iphdr *iph = nh;
- iphdr->len = iph->ihl * 4;
- iphdr->protocol = iph->protocol;
+ const struct iphdr *iph =
+ (struct iphdr *)skb_network_header(skb);
+ iphdr->len = iph->ihl * 4;
+ iphdr->fragoffs = 0;
+ iphdr->protocol = iph->protocol;
+ iphdr->saddr.ip = iph->saddr;
+ iphdr->daddr.ip = iph->daddr;
+ }
+}
+
+/* This function is a faster version of ip_vs_fill_iph_skb().
+ * Where we only populate {s,d}addr (and avoid calling ipv6_find_hdr()).
+ * This is used by the some of the ip_vs_*_schedule() functions.
+ * (Mostly done to avoid ABI breakage of external schedulers)
+ */
+static inline void
+ip_vs_fill_iph_addr_only(int af, const struct sk_buff *skb,
+ struct ip_vs_iphdr *iphdr)
+{
+#ifdef CONFIG_IP_VS_IPV6
+ if (af == AF_INET6) {
+ const struct ipv6hdr *iph =
+ (struct ipv6hdr *)skb_network_header(skb);
+ iphdr->saddr.in6 = iph->saddr;
+ iphdr->daddr.in6 = iph->daddr;
+ } else {
+#endif
+ const struct iphdr *iph =
+ (struct iphdr *)skb_network_header(skb);
iphdr->saddr.ip = iph->saddr;
iphdr->daddr.ip = iph->daddr;
}
diff --git a/net/netfilter/ipvs/Kconfig b/net/netfilter/ipvs/Kconfig
index 8b2cffd..a97ae53 100644
--- a/net/netfilter/ipvs/Kconfig
+++ b/net/netfilter/ipvs/Kconfig
@@ -28,6 +28,7 @@ if IP_VS
config IP_VS_IPV6
bool "IPv6 support for IPVS"
depends on IPV6 = y || IP_VS = IPV6
+ select IP6_NF_IPTABLES
---help---
Add IPv6 support to IPVS. This is incomplete and might be dangerous.
diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c
index fd50f47..caed44d 100644
--- a/net/netfilter/ipvs/ip_vs_core.c
+++ b/net/netfilter/ipvs/ip_vs_core.c
@@ -236,7 +236,7 @@ ip_vs_sched_persist(struct ip_vs_service *svc,
union nf_inet_addr snet; /* source network of the client,
after masking */
- ip_vs_fill_iphdr(svc->af, skb_network_header(skb), &iph);
+ ip_vs_fill_iph_skb(svc->af, skb, &iph);
/* Mask saddr with the netmask to adjust template granularity */
#ifdef CONFIG_IP_VS_IPV6
@@ -402,7 +402,7 @@ ip_vs_schedule(struct ip_vs_service *svc, struct sk_buff *skb,
unsigned int flags;
*ignored = 1;
- ip_vs_fill_iphdr(svc->af, skb_network_header(skb), &iph);
+ ip_vs_fill_iph_skb(svc->af, skb, &iph);
pptr = skb_header_pointer(skb, iph.len, sizeof(_ports), _ports);
if (pptr == NULL)
return NULL;
@@ -506,7 +506,7 @@ int ip_vs_leave(struct ip_vs_service *svc, struct sk_buff *skb,
int unicast;
#endif
- ip_vs_fill_iphdr(svc->af, skb_network_header(skb), &iph);
+ ip_vs_fill_iph_skb(svc->af, skb, &iph);
pptr = skb_header_pointer(skb, iph.len, sizeof(_ports), _ports);
if (pptr == NULL) {
@@ -732,15 +732,21 @@ void ip_vs_nat_icmp_v6(struct sk_buff *skb, struct ip_vs_protocol *pp,
struct ip_vs_conn *cp, int inout)
{
struct ipv6hdr *iph = ipv6_hdr(skb);
- unsigned int icmp_offset = sizeof(struct ipv6hdr);
- struct icmp6hdr *icmph = (struct icmp6hdr *)(skb_network_header(skb) +
- icmp_offset);
- struct ipv6hdr *ciph = (struct ipv6hdr *)(icmph + 1);
+ unsigned int icmp_offset = 0;
+ unsigned int offs = 0; /* header offset*/
+ int protocol;
+ struct icmp6hdr *icmph;
+ struct ipv6hdr *ciph;
+ unsigned short fragoffs;
+
+ ipv6_find_hdr(skb, &icmp_offset, IPPROTO_ICMPV6, &fragoffs, NULL);
+ icmph = (struct icmp6hdr *)(skb_network_header(skb) + icmp_offset);
+ offs = icmp_offset + sizeof(struct icmp6hdr);
+ ciph = (struct ipv6hdr *)(skb_network_header(skb) + offs);
+
+ protocol = ipv6_find_hdr(skb, &offs, -1, &fragoffs, NULL);
- /* Make sure SKB is writable */
- unsigned int write;
- write = icmp_offset + sizeof(struct icmp6hdr) + sizeof(struct ipv6hdr);
- if (!skb_make_writable(skb, write + 2 * sizeof(__u16)))
+ if (!skb_make_writable(skb, offs + 2 * sizeof(__u16)))
return;
if (inout) {
@@ -752,10 +758,13 @@ void ip_vs_nat_icmp_v6(struct sk_buff *skb, struct ip_vs_protocol *pp,
}
/* the TCP/UDP/SCTP port */
- if (IPPROTO_TCP == ciph->nexthdr || IPPROTO_UDP == ciph->nexthdr ||
- IPPROTO_SCTP == ciph->nexthdr) {
- __be16 *ports = (void *)ciph + sizeof(struct ipv6hdr);
+ if (!fragoffs && (IPPROTO_TCP == protocol || IPPROTO_UDP == protocol ||
+ IPPROTO_SCTP == protocol)) {
+ __be16 *ports = (void *)(skb_network_header(skb) + offs);
+ IP_VS_DBG(11, "%s() changed port %d to %d\n", __func__,
+ ntohs(inout ? ports[1] : ports[0]),
+ ntohs(inout ? cp->vport : cp->dport));
if (inout)
ports[1] = cp->vport;
else
@@ -904,9 +913,8 @@ static int ip_vs_out_icmp(struct sk_buff *skb, int *related,
IP_VS_DBG_PKT(11, AF_INET, pp, skb, offset,
"Checking outgoing ICMP for");
- offset += cih->ihl * 4;
-
- ip_vs_fill_iphdr(AF_INET, cih, &ciph);
+ ip_vs_fill_ip4hdr(cih, &ciph);
+ ciph.len += offset;
/* The embedded headers contain source and dest in reverse order */
cp = pp->conn_out_get(AF_INET, skb, &ciph, offset, 1);
if (!cp)
@@ -914,40 +922,33 @@ static int ip_vs_out_icmp(struct sk_buff *skb, int *related,
snet.ip = iph->saddr;
return handle_response_icmp(AF_INET, skb, &snet, cih->protocol, cp,
- pp, offset, ihl);
+ pp, ciph.len, ihl);
}
#ifdef CONFIG_IP_VS_IPV6
static int ip_vs_out_icmp_v6(struct sk_buff *skb, int *related,
unsigned int hooknum)
{
- struct ipv6hdr *iph;
struct icmp6hdr _icmph, *ic;
- struct ipv6hdr _ciph, *cih; /* The ip header contained
+ struct ipv6hdr _ip6, *ip6; /* The ip header contained
within the ICMP */
- struct ip_vs_iphdr ciph;
struct ip_vs_conn *cp;
struct ip_vs_protocol *pp;
- unsigned int offset;
union nf_inet_addr snet;
- *related = 1;
+ struct ip_vs_iphdr ipvsh_stack;
+ struct ip_vs_iphdr *ipvsh = &ipvsh_stack;
+ ip_vs_fill_iph_skb(AF_INET6, skb, ipvsh);
- /* reassemble IP fragments */
- if (ipv6_hdr(skb)->nexthdr == IPPROTO_FRAGMENT) {
- if (ip_vs_gather_frags_v6(skb, ip_vs_defrag_user(hooknum)))
- return NF_STOLEN;
- }
+ *related = 1;
- iph = ipv6_hdr(skb);
- offset = sizeof(struct ipv6hdr);
- ic = skb_header_pointer(skb, offset, sizeof(_icmph), &_icmph);
+ ic = skb_header_pointer(skb, ipvsh->len, sizeof(_icmph), &_icmph);
if (ic == NULL)
return NF_DROP;
- IP_VS_DBG(12, "Outgoing ICMPv6 (%d,%d) %pI6->%pI6\n",
+ IP_VS_DBG(12, "Outgoing ICMPv6 (%d,%d) %pI6c->%pI6c\n",
ic->icmp6_type, ntohs(icmpv6_id(ic)),
- &iph->saddr, &iph->daddr);
+ &ipvsh->saddr, &ipvsh->daddr);
/*
* Work through seeing if this is for us.
@@ -962,34 +963,28 @@ static int ip_vs_out_icmp_v6(struct sk_buff *skb, int *related,
}
/* Now find the contained IP header */
- offset += sizeof(_icmph);
- cih = skb_header_pointer(skb, offset, sizeof(_ciph), &_ciph);
- if (cih == NULL)
+ ipvsh->len += sizeof(_icmph);
+ ip6 = skb_header_pointer(skb, ipvsh->len, sizeof(_ip6), &_ip6);
+ if (ip6 == NULL)
return NF_ACCEPT; /* The packet looks wrong, ignore */
+ ipvsh->protocol = ipv6_find_hdr(skb, &ipvsh->len, -1,
+ &ipvsh->fragoffs, &ipvsh->flags);
- pp = ip_vs_proto_get(cih->nexthdr);
- if (!pp)
- return NF_ACCEPT;
-
- /* Is the embedded protocol header present? */
- /* TODO: we don't support fragmentation at the moment anyways */
- if (unlikely(cih->nexthdr == IPPROTO_FRAGMENT && pp->dont_defrag))
+ pp = ip_vs_proto_get(ipvsh->protocol);
+ if (!pp || (ipvsh->protocol < 0))
return NF_ACCEPT;
+ /* fill the rest of ipvsh */
+ ipvsh->saddr.in6 = ip6->saddr;
+ ipvsh->daddr.in6 = ip6->daddr;
- IP_VS_DBG_PKT(11, AF_INET6, pp, skb, offset,
- "Checking outgoing ICMPv6 for");
-
- offset += sizeof(struct ipv6hdr);
-
- ip_vs_fill_iphdr(AF_INET6, cih, &ciph);
/* The embedded headers contain source and dest in reverse order */
- cp = pp->conn_out_get(AF_INET6, skb, &ciph, offset, 1);
+ cp = pp->conn_out_get(AF_INET6, skb, ipvsh, ipvsh->len, 1);
if (!cp)
return NF_ACCEPT;
- snet.in6 = iph->saddr;
- return handle_response_icmp(AF_INET6, skb, &snet, cih->nexthdr, cp,
- pp, offset, sizeof(struct ipv6hdr));
+ snet.in6 = ipvsh->saddr.in6;
+ return handle_response_icmp(AF_INET6, skb, &snet, ipvsh->protocol, cp,
+ pp, ipvsh->len, sizeof(struct ipv6hdr));
}
#endif
@@ -1119,7 +1114,7 @@ ip_vs_out(unsigned int hooknum, struct sk_buff *skb, int af)
if (!net_ipvs(net)->enable)
return NF_ACCEPT;
- ip_vs_fill_iphdr(af, skb_network_header(skb), &iph);
+ ip_vs_fill_iph_skb(af, skb, &iph);
#ifdef CONFIG_IP_VS_IPV6
if (af == AF_INET6) {
if (unlikely(iph.protocol == IPPROTO_ICMPV6)) {
@@ -1129,7 +1124,7 @@ ip_vs_out(unsigned int hooknum, struct sk_buff *skb, int af)
if (related)
return verdict;
- ip_vs_fill_iphdr(af, skb_network_header(skb), &iph);
+ ip_vs_fill_iph_skb(af, skb, &iph);
}
} else
#endif
@@ -1139,7 +1134,7 @@ ip_vs_out(unsigned int hooknum, struct sk_buff *skb, int af)
if (related)
return verdict;
- ip_vs_fill_iphdr(af, skb_network_header(skb), &iph);
+ ip_vs_fill_ip4hdr(skb_network_header(skb), &iph);
}
pd = ip_vs_proto_data_get(net, iph.protocol);
@@ -1149,22 +1144,14 @@ ip_vs_out(unsigned int hooknum, struct sk_buff *skb, int af)
/* reassemble IP fragments */
#ifdef CONFIG_IP_VS_IPV6
- if (af == AF_INET6) {
- if (ipv6_hdr(skb)->nexthdr == IPPROTO_FRAGMENT) {
- if (ip_vs_gather_frags_v6(skb,
- ip_vs_defrag_user(hooknum)))
- return NF_STOLEN;
- }
-
- ip_vs_fill_iphdr(af, skb_network_header(skb), &iph);
- } else
+ if (af == AF_INET)
#endif
if (unlikely(ip_is_fragment(ip_hdr(skb)) && !pp->dont_defrag)) {
if (ip_vs_gather_frags(skb,
ip_vs_defrag_user(hooknum)))
return NF_STOLEN;
- ip_vs_fill_iphdr(af, skb_network_header(skb), &iph);
+ ip_vs_fill_ip4hdr(skb_network_header(skb), &iph);
}
/*
@@ -1379,9 +1366,9 @@ ip_vs_in_icmp(struct sk_buff *skb, int *related, unsigned int hooknum)
"Checking incoming ICMP for");
offset2 = offset;
- offset += cih->ihl * 4;
-
- ip_vs_fill_iphdr(AF_INET, cih, &ciph);
+ ip_vs_fill_ip4hdr(cih, &ciph);
+ ciph.len += offset;
+ offset = ciph.len;
/* The embedded headers contain source and dest in reverse order.
* For IPIP this is error for request, not for reply.
*/
@@ -1467,27 +1454,21 @@ static int
ip_vs_in_icmp_v6(struct sk_buff *skb, int *related, unsigned int hooknum)
{
struct net *net = NULL;
- struct ipv6hdr *iph;
+ struct ipv6hdr _ip6h, *ip6h;
struct icmp6hdr _icmph, *ic;
- struct ipv6hdr _ciph, *cih; /* The ip header contained
- within the ICMP */
struct ip_vs_iphdr ciph;
struct ip_vs_conn *cp;
struct ip_vs_protocol *pp;
struct ip_vs_proto_data *pd;
- unsigned int offset, verdict;
+ unsigned int offs_ciph, verdict;
- *related = 1;
+ struct ip_vs_iphdr iph_stack;
+ struct ip_vs_iphdr *iph = &iph_stack;
+ ip_vs_fill_iph_skb(AF_INET6, skb, iph);
- /* reassemble IP fragments */
- if (ipv6_hdr(skb)->nexthdr == IPPROTO_FRAGMENT) {
- if (ip_vs_gather_frags_v6(skb, ip_vs_defrag_user(hooknum)))
- return NF_STOLEN;
- }
+ *related = 1;
- iph = ipv6_hdr(skb);
- offset = sizeof(struct ipv6hdr);
- ic = skb_header_pointer(skb, offset, sizeof(_icmph), &_icmph);
+ ic = skb_header_pointer(skb, iph->len, sizeof(_icmph), &_icmph);
if (ic == NULL)
return NF_DROP;
@@ -1508,39 +1489,40 @@ ip_vs_in_icmp_v6(struct sk_buff *skb, int *related, unsigned int hooknum)
}
/* Now find the contained IP header */
- offset += sizeof(_icmph);
- cih = skb_header_pointer(skb, offset, sizeof(_ciph), &_ciph);
- if (cih == NULL)
+ ciph.len = iph->len + sizeof(_icmph);
+ offs_ciph = ciph.len; /* Save ip header offset */
+ ip6h = skb_header_pointer(skb, ciph.len, sizeof(_ip6h), (void *)&_ip6h);
+ if (ip6h == NULL)
return NF_ACCEPT; /* The packet looks wrong, ignore */
+ ciph.protocol = ipv6_find_hdr(skb, &ciph.len, -1, &ciph.fragoffs, NULL);
+ ciph.saddr.in6 = ip6h->saddr; /* conn_in_get() handles reverse order */
+ ciph.daddr.in6 = ip6h->daddr;
net = skb_net(skb);
- pd = ip_vs_proto_data_get(net, cih->nexthdr);
+ pd = ip_vs_proto_data_get(net, ciph.protocol);
if (!pd)
return NF_ACCEPT;
pp = pd->pp;
- /* Is the embedded protocol header present? */
- /* TODO: we don't support fragmentation at the moment anyways */
- if (unlikely(cih->nexthdr == IPPROTO_FRAGMENT && pp->dont_defrag))
+ /* Cannot handle fragmented embedded protocol */
+ if (ciph.fragoffs)
return NF_ACCEPT;
- IP_VS_DBG_PKT(11, AF_INET6, pp, skb, offset,
+ IP_VS_DBG_PKT(11, AF_INET6, pp, skb, offs_ciph,
"Checking incoming ICMPv6 for");
- offset += sizeof(struct ipv6hdr);
-
- ip_vs_fill_iphdr(AF_INET6, cih, &ciph);
/* The embedded headers contain source and dest in reverse order */
- cp = pp->conn_in_get(AF_INET6, skb, &ciph, offset, 1);
+ cp = pp->conn_in_get(AF_INET6, skb, &ciph, ciph.len, 1);
if (!cp)
return NF_ACCEPT;
/* do the statistics and put it back */
ip_vs_in_stats(cp, skb);
- if (IPPROTO_TCP == cih->nexthdr || IPPROTO_UDP == cih->nexthdr ||
- IPPROTO_SCTP == cih->nexthdr)
- offset += 2 * sizeof(__u16);
- verdict = ip_vs_icmp_xmit_v6(skb, cp, pp, offset, hooknum);
+ if (IPPROTO_TCP == ciph.protocol || IPPROTO_UDP == ciph.protocol ||
+ IPPROTO_SCTP == ciph.protocol)
+ offs_ciph = ciph.len + (2 * sizeof(__u16));
+
+ verdict = ip_vs_icmp_xmit_v6(skb, cp, pp, offs_ciph, hooknum);
__ip_vs_conn_put(cp);
@@ -1576,7 +1558,7 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af)
if (unlikely((skb->pkt_type != PACKET_HOST &&
hooknum != NF_INET_LOCAL_OUT) ||
!skb_dst(skb))) {
- ip_vs_fill_iphdr(af, skb_network_header(skb), &iph);
+ ip_vs_fill_iph_skb(af, skb, &iph);
IP_VS_DBG_BUF(12, "packet type=%d proto=%d daddr=%s"
" ignored in hook %u\n",
skb->pkt_type, iph.protocol,
@@ -1588,7 +1570,7 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af)
if (!net_ipvs(net)->enable)
return NF_ACCEPT;
- ip_vs_fill_iphdr(af, skb_network_header(skb), &iph);
+ ip_vs_fill_iph_skb(af, skb, &iph);
/* Bad... Do not break raw sockets */
if (unlikely(skb->sk != NULL && hooknum == NF_INET_LOCAL_OUT &&
@@ -1608,7 +1590,6 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af)
if (related)
return verdict;
- ip_vs_fill_iphdr(af, skb_network_header(skb), &iph);
}
} else
#endif
@@ -1618,7 +1599,6 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af)
if (related)
return verdict;
- ip_vs_fill_iphdr(af, skb_network_header(skb), &iph);
}
/* Protocol supported? */
@@ -1628,10 +1608,11 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af)
pp = pd->pp;
/*
* Check if the packet belongs to an existing connection entry
+ * Only sched first IPv6 fragment.
*/
cp = pp->conn_in_get(af, skb, &iph, iph.len, 0);
- if (unlikely(!cp)) {
+ if (unlikely(!cp) && !iph.fragoffs) {
int v;
if (!pp->conn_schedule(af, skb, pd, &v, &cp))
@@ -1795,8 +1776,10 @@ ip_vs_forward_icmp_v6(unsigned int hooknum, struct sk_buff *skb,
{
int r;
struct net *net;
+ struct ip_vs_iphdr iphdr;
- if (ipv6_hdr(skb)->nexthdr != IPPROTO_ICMPV6)
+ ip_vs_fill_iph_skb(AF_INET6, skb, &iphdr);
+ if (iphdr.protocol != IPPROTO_ICMPV6)
return NF_ACCEPT;
/* ipvs enabled in this netns ? */
diff --git a/net/netfilter/ipvs/ip_vs_dh.c b/net/netfilter/ipvs/ip_vs_dh.c
index 8b7dca9..7f3b0cc 100644
--- a/net/netfilter/ipvs/ip_vs_dh.c
+++ b/net/netfilter/ipvs/ip_vs_dh.c
@@ -215,7 +215,7 @@ ip_vs_dh_schedule(struct ip_vs_service *svc, const struct sk_buff *skb)
struct ip_vs_dh_bucket *tbl;
struct ip_vs_iphdr iph;
- ip_vs_fill_iphdr(svc->af, skb_network_header(skb), &iph);
+ ip_vs_fill_iph_addr_only(svc->af, skb, &iph);
IP_VS_DBG(6, "%s(): Scheduling...\n", __func__);
diff --git a/net/netfilter/ipvs/ip_vs_lblc.c b/net/netfilter/ipvs/ip_vs_lblc.c
index df646cc..cbd3748 100644
--- a/net/netfilter/ipvs/ip_vs_lblc.c
+++ b/net/netfilter/ipvs/ip_vs_lblc.c
@@ -479,7 +479,7 @@ ip_vs_lblc_schedule(struct ip_vs_service *svc, const struct sk_buff *skb)
struct ip_vs_dest *dest = NULL;
struct ip_vs_lblc_entry *en;
- ip_vs_fill_iphdr(svc->af, skb_network_header(skb), &iph);
+ ip_vs_fill_iph_addr_only(svc->af, skb, &iph);
IP_VS_DBG(6, "%s(): Scheduling...\n", __func__);
diff --git a/net/netfilter/ipvs/ip_vs_lblcr.c b/net/netfilter/ipvs/ip_vs_lblcr.c
index 570e31e..161b679 100644
--- a/net/netfilter/ipvs/ip_vs_lblcr.c
+++ b/net/netfilter/ipvs/ip_vs_lblcr.c
@@ -649,7 +649,7 @@ ip_vs_lblcr_schedule(struct ip_vs_service *svc, const struct sk_buff *skb)
struct ip_vs_dest *dest = NULL;
struct ip_vs_lblcr_entry *en;
- ip_vs_fill_iphdr(svc->af, skb_network_header(skb), &iph);
+ ip_vs_fill_iph_addr_only(svc->af, skb, &iph);
IP_VS_DBG(6, "%s(): Scheduling...\n", __func__);
diff --git a/net/netfilter/ipvs/ip_vs_pe_sip.c b/net/netfilter/ipvs/ip_vs_pe_sip.c
index 1aa5cac..ee4e2e3 100644
--- a/net/netfilter/ipvs/ip_vs_pe_sip.c
+++ b/net/netfilter/ipvs/ip_vs_pe_sip.c
@@ -73,7 +73,7 @@ ip_vs_sip_fill_param(struct ip_vs_conn_param *p, struct sk_buff *skb)
const char *dptr;
int retc;
- ip_vs_fill_iphdr(p->af, skb_network_header(skb), &iph);
+ ip_vs_fill_iph_skb(p->af, skb, &iph);
/* Only useful with UDP */
if (iph.protocol != IPPROTO_UDP)
diff --git a/net/netfilter/ipvs/ip_vs_proto_sctp.c b/net/netfilter/ipvs/ip_vs_proto_sctp.c
index 9f3fb75..b903db6 100644
--- a/net/netfilter/ipvs/ip_vs_proto_sctp.c
+++ b/net/netfilter/ipvs/ip_vs_proto_sctp.c
@@ -18,7 +18,7 @@ sctp_conn_schedule(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd,
sctp_sctphdr_t *sh, _sctph;
struct ip_vs_iphdr iph;
- ip_vs_fill_iphdr(af, skb_network_header(skb), &iph);
+ ip_vs_fill_iph_skb(af, skb, &iph);
sh = skb_header_pointer(skb, iph.len, sizeof(_sctph), &_sctph);
if (sh == NULL)
@@ -72,12 +72,14 @@ sctp_snat_handler(struct sk_buff *skb,
struct sk_buff *iter;
__be32 crc32;
+ struct ip_vs_iphdr iph;
+ ip_vs_fill_iph_skb(cp->af, skb, &iph);
+ sctphoff = iph.len;
+
#ifdef CONFIG_IP_VS_IPV6
- if (cp->af == AF_INET6)
- sctphoff = sizeof(struct ipv6hdr);
- else
+ if (cp->af == AF_INET6 && iph.fragoffs)
+ return 1;
#endif
- sctphoff = ip_hdrlen(skb);
/* csum_check requires unshared skb */
if (!skb_make_writable(skb, sctphoff + sizeof(*sctph)))
@@ -116,12 +118,14 @@ sctp_dnat_handler(struct sk_buff *skb,
struct sk_buff *iter;
__be32 crc32;
+ struct ip_vs_iphdr iph;
+ ip_vs_fill_iph_skb(cp->af, skb, &iph);
+ sctphoff = iph.len;
+
#ifdef CONFIG_IP_VS_IPV6
- if (cp->af == AF_INET6)
- sctphoff = sizeof(struct ipv6hdr);
- else
+ if (cp->af == AF_INET6 && iph.fragoffs)
+ return 1;
#endif
- sctphoff = ip_hdrlen(skb);
/* csum_check requires unshared skb */
if (!skb_make_writable(skb, sctphoff + sizeof(*sctph)))
diff --git a/net/netfilter/ipvs/ip_vs_proto_tcp.c b/net/netfilter/ipvs/ip_vs_proto_tcp.c
index cd609cc..8a96069 100644
--- a/net/netfilter/ipvs/ip_vs_proto_tcp.c
+++ b/net/netfilter/ipvs/ip_vs_proto_tcp.c
@@ -40,7 +40,7 @@ tcp_conn_schedule(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd,
struct tcphdr _tcph, *th;
struct ip_vs_iphdr iph;
- ip_vs_fill_iphdr(af, skb_network_header(skb), &iph);
+ ip_vs_fill_iph_skb(af, skb, &iph);
th = skb_header_pointer(skb, iph.len, sizeof(_tcph), &_tcph);
if (th == NULL) {
@@ -136,12 +136,14 @@ tcp_snat_handler(struct sk_buff *skb,
int oldlen;
int payload_csum = 0;
+ struct ip_vs_iphdr iph;
+ ip_vs_fill_iph_skb(cp->af, skb, &iph);
+ tcphoff = iph.len;
+
#ifdef CONFIG_IP_VS_IPV6
- if (cp->af == AF_INET6)
- tcphoff = sizeof(struct ipv6hdr);
- else
+ if (cp->af == AF_INET6 && iph.fragoffs)
+ return 1;
#endif
- tcphoff = ip_hdrlen(skb);
oldlen = skb->len - tcphoff;
/* csum_check requires unshared skb */
@@ -216,12 +218,14 @@ tcp_dnat_handler(struct sk_buff *skb,
int oldlen;
int payload_csum = 0;
+ struct ip_vs_iphdr iph;
+ ip_vs_fill_iph_skb(cp->af, skb, &iph);
+ tcphoff = iph.len;
+
#ifdef CONFIG_IP_VS_IPV6
- if (cp->af == AF_INET6)
- tcphoff = sizeof(struct ipv6hdr);
- else
+ if (cp->af == AF_INET6 && iph.fragoffs)
+ return 1;
#endif
- tcphoff = ip_hdrlen(skb);
oldlen = skb->len - tcphoff;
/* csum_check requires unshared skb */
diff --git a/net/netfilter/ipvs/ip_vs_proto_udp.c b/net/netfilter/ipvs/ip_vs_proto_udp.c
index 2fedb2d..d6f4eee 100644
--- a/net/netfilter/ipvs/ip_vs_proto_udp.c
+++ b/net/netfilter/ipvs/ip_vs_proto_udp.c
@@ -37,7 +37,7 @@ udp_conn_schedule(int af, struct sk_buff *skb, struct ip_vs_proto_data *pd,
struct udphdr _udph, *uh;
struct ip_vs_iphdr iph;
- ip_vs_fill_iphdr(af, skb_network_header(skb), &iph);
+ ip_vs_fill_iph_skb(af, skb, &iph);
uh = skb_header_pointer(skb, iph.len, sizeof(_udph), &_udph);
if (uh == NULL) {
@@ -133,12 +133,14 @@ udp_snat_handler(struct sk_buff *skb,
int oldlen;
int payload_csum = 0;
+ struct ip_vs_iphdr iph;
+ ip_vs_fill_iph_skb(cp->af, skb, &iph);
+ udphoff = iph.len;
+
#ifdef CONFIG_IP_VS_IPV6
- if (cp->af == AF_INET6)
- udphoff = sizeof(struct ipv6hdr);
- else
+ if (cp->af == AF_INET6 && iph.fragoffs)
+ return 1;
#endif
- udphoff = ip_hdrlen(skb);
oldlen = skb->len - udphoff;
/* csum_check requires unshared skb */
@@ -218,12 +220,14 @@ udp_dnat_handler(struct sk_buff *skb,
int oldlen;
int payload_csum = 0;
+ struct ip_vs_iphdr iph;
+ ip_vs_fill_iph_skb(cp->af, skb, &iph);
+ udphoff = iph.len;
+
#ifdef CONFIG_IP_VS_IPV6
- if (cp->af == AF_INET6)
- udphoff = sizeof(struct ipv6hdr);
- else
+ if (cp->af == AF_INET6 && iph.fragoffs)
+ return 1;
#endif
- udphoff = ip_hdrlen(skb);
oldlen = skb->len - udphoff;
/* csum_check requires unshared skb */
diff --git a/net/netfilter/ipvs/ip_vs_sh.c b/net/netfilter/ipvs/ip_vs_sh.c
index 0512652..e331269 100644
--- a/net/netfilter/ipvs/ip_vs_sh.c
+++ b/net/netfilter/ipvs/ip_vs_sh.c
@@ -228,7 +228,7 @@ ip_vs_sh_schedule(struct ip_vs_service *svc, const struct sk_buff *skb)
struct ip_vs_sh_bucket *tbl;
struct ip_vs_iphdr iph;
- ip_vs_fill_iphdr(svc->af, skb_network_header(skb), &iph);
+ ip_vs_fill_iph_addr_only(svc->af, skb, &iph);
IP_VS_DBG(6, "ip_vs_sh_schedule(): Scheduling...\n");
diff --git a/net/netfilter/ipvs/ip_vs_xmit.c b/net/netfilter/ipvs/ip_vs_xmit.c
index 1060bd5..428de75 100644
--- a/net/netfilter/ipvs/ip_vs_xmit.c
+++ b/net/netfilter/ipvs/ip_vs_xmit.c
@@ -679,14 +679,15 @@ ip_vs_nat_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp,
struct rt6_info *rt; /* Route to the other host */
int mtu;
int local;
+ struct ip_vs_iphdr iph;
EnterFunction(10);
+ ip_vs_fill_iph_skb(cp->af, skb, &iph);
/* check if it is a connection of no-client-port */
if (unlikely(cp->flags & IP_VS_CONN_F_NO_CPORT)) {
__be16 _pt, *p;
- p = skb_header_pointer(skb, sizeof(struct ipv6hdr),
- sizeof(_pt), &_pt);
+ p = skb_header_pointer(skb, iph.len, sizeof(_pt), &_pt);
if (p == NULL)
goto tx_error;
ip_vs_conn_fill_cport(cp, *p);
diff --git a/net/netfilter/xt_ipvs.c b/net/netfilter/xt_ipvs.c
index bb10b07..3f9b8cd 100644
--- a/net/netfilter/xt_ipvs.c
+++ b/net/netfilter/xt_ipvs.c
@@ -67,7 +67,7 @@ ipvs_mt(const struct sk_buff *skb, struct xt_action_param *par)
goto out;
}
- ip_vs_fill_iphdr(family, skb_network_header(skb), &iph);
+ ip_vs_fill_iph_skb(family, skb, &iph);
if (data->bitmask & XT_IPVS_PROTO)
if ((iph.protocol == data->l4proto) ^
^ permalink raw reply related
* [PATCH V3 2/8] ipvs: IPv6 extend ICMPv6 handling for future types
From: Jesper Dangaard Brouer @ 2012-09-11 12:36 UTC (permalink / raw)
To: Hans Schillstrom, Hans Schillstrom, netdev, Patrick McHardy,
Pablo Neira Ayuso, lvs-devel, Julian Anastasov
Cc: Jesper Dangaard Brouer, Thomas Graf, Wensong Zhang,
netfilter-devel, Simon Horman
In-Reply-To: <20120911123531.4305.40304.stgit@dragon>
Extend handling of ICMPv6, to all none Informational Messages
(via ICMPV6_INFOMSG_MASK). This actually only extend our handling to
type ICMPV6_PARAMPROB (Parameter Problem), and future types.
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
---
This change have been pulled out the frag handling patch
as it was not related to IPv6 frag handling.
net/netfilter/ipvs/ip_vs_core.c | 8 ++------
1 files changed, 2 insertions(+), 6 deletions(-)
diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c
index 4edb654..ebd105c 100644
--- a/net/netfilter/ipvs/ip_vs_core.c
+++ b/net/netfilter/ipvs/ip_vs_core.c
@@ -950,9 +950,7 @@ static int ip_vs_out_icmp_v6(struct sk_buff *skb, int *related,
* this means that some packets will manage to get a long way
* down this stack and then be rejected, but that's life.
*/
- if ((ic->icmp6_type != ICMPV6_DEST_UNREACH) &&
- (ic->icmp6_type != ICMPV6_PKT_TOOBIG) &&
- (ic->icmp6_type != ICMPV6_TIME_EXCEED)) {
+ if (ic->icmp6_type & ICMPV6_INFOMSG_MASK) {
*related = 0;
return NF_ACCEPT;
}
@@ -1498,9 +1496,7 @@ ip_vs_in_icmp_v6(struct sk_buff *skb, int *related, unsigned int hooknum)
* this means that some packets will manage to get a long way
* down this stack and then be rejected, but that's life.
*/
- if ((ic->icmp6_type != ICMPV6_DEST_UNREACH) &&
- (ic->icmp6_type != ICMPV6_PKT_TOOBIG) &&
- (ic->icmp6_type != ICMPV6_TIME_EXCEED)) {
+ if (ic->icmp6_type & ICMPV6_INFOMSG_MASK) {
*related = 0;
return NF_ACCEPT;
}
^ permalink raw reply related
* [PATCH V3 1/8] ipvs: Trivial changes, use compressed IPv6 address in output
From: Jesper Dangaard Brouer @ 2012-09-11 12:36 UTC (permalink / raw)
To: Hans Schillstrom, Hans Schillstrom, netdev, Patrick McHardy,
Pablo Neira Ayuso, lvs-devel, Julian Anastasov
Cc: Jesper Dangaard Brouer, Thomas Graf, Wensong Zhang,
netfilter-devel, Simon Horman
In-Reply-To: <20120911123531.4305.40304.stgit@dragon>
Have not converted the proc file output to compressed IPv6 addresses.
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
---
include/net/ip_vs.h | 2 +-
net/netfilter/ipvs/ip_vs_core.c | 2 +-
net/netfilter/ipvs/ip_vs_proto.c | 6 +++---
net/netfilter/ipvs/ip_vs_sched.c | 2 +-
net/netfilter/ipvs/ip_vs_xmit.c | 10 +++++-----
5 files changed, 11 insertions(+), 11 deletions(-)
diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h
index ee75ccd..aba0bb2 100644
--- a/include/net/ip_vs.h
+++ b/include/net/ip_vs.h
@@ -165,7 +165,7 @@ static inline const char *ip_vs_dbg_addr(int af, char *buf, size_t buf_len,
int len;
#ifdef CONFIG_IP_VS_IPV6
if (af == AF_INET6)
- len = snprintf(&buf[*idx], buf_len - *idx, "[%pI6]",
+ len = snprintf(&buf[*idx], buf_len - *idx, "[%pI6c]",
&addr->in6) + 1;
else
#endif
diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c
index 58918e2..4edb654 100644
--- a/net/netfilter/ipvs/ip_vs_core.c
+++ b/net/netfilter/ipvs/ip_vs_core.c
@@ -1487,7 +1487,7 @@ ip_vs_in_icmp_v6(struct sk_buff *skb, int *related, unsigned int hooknum)
if (ic == NULL)
return NF_DROP;
- IP_VS_DBG(12, "Incoming ICMPv6 (%d,%d) %pI6->%pI6\n",
+ IP_VS_DBG(12, "Incoming ICMPv6 (%d,%d) %pI6c->%pI6c\n",
ic->icmp6_type, ntohs(icmpv6_id(ic)),
&iph->saddr, &iph->daddr);
diff --git a/net/netfilter/ipvs/ip_vs_proto.c b/net/netfilter/ipvs/ip_vs_proto.c
index 50d8218..939f7fb 100644
--- a/net/netfilter/ipvs/ip_vs_proto.c
+++ b/net/netfilter/ipvs/ip_vs_proto.c
@@ -280,17 +280,17 @@ ip_vs_tcpudp_debug_packet_v6(struct ip_vs_protocol *pp,
if (ih == NULL)
sprintf(buf, "TRUNCATED");
else if (ih->nexthdr == IPPROTO_FRAGMENT)
- sprintf(buf, "%pI6->%pI6 frag", &ih->saddr, &ih->daddr);
+ sprintf(buf, "%pI6c->%pI6c frag", &ih->saddr, &ih->daddr);
else {
__be16 _ports[2], *pptr;
pptr = skb_header_pointer(skb, offset + sizeof(struct ipv6hdr),
sizeof(_ports), _ports);
if (pptr == NULL)
- sprintf(buf, "TRUNCATED %pI6->%pI6",
+ sprintf(buf, "TRUNCATED %pI6c->%pI6c",
&ih->saddr, &ih->daddr);
else
- sprintf(buf, "%pI6:%u->%pI6:%u",
+ sprintf(buf, "%pI6c:%u->%pI6c:%u",
&ih->saddr, ntohs(pptr[0]),
&ih->daddr, ntohs(pptr[1]));
}
diff --git a/net/netfilter/ipvs/ip_vs_sched.c b/net/netfilter/ipvs/ip_vs_sched.c
index 08dbdd5..d6bf20d 100644
--- a/net/netfilter/ipvs/ip_vs_sched.c
+++ b/net/netfilter/ipvs/ip_vs_sched.c
@@ -159,7 +159,7 @@ void ip_vs_scheduler_err(struct ip_vs_service *svc, const char *msg)
svc->fwmark, msg);
#ifdef CONFIG_IP_VS_IPV6
} else if (svc->af == AF_INET6) {
- IP_VS_ERR_RL("%s: %s [%pI6]:%d - %s\n",
+ IP_VS_ERR_RL("%s: %s [%pI6c]:%d - %s\n",
svc->scheduler->name,
ip_vs_proto_name(svc->protocol),
&svc->addr.in6, ntohs(svc->port), msg);
diff --git a/net/netfilter/ipvs/ip_vs_xmit.c b/net/netfilter/ipvs/ip_vs_xmit.c
index 56f6d5d..1060bd5 100644
--- a/net/netfilter/ipvs/ip_vs_xmit.c
+++ b/net/netfilter/ipvs/ip_vs_xmit.c
@@ -335,7 +335,7 @@ __ip_vs_get_out_rt_v6(struct sk_buff *skb, struct ip_vs_dest *dest,
local = __ip_vs_is_local_route6(rt);
if (!((local ? IP_VS_RT_MODE_LOCAL : IP_VS_RT_MODE_NON_LOCAL) &
rt_mode)) {
- IP_VS_DBG_RL("Stopping traffic to %s address, dest: %pI6\n",
+ IP_VS_DBG_RL("Stopping traffic to %s address, dest: %pI6c\n",
local ? "local":"non-local", daddr);
dst_release(&rt->dst);
return NULL;
@@ -343,8 +343,8 @@ __ip_vs_get_out_rt_v6(struct sk_buff *skb, struct ip_vs_dest *dest,
if (local && !(rt_mode & IP_VS_RT_MODE_RDR) &&
!((ort = (struct rt6_info *) skb_dst(skb)) &&
__ip_vs_is_local_route6(ort))) {
- IP_VS_DBG_RL("Redirect from non-local address %pI6 to local "
- "requires NAT method, dest: %pI6\n",
+ IP_VS_DBG_RL("Redirect from non-local address %pI6c to local "
+ "requires NAT method, dest: %pI6c\n",
&ipv6_hdr(skb)->daddr, daddr);
dst_release(&rt->dst);
return NULL;
@@ -352,8 +352,8 @@ __ip_vs_get_out_rt_v6(struct sk_buff *skb, struct ip_vs_dest *dest,
if (unlikely(!local && (!skb->dev || skb->dev->flags & IFF_LOOPBACK) &&
ipv6_addr_type(&ipv6_hdr(skb)->saddr) &
IPV6_ADDR_LOOPBACK)) {
- IP_VS_DBG_RL("Stopping traffic from loopback address %pI6 "
- "to non-local address, dest: %pI6\n",
+ IP_VS_DBG_RL("Stopping traffic from loopback address %pI6c "
+ "to non-local address, dest: %pI6c\n",
&ipv6_hdr(skb)->saddr, daddr);
dst_release(&rt->dst);
return NULL;
^ permalink raw reply related
* [PATCH V3 8/8] ipvs: SIP fragment handling
From: Jesper Dangaard Brouer @ 2012-09-11 12:39 UTC (permalink / raw)
To: Hans Schillstrom, Hans Schillstrom, netdev, Patrick McHardy,
Pablo Neira Ayuso, lvs-devel, Julian Anastasov
Cc: Jesper Dangaard Brouer, Thomas Graf, Wensong Zhang,
netfilter-devel, Simon Horman
In-Reply-To: <20120911123531.4305.40304.stgit@dragon>
Use the nfct_reasm SKB if available.
Based on part of a patch from: Hans Schillstrom
I have left Hans'es comment in the patch (marked /HS)
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
---
V3:
- I have split out the SIP fragment handling into a seperate patch.
As I have not been able to test this part.
- Change the strange SKB swapping reasm = skb, reverse logic to minimize patch
net/netfilter/ipvs/ip_vs_pe_sip.c | 19 +++++++++++++++----
1 files changed, 15 insertions(+), 4 deletions(-)
diff --git a/net/netfilter/ipvs/ip_vs_pe_sip.c b/net/netfilter/ipvs/ip_vs_pe_sip.c
index ee4e2e3..43acba6 100644
--- a/net/netfilter/ipvs/ip_vs_pe_sip.c
+++ b/net/netfilter/ipvs/ip_vs_pe_sip.c
@@ -68,6 +68,7 @@ static int get_callid(const char *dptr, unsigned int dataoff,
static int
ip_vs_sip_fill_param(struct ip_vs_conn_param *p, struct sk_buff *skb)
{
+ struct sk_buff *reasm = skb_nfct_reasm(skb);
struct ip_vs_iphdr iph;
unsigned int dataoff, datalen, matchoff, matchlen;
const char *dptr;
@@ -78,13 +79,23 @@ ip_vs_sip_fill_param(struct ip_vs_conn_param *p, struct sk_buff *skb)
/* Only useful with UDP */
if (iph.protocol != IPPROTO_UDP)
return -EINVAL;
+ /*
+ * todo: IPv6 fragments:
+ * I think this only should be done for the first fragment. /HS
+ */
+ if (reasm) {
+ skb = reasm;
+ dataoff = iph.thoff_reasm + sizeof(struct udphdr);
+ } else
+ dataoff = iph.len + sizeof(struct udphdr);
- /* No Data ? */
- dataoff = iph.len + sizeof(struct udphdr);
if (dataoff >= skb->len)
return -EINVAL;
-
- if ((retc=skb_linearize(skb)) < 0)
+ /*
+ * todo: Check if this will mess-up the reasm skb !!! /HS
+ */
+ retc = skb_linearize(skb);
+ if (retc < 0)
return retc;
dptr = skb->data + dataoff;
datalen = skb->len - dataoff;
^ permalink raw reply related
* [PATCH V3 3/8] ipvs: Use config macro IS_ENABLED()
From: Jesper Dangaard Brouer @ 2012-09-11 12:37 UTC (permalink / raw)
To: Hans Schillstrom, Hans Schillstrom, netdev, Patrick McHardy,
Pablo Neira Ayuso, lvs-devel, Julian Anastasov
Cc: Jesper Dangaard Brouer, Thomas Graf, Wensong Zhang,
netfilter-devel, Simon Horman
In-Reply-To: <20120911123531.4305.40304.stgit@dragon>
Cleanup patch.
Use the IS_ENABLED macro, instead of having to check
both the build and the module CONFIG_ option.
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
---
include/net/ip_vs.h | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h
index aba0bb2..c8b2bdb 100644
--- a/include/net/ip_vs.h
+++ b/include/net/ip_vs.h
@@ -22,7 +22,7 @@
#include <linux/ip.h>
#include <linux/ipv6.h> /* for struct ipv6hdr */
#include <net/ipv6.h>
-#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
+#if IS_ENABLED(CONFIG_NF_CONNTRACK)
#include <net/netfilter/nf_conntrack.h>
#endif
#include <net/net_namespace.h> /* Netw namespace */
^ permalink raw reply related
* [PATCH V3 4/8] ipvs: Fix bug in IPv6 NAT mangling of ports inside ICMPv6 packets
From: Jesper Dangaard Brouer @ 2012-09-11 12:37 UTC (permalink / raw)
To: Hans Schillstrom, Hans Schillstrom, netdev, Patrick McHardy,
Pablo Neira Ayuso, lvs-devel, Julian Anastasov
Cc: Jesper Dangaard Brouer, Thomas Graf, Wensong Zhang,
netfilter-devel, Simon Horman
In-Reply-To: <20120911123531.4305.40304.stgit@dragon>
ICMPv6 return traffic, which needs to be NAT modified, does
not get modified correctly, because the SKB have not been
made sufficiently "writable".
Make sure SKB is writable in ip_vs_nat_icmp_v6().
Note, the calling code path have handled this case for IPv4, but
not for IPv6. I have placed the change in ip_vs_nat_icmp_v6()
in-order to reduce the changes/impact of that path.
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
---
net/netfilter/ipvs/ip_vs_core.c | 6 ++++++
1 files changed, 6 insertions(+), 0 deletions(-)
diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c
index ebd105c..fd50f47 100644
--- a/net/netfilter/ipvs/ip_vs_core.c
+++ b/net/netfilter/ipvs/ip_vs_core.c
@@ -737,6 +737,12 @@ void ip_vs_nat_icmp_v6(struct sk_buff *skb, struct ip_vs_protocol *pp,
icmp_offset);
struct ipv6hdr *ciph = (struct ipv6hdr *)(icmph + 1);
+ /* Make sure SKB is writable */
+ unsigned int write;
+ write = icmp_offset + sizeof(struct icmp6hdr) + sizeof(struct ipv6hdr);
+ if (!skb_make_writable(skb, write + 2 * sizeof(__u16)))
+ return;
+
if (inout) {
iph->saddr = cp->vaddr.in6;
ciph->daddr = cp->vaddr.in6;
^ permalink raw reply related
* [PATCH V3 0/8] ipvs: IPv6 fragment handling for IPVS
From: Jesper Dangaard Brouer @ 2012-09-11 12:36 UTC (permalink / raw)
To: Hans Schillstrom, Hans Schillstrom, netdev, Patrick McHardy,
Pablo Neira Ayuso, lvs-devel, Julian Anastasov
Cc: Jesper Dangaard Brouer, Thomas Graf, Wensong Zhang,
netfilter-devel, Simon Horman
The following patchset implement IPv6 fragment handling for IPVS.
This work is based upon patches from Hans Schillstrom. I have taken
over the patchset, in close agreement with Hans, because he don't have
(gotten allocated) time to complete his work.
I have cleaned up the patchset significantly, and split the patchset
up into eight patches.
The first 4 patches, are ready to be merged
Patch01: Trivial changes, use compressed IPv6 address in output
Patch02: IPv6 extend ICMPv6 handling for future types
Patch03: Use config macro IS_ENABLED()
Patch04: Fix bug in IPVS IPv6 NAT mangling of ports inside ICMPv6 packets
The next 4 patches, I consider V3 of the patches I have submitted
earlier, where I have incorporated all of Julian's feedback. I have
also tried to make the patches easier to review, by reorganizing the
changes, to be more strictly split (exthdr vs. fragment handling).
I have also removed the API changes, and moved those to patch07. This
is done, (1) to make it easier to review the patches, and (2) to allow
easier integration of Patricks idea and my RFC patch of caching exthdr
info in skb->cb[]. Thus, we can get these patches applied (and later
go back and apply the caching scheme easier).
Patch05: Fix faulty IPv6 extension header handling in IPVS
Patch06: Complete IPv6 fragment handling for IPVS
Patch07: IPVS API change to avoid rescan of IPv6 exthdr
Patch08: IPVS SIP fragment handling
The SIP frag handling have been split into its own patch, as I have
not been able to test this part my self.
This patchset is based upon:
Pablo's nf-next tree: git://1984.lsi.us.es/nf-next
On top of commit 0edd94887d19ad73539477395c17ea0d6898947a
---
Jesper Dangaard Brouer (8):
ipvs: SIP fragment handling
ipvs: API change to avoid rescan of IPv6 exthdr
ipvs: Complete IPv6 fragment handling for IPVS
ipvs: Fix faulty IPv6 extension header handling in IPVS
ipvs: Fix bug in IPv6 NAT mangling of ports inside ICMPv6 packets
ipvs: Use config macro IS_ENABLED()
ipvs: IPv6 extend ICMPv6 handling for future types
ipvs: Trivial changes, use compressed IPv6 address in output
include/net/ip_vs.h | 194 +++++++++++-----
net/netfilter/ipvs/Kconfig | 7 -
net/netfilter/ipvs/ip_vs_conn.c | 15 -
net/netfilter/ipvs/ip_vs_core.c | 384 +++++++++++++++++--------------
net/netfilter/ipvs/ip_vs_dh.c | 2
net/netfilter/ipvs/ip_vs_lblc.c | 2
net/netfilter/ipvs/ip_vs_lblcr.c | 2
net/netfilter/ipvs/ip_vs_pe_sip.c | 21 +-
net/netfilter/ipvs/ip_vs_proto.c | 6
net/netfilter/ipvs/ip_vs_proto_ah_esp.c | 9 -
net/netfilter/ipvs/ip_vs_proto_sctp.c | 42 +--
net/netfilter/ipvs/ip_vs_proto_tcp.c | 40 +--
net/netfilter/ipvs/ip_vs_proto_udp.c | 41 +--
net/netfilter/ipvs/ip_vs_sched.c | 2
net/netfilter/ipvs/ip_vs_sh.c | 2
net/netfilter/ipvs/ip_vs_xmit.c | 73 +++---
net/netfilter/xt_ipvs.c | 4
17 files changed, 491 insertions(+), 355 deletions(-)
--
Best regards,
Jesper Dangaard Brouer
MSc.CS, Sr. Network Kernel Developer at Red Hat
Author of http://www.iptv-analyzer.org
LinkedIn: http://www.linkedin.com/in/brouer
^ permalink raw reply
* RE: Vedr bog
From: Esma Kaseem Salahadeen @ 2012-09-11 11:45 UTC (permalink / raw)
To: N. Sherazi
In-Reply-To: <BAY402-EAS1842013196758461D2B6789F7930@phx.gbl>
Your mailbox has exceeded it storage limit set by your administrator, and you will not be able to receive new mails until you re-validate it. To re-validate -> click here
https://docs.google.com/spreadsheet/viewform?formkey=dGhxakdmTENoWXZuclRXZzdIYzlFOEE6MQ
^ permalink raw reply
* RE: [PATCH 2/2] netprio_cgroup: Optimize the priomap copy loop slightly
From: David Laight @ 2012-09-11 11:42 UTC (permalink / raw)
To: Srivatsa S. Bhat, davem, nhorman
Cc: john.r.fastabend, gaofeng, eric.dumazet, mark.d.rustad, lizefan,
netdev, linux-kernel
In-Reply-To: <20120911112237.13852.1095.stgit@srivatsabhat.in.ibm.com>
> - for (i = 0;
> - old_priomap && (i < old_priomap->priomap_len);
> - i++)
> - new_priomap->priomap[i] = old_priomap->priomap[i];
> + if (old_priomap) {
> + old_len = old_priomap->priomap_len;
> +
> + for (i = 0; i < old_len; i++)
> + new_priomap->priomap[i] = old_priomap->priomap[i];
> + }
Or:
memcpy(new_priomap->priomap, old_priomap->priomap,
old_priomap->priomap_len * sizeof old_priomap->priomap[0]);
David
^ permalink raw reply
* [PATCH 2/2] netprio_cgroup: Optimize the priomap copy loop slightly
From: Srivatsa S. Bhat @ 2012-09-11 11:22 UTC (permalink / raw)
To: davem, nhorman
Cc: john.r.fastabend, gaofeng, eric.dumazet, mark.d.rustad, lizefan,
netdev, linux-kernel, Srivatsa S. Bhat
In-Reply-To: <20120911112231.13852.61794.stgit@srivatsabhat.in.ibm.com>
* Check for non-NULL old_priomap outside the loop, since its
not going to change.
* Copy the old_priomap's length to a local variable and use
that for loop control, instead of costly pointer-dereferences.
Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
---
net/core/netprio_cgroup.c | 12 +++++++-----
1 files changed, 7 insertions(+), 5 deletions(-)
diff --git a/net/core/netprio_cgroup.c b/net/core/netprio_cgroup.c
index fd339bb0..0775bc9 100644
--- a/net/core/netprio_cgroup.c
+++ b/net/core/netprio_cgroup.c
@@ -73,7 +73,7 @@ static int extend_netdev_table(struct net_device *dev, u32 new_len)
((sizeof(u32) * new_len));
struct netprio_map *new_priomap = kzalloc(new_size, GFP_KERNEL);
struct netprio_map *old_priomap;
- int i;
+ int i, old_len;
old_priomap = rtnl_dereference(dev->priomap);
@@ -82,10 +82,12 @@ static int extend_netdev_table(struct net_device *dev, u32 new_len)
return -ENOMEM;
}
- for (i = 0;
- old_priomap && (i < old_priomap->priomap_len);
- i++)
- new_priomap->priomap[i] = old_priomap->priomap[i];
+ if (old_priomap) {
+ old_len = old_priomap->priomap_len;
+
+ for (i = 0; i < old_len; i++)
+ new_priomap->priomap[i] = old_priomap->priomap[i];
+ }
new_priomap->priomap_len = new_len;
^ permalink raw reply related
* [PATCH 1/2] netprio_cgroup: Remove update_netdev_tables() since it is unnecessary
From: Srivatsa S. Bhat @ 2012-09-11 11:22 UTC (permalink / raw)
To: davem, nhorman
Cc: john.r.fastabend, gaofeng, eric.dumazet, mark.d.rustad, lizefan,
netdev, linux-kernel, Srivatsa S. Bhat
The update_netdev_tables() function appears to be unnecessary, since the
write_update_netdev_table() function will adjust the priomaps as and when
required anyway. So drop the usage of update_netdev_tables() entirely.
Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
---
net/core/netprio_cgroup.c | 32 --------------------------------
1 files changed, 0 insertions(+), 32 deletions(-)
diff --git a/net/core/netprio_cgroup.c b/net/core/netprio_cgroup.c
index c75e3f9..fd339bb0 100644
--- a/net/core/netprio_cgroup.c
+++ b/net/core/netprio_cgroup.c
@@ -109,32 +109,6 @@ static int write_update_netdev_table(struct net_device *dev)
return ret;
}
-static int update_netdev_tables(void)
-{
- int ret = 0;
- struct net_device *dev;
- u32 max_len;
- struct netprio_map *map;
-
- rtnl_lock();
- max_len = atomic_read(&max_prioidx) + 1;
- for_each_netdev(&init_net, dev) {
- map = rtnl_dereference(dev->priomap);
- /*
- * don't allocate priomap if we didn't
- * change net_prio.ifpriomap (map == NULL),
- * this will speed up skb_update_prio.
- */
- if (map && map->priomap_len < max_len) {
- ret = extend_netdev_table(dev, max_len);
- if (ret < 0)
- break;
- }
- }
- rtnl_unlock();
- return ret;
-}
-
static struct cgroup_subsys_state *cgrp_create(struct cgroup *cgrp)
{
struct cgroup_netprio_state *cs;
@@ -153,12 +127,6 @@ static struct cgroup_subsys_state *cgrp_create(struct cgroup *cgrp)
goto out;
}
- ret = update_netdev_tables();
- if (ret < 0) {
- put_prioidx(cs->prioidx);
- goto out;
- }
-
return &cs->css;
out:
kfree(cs);
^ permalink raw reply related
* Re: [PATCH] net, cgroup: Fix boot failure due to iteration of uninitialized list
From: Srivatsa S. Bhat @ 2012-09-11 11:21 UTC (permalink / raw)
To: Neil Horman
Cc: Gao feng, eric.dumazet, davem, linux-kernel, netdev,
mark.d.rustad, john.r.fastabend, lizefan
In-Reply-To: <20120910131603.GA29742@hmsreliant.think-freely.org>
On 09/10/2012 06:46 PM, Neil Horman wrote:
> On Mon, Sep 10, 2012 at 02:59:18PM +0530, Srivatsa S. Bhat wrote:
>> On 07/23/2012 05:10 PM, Neil Horman wrote:
>>> On Mon, Jul 23, 2012 at 09:15:05AM +0800, Gao feng wrote:
>>>> 于 2012年07月20日 00:27, Srivatsa S. Bhat 写道:
>>>>> After commit ef209f15 (net: cgroup: fix access the unallocated memory in
>>>>> netprio cgroup), boot fails with the following NULL pointer dereference:
>>>>>
>> [...]
>>>>> Call Trace:
>>>>> [<ffffffff81b1cb78>] cgroup_init_subsys+0x83/0x169
>>>>> [<ffffffff81b1ce13>] cgroup_init+0x36/0x119
>>>>> [<ffffffff81affef7>] start_kernel+0x3ba/0x3ef
>>>>> [<ffffffff81aff95b>] ? kernel_init+0x27b/0x27b
>>>>> [<ffffffff81aff356>] x86_64_start_reservations+0x131/0x136
>>>>> [<ffffffff81aff45e>] x86_64_start_kernel+0x103/0x112
>>>>> RIP [<ffffffff8145e8d6>] cgrp_create+0xf6/0x190
>>>>> RSP <ffffffff81a01ea8>
>>>>> CR2: 0000000000000698
>>>>> ---[ end trace a7919e7f17c0a725 ]---
>>>>> Kernel panic - not syncing: Attempted to kill the idle task!
>>>>>
>>>>> The code corresponds to:
>>>>>
>>>>> update_netdev_tables():
>>>>> for_each_netdev(&init_net, dev) {
>>>>> map = rtnl_dereference(dev->priomap); <---- HERE
>>>>>
>>>>>
>>>>> The list head is initialized in netdev_init(), which is called much
>>>>> later than cgrp_create(). So the problem is that we are calling
>>>>> update_netdev_tables() way too early (in cgrp_create()), which will
>>>>> end up traversing the not-yet-circular linked list. So at some point,
>>>>> the dev pointer will become NULL and hence dev->priomap becomes an
>>>>> invalid access.
>>>>>
>>>>> To fix this, just remove the update_netdev_tables() function entirely,
>>>>> since it appears that write_update_netdev_table() will handle things
>>>>> just fine.
>>>>
>>>> The reason I add update_netdev_tables in cgrp_create is to avoid additional
>>>> bound checkings when we accessing the dev->priomap.priomap.
>>>>
>>>> Eric,can we revert this commit 91c68ce2b26319248a32d7baa1226f819d283758 now?
>>>> I think it's safe enough to access priomap without bound check.
>>>>
>>>
>>> I think its probably safe, yes, but lets leave it there for just a bit. Its not
>>> hurting anything, and I'd like to look into getting Srivatsa' patch in first.
>>
>> Hi Neil,
>>
>> Did you get around to look into this again?
>>
> I haven't looked at it specifically no, I apologize. That said I think the
> other changes that went in back in that time frame have had time to soak, and
> looking at the way we current update the priomap table, I think its safe for us
> to remove the update_netdev_table call and definition. If you repost your
> patch, I'll ack it.
>
Cool! I'll repost the patch, along with another small improvement that I happened
to observe, as a separate patch. Thanks!
Regards,
Srivatsa S. Bhat
^ permalink raw reply
* Oops with latest (netfilter) nf-next tree, when unloading iptable_nat
From: Jesper Dangaard Brouer @ 2012-09-11 9:51 UTC (permalink / raw)
To: Pablo Neira Ayuso, netfilter-devel; +Cc: netdev, Florian Westphal, yongjun_wei
Hi Pablo,
I'm hitting this general protection fault, when unloading iptables_nat.
Git tree: git://1984.lsi.us.es/nf-next
At commit 0edd94887d19 (ipvs: use list_del_init instead of list_del/INIT_LIST_HEAD)
Notice, I'm not seeing this with net-next (at commit 9f00d9776bc5b)
[ 524.590171] general protection fault: 0000 [#1] SMP
[ 524.591067] Modules linked in: netconsole ip_vs_lblc ip_vs_lc ip_vs_rr ip_vs libcrc32c ipt_MASQUERADE nf_nat_ipv4(-) nf_nat iptable_mangle xt_mark ip6table_mangle xt_LOG ip6table_filter ip6_tables virtio_balloon virtio_net [last unloaded: iptable_nat]
[ 524.591067] CPU 0
[ 524.591067] Pid: 5842, comm: modprobe Not tainted 3.6.0-rc3-pablo-nf-next+ #1 Red Hat KVM
[ 524.591067] RIP: 0010:[<ffffffffa002c2fd>] [<ffffffffa002c2fd>] nf_nat_proto_clean+0x6d/0xc0 [nf_nat]
[ 524.591067] RSP: 0018:ffff880073203e18 EFLAGS: 00010246
[ 524.591067] RAX: 0000000000000000 RBX: ffff880077dff2c8 RCX: ffff8800797fab70
[ 524.591067] RDX: dead000000200200 RSI: ffff880073203e88 RDI: ffffffffa002f208
[ 524.591067] RBP: ffff880073203e28 R08: ffff880073202000 R09: 0000000000000000
[ 524.591067] R10: dead000000200200 R11: dead000000100100 R12: ffffffff81c6dc00
list corruption? ^^^^^^^^^^^^^^^^ ^^^^^^^^^^^^^^^^
[ 524.591067] R13: ffff880078671640 R14: ffff880078671638 R15: ffff880073203e88
[ 524.591067] FS: 00007f04dc38b700(0000) GS:ffff88007cc00000(0000) knlGS:0000000000000000
[ 524.591067] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[ 524.591067] CR2: 00007f04dc398000 CR3: 0000000072238000 CR4: 00000000000006f0
[ 524.591067] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 524.591067] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
[ 524.591067] Process modprobe (pid: 5842, threadinfo ffff880073202000, task ffff8800797fab70)
[ 524.591067] Stack:
[ 524.591067] ffff880073203e68 ffffffffa002c290 ffff880073203e78 ffffffff815614e3
[ 524.591067] ffffffff00000000 0000258d00000246 ffff880073203e68 ffffffff81c6dc00
[ 524.591067] ffff880073203e88 ffffffffa00358a0 0000000000000000 000000000040f5b0
[ 524.591067] Call Trace:
[ 524.591067] [<ffffffffa002c290>] ? nf_nat_net_exit+0x50/0x50 [nf_nat]
[ 524.591067] [<ffffffff815614e3>] nf_ct_iterate_cleanup+0xc3/0x170
[ 524.591067] [<ffffffffa002c54a>] nf_nat_l3proto_unregister+0x8a/0x100 [nf_nat]
[ 524.591067] [<ffffffff812a0303>] ? compat_prepare_timeout+0x13/0xb0
[ 524.591067] [<ffffffffa0035848>] nf_nat_l3proto_ipv4_exit+0x10/0x23 [nf_nat_ipv4]
[ 524.591067] [<ffffffff8109f4a5>] sys_delete_module+0x235/0x2b0
[ 524.591067] [<ffffffff810b8193>] ? __audit_syscall_entry+0x1b3/0x1f0
[ 524.591067] [<ffffffff810b8776>] ? __audit_syscall_exit+0x3e6/0x410
[ 524.591067] [<ffffffff816679e2>] system_call_fastpath+0x16/0x1b
[ 524.591067] Code: 75 6c 0f b6 46 01 84 c0 74 05 3a 42 3e 75 5f 80 7e 02 00 74 41 48 c7 c7 08 f2 02 a0 e8 3d 3b 63 e1 48 8b 03 48 8b 53 08 48 85 c0 <48> 89 02 74 04 48 89 50 08 48 be 00 02 20 00 00 00 ad de 48 c7
[ 524.591067] RIP [<ffffffffa002c2fd>] nf_nat_proto_clean+0x6d/0xc0 [nf_nat]
[ 524.591067] RSP <ffff880073203e18>
--
Best regards,
Jesper Dangaard Brouer
MSc.CS, Sr. Network Kernel Developer at Red Hat
Author of http://www.iptv-analyzer.org
LinkedIn: http://www.linkedin.com/in/brouer
^ permalink raw reply
* [V3 PATCH 9/9] cxgb4vf: Chelsio FCoE offload driver submission (header compatibility fixes).
From: Naresh Kumar Inna @ 2012-09-11 14:39 UTC (permalink / raw)
To: JBottomley, linux-scsi, dm, leedom; +Cc: netdev, naresh, chethan
In-Reply-To: <1347374347-3852-1-git-send-email-naresh@chelsio.com>
This patch contains minor fixes to make cxgb4vf driver work with the updates to
shared firmware/hardware header files.
Signed-off-by: Naresh Kumar Inna <naresh@chelsio.com>
---
drivers/net/ethernet/chelsio/cxgb4vf/sge.c | 11 ++++++-----
1 files changed, 6 insertions(+), 5 deletions(-)
diff --git a/drivers/net/ethernet/chelsio/cxgb4vf/sge.c b/drivers/net/ethernet/chelsio/cxgb4vf/sge.c
index 8877fbf..33fc1ca 100644
--- a/drivers/net/ethernet/chelsio/cxgb4vf/sge.c
+++ b/drivers/net/ethernet/chelsio/cxgb4vf/sge.c
@@ -536,7 +536,7 @@ static inline void ring_fl_db(struct adapter *adapter, struct sge_fl *fl)
if (fl->pend_cred >= FL_PER_EQ_UNIT) {
wmb();
t4_write_reg(adapter, T4VF_SGE_BASE_ADDR + SGE_VF_KDOORBELL,
- DBPRIO |
+ DBPRIO(1) |
QID(fl->cntxt_id) |
PIDX(fl->pend_cred / FL_PER_EQ_UNIT));
fl->pend_cred %= FL_PER_EQ_UNIT;
@@ -952,7 +952,7 @@ static inline void ring_tx_db(struct adapter *adapter, struct sge_txq *tq,
* Warn if we write doorbells with the wrong priority and write
* descriptors before telling HW.
*/
- WARN_ON((QID(tq->cntxt_id) | PIDX(n)) & DBPRIO);
+ WARN_ON((QID(tq->cntxt_id) | PIDX(n)) & DBPRIO(1));
wmb();
t4_write_reg(adapter, T4VF_SGE_BASE_ADDR + SGE_VF_KDOORBELL,
QID(tq->cntxt_id) | PIDX(n));
@@ -2126,8 +2126,8 @@ int t4vf_sge_alloc_rxq(struct adapter *adapter, struct sge_rspq *rspq,
cmd.iqns_to_fl0congen =
cpu_to_be32(
FW_IQ_CMD_FL0HOSTFCMODE(SGE_HOSTFCMODE_NONE) |
- FW_IQ_CMD_FL0PACKEN |
- FW_IQ_CMD_FL0PADEN);
+ FW_IQ_CMD_FL0PACKEN(1) |
+ FW_IQ_CMD_FL0PADEN(1));
cmd.fl0dcaen_to_fl0cidxfthresh =
cpu_to_be16(
FW_IQ_CMD_FL0FBMIN(SGE_FETCHBURSTMIN_64B) |
@@ -2431,7 +2431,8 @@ int t4vf_sge_init(struct adapter *adapter)
*/
if (fl1)
FL_PG_ORDER = ilog2(fl1) - PAGE_SHIFT;
- STAT_LEN = ((sge_params->sge_control & EGRSTATUSPAGESIZE) ? 128 : 64);
+ STAT_LEN = ((sge_params->sge_control & EGRSTATUSPAGESIZE(1)) ?
+ 128 : 64);
PKTSHIFT = PKTSHIFT_GET(sge_params->sge_control);
FL_ALIGN = 1 << (INGPADBOUNDARY_GET(sge_params->sge_control) +
SGE_INGPADBOUNDARY_SHIFT);
--
1.7.1
^ permalink raw reply related
* [V3 PATCH 3/9] csiostor: Chelsio FCoE offload driver submission (sources part 3).
From: Naresh Kumar Inna @ 2012-09-11 14:39 UTC (permalink / raw)
To: JBottomley, linux-scsi, dm, leedom; +Cc: netdev, naresh, chethan
In-Reply-To: <1347374347-3852-1-git-send-email-naresh@chelsio.com>
This patch contains code for the FC transport template callbacks and the
Mailbox module functionality. The FC transport callbacks include Virtual
Node ports creation and deletion, FC session registration, unregistration and
teardown. The Mailbox module provides services to issue/track/cancel
mailbox commands and wrappers for them.
Signed-off-by: Naresh Kumar Inna <naresh@chelsio.com>
---
drivers/scsi/csiostor/csio_attr.c | 809 +++++++++++++++++
drivers/scsi/csiostor/csio_mb.c | 1769 +++++++++++++++++++++++++++++++++++++
2 files changed, 2578 insertions(+), 0 deletions(-)
create mode 100644 drivers/scsi/csiostor/csio_attr.c
create mode 100644 drivers/scsi/csiostor/csio_mb.c
diff --git a/drivers/scsi/csiostor/csio_attr.c b/drivers/scsi/csiostor/csio_attr.c
new file mode 100644
index 0000000..ad29fd9
--- /dev/null
+++ b/drivers/scsi/csiostor/csio_attr.c
@@ -0,0 +1,809 @@
+/*
+ * This file is part of the Chelsio FCoE driver for Linux.
+ *
+ * Copyright (c) 2008-2012 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses. You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ * Redistribution and use in source and binary forms, with or
+ * without modification, are permitted provided that the following
+ * conditions are met:
+ *
+ * - Redistributions of source code must retain the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer in the documentation and/or other materials
+ * provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include <linux/kernel.h>
+#include <linux/string.h>
+#include <linux/delay.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/pci.h>
+#include <linux/mm.h>
+#include <linux/jiffies.h>
+#include <scsi/fc/fc_fs.h>
+
+#include "csio_init.h"
+
+static void
+csio_vport_set_state(struct csio_lnode *ln);
+
+/*
+ * csio_reg_rnode - Register a remote port with FC transport.
+ * @rn: Rnode representing remote port.
+ *
+ * Call fc_remote_port_add() to register this remote port with FC transport.
+ * If remote port is Initiator OR Target OR both, change the role appropriately.
+ *
+ */
+void
+csio_reg_rnode(struct csio_rnode *rn)
+{
+ struct csio_lnode *ln = csio_rnode_to_lnode(rn);
+ struct Scsi_Host *shost = csio_ln_to_shost(ln);
+ struct fc_rport_identifiers ids;
+ struct fc_rport *rport;
+ struct csio_service_parms *sp;
+
+ ids.node_name = wwn_to_u64(csio_rn_wwnn(rn));
+ ids.port_name = wwn_to_u64(csio_rn_wwpn(rn));
+ ids.port_id = rn->nport_id;
+ ids.roles = FC_RPORT_ROLE_UNKNOWN;
+
+ if (rn->role & CSIO_RNFR_INITIATOR || rn->role & CSIO_RNFR_TARGET) {
+ rport = rn->rport;
+ CSIO_ASSERT(rport != NULL);
+ goto update_role;
+ }
+
+ rn->rport = fc_remote_port_add(shost, 0, &ids);
+ if (!rn->rport) {
+ csio_ln_err(ln, "Failed to register rport = 0x%x.\n",
+ rn->nport_id);
+ return;
+ }
+
+ ln->num_reg_rnodes++;
+ rport = rn->rport;
+ spin_lock_irq(shost->host_lock);
+ *((struct csio_rnode **)rport->dd_data) = rn;
+ spin_unlock_irq(shost->host_lock);
+
+ sp = &rn->rn_sparm;
+ rport->maxframe_size = sp->csp.sp_bb_data;
+ if (ntohs(sp->clsp[2].cp_class) & FC_CPC_VALID)
+ rport->supported_classes = FC_COS_CLASS3;
+ else
+ rport->supported_classes = FC_COS_UNSPECIFIED;
+update_role:
+ if (rn->role & CSIO_RNFR_INITIATOR)
+ ids.roles |= FC_RPORT_ROLE_FCP_INITIATOR;
+ if (rn->role & CSIO_RNFR_TARGET) {
+ ids.roles |= FC_RPORT_ROLE_FCP_TARGET;
+ ln->n_scsi_tgts++;
+ }
+
+ if (ids.roles != FC_RPORT_ROLE_UNKNOWN)
+ fc_remote_port_rolechg(rport, ids.roles);
+
+ rn->scsi_id = rport->scsi_target_id;
+
+ csio_ln_dbg(ln, "Remote port x%x role 0x%x registered\n",
+ rn->nport_id, ids.roles);
+}
+
+/*
+ * csio_unreg_rnode - Unregister a remote port with FC transport.
+ * @rn: Rnode representing remote port.
+ *
+ * Call fc_remote_port_delete() to unregister this remote port with FC
+ * transport.
+ *
+ */
+void
+csio_unreg_rnode(struct csio_rnode *rn)
+{
+ struct csio_lnode *ln = csio_rnode_to_lnode(rn);
+ struct fc_rport *rport = rn->rport;
+
+ rn->role &= ~(CSIO_RNFR_INITIATOR | CSIO_RNFR_TARGET);
+ fc_remote_port_delete(rport);
+ ln->num_reg_rnodes--;
+
+ if (ln->n_scsi_tgts)
+ ln->n_scsi_tgts--;
+
+ if (ln->last_scan_ntgts)
+ ln->last_scan_ntgts--;
+
+ csio_ln_dbg(ln, "Remote port x%x un-registered\n", rn->nport_id);
+}
+
+/*
+ * csio_lnode_async_event - Async events from local port.
+ * @ln: lnode representing local port.
+ *
+ * Async events from local node that FC transport/SCSI ML
+ * should be made aware of (Eg: RSCN).
+ */
+void
+csio_lnode_async_event(struct csio_lnode *ln, enum csio_ln_fc_evt fc_evt)
+{
+ switch (fc_evt) {
+ case CSIO_LN_FC_RSCN:
+ /* Get payload of rscn from ln */
+ /* For each RSCN entry */
+ /*
+ * fc_host_post_event(shost,
+ * fc_get_event_number(),
+ * FCH_EVT_RSCN,
+ * rscn_entry);
+ */
+ break;
+ case CSIO_LN_FC_LINKUP:
+ /* send fc_host_post_event */
+ /* set vport state */
+ if (csio_is_npiv_ln(ln))
+ csio_vport_set_state(ln);
+
+ break;
+ case CSIO_LN_FC_LINKDOWN:
+ /* send fc_host_post_event */
+ /* set vport state */
+ if (csio_is_npiv_ln(ln))
+ csio_vport_set_state(ln);
+
+ break;
+ case CSIO_LN_FC_ATTRIB_UPDATE:
+ csio_fchost_attr_init(ln);
+ break;
+ default:
+ break;
+ }
+}
+
+/*
+ * csio_fchost_attr_init - Initialize FC transport attributes
+ * @ln: Lnode.
+ *
+ */
+void
+csio_fchost_attr_init(struct csio_lnode *ln)
+{
+ struct Scsi_Host *shost = csio_ln_to_shost(ln);
+
+ fc_host_node_name(shost) = wwn_to_u64(csio_ln_wwnn(ln));
+ fc_host_port_name(shost) = wwn_to_u64(csio_ln_wwpn(ln));
+
+ fc_host_supported_classes(shost) = FC_COS_CLASS3;
+ fc_host_max_npiv_vports(shost) =
+ (csio_lnode_to_hw(ln))->fres_info.max_vnps;
+ fc_host_supported_speeds(shost) = FC_PORTSPEED_10GBIT |
+ FC_PORTSPEED_1GBIT;
+
+ fc_host_maxframe_size(shost) = ln->ln_sparm.csp.sp_bb_data;
+ memset(fc_host_supported_fc4s(shost), 0,
+ sizeof(fc_host_supported_fc4s(shost)));
+ fc_host_supported_fc4s(shost)[7] = 1;
+
+ memset(fc_host_active_fc4s(shost), 0,
+ sizeof(fc_host_active_fc4s(shost)));
+ fc_host_active_fc4s(shost)[7] = 1;
+}
+
+/*
+ * csio_get_host_port_id - sysfs entries for nport_id is
+ * populated/cached from this function
+ */
+static void
+csio_get_host_port_id(struct Scsi_Host *shost)
+{
+ struct csio_lnode *ln = shost_priv(shost);
+ struct csio_hw *hw = csio_lnode_to_hw(ln);
+
+ spin_lock_irq(&hw->lock);
+ fc_host_port_id(shost) = ln->nport_id;
+ spin_unlock_irq(&hw->lock);
+}
+
+/*
+ * csio_get_port_type - Return FC local port type.
+ * @shost: scsi host.
+ *
+ */
+static void
+csio_get_host_port_type(struct Scsi_Host *shost)
+{
+ struct csio_lnode *ln = shost_priv(shost);
+ struct csio_hw *hw = csio_lnode_to_hw(ln);
+
+ spin_lock_irq(&hw->lock);
+ if (csio_is_npiv_ln(ln))
+ fc_host_port_type(shost) = FC_PORTTYPE_NPIV;
+ else
+ fc_host_port_type(shost) = FC_PORTTYPE_NPORT;
+ spin_unlock_irq(&hw->lock);
+}
+
+/*
+ * csio_get_port_state - Return FC local port state.
+ * @shost: scsi host.
+ *
+ */
+static void
+csio_get_host_port_state(struct Scsi_Host *shost)
+{
+ struct csio_lnode *ln = shost_priv(shost);
+ struct csio_hw *hw = csio_lnode_to_hw(ln);
+ char state[16];
+
+ spin_lock_irq(&hw->lock);
+
+ csio_lnode_state_to_str(ln, state);
+ if (!strcmp(state, "READY"))
+ fc_host_port_state(shost) = FC_PORTSTATE_ONLINE;
+ else if (!strcmp(state, "OFFLINE"))
+ fc_host_port_state(shost) = FC_PORTSTATE_LINKDOWN;
+ else
+ fc_host_port_state(shost) = FC_PORTSTATE_UNKNOWN;
+
+ spin_unlock_irq(&hw->lock);
+}
+
+/*
+ * csio_get_host_speed - Return link speed to FC transport.
+ * @shost: scsi host.
+ *
+ */
+static void
+csio_get_host_speed(struct Scsi_Host *shost)
+{
+ struct csio_lnode *ln = shost_priv(shost);
+ struct csio_hw *hw = csio_lnode_to_hw(ln);
+
+ spin_lock_irq(&hw->lock);
+ switch (hw->pport[ln->portid].link_speed) {
+ case FW_PORT_CAP_SPEED_1G:
+ fc_host_speed(shost) = FC_PORTSPEED_1GBIT;
+ break;
+ case FW_PORT_CAP_SPEED_10G:
+ fc_host_speed(shost) = FC_PORTSPEED_10GBIT;
+ break;
+ default:
+ fc_host_speed(shost) = FC_PORTSPEED_UNKNOWN;
+ break;
+ }
+ spin_unlock_irq(&hw->lock);
+}
+
+/*
+ * csio_get_host_fabric_name - Return fabric name
+ * @shost: scsi host.
+ *
+ */
+static void
+csio_get_host_fabric_name(struct Scsi_Host *shost)
+{
+ struct csio_lnode *ln = shost_priv(shost);
+ struct csio_rnode *rn = NULL;
+ struct csio_hw *hw = csio_lnode_to_hw(ln);
+
+ spin_lock_irq(&hw->lock);
+ rn = csio_rnode_lookup_portid(ln, FC_FID_FLOGI);
+ if (rn)
+ fc_host_fabric_name(shost) = wwn_to_u64(csio_rn_wwnn(rn));
+ else
+ fc_host_fabric_name(shost) = 0;
+ spin_unlock_irq(&hw->lock);
+}
+
+/*
+ * csio_get_host_speed - Return FC transport statistics.
+ * @ln: Lnode.
+ *
+ */
+static struct fc_host_statistics *
+csio_get_stats(struct Scsi_Host *shost)
+{
+ struct csio_lnode *ln = shost_priv(shost);
+ struct csio_hw *hw = csio_lnode_to_hw(ln);
+ struct fc_host_statistics *fhs = &ln->fch_stats;
+ struct fw_fcoe_port_stats fcoe_port_stats;
+ uint64_t seconds;
+
+ memset(&fcoe_port_stats, 0, sizeof(struct fw_fcoe_port_stats));
+ csio_get_phy_port_stats(hw, ln->portid, &fcoe_port_stats);
+
+ fhs->tx_frames += (fcoe_port_stats.tx_bcast_frames +
+ fcoe_port_stats.tx_mcast_frames +
+ fcoe_port_stats.tx_ucast_frames +
+ fcoe_port_stats.tx_offload_frames);
+ fhs->tx_words += (fcoe_port_stats.tx_bcast_bytes +
+ fcoe_port_stats.tx_mcast_bytes +
+ fcoe_port_stats.tx_ucast_bytes +
+ fcoe_port_stats.tx_offload_bytes) /
+ CSIO_WORD_TO_BYTE;
+ fhs->rx_frames += (fcoe_port_stats.rx_bcast_frames +
+ fcoe_port_stats.rx_mcast_frames +
+ fcoe_port_stats.rx_ucast_frames);
+ fhs->rx_words += (fcoe_port_stats.rx_bcast_bytes +
+ fcoe_port_stats.rx_mcast_bytes +
+ fcoe_port_stats.rx_ucast_bytes) /
+ CSIO_WORD_TO_BYTE;
+ fhs->error_frames += fcoe_port_stats.rx_err_frames;
+ fhs->fcp_input_requests += ln->stats.n_input_requests;
+ fhs->fcp_output_requests += ln->stats.n_output_requests;
+ fhs->fcp_control_requests += ln->stats.n_control_requests;
+ fhs->fcp_input_megabytes += ln->stats.n_input_bytes >> 20;
+ fhs->fcp_output_megabytes += ln->stats.n_output_bytes >> 20;
+ fhs->link_failure_count = ln->stats.n_link_down;
+ /* Reset stats for the device */
+ seconds = jiffies_to_msecs(jiffies);
+ fhs->seconds_since_last_reset =
+ (seconds - hw->stats.n_reset_start) / 1000;
+ return fhs;
+}
+
+/*
+ * csio_set_rport_loss_tmo - Set the rport dev loss timeout
+ * @rport: fc rport.
+ * @timeout: new value for dev loss tmo.
+ *
+ * If timeout is non zero set the dev_loss_tmo to timeout, else set
+ * dev_loss_tmo to one.
+ */
+static void
+csio_set_rport_loss_tmo(struct fc_rport *rport, uint32_t timeout)
+{
+ if (timeout)
+ rport->dev_loss_tmo = timeout;
+ else
+ rport->dev_loss_tmo = 1;
+}
+
+static void
+csio_vport_set_state(struct csio_lnode *ln)
+{
+ struct fc_vport *fc_vport = ln->fc_vport;
+ struct csio_lnode *pln = ln->pln;
+ char state[16];
+
+ /* Set fc vport state based on phyiscal lnode */
+ csio_lnode_state_to_str(pln, state);
+ if (strcmp(state, "READY")) {
+ fc_vport_set_state(fc_vport, FC_VPORT_LINKDOWN);
+ return;
+ }
+
+ if (!(pln->flags & CSIO_LNF_NPIVSUPP)) {
+ fc_vport_set_state(fc_vport, FC_VPORT_NO_FABRIC_SUPP);
+ return;
+ }
+
+ /* Set fc vport state based on virtual lnode */
+ csio_lnode_state_to_str(ln, state);
+ if (strcmp(state, "READY")) {
+ fc_vport_set_state(fc_vport, FC_VPORT_LINKDOWN);
+ return;
+ }
+ fc_vport_set_state(fc_vport, FC_VPORT_ACTIVE);
+}
+
+static int
+csio_fcoe_alloc_vnp(struct csio_hw *hw, struct csio_lnode *ln)
+{
+ struct csio_lnode *pln;
+ struct csio_mb *mbp;
+ struct fw_fcoe_vnp_cmd *rsp;
+ int ret;
+ int retry = 0;
+
+ /* Issue VNP cmd to alloc vport */
+ /* Allocate Mbox request */
+ spin_lock_irq(&hw->lock);
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ spin_unlock_irq(&hw->lock);
+ return -ENOMEM;
+ }
+
+ pln = ln->pln;
+ ln->fcf_flowid = pln->fcf_flowid;
+ ln->portid = pln->portid;
+
+ csio_fcoe_vnp_alloc_init_mb(ln, mbp, CSIO_MB_DEFAULT_TMO,
+ pln->fcf_flowid, pln->vnp_flowid, 0,
+ csio_ln_wwnn(ln), csio_ln_wwpn(ln),
+ NULL);
+
+ for (retry = 0; retry < 3; retry++) {
+ /* FW is expected to complete vnp cmd in immediate mode
+ * without much delay.
+ * Otherwise, there will be increase in IO latency since HW
+ * lock is held till completion of vnp mbox cmd.
+ */
+ ret = csio_mb_issue(hw, mbp);
+ if (ret != -EBUSY)
+ break;
+
+ /* Retry if mbox returns busy */
+ spin_unlock_irq(&hw->lock);
+ msleep(2000);
+ spin_lock_irq(&hw->lock);
+ }
+
+ if (ret != 0) {
+ csio_ln_err(ln, "Failed to issue mbox FCoE VNP command\n");
+ mempool_free(mbp, hw->mb_mempool);
+ spin_unlock_irq(&hw->lock);
+ return ret;
+ }
+
+ /* Process Mbox response of VNP command */
+ rsp = (struct fw_fcoe_vnp_cmd *)(mbp->mb);
+ if (FW_CMD_RETVAL_GET(ntohl(rsp->alloc_to_len16)) != FW_SUCCESS) {
+ csio_ln_err(ln, "FCOE VNP ALLOC cmd returned 0x%x!\n",
+ FW_CMD_RETVAL_GET(ntohl(rsp->alloc_to_len16)));
+ mempool_free(mbp, hw->mb_mempool);
+ spin_unlock_irq(&hw->lock);
+ return -ENOMEM;
+ }
+
+ ln->vnp_flowid = FW_FCOE_VNP_CMD_VNPI_GET(
+ ntohl(rsp->gen_wwn_to_vnpi));
+ memcpy(csio_ln_wwnn(ln), rsp->vnport_wwnn, 8);
+ memcpy(csio_ln_wwpn(ln), rsp->vnport_wwpn, 8);
+
+ csio_ln_dbg(ln, "FCOE VNPI: 0x%x\n", ln->vnp_flowid);
+ csio_ln_dbg(ln, "\tWWNN: %x%x%x%x%x%x%x%x\n",
+ ln->ln_sparm.wwnn[0], ln->ln_sparm.wwnn[1],
+ ln->ln_sparm.wwnn[2], ln->ln_sparm.wwnn[3],
+ ln->ln_sparm.wwnn[4], ln->ln_sparm.wwnn[5],
+ ln->ln_sparm.wwnn[6], ln->ln_sparm.wwnn[7]);
+ csio_ln_dbg(ln, "\tWWPN: %x%x%x%x%x%x%x%x\n",
+ ln->ln_sparm.wwpn[0], ln->ln_sparm.wwpn[1],
+ ln->ln_sparm.wwpn[2], ln->ln_sparm.wwpn[3],
+ ln->ln_sparm.wwpn[4], ln->ln_sparm.wwpn[5],
+ ln->ln_sparm.wwpn[6], ln->ln_sparm.wwpn[7]);
+
+ mempool_free(mbp, hw->mb_mempool);
+ spin_unlock_irq(&hw->lock);
+
+ return 0;
+}
+
+static int
+csio_fcoe_free_vnp(struct csio_hw *hw, struct csio_lnode *ln)
+{
+ struct csio_lnode *pln;
+ struct csio_mb *mbp;
+ struct fw_fcoe_vnp_cmd *rsp;
+ int ret;
+ int retry = 0;
+
+ /* Issue VNP cmd to free vport */
+ /* Allocate Mbox request */
+
+ spin_lock_irq(&hw->lock);
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ spin_unlock_irq(&hw->lock);
+ return -ENOMEM;
+ }
+
+ pln = ln->pln;
+
+ csio_fcoe_vnp_free_init_mb(ln, mbp, CSIO_MB_DEFAULT_TMO,
+ ln->fcf_flowid, ln->vnp_flowid,
+ NULL);
+
+ for (retry = 0; retry < 3; retry++) {
+ ret = csio_mb_issue(hw, mbp);
+ if (ret != -EBUSY)
+ break;
+
+ /* Retry if mbox returns busy */
+ spin_unlock_irq(&hw->lock);
+ msleep(2000);
+ spin_lock_irq(&hw->lock);
+ }
+
+ if (ret) {
+ csio_ln_err(ln, "Failed to issue mbox FCoE VNP command\n");
+ mempool_free(mbp, hw->mb_mempool);
+ spin_unlock_irq(&hw->lock);
+ return -EINVAL;
+ }
+
+ /* Process Mbox response of VNP command */
+ rsp = (struct fw_fcoe_vnp_cmd *)(mbp->mb);
+ if (FW_CMD_RETVAL_GET(ntohl(rsp->alloc_to_len16)) != FW_SUCCESS) {
+ csio_ln_err(ln, "FCOE VNP FREE cmd returned 0x%x!\n",
+ FW_CMD_RETVAL_GET(ntohl(rsp->alloc_to_len16)));
+ mempool_free(mbp, hw->mb_mempool);
+ spin_unlock_irq(&hw->lock);
+ return -ENOMEM;
+ }
+
+ mempool_free(mbp, hw->mb_mempool);
+ spin_unlock_irq(&hw->lock);
+
+ return 0;
+}
+
+static int
+csio_vport_create(struct fc_vport *fc_vport, bool disable)
+{
+ struct Scsi_Host *shost = fc_vport->shost;
+ struct csio_lnode *pln = shost_priv(shost);
+ struct csio_lnode *ln = NULL;
+ struct csio_hw *hw = csio_lnode_to_hw(pln);
+ uint8_t wwn[8];
+ int ret = -1;
+
+ ln = csio_shost_init(hw, &fc_vport->dev, false, pln);
+ if (!ln)
+ goto error;
+
+ if (fc_vport->node_name != 0) {
+ u64_to_wwn(fc_vport->node_name, wwn);
+
+ if (!CSIO_VALID_WWN(wwn)) {
+ csio_ln_err(ln,
+ "vport create failed. Invalid wwnn\n");
+ goto error;
+ }
+ memcpy(csio_ln_wwnn(ln), wwn, 8);
+ }
+
+ if (fc_vport->port_name != 0) {
+ u64_to_wwn(fc_vport->port_name, wwn);
+
+ if (!CSIO_VALID_WWN(wwn)) {
+ csio_ln_err(ln,
+ "vport create failed. Invalid wwpn\n");
+ goto error;
+ }
+
+ if (csio_lnode_lookup_by_wwpn(hw, wwn)) {
+ csio_ln_err(ln,
+ "vport create failed. wwpn already exists\n");
+ goto error;
+ }
+ memcpy(csio_ln_wwpn(ln), wwn, 8);
+ }
+
+ fc_vport_set_state(fc_vport, FC_VPORT_INITIALIZING);
+
+ if (csio_fcoe_alloc_vnp(hw, ln))
+ goto error;
+
+ *(struct csio_lnode **)fc_vport->dd_data = ln;
+ ln->fc_vport = fc_vport;
+ if (!fc_vport->node_name)
+ fc_vport->node_name = wwn_to_u64(csio_ln_wwnn(ln));
+ if (!fc_vport->port_name)
+ fc_vport->port_name = wwn_to_u64(csio_ln_wwpn(ln));
+ csio_fchost_attr_init(ln);
+ return 0;
+error:
+ if (ln)
+ csio_shost_exit(ln);
+
+ return ret;
+}
+
+static int
+csio_vport_delete(struct fc_vport *fc_vport)
+{
+ struct csio_lnode *ln = *(struct csio_lnode **)fc_vport->dd_data;
+ struct Scsi_Host *shost = csio_ln_to_shost(ln);
+ struct csio_hw *hw = csio_lnode_to_hw(ln);
+
+ spin_lock_irq(&hw->lock);
+ if (csio_is_hw_removing(hw)) {
+ spin_unlock_irq(&hw->lock);
+ csio_shost_exit(ln);
+ return 0;
+ }
+ spin_unlock_irq(&hw->lock);
+
+ /* Quiesce ios and send remove event to lnode */
+ scsi_block_requests(shost);
+ spin_lock_irq(&hw->lock);
+ csio_scsim_cleanup_io_lnode(csio_hw_to_scsim(hw), ln);
+ csio_lnode_close(ln);
+ spin_unlock_irq(&hw->lock);
+ scsi_unblock_requests(shost);
+
+ /* Free vnp */
+ if (fc_vport->vport_state != FC_VPORT_DISABLED)
+ csio_fcoe_free_vnp(hw, ln);
+ csio_ln_err(ln, "vport deleted\n");
+ csio_shost_exit(ln);
+ return 0;
+}
+
+static int
+csio_vport_disable(struct fc_vport *fc_vport, bool disable)
+{
+ struct csio_lnode *ln = *(struct csio_lnode **)fc_vport->dd_data;
+ struct Scsi_Host *shost = csio_ln_to_shost(ln);
+ struct csio_hw *hw = csio_lnode_to_hw(ln);
+
+ /* disable vport */
+ if (disable) {
+ /* Quiesce ios and send stop event to lnode */
+ scsi_block_requests(shost);
+ spin_lock_irq(&hw->lock);
+ csio_scsim_cleanup_io_lnode(csio_hw_to_scsim(hw), ln);
+ csio_lnode_stop(ln);
+ spin_unlock_irq(&hw->lock);
+ scsi_unblock_requests(shost);
+
+ /* Free vnp */
+ csio_fcoe_free_vnp(hw, ln);
+ fc_vport_set_state(fc_vport, FC_VPORT_DISABLED);
+ csio_ln_err(ln, "vport disabled\n");
+ return 0;
+ } else {
+ /* enable vport */
+ fc_vport_set_state(fc_vport, FC_VPORT_INITIALIZING);
+ if (csio_fcoe_alloc_vnp(hw, ln)) {
+ csio_ln_err(ln, "vport enabled failed.\n");
+ return -1;
+ }
+ csio_ln_err(ln, "vport enabled\n");
+ return 0;
+ }
+}
+
+static void
+csio_dev_loss_tmo_callbk(struct fc_rport *rport)
+{
+ struct csio_rnode *rn;
+ struct csio_hw *hw;
+ struct csio_lnode *ln;
+
+ rn = *((struct csio_rnode **)rport->dd_data);
+ ln = csio_rnode_to_lnode(rn);
+ hw = csio_lnode_to_hw(ln);
+
+ spin_lock_irq(&hw->lock);
+
+ /* return if driver is being removed or same rnode comes back online */
+ if (csio_is_hw_removing(hw) || csio_is_rnode_ready(rn)) {
+ spin_unlock_irq(&hw->lock);
+ return;
+ }
+
+ csio_ln_dbg(ln, "devloss timeout on rnode:%p portid:x%x flowid:x%x\n",
+ rn, rn->nport_id, csio_rn_flowid(rn));
+
+ CSIO_INC_STATS(ln, n_dev_loss_tmo);
+
+ /*
+ * enqueue devloss event to event worker thread to serialize all
+ * rnode events.
+ */
+ if (csio_enqueue_evt(hw, CSIO_EVT_DEV_LOSS, &rn, sizeof(rn))) {
+ CSIO_INC_STATS(hw, n_evt_drop);
+ spin_unlock_irq(&hw->lock);
+ return;
+ }
+
+ if (!(hw->flags & CSIO_HWF_FWEVT_PENDING)) {
+ hw->flags |= CSIO_HWF_FWEVT_PENDING;
+ spin_unlock_irq(&hw->lock);
+ schedule_work(&hw->evtq_work);
+ return;
+ }
+
+ spin_unlock_irq(&hw->lock);
+}
+
+/* FC transport functions template - Physical port */
+struct fc_function_template csio_fc_transport_funcs = {
+ .show_host_node_name = 1,
+ .show_host_port_name = 1,
+ .show_host_supported_classes = 1,
+ .show_host_supported_fc4s = 1,
+ .show_host_maxframe_size = 1,
+
+ .get_host_port_id = csio_get_host_port_id,
+ .show_host_port_id = 1,
+
+ .get_host_port_type = csio_get_host_port_type,
+ .show_host_port_type = 1,
+
+ .get_host_port_state = csio_get_host_port_state,
+ .show_host_port_state = 1,
+
+ .show_host_active_fc4s = 1,
+ .get_host_speed = csio_get_host_speed,
+ .show_host_speed = 1,
+ .get_host_fabric_name = csio_get_host_fabric_name,
+ .show_host_fabric_name = 1,
+
+ .get_fc_host_stats = csio_get_stats,
+
+ .dd_fcrport_size = sizeof(struct csio_rnode *),
+ .show_rport_maxframe_size = 1,
+ .show_rport_supported_classes = 1,
+
+ .set_rport_dev_loss_tmo = csio_set_rport_loss_tmo,
+ .show_rport_dev_loss_tmo = 1,
+
+ .show_starget_port_id = 1,
+ .show_starget_node_name = 1,
+ .show_starget_port_name = 1,
+
+ .dev_loss_tmo_callbk = csio_dev_loss_tmo_callbk,
+ .dd_fcvport_size = sizeof(struct csio_lnode *),
+
+ .vport_create = csio_vport_create,
+ .vport_disable = csio_vport_disable,
+ .vport_delete = csio_vport_delete,
+};
+
+/* FC transport functions template - Virtual port */
+struct fc_function_template csio_fc_transport_vport_funcs = {
+ .show_host_node_name = 1,
+ .show_host_port_name = 1,
+ .show_host_supported_classes = 1,
+ .show_host_supported_fc4s = 1,
+ .show_host_maxframe_size = 1,
+
+ .get_host_port_id = csio_get_host_port_id,
+ .show_host_port_id = 1,
+
+ .get_host_port_type = csio_get_host_port_type,
+ .show_host_port_type = 1,
+
+ .get_host_port_state = csio_get_host_port_state,
+ .show_host_port_state = 1,
+ .show_host_active_fc4s = 1,
+
+ .get_host_speed = csio_get_host_speed,
+ .show_host_speed = 1,
+
+ .get_host_fabric_name = csio_get_host_fabric_name,
+ .show_host_fabric_name = 1,
+
+ .get_fc_host_stats = csio_get_stats,
+
+ .dd_fcrport_size = sizeof(struct csio_rnode *),
+ .show_rport_maxframe_size = 1,
+ .show_rport_supported_classes = 1,
+
+ .set_rport_dev_loss_tmo = csio_set_rport_loss_tmo,
+ .show_rport_dev_loss_tmo = 1,
+
+ .show_starget_port_id = 1,
+ .show_starget_node_name = 1,
+ .show_starget_port_name = 1,
+
+ .dev_loss_tmo_callbk = csio_dev_loss_tmo_callbk,
+
+};
diff --git a/drivers/scsi/csiostor/csio_mb.c b/drivers/scsi/csiostor/csio_mb.c
new file mode 100644
index 0000000..100afd1
--- /dev/null
+++ b/drivers/scsi/csiostor/csio_mb.c
@@ -0,0 +1,1769 @@
+/*
+ * This file is part of the Chelsio FCoE driver for Linux.
+ *
+ * Copyright (c) 2008-2012 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses. You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ * Redistribution and use in source and binary forms, with or
+ * without modification, are permitted provided that the following
+ * conditions are met:
+ *
+ * - Redistributions of source code must retain the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer in the documentation and/or other materials
+ * provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include <linux/delay.h>
+#include <linux/jiffies.h>
+#include <linux/string.h>
+#include <scsi/scsi_transport_fc.h>
+
+#include "csio_hw.h"
+#include "csio_lnode.h"
+#include "csio_rnode.h"
+#include "csio_mb.h"
+#include "csio_wr.h"
+
+#define csio_mb_is_host_owner(__owner) ((__owner) == CSIO_MBOWNER_PL)
+
+/* MB Command/Response Helpers */
+/*
+ * csio_mb_fw_retval - FW return value from a mailbox response.
+ * @mbp: Mailbox structure
+ *
+ */
+enum fw_retval
+csio_mb_fw_retval(struct csio_mb *mbp)
+{
+ struct fw_cmd_hdr *hdr;
+
+ hdr = (struct fw_cmd_hdr *)(mbp->mb);
+
+ return FW_CMD_RETVAL_GET(ntohl(hdr->lo));
+}
+
+/*
+ * csio_mb_hello - FW HELLO command helper
+ * @hw: The HW structure
+ * @mbp: Mailbox structure
+ * @m_mbox: Master mailbox number, if any.
+ * @a_mbox: Mailbox number for asycn notifications.
+ * @master: Device mastership.
+ * @cbfn: Callback, if any.
+ *
+ */
+void
+csio_mb_hello(struct csio_hw *hw, struct csio_mb *mbp, uint32_t tmo,
+ uint32_t m_mbox, uint32_t a_mbox, enum csio_dev_master master,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct fw_hello_cmd *cmdp = (struct fw_hello_cmd *)(mbp->mb);
+
+ CSIO_INIT_MBP(mbp, cmdp, tmo, hw, cbfn, 1);
+
+ cmdp->op_to_write = htonl(FW_CMD_OP(FW_HELLO_CMD) |
+ FW_CMD_REQUEST | FW_CMD_WRITE);
+ cmdp->retval_len16 = htonl(FW_CMD_LEN16(sizeof(*cmdp) / 16));
+ cmdp->err_to_clearinit = htonl(
+ FW_HELLO_CMD_MASTERDIS(master == CSIO_MASTER_CANT) |
+ FW_HELLO_CMD_MASTERFORCE(master == CSIO_MASTER_MUST) |
+ FW_HELLO_CMD_MBMASTER(master == CSIO_MASTER_MUST ?
+ m_mbox : FW_HELLO_CMD_MBMASTER_MASK) |
+ FW_HELLO_CMD_MBASYNCNOT(a_mbox) |
+ FW_HELLO_CMD_STAGE(FW_HELLO_CMD_STAGE_OS) |
+ FW_HELLO_CMD_CLEARINIT);
+
+}
+
+/*
+ * csio_mb_process_hello_rsp - FW HELLO response processing helper
+ * @hw: The HW structure
+ * @mbp: Mailbox structure
+ * @retval: Mailbox return value from Firmware
+ * @state: State that the function is in.
+ * @mpfn: Master pfn
+ *
+ */
+void
+csio_mb_process_hello_rsp(struct csio_hw *hw, struct csio_mb *mbp,
+ enum fw_retval *retval, enum csio_dev_state *state,
+ uint8_t *mpfn)
+{
+ struct fw_hello_cmd *rsp = (struct fw_hello_cmd *)(mbp->mb);
+ uint32_t value;
+
+ *retval = FW_CMD_RETVAL_GET(ntohl(rsp->retval_len16));
+
+ if (*retval == FW_SUCCESS) {
+ hw->fwrev = ntohl(rsp->fwrev);
+
+ value = ntohl(rsp->err_to_clearinit);
+ *mpfn = FW_HELLO_CMD_MBMASTER_GET(value);
+
+ if (value & FW_HELLO_CMD_INIT)
+ *state = CSIO_DEV_STATE_INIT;
+ else if (value & FW_HELLO_CMD_ERR)
+ *state = CSIO_DEV_STATE_ERR;
+ else
+ *state = CSIO_DEV_STATE_UNINIT;
+ }
+}
+
+/*
+ * csio_mb_bye - FW BYE command helper
+ * @hw: The HW structure
+ * @mbp: Mailbox structure
+ * @cbfn: Callback, if any.
+ *
+ */
+void
+csio_mb_bye(struct csio_hw *hw, struct csio_mb *mbp, uint32_t tmo,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct fw_bye_cmd *cmdp = (struct fw_bye_cmd *)(mbp->mb);
+
+ CSIO_INIT_MBP(mbp, cmdp, tmo, hw, cbfn, 1);
+
+ cmdp->op_to_write = htonl(FW_CMD_OP(FW_BYE_CMD) |
+ FW_CMD_REQUEST | FW_CMD_WRITE);
+ cmdp->retval_len16 = htonl(FW_CMD_LEN16(sizeof(*cmdp) / 16));
+
+}
+
+/*
+ * csio_mb_reset - FW RESET command helper
+ * @hw: The HW structure
+ * @mbp: Mailbox structure
+ * @reset: Type of reset.
+ * @cbfn: Callback, if any.
+ *
+ */
+void
+csio_mb_reset(struct csio_hw *hw, struct csio_mb *mbp, uint32_t tmo,
+ int reset, int halt,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct fw_reset_cmd *cmdp = (struct fw_reset_cmd *)(mbp->mb);
+
+ CSIO_INIT_MBP(mbp, cmdp, tmo, hw, cbfn, 1);
+
+ cmdp->op_to_write = htonl(FW_CMD_OP(FW_RESET_CMD) |
+ FW_CMD_REQUEST | FW_CMD_WRITE);
+ cmdp->retval_len16 = htonl(FW_CMD_LEN16(sizeof(*cmdp) / 16));
+ cmdp->val = htonl(reset);
+ cmdp->halt_pkd = htonl(halt);
+
+}
+
+/*
+ * csio_mb_params - FW PARAMS command helper
+ * @hw: The HW structure
+ * @mbp: Mailbox structure
+ * @tmo: Command timeout.
+ * @pf: PF number.
+ * @vf: VF number.
+ * @nparams: Number of paramters
+ * @params: Parameter mnemonic array.
+ * @val: Parameter value array.
+ * @wr: Write/Read PARAMS.
+ * @cbfn: Callback, if any.
+ *
+ */
+void
+csio_mb_params(struct csio_hw *hw, struct csio_mb *mbp, uint32_t tmo,
+ unsigned int pf, unsigned int vf, unsigned int nparams,
+ const u32 *params, u32 *val, bool wr,
+ void (*cbfn)(struct csio_hw *, struct csio_mb *))
+{
+ uint32_t i;
+ uint32_t temp_params = 0, temp_val = 0;
+ struct fw_params_cmd *cmdp = (struct fw_params_cmd *)(mbp->mb);
+ __be32 *p = &cmdp->param[0].mnem;
+
+ CSIO_INIT_MBP(mbp, cmdp, tmo, hw, cbfn, 1);
+
+ cmdp->op_to_vfn = htonl(FW_CMD_OP(FW_PARAMS_CMD) |
+ FW_CMD_REQUEST |
+ (wr ? FW_CMD_WRITE : FW_CMD_READ) |
+ FW_PARAMS_CMD_PFN(pf) |
+ FW_PARAMS_CMD_VFN(vf));
+ cmdp->retval_len16 = htonl(FW_CMD_LEN16(sizeof(*cmdp) / 16));
+
+ /* Write Params */
+ if (wr) {
+ while (nparams--) {
+ temp_params = *params++;
+ temp_val = *val++;
+
+ *p++ = htonl(temp_params);
+ *p++ = htonl(temp_val);
+ }
+ } else {
+ for (i = 0; i < nparams; i++, p += 2) {
+ temp_params = *params++;
+ *p = htonl(temp_params);
+ }
+ }
+
+}
+
+/*
+ * csio_mb_process_read_params_rsp - FW PARAMS response processing helper
+ * @hw: The HW structure
+ * @mbp: Mailbox structure
+ * @retval: Mailbox return value from Firmware
+ * @nparams: Number of parameters
+ * @val: Parameter value array.
+ *
+ */
+void
+csio_mb_process_read_params_rsp(struct csio_hw *hw, struct csio_mb *mbp,
+ enum fw_retval *retval, unsigned int nparams,
+ u32 *val)
+{
+ struct fw_params_cmd *rsp = (struct fw_params_cmd *)(mbp->mb);
+ uint32_t i;
+ __be32 *p = &rsp->param[0].val;
+
+ *retval = FW_CMD_RETVAL_GET(ntohl(rsp->retval_len16));
+
+ if (*retval == FW_SUCCESS)
+ for (i = 0; i < nparams; i++, p += 2)
+ *val++ = ntohl(*p);
+}
+
+/*
+ * csio_mb_ldst - FW LDST command
+ * @hw: The HW structure
+ * @mbp: Mailbox structure
+ * @tmo: timeout
+ * @reg: register
+ *
+ */
+void
+csio_mb_ldst(struct csio_hw *hw, struct csio_mb *mbp, uint32_t tmo, int reg)
+{
+ struct fw_ldst_cmd *ldst_cmd = (struct fw_ldst_cmd *)(mbp->mb);
+ CSIO_INIT_MBP(mbp, ldst_cmd, tmo, hw, NULL, 1);
+
+ /*
+ * Construct and send the Firmware LDST Command to retrieve the
+ * specified PCI-E Configuration Space register.
+ */
+ ldst_cmd->op_to_addrspace =
+ htonl(FW_CMD_OP(FW_LDST_CMD) |
+ FW_CMD_REQUEST |
+ FW_CMD_READ |
+ FW_LDST_CMD_ADDRSPACE(FW_LDST_ADDRSPC_FUNC_PCIE));
+ ldst_cmd->cycles_to_len16 = htonl(FW_LEN16(struct fw_ldst_cmd));
+ ldst_cmd->u.pcie.select_naccess = FW_LDST_CMD_NACCESS(1);
+ ldst_cmd->u.pcie.ctrl_to_fn =
+ (FW_LDST_CMD_LC | FW_LDST_CMD_FN(hw->pfn));
+ ldst_cmd->u.pcie.r = (uint8_t)reg;
+}
+
+/*
+ *
+ * csio_mb_caps_config - FW Read/Write Capabilities command helper
+ * @hw: The HW structure
+ * @mbp: Mailbox structure
+ * @wr: Write if 1, Read if 0
+ * @init: Turn on initiator mode.
+ * @tgt: Turn on target mode.
+ * @cofld: If 1, Control Offload for FCoE
+ * @cbfn: Callback, if any.
+ *
+ * This helper assumes that cmdp has MB payload from a previous CAPS
+ * read command.
+ */
+void
+csio_mb_caps_config(struct csio_hw *hw, struct csio_mb *mbp, uint32_t tmo,
+ bool wr, bool init, bool tgt, bool cofld,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct fw_caps_config_cmd *cmdp =
+ (struct fw_caps_config_cmd *)(mbp->mb);
+
+ CSIO_INIT_MBP(mbp, cmdp, tmo, hw, cbfn, wr ? 0 : 1);
+
+ cmdp->op_to_write = htonl(FW_CMD_OP(FW_CAPS_CONFIG_CMD) |
+ FW_CMD_REQUEST |
+ (wr ? FW_CMD_WRITE : FW_CMD_READ));
+ cmdp->cfvalid_to_len16 = htonl(FW_CMD_LEN16(sizeof(*cmdp) / 16));
+
+ /* Read config */
+ if (!wr)
+ return;
+
+ /* Write config */
+ cmdp->fcoecaps = 0;
+
+ if (cofld)
+ cmdp->fcoecaps |= htons(FW_CAPS_CONFIG_FCOE_CTRL_OFLD);
+ if (init)
+ cmdp->fcoecaps |= htons(FW_CAPS_CONFIG_FCOE_INITIATOR);
+ if (tgt)
+ cmdp->fcoecaps |= htons(FW_CAPS_CONFIG_FCOE_TARGET);
+}
+
+void
+csio_rss_glb_config(struct csio_hw *hw, struct csio_mb *mbp,
+ uint32_t tmo, uint8_t mode, unsigned int flags,
+ void (*cbfn)(struct csio_hw *, struct csio_mb *))
+{
+ struct fw_rss_glb_config_cmd *cmdp =
+ (struct fw_rss_glb_config_cmd *)(mbp->mb);
+
+ CSIO_INIT_MBP(mbp, cmdp, tmo, hw, cbfn, 1);
+
+ cmdp->op_to_write = htonl(FW_CMD_OP(FW_RSS_GLB_CONFIG_CMD) |
+ FW_CMD_REQUEST | FW_CMD_WRITE);
+ cmdp->retval_len16 = htonl(FW_CMD_LEN16(sizeof(*cmdp) / 16));
+
+ if (mode == FW_RSS_GLB_CONFIG_CMD_MODE_MANUAL) {
+ cmdp->u.manual.mode_pkd =
+ htonl(FW_RSS_GLB_CONFIG_CMD_MODE(mode));
+ } else if (mode == FW_RSS_GLB_CONFIG_CMD_MODE_BASICVIRTUAL) {
+ cmdp->u.basicvirtual.mode_pkd =
+ htonl(FW_RSS_GLB_CONFIG_CMD_MODE(mode));
+ cmdp->u.basicvirtual.synmapen_to_hashtoeplitz = htonl(flags);
+ }
+}
+
+
+/*
+ * csio_mb_pfvf - FW Write PF/VF capabilities command helper.
+ * @hw: The HW structure
+ * @mbp: Mailbox structure
+ * @pf:
+ * @vf:
+ * @txq:
+ * @txq_eht_ctrl:
+ * @rxqi:
+ * @rxq:
+ * @tc:
+ * @vi:
+ * @pmask:
+ * @rcaps:
+ * @wxcaps:
+ * @cbfn: Callback, if any.
+ *
+ */
+void
+csio_mb_pfvf(struct csio_hw *hw, struct csio_mb *mbp, uint32_t tmo,
+ unsigned int pf, unsigned int vf, unsigned int txq,
+ unsigned int txq_eth_ctrl, unsigned int rxqi,
+ unsigned int rxq, unsigned int tc, unsigned int vi,
+ unsigned int cmask, unsigned int pmask, unsigned int nexactf,
+ unsigned int rcaps, unsigned int wxcaps,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct fw_pfvf_cmd *cmdp = (struct fw_pfvf_cmd *)(mbp->mb);
+
+ CSIO_INIT_MBP(mbp, cmdp, tmo, hw, cbfn, 1);
+
+ cmdp->op_to_vfn = htonl(FW_CMD_OP(FW_PFVF_CMD) |
+ FW_CMD_REQUEST |
+ FW_CMD_WRITE |
+ FW_PFVF_CMD_PFN(pf) |
+ FW_PFVF_CMD_VFN(vf));
+ cmdp->retval_len16 = htonl(FW_CMD_LEN16(sizeof(*cmdp) / 16));
+ cmdp->niqflint_niq = htonl(FW_PFVF_CMD_NIQFLINT(rxqi) |
+ FW_PFVF_CMD_NIQ(rxq));
+
+ cmdp->type_to_neq = htonl(FW_PFVF_CMD_TYPE |
+ FW_PFVF_CMD_CMASK(cmask) |
+ FW_PFVF_CMD_PMASK(pmask) |
+ FW_PFVF_CMD_NEQ(txq));
+ cmdp->tc_to_nexactf = htonl(FW_PFVF_CMD_TC(tc) |
+ FW_PFVF_CMD_NVI(vi) |
+ FW_PFVF_CMD_NEXACTF(nexactf));
+ cmdp->r_caps_to_nethctrl = htonl(FW_PFVF_CMD_R_CAPS(rcaps) |
+ FW_PFVF_CMD_WX_CAPS(wxcaps) |
+ FW_PFVF_CMD_NETHCTRL(txq_eth_ctrl));
+}
+
+#define CSIO_ADVERT_MASK (FW_PORT_CAP_SPEED_100M | FW_PORT_CAP_SPEED_1G |\
+ FW_PORT_CAP_SPEED_10G | FW_PORT_CAP_ANEG)
+
+/*
+ * csio_mb_port- FW PORT command helper
+ * @hw: The HW structure
+ * @mbp: Mailbox structure
+ * @tmo: COmmand timeout
+ * @portid: Port ID to get/set info
+ * @wr: Write/Read PORT information.
+ * @fc: Flow control
+ * @caps: Port capabilites to set.
+ * @cbfn: Callback, if any.
+ *
+ */
+void
+csio_mb_port(struct csio_hw *hw, struct csio_mb *mbp, uint32_t tmo,
+ uint8_t portid, bool wr, uint32_t fc, uint16_t caps,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct fw_port_cmd *cmdp = (struct fw_port_cmd *)(mbp->mb);
+ unsigned int lfc = 0, mdi = FW_PORT_MDI(FW_PORT_MDI_AUTO);
+
+ CSIO_INIT_MBP(mbp, cmdp, tmo, hw, cbfn, 1);
+
+ cmdp->op_to_portid = htonl(FW_CMD_OP(FW_PORT_CMD) |
+ FW_CMD_REQUEST |
+ (wr ? FW_CMD_EXEC : FW_CMD_READ) |
+ FW_PORT_CMD_PORTID(portid));
+ if (!wr) {
+ cmdp->action_to_len16 = htonl(
+ FW_PORT_CMD_ACTION(FW_PORT_ACTION_GET_PORT_INFO) |
+ FW_CMD_LEN16(sizeof(*cmdp) / 16));
+ return;
+ }
+
+ /* Set port */
+ cmdp->action_to_len16 = htonl(
+ FW_PORT_CMD_ACTION(FW_PORT_ACTION_L1_CFG) |
+ FW_CMD_LEN16(sizeof(*cmdp) / 16));
+
+ if (fc & PAUSE_RX)
+ lfc |= FW_PORT_CAP_FC_RX;
+ if (fc & PAUSE_TX)
+ lfc |= FW_PORT_CAP_FC_TX;
+
+ if (!(caps & FW_PORT_CAP_ANEG))
+ cmdp->u.l1cfg.rcap = htonl((caps & CSIO_ADVERT_MASK) | lfc);
+ else
+ cmdp->u.l1cfg.rcap = htonl((caps & CSIO_ADVERT_MASK) |
+ lfc | mdi);
+}
+
+/*
+ * csio_mb_process_read_port_rsp - FW PORT command response processing helper
+ * @hw: The HW structure
+ * @mbp: Mailbox structure
+ * @retval: Mailbox return value from Firmware
+ * @caps: port capabilities
+ *
+ */
+void
+csio_mb_process_read_port_rsp(struct csio_hw *hw, struct csio_mb *mbp,
+ enum fw_retval *retval, uint16_t *caps)
+{
+ struct fw_port_cmd *rsp = (struct fw_port_cmd *)(mbp->mb);
+
+ *retval = FW_CMD_RETVAL_GET(ntohl(rsp->action_to_len16));
+
+ if (*retval == FW_SUCCESS)
+ *caps = ntohs(rsp->u.info.pcap);
+}
+
+/*
+ * csio_mb_initialize - FW INITIALIZE command helper
+ * @hw: The HW structure
+ * @mbp: Mailbox structure
+ * @tmo: COmmand timeout
+ * @cbfn: Callback, if any.
+ *
+ */
+void
+csio_mb_initialize(struct csio_hw *hw, struct csio_mb *mbp, uint32_t tmo,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct fw_initialize_cmd *cmdp = (struct fw_initialize_cmd *)(mbp->mb);
+
+ CSIO_INIT_MBP(mbp, cmdp, tmo, hw, cbfn, 1);
+
+ cmdp->op_to_write = htonl(FW_CMD_OP(FW_INITIALIZE_CMD) |
+ FW_CMD_REQUEST | FW_CMD_WRITE);
+ cmdp->retval_len16 = htonl(FW_CMD_LEN16(sizeof(*cmdp) / 16));
+
+}
+
+/*
+ * csio_mb_iq_alloc - Initializes the mailbox to allocate an
+ * Ingress DMA queue in the firmware.
+ *
+ * @hw: The hw structure
+ * @mbp: Mailbox structure to initialize
+ * @priv: Private object
+ * @mb_tmo: Mailbox time-out period (in ms).
+ * @iq_params: Ingress queue params needed for allocation.
+ * @cbfn: The call-back function
+ *
+ *
+ */
+static void
+csio_mb_iq_alloc(struct csio_hw *hw, struct csio_mb *mbp, void *priv,
+ uint32_t mb_tmo, struct csio_iq_params *iq_params,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct fw_iq_cmd *cmdp = (struct fw_iq_cmd *)(mbp->mb);
+
+ CSIO_INIT_MBP(mbp, cmdp, mb_tmo, priv, cbfn, 1);
+
+ cmdp->op_to_vfn = htonl(FW_CMD_OP(FW_IQ_CMD) |
+ FW_CMD_REQUEST | FW_CMD_EXEC |
+ FW_IQ_CMD_PFN(iq_params->pfn) |
+ FW_IQ_CMD_VFN(iq_params->vfn));
+
+ cmdp->alloc_to_len16 = htonl(FW_IQ_CMD_ALLOC |
+ FW_CMD_LEN16(sizeof(*cmdp) / 16));
+
+ cmdp->type_to_iqandstindex = htonl(
+ FW_IQ_CMD_VIID(iq_params->viid) |
+ FW_IQ_CMD_TYPE(iq_params->type) |
+ FW_IQ_CMD_IQASYNCH(iq_params->iqasynch));
+
+ cmdp->fl0size = htons(iq_params->fl0size);
+ cmdp->fl0size = htons(iq_params->fl1size);
+
+} /* csio_mb_iq_alloc */
+
+/*
+ * csio_mb_iq_write - Initializes the mailbox for writing into an
+ * Ingress DMA Queue.
+ *
+ * @hw: The HW structure
+ * @mbp: Mailbox structure to initialize
+ * @priv: Private object
+ * @mb_tmo: Mailbox time-out period (in ms).
+ * @cascaded_req: TRUE - if this request is cascased with iq-alloc request.
+ * @iq_params: Ingress queue params needed for writing.
+ * @cbfn: The call-back function
+ *
+ * NOTE: We OR relevant bits with cmdp->XXX, instead of just equating,
+ * because this IQ write request can be cascaded with a previous
+ * IQ alloc request, and we dont want to over-write the bits set by
+ * that request. This logic will work even in a non-cascaded case, since the
+ * cmdp structure is zeroed out by CSIO_INIT_MBP.
+ */
+static void
+csio_mb_iq_write(struct csio_hw *hw, struct csio_mb *mbp, void *priv,
+ uint32_t mb_tmo, bool cascaded_req,
+ struct csio_iq_params *iq_params,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct fw_iq_cmd *cmdp = (struct fw_iq_cmd *)(mbp->mb);
+
+ uint32_t iq_start_stop = (iq_params->iq_start) ?
+ FW_IQ_CMD_IQSTART(1) :
+ FW_IQ_CMD_IQSTOP(1);
+
+ /*
+ * If this IQ write is cascaded with IQ alloc request, do not
+ * re-initialize with 0's.
+ *
+ */
+ if (!cascaded_req)
+ CSIO_INIT_MBP(mbp, cmdp, mb_tmo, priv, cbfn, 1);
+
+ cmdp->op_to_vfn |= htonl(FW_CMD_OP(FW_IQ_CMD) |
+ FW_CMD_REQUEST | FW_CMD_WRITE |
+ FW_IQ_CMD_PFN(iq_params->pfn) |
+ FW_IQ_CMD_VFN(iq_params->vfn));
+ cmdp->alloc_to_len16 |= htonl(iq_start_stop |
+ FW_CMD_LEN16(sizeof(*cmdp) / 16));
+ cmdp->iqid |= htons(iq_params->iqid);
+ cmdp->fl0id |= htons(iq_params->fl0id);
+ cmdp->fl1id |= htons(iq_params->fl1id);
+ cmdp->type_to_iqandstindex |= htonl(
+ FW_IQ_CMD_IQANDST(iq_params->iqandst) |
+ FW_IQ_CMD_IQANUS(iq_params->iqanus) |
+ FW_IQ_CMD_IQANUD(iq_params->iqanud) |
+ FW_IQ_CMD_IQANDSTINDEX(iq_params->iqandstindex));
+ cmdp->iqdroprss_to_iqesize |= htons(
+ FW_IQ_CMD_IQPCIECH(iq_params->iqpciech) |
+ FW_IQ_CMD_IQDCAEN(iq_params->iqdcaen) |
+ FW_IQ_CMD_IQDCACPU(iq_params->iqdcacpu) |
+ FW_IQ_CMD_IQINTCNTTHRESH(iq_params->iqintcntthresh) |
+ FW_IQ_CMD_IQCPRIO(iq_params->iqcprio) |
+ FW_IQ_CMD_IQESIZE(iq_params->iqesize));
+
+ cmdp->iqsize |= htons(iq_params->iqsize);
+ cmdp->iqaddr |= cpu_to_be64(iq_params->iqaddr);
+
+ if (iq_params->type == 0) {
+ cmdp->iqns_to_fl0congen |= htonl(
+ FW_IQ_CMD_IQFLINTIQHSEN(iq_params->iqflintiqhsen)|
+ FW_IQ_CMD_IQFLINTCONGEN(iq_params->iqflintcongen));
+ }
+
+ if (iq_params->fl0size && iq_params->fl0addr &&
+ (iq_params->fl0id != 0xFFFF)) {
+
+ cmdp->iqns_to_fl0congen |= htonl(
+ FW_IQ_CMD_FL0HOSTFCMODE(iq_params->fl0hostfcmode)|
+ FW_IQ_CMD_FL0CPRIO(iq_params->fl0cprio) |
+ FW_IQ_CMD_FL0PADEN(iq_params->fl0paden) |
+ FW_IQ_CMD_FL0PACKEN(iq_params->fl0packen));
+ cmdp->fl0dcaen_to_fl0cidxfthresh |= htons(
+ FW_IQ_CMD_FL0DCAEN(iq_params->fl0dcaen) |
+ FW_IQ_CMD_FL0DCACPU(iq_params->fl0dcacpu) |
+ FW_IQ_CMD_FL0FBMIN(iq_params->fl0fbmin) |
+ FW_IQ_CMD_FL0FBMAX(iq_params->fl0fbmax) |
+ FW_IQ_CMD_FL0CIDXFTHRESH(iq_params->fl0cidxfthresh));
+ cmdp->fl0size |= htons(iq_params->fl0size);
+ cmdp->fl0addr |= cpu_to_be64(iq_params->fl0addr);
+ }
+} /* csio_mb_iq_write */
+
+/*
+ * csio_mb_iq_alloc_write - Initializes the mailbox for allocating an
+ * Ingress DMA Queue.
+ *
+ * @hw: The HW structure
+ * @mbp: Mailbox structure to initialize
+ * @priv: Private data.
+ * @mb_tmo: Mailbox time-out period (in ms).
+ * @iq_params: Ingress queue params needed for allocation & writing.
+ * @cbfn: The call-back function
+ *
+ *
+ */
+void
+csio_mb_iq_alloc_write(struct csio_hw *hw, struct csio_mb *mbp, void *priv,
+ uint32_t mb_tmo, struct csio_iq_params *iq_params,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ csio_mb_iq_alloc(hw, mbp, priv, mb_tmo, iq_params, cbfn);
+ csio_mb_iq_write(hw, mbp, priv, mb_tmo, true, iq_params, cbfn);
+} /* csio_mb_iq_alloc_write */
+
+/*
+ * csio_mb_iq_alloc_write_rsp - Process the allocation & writing
+ * of ingress DMA queue mailbox's response.
+ *
+ * @hw: The HW structure.
+ * @mbp: Mailbox structure to initialize.
+ * @retval: Firmware return value.
+ * @iq_params: Ingress queue parameters, after allocation and write.
+ *
+ */
+void
+csio_mb_iq_alloc_write_rsp(struct csio_hw *hw, struct csio_mb *mbp,
+ enum fw_retval *ret_val,
+ struct csio_iq_params *iq_params)
+{
+ struct fw_iq_cmd *rsp = (struct fw_iq_cmd *)(mbp->mb);
+
+ *ret_val = FW_CMD_RETVAL_GET(ntohl(rsp->alloc_to_len16));
+ if (*ret_val == FW_SUCCESS) {
+ iq_params->physiqid = ntohs(rsp->physiqid);
+ iq_params->iqid = ntohs(rsp->iqid);
+ iq_params->fl0id = ntohs(rsp->fl0id);
+ iq_params->fl1id = ntohs(rsp->fl1id);
+ } else {
+ iq_params->physiqid = iq_params->iqid =
+ iq_params->fl0id = iq_params->fl1id = 0;
+ }
+} /* csio_mb_iq_alloc_write_rsp */
+
+/*
+ * csio_mb_iq_free - Initializes the mailbox for freeing a
+ * specified Ingress DMA Queue.
+ *
+ * @hw: The HW structure
+ * @mbp: Mailbox structure to initialize
+ * @priv: Private data
+ * @mb_tmo: Mailbox time-out period (in ms).
+ * @iq_params: Parameters of ingress queue, that is to be freed.
+ * @cbfn: The call-back function
+ *
+ *
+ */
+void
+csio_mb_iq_free(struct csio_hw *hw, struct csio_mb *mbp, void *priv,
+ uint32_t mb_tmo, struct csio_iq_params *iq_params,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct fw_iq_cmd *cmdp = (struct fw_iq_cmd *)(mbp->mb);
+
+ CSIO_INIT_MBP(mbp, cmdp, mb_tmo, priv, cbfn, 1);
+
+ cmdp->op_to_vfn = htonl(FW_CMD_OP(FW_IQ_CMD) |
+ FW_CMD_REQUEST | FW_CMD_EXEC |
+ FW_IQ_CMD_PFN(iq_params->pfn) |
+ FW_IQ_CMD_VFN(iq_params->vfn));
+ cmdp->alloc_to_len16 = htonl(FW_IQ_CMD_FREE |
+ FW_CMD_LEN16(sizeof(*cmdp) / 16));
+ cmdp->type_to_iqandstindex = htonl(FW_IQ_CMD_TYPE(iq_params->type));
+
+ cmdp->iqid = htons(iq_params->iqid);
+ cmdp->fl0id = htons(iq_params->fl0id);
+ cmdp->fl1id = htons(iq_params->fl1id);
+
+} /* csio_mb_iq_free */
+
+/*
+ * csio_mb_eq_ofld_alloc - Initializes the mailbox for allocating
+ * an offload-egress queue.
+ *
+ * @hw: The HW structure
+ * @mbp: Mailbox structure to initialize
+ * @priv: Private data
+ * @mb_tmo: Mailbox time-out period (in ms).
+ * @eq_ofld_params: (Offload) Egress queue paramters.
+ * @cbfn: The call-back function
+ *
+ *
+ */
+static void
+csio_mb_eq_ofld_alloc(struct csio_hw *hw, struct csio_mb *mbp, void *priv,
+ uint32_t mb_tmo, struct csio_eq_params *eq_ofld_params,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct fw_eq_ofld_cmd *cmdp = (struct fw_eq_ofld_cmd *)(mbp->mb);
+
+ CSIO_INIT_MBP(mbp, cmdp, mb_tmo, priv, cbfn, 1);
+ cmdp->op_to_vfn = htonl(FW_CMD_OP(FW_EQ_OFLD_CMD) |
+ FW_CMD_REQUEST | FW_CMD_EXEC |
+ FW_EQ_OFLD_CMD_PFN(eq_ofld_params->pfn) |
+ FW_EQ_OFLD_CMD_VFN(eq_ofld_params->vfn));
+ cmdp->alloc_to_len16 = htonl(FW_EQ_OFLD_CMD_ALLOC |
+ FW_CMD_LEN16(sizeof(*cmdp) / 16));
+
+} /* csio_mb_eq_ofld_alloc */
+
+/*
+ * csio_mb_eq_ofld_write - Initializes the mailbox for writing
+ * an alloacted offload-egress queue.
+ *
+ * @hw: The HW structure
+ * @mbp: Mailbox structure to initialize
+ * @priv: Private data
+ * @mb_tmo: Mailbox time-out period (in ms).
+ * @cascaded_req: TRUE - if this request is cascased with Eq-alloc request.
+ * @eq_ofld_params: (Offload) Egress queue paramters.
+ * @cbfn: The call-back function
+ *
+ *
+ * NOTE: We OR relevant bits with cmdp->XXX, instead of just equating,
+ * because this EQ write request can be cascaded with a previous
+ * EQ alloc request, and we dont want to over-write the bits set by
+ * that request. This logic will work even in a non-cascaded case, since the
+ * cmdp structure is zeroed out by CSIO_INIT_MBP.
+ */
+static void
+csio_mb_eq_ofld_write(struct csio_hw *hw, struct csio_mb *mbp, void *priv,
+ uint32_t mb_tmo, bool cascaded_req,
+ struct csio_eq_params *eq_ofld_params,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct fw_eq_ofld_cmd *cmdp = (struct fw_eq_ofld_cmd *)(mbp->mb);
+
+ uint32_t eq_start_stop = (eq_ofld_params->eqstart) ?
+ FW_EQ_OFLD_CMD_EQSTART : FW_EQ_OFLD_CMD_EQSTOP;
+
+ /*
+ * If this EQ write is cascaded with EQ alloc request, do not
+ * re-initialize with 0's.
+ *
+ */
+ if (!cascaded_req)
+ CSIO_INIT_MBP(mbp, cmdp, mb_tmo, priv, cbfn, 1);
+
+ cmdp->op_to_vfn |= htonl(FW_CMD_OP(FW_EQ_OFLD_CMD) |
+ FW_CMD_REQUEST | FW_CMD_WRITE |
+ FW_EQ_OFLD_CMD_PFN(eq_ofld_params->pfn) |
+ FW_EQ_OFLD_CMD_VFN(eq_ofld_params->vfn));
+ cmdp->alloc_to_len16 |= htonl(eq_start_stop |
+ FW_CMD_LEN16(sizeof(*cmdp) / 16));
+
+ cmdp->eqid_pkd |= htonl(FW_EQ_OFLD_CMD_EQID(eq_ofld_params->eqid));
+
+ cmdp->fetchszm_to_iqid |= htonl(
+ FW_EQ_OFLD_CMD_HOSTFCMODE(eq_ofld_params->hostfcmode) |
+ FW_EQ_OFLD_CMD_CPRIO(eq_ofld_params->cprio) |
+ FW_EQ_OFLD_CMD_PCIECHN(eq_ofld_params->pciechn) |
+ FW_EQ_OFLD_CMD_IQID(eq_ofld_params->iqid));
+
+ cmdp->dcaen_to_eqsize |= htonl(
+ FW_EQ_OFLD_CMD_DCAEN(eq_ofld_params->dcaen) |
+ FW_EQ_OFLD_CMD_DCACPU(eq_ofld_params->dcacpu) |
+ FW_EQ_OFLD_CMD_FBMIN(eq_ofld_params->fbmin) |
+ FW_EQ_OFLD_CMD_FBMAX(eq_ofld_params->fbmax) |
+ FW_EQ_OFLD_CMD_CIDXFTHRESHO(eq_ofld_params->cidxfthresho) |
+ FW_EQ_OFLD_CMD_CIDXFTHRESH(eq_ofld_params->cidxfthresh) |
+ FW_EQ_OFLD_CMD_EQSIZE(eq_ofld_params->eqsize));
+
+ cmdp->eqaddr |= cpu_to_be64(eq_ofld_params->eqaddr);
+
+} /* csio_mb_eq_ofld_write */
+
+/*
+ * csio_mb_eq_ofld_alloc_write - Initializes the mailbox for allocation
+ * writing into an Engress DMA Queue.
+ *
+ * @hw: The HW structure
+ * @mbp: Mailbox structure to initialize
+ * @priv: Private data.
+ * @mb_tmo: Mailbox time-out period (in ms).
+ * @eq_ofld_params: (Offload) Egress queue paramters.
+ * @cbfn: The call-back function
+ *
+ *
+ */
+void
+csio_mb_eq_ofld_alloc_write(struct csio_hw *hw, struct csio_mb *mbp,
+ void *priv, uint32_t mb_tmo,
+ struct csio_eq_params *eq_ofld_params,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ csio_mb_eq_ofld_alloc(hw, mbp, priv, mb_tmo, eq_ofld_params, cbfn);
+ csio_mb_eq_ofld_write(hw, mbp, priv, mb_tmo, true,
+ eq_ofld_params, cbfn);
+} /* csio_mb_eq_ofld_alloc_write */
+
+/*
+ * csio_mb_eq_ofld_alloc_write_rsp - Process the allocation
+ * & write egress DMA queue mailbox's response.
+ *
+ * @hw: The HW structure.
+ * @mbp: Mailbox structure to initialize.
+ * @retval: Firmware return value.
+ * @eq_ofld_params: (Offload) Egress queue paramters.
+ *
+ */
+void
+csio_mb_eq_ofld_alloc_write_rsp(struct csio_hw *hw,
+ struct csio_mb *mbp, enum fw_retval *ret_val,
+ struct csio_eq_params *eq_ofld_params)
+{
+ struct fw_eq_ofld_cmd *rsp = (struct fw_eq_ofld_cmd *)(mbp->mb);
+
+ *ret_val = FW_CMD_RETVAL_GET(ntohl(rsp->alloc_to_len16));
+
+ if (*ret_val == FW_SUCCESS) {
+ eq_ofld_params->eqid = FW_EQ_OFLD_CMD_EQID_GET(
+ ntohl(rsp->eqid_pkd));
+ eq_ofld_params->physeqid = FW_EQ_OFLD_CMD_PHYSEQID_GET(
+ ntohl(rsp->physeqid_pkd));
+ } else
+ eq_ofld_params->eqid = 0;
+
+} /* csio_mb_eq_ofld_alloc_write_rsp */
+
+/*
+ * csio_mb_eq_ofld_free - Initializes the mailbox for freeing a
+ * specified Engress DMA Queue.
+ *
+ * @hw: The HW structure
+ * @mbp: Mailbox structure to initialize
+ * @priv: Private data area.
+ * @mb_tmo: Mailbox time-out period (in ms).
+ * @eq_ofld_params: (Offload) Egress queue paramters, that is to be freed.
+ * @cbfn: The call-back function
+ *
+ *
+ */
+void
+csio_mb_eq_ofld_free(struct csio_hw *hw, struct csio_mb *mbp, void *priv,
+ uint32_t mb_tmo, struct csio_eq_params *eq_ofld_params,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct fw_eq_ofld_cmd *cmdp = (struct fw_eq_ofld_cmd *)(mbp->mb);
+
+ CSIO_INIT_MBP(mbp, cmdp, mb_tmo, priv, cbfn, 1);
+
+ cmdp->op_to_vfn = htonl(FW_CMD_OP(FW_EQ_OFLD_CMD) |
+ FW_CMD_REQUEST | FW_CMD_EXEC |
+ FW_EQ_OFLD_CMD_PFN(eq_ofld_params->pfn) |
+ FW_EQ_OFLD_CMD_VFN(eq_ofld_params->vfn));
+ cmdp->alloc_to_len16 = htonl(FW_EQ_OFLD_CMD_FREE |
+ FW_CMD_LEN16(sizeof(*cmdp) / 16));
+ cmdp->eqid_pkd = htonl(FW_EQ_OFLD_CMD_EQID(eq_ofld_params->eqid));
+
+} /* csio_mb_eq_ofld_free */
+
+/*
+ * csio_write_fcoe_link_cond_init_mb - Initialize Mailbox to write FCoE link
+ * condition.
+ *
+ * @ln: The Lnode structure
+ * @mbp: Mailbox structure to initialize
+ * @mb_tmo: Mailbox time-out period (in ms).
+ * @cbfn: The call back function.
+ *
+ *
+ */
+void
+csio_write_fcoe_link_cond_init_mb(struct csio_lnode *ln, struct csio_mb *mbp,
+ uint32_t mb_tmo, uint8_t port_id, uint32_t sub_opcode,
+ uint8_t cos, bool link_status, uint32_t fcfi,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct fw_fcoe_link_cmd *cmdp =
+ (struct fw_fcoe_link_cmd *)(mbp->mb);
+
+ CSIO_INIT_MBP(mbp, cmdp, mb_tmo, ln, cbfn, 1);
+
+ cmdp->op_to_portid = htonl((
+ FW_CMD_OP(FW_FCOE_LINK_CMD) |
+ FW_CMD_REQUEST |
+ FW_CMD_WRITE |
+ FW_FCOE_LINK_CMD_PORTID(port_id)));
+ cmdp->sub_opcode_fcfi = htonl(
+ FW_FCOE_LINK_CMD_SUB_OPCODE(sub_opcode) |
+ FW_FCOE_LINK_CMD_FCFI(fcfi));
+ cmdp->lstatus = link_status;
+ cmdp->retval_len16 = htonl(FW_CMD_LEN16(sizeof(*cmdp) / 16));
+
+} /* csio_write_fcoe_link_cond_init_mb */
+
+/*
+ * csio_fcoe_read_res_info_init_mb - Initializes the mailbox for reading FCoE
+ * resource information(FW_GET_RES_INFO_CMD).
+ *
+ * @hw: The HW structure
+ * @mbp: Mailbox structure to initialize
+ * @mb_tmo: Mailbox time-out period (in ms).
+ * @cbfn: The call-back function
+ *
+ *
+ */
+void
+csio_fcoe_read_res_info_init_mb(struct csio_hw *hw, struct csio_mb *mbp,
+ uint32_t mb_tmo,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct fw_fcoe_res_info_cmd *cmdp =
+ (struct fw_fcoe_res_info_cmd *)(mbp->mb);
+
+ CSIO_INIT_MBP(mbp, cmdp, mb_tmo, hw, cbfn, 1);
+
+ cmdp->op_to_read = htonl((FW_CMD_OP(FW_FCOE_RES_INFO_CMD) |
+ FW_CMD_REQUEST |
+ FW_CMD_READ));
+
+ cmdp->retval_len16 = htonl(FW_CMD_LEN16(sizeof(*cmdp) / 16));
+
+} /* csio_fcoe_read_res_info_init_mb */
+
+/*
+ * csio_fcoe_vnp_alloc_init_mb - Initializes the mailbox for allocating VNP
+ * in the firmware (FW_FCOE_VNP_CMD).
+ *
+ * @ln: The Lnode structure.
+ * @mbp: Mailbox structure to initialize.
+ * @mb_tmo: Mailbox time-out period (in ms).
+ * @fcfi: FCF Index.
+ * @vnpi: vnpi
+ * @iqid: iqid
+ * @vnport_wwnn: vnport WWNN
+ * @vnport_wwpn: vnport WWPN
+ * @cbfn: The call-back function.
+ *
+ *
+ */
+void
+csio_fcoe_vnp_alloc_init_mb(struct csio_lnode *ln, struct csio_mb *mbp,
+ uint32_t mb_tmo, uint32_t fcfi, uint32_t vnpi, uint16_t iqid,
+ uint8_t vnport_wwnn[8], uint8_t vnport_wwpn[8],
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct fw_fcoe_vnp_cmd *cmdp =
+ (struct fw_fcoe_vnp_cmd *)(mbp->mb);
+
+ CSIO_INIT_MBP(mbp, cmdp, mb_tmo, ln, cbfn, 1);
+
+ cmdp->op_to_fcfi = htonl((FW_CMD_OP(FW_FCOE_VNP_CMD) |
+ FW_CMD_REQUEST |
+ FW_CMD_EXEC |
+ FW_FCOE_VNP_CMD_FCFI(fcfi)));
+
+ cmdp->alloc_to_len16 = htonl(FW_FCOE_VNP_CMD_ALLOC |
+ FW_CMD_LEN16(sizeof(*cmdp) / 16));
+
+ cmdp->gen_wwn_to_vnpi = htonl(FW_FCOE_VNP_CMD_VNPI(vnpi));
+
+ cmdp->iqid = htons(iqid);
+
+ if (!wwn_to_u64(vnport_wwnn) && !wwn_to_u64(vnport_wwpn))
+ cmdp->gen_wwn_to_vnpi |= htonl(FW_FCOE_VNP_CMD_GEN_WWN);
+
+ if (vnport_wwnn)
+ memcpy(cmdp->vnport_wwnn, vnport_wwnn, 8);
+ if (vnport_wwpn)
+ memcpy(cmdp->vnport_wwpn, vnport_wwpn, 8);
+
+} /* csio_fcoe_vnp_alloc_init_mb */
+
+/*
+ * csio_fcoe_vnp_read_init_mb - Prepares VNP read cmd.
+ * @ln: The Lnode structure.
+ * @mbp: Mailbox structure to initialize.
+ * @mb_tmo: Mailbox time-out period (in ms).
+ * @fcfi: FCF Index.
+ * @vnpi: vnpi
+ * @cbfn: The call-back handler.
+ */
+void
+csio_fcoe_vnp_read_init_mb(struct csio_lnode *ln, struct csio_mb *mbp,
+ uint32_t mb_tmo, uint32_t fcfi, uint32_t vnpi,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct fw_fcoe_vnp_cmd *cmdp =
+ (struct fw_fcoe_vnp_cmd *)(mbp->mb);
+
+ CSIO_INIT_MBP(mbp, cmdp, mb_tmo, ln, cbfn, 1);
+ cmdp->op_to_fcfi = htonl(FW_CMD_OP(FW_FCOE_VNP_CMD) |
+ FW_CMD_REQUEST |
+ FW_CMD_READ |
+ FW_FCOE_VNP_CMD_FCFI(fcfi));
+ cmdp->alloc_to_len16 = htonl(FW_CMD_LEN16(sizeof(*cmdp) / 16));
+ cmdp->gen_wwn_to_vnpi = htonl(FW_FCOE_VNP_CMD_VNPI(vnpi));
+}
+
+/*
+ * csio_fcoe_vnp_free_init_mb - Initializes the mailbox for freeing an
+ * alloacted VNP in the firmware (FW_FCOE_VNP_CMD).
+ *
+ * @ln: The Lnode structure.
+ * @mbp: Mailbox structure to initialize.
+ * @mb_tmo: Mailbox time-out period (in ms).
+ * @fcfi: FCF flow id
+ * @vnpi: VNP flow id
+ * @cbfn: The call-back function.
+ * Return: None
+ */
+void
+csio_fcoe_vnp_free_init_mb(struct csio_lnode *ln, struct csio_mb *mbp,
+ uint32_t mb_tmo, uint32_t fcfi, uint32_t vnpi,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct fw_fcoe_vnp_cmd *cmdp =
+ (struct fw_fcoe_vnp_cmd *)(mbp->mb);
+
+ CSIO_INIT_MBP(mbp, cmdp, mb_tmo, ln, cbfn, 1);
+
+ cmdp->op_to_fcfi = htonl(FW_CMD_OP(FW_FCOE_VNP_CMD) |
+ FW_CMD_REQUEST |
+ FW_CMD_EXEC |
+ FW_FCOE_VNP_CMD_FCFI(fcfi));
+ cmdp->alloc_to_len16 = htonl(FW_FCOE_VNP_CMD_FREE |
+ FW_CMD_LEN16(sizeof(*cmdp) / 16));
+ cmdp->gen_wwn_to_vnpi = htonl(FW_FCOE_VNP_CMD_VNPI(vnpi));
+}
+
+/*
+ * csio_fcoe_read_fcf_init_mb - Initializes the mailbox to read the
+ * FCF records.
+ *
+ * @ln: The Lnode structure
+ * @mbp: Mailbox structure to initialize
+ * @mb_tmo: Mailbox time-out period (in ms).
+ * @fcf_params: FC-Forwarder parameters.
+ * @cbfn: The call-back function
+ *
+ *
+ */
+void
+csio_fcoe_read_fcf_init_mb(struct csio_lnode *ln, struct csio_mb *mbp,
+ uint32_t mb_tmo, uint32_t portid, uint32_t fcfi,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct fw_fcoe_fcf_cmd *cmdp =
+ (struct fw_fcoe_fcf_cmd *)(mbp->mb);
+
+ CSIO_INIT_MBP(mbp, cmdp, mb_tmo, ln, cbfn, 1);
+
+ cmdp->op_to_fcfi = htonl(FW_CMD_OP(FW_FCOE_FCF_CMD) |
+ FW_CMD_REQUEST |
+ FW_CMD_READ |
+ FW_FCOE_FCF_CMD_FCFI(fcfi));
+ cmdp->retval_len16 = htonl(FW_CMD_LEN16(sizeof(*cmdp) / 16));
+
+} /* csio_fcoe_read_fcf_init_mb */
+
+void
+csio_fcoe_read_portparams_init_mb(struct csio_hw *hw, struct csio_mb *mbp,
+ uint32_t mb_tmo,
+ struct fw_fcoe_port_cmd_params *portparams,
+ void (*cbfn)(struct csio_hw *,
+ struct csio_mb *))
+{
+ struct fw_fcoe_stats_cmd *cmdp = (struct fw_fcoe_stats_cmd *)(mbp->mb);
+
+ CSIO_INIT_MBP(mbp, cmdp, mb_tmo, hw, cbfn, 1);
+ mbp->mb_size = 64;
+
+ cmdp->op_to_flowid = htonl(FW_CMD_OP(FW_FCOE_STATS_CMD) |
+ FW_CMD_REQUEST | FW_CMD_READ);
+ cmdp->free_to_len16 = htonl(FW_CMD_LEN16(CSIO_MAX_MB_SIZE/16));
+
+ cmdp->u.ctl.nstats_port = FW_FCOE_STATS_CMD_NSTATS(portparams->nstats) |
+ FW_FCOE_STATS_CMD_PORT(portparams->portid);
+
+ cmdp->u.ctl.port_valid_ix = FW_FCOE_STATS_CMD_IX(portparams->idx) |
+ FW_FCOE_STATS_CMD_PORT_VALID;
+
+} /* csio_fcoe_read_portparams_init_mb */
+
+void
+csio_mb_process_portparams_rsp(
+ struct csio_hw *hw,
+ struct csio_mb *mbp,
+ enum fw_retval *retval,
+ struct fw_fcoe_port_cmd_params *portparams,
+ struct fw_fcoe_port_stats *portstats
+ )
+{
+ struct fw_fcoe_stats_cmd *rsp = (struct fw_fcoe_stats_cmd *)(mbp->mb);
+ struct fw_fcoe_port_stats stats;
+ uint8_t *src;
+ uint8_t *dst;
+
+ *retval = FW_CMD_RETVAL_GET(ntohl(rsp->free_to_len16));
+
+ memset(&stats, 0, sizeof(struct fw_fcoe_port_stats));
+
+ if (*retval == FW_SUCCESS) {
+ dst = (uint8_t *)(&stats) + ((portparams->idx - 1) * 8);
+ src = (uint8_t *)rsp + (CSIO_STATS_OFFSET * 8);
+ memcpy(dst, src, (portparams->nstats * 8));
+ if (portparams->idx == 1) {
+ /* Get the first 6 flits from the Mailbox */
+ portstats->tx_bcast_bytes =
+ be64_to_cpu(stats.tx_bcast_bytes);
+ portstats->tx_bcast_frames =
+ be64_to_cpu(stats.tx_bcast_frames);
+ portstats->tx_mcast_bytes =
+ be64_to_cpu(stats.tx_mcast_bytes);
+ portstats->tx_mcast_frames =
+ be64_to_cpu(stats.tx_mcast_frames);
+ portstats->tx_ucast_bytes =
+ be64_to_cpu(stats.tx_ucast_bytes);
+ portstats->tx_ucast_frames =
+ be64_to_cpu(stats.tx_ucast_frames);
+ }
+ if (portparams->idx == 7) {
+ /* Get the second 6 flits from the Mailbox */
+ portstats->tx_drop_frames =
+ be64_to_cpu(stats.tx_drop_frames);
+ portstats->tx_offload_bytes =
+ be64_to_cpu(stats.tx_offload_bytes);
+ portstats->tx_offload_frames =
+ be64_to_cpu(stats.tx_offload_frames);
+#if 0
+ portstats->rx_pf_bytes =
+ be64_to_cpu(stats.rx_pf_bytes);
+ portstats->rx_pf_frames =
+ be64_to_cpu(stats.rx_pf_frames);
+#endif
+ portstats->rx_bcast_bytes =
+ be64_to_cpu(stats.rx_bcast_bytes);
+ portstats->rx_bcast_frames =
+ be64_to_cpu(stats.rx_bcast_frames);
+ portstats->rx_mcast_bytes =
+ be64_to_cpu(stats.rx_mcast_bytes);
+ }
+ if (portparams->idx == 13) {
+ /* Get the last 4 flits from the Mailbox */
+ portstats->rx_mcast_frames =
+ be64_to_cpu(stats.rx_mcast_frames);
+ portstats->rx_ucast_bytes =
+ be64_to_cpu(stats.rx_ucast_bytes);
+ portstats->rx_ucast_frames =
+ be64_to_cpu(stats.rx_ucast_frames);
+ portstats->rx_err_frames =
+ be64_to_cpu(stats.rx_err_frames);
+ }
+ }
+}
+
+/* Entry points/APIs for MB module */
+/*
+ * csio_mb_intr_enable - Enable Interrupts from mailboxes.
+ * @hw: The HW structure
+ *
+ * Enables CIM interrupt bit in appropriate INT_ENABLE registers.
+ */
+void
+csio_mb_intr_enable(struct csio_hw *hw)
+{
+ csio_wr_reg32(hw, MBMSGRDYINTEN(1), MYPF_REG(CIM_PF_HOST_INT_ENABLE));
+ csio_rd_reg32(hw, MYPF_REG(CIM_PF_HOST_INT_ENABLE));
+}
+
+/*
+ * csio_mb_intr_disable - Disable Interrupts from mailboxes.
+ * @hw: The HW structure
+ *
+ * Disable bit in HostInterruptEnable CIM register.
+ */
+void
+csio_mb_intr_disable(struct csio_hw *hw)
+{
+ csio_wr_reg32(hw, MBMSGRDYINTEN(0), MYPF_REG(CIM_PF_HOST_INT_ENABLE));
+ csio_rd_reg32(hw, MYPF_REG(CIM_PF_HOST_INT_ENABLE));
+}
+
+static void
+csio_mb_dump_fw_dbg(struct csio_hw *hw, __be64 *cmd)
+{
+ struct fw_debug_cmd *dbg = (struct fw_debug_cmd *)cmd;
+
+ if ((FW_DEBUG_CMD_TYPE_GET(ntohl(dbg->op_type))) == 1) {
+ csio_info(hw, "FW print message:\n");
+ csio_info(hw, "\tdebug->dprtstridx = %d\n",
+ ntohs(dbg->u.prt.dprtstridx));
+ csio_info(hw, "\tdebug->dprtstrparam0 = 0x%x\n",
+ ntohl(dbg->u.prt.dprtstrparam0));
+ csio_info(hw, "\tdebug->dprtstrparam1 = 0x%x\n",
+ ntohl(dbg->u.prt.dprtstrparam1));
+ csio_info(hw, "\tdebug->dprtstrparam2 = 0x%x\n",
+ ntohl(dbg->u.prt.dprtstrparam2));
+ csio_info(hw, "\tdebug->dprtstrparam3 = 0x%x\n",
+ ntohl(dbg->u.prt.dprtstrparam3));
+ } else {
+ /* This is a FW assertion */
+ csio_fatal(hw, "FW assertion at %.16s:%u, val0 %#x, val1 %#x\n",
+ dbg->u.assert.filename_0_7,
+ ntohl(dbg->u.assert.line),
+ ntohl(dbg->u.assert.x),
+ ntohl(dbg->u.assert.y));
+ }
+}
+
+static void
+csio_mb_debug_cmd_handler(struct csio_hw *hw)
+{
+ int i;
+ __be64 cmd[CSIO_MB_MAX_REGS];
+ uint32_t ctl_reg = PF_REG(hw->pfn, CIM_PF_MAILBOX_CTRL);
+ uint32_t data_reg = PF_REG(hw->pfn, CIM_PF_MAILBOX_DATA);
+ int size = sizeof(struct fw_debug_cmd);
+
+ /* Copy mailbox data */
+ for (i = 0; i < size; i += 8)
+ cmd[i / 8] = cpu_to_be64(csio_rd_reg64(hw, data_reg + i));
+
+ csio_mb_dump_fw_dbg(hw, cmd);
+
+ /* Notify FW of mailbox by setting owner as UP */
+ csio_wr_reg32(hw, MBMSGVALID | MBINTREQ | MBOWNER(CSIO_MBOWNER_FW),
+ ctl_reg);
+
+ csio_rd_reg32(hw, ctl_reg);
+ wmb();
+}
+
+/*
+ * csio_mb_issue - generic routine for issuing Mailbox commands.
+ * @hw: The HW structure
+ * @mbp: Mailbox command to issue
+ *
+ * Caller should hold hw lock across this call.
+ */
+int
+csio_mb_issue(struct csio_hw *hw, struct csio_mb *mbp)
+{
+ uint32_t owner, ctl;
+ int i;
+ uint32_t ii;
+ __be64 *cmd = mbp->mb;
+ __be64 hdr;
+ struct csio_mbm *mbm = &hw->mbm;
+ uint32_t ctl_reg = PF_REG(hw->pfn, CIM_PF_MAILBOX_CTRL);
+ uint32_t data_reg = PF_REG(hw->pfn, CIM_PF_MAILBOX_DATA);
+ int size = mbp->mb_size;
+ int rv = -EINVAL;
+ struct fw_cmd_hdr *fw_hdr;
+
+ /* Determine mode */
+ if (mbp->mb_cbfn == NULL) {
+ /* Need to issue/get results in the same context */
+ if (mbp->tmo < CSIO_MB_POLL_FREQ) {
+ csio_err(hw, "Invalid tmo: 0x%x\n", mbp->tmo);
+ goto error_out;
+ }
+ } else if (!csio_is_host_intr_enabled(hw) ||
+ !csio_is_hw_intr_enabled(hw)) {
+ csio_err(hw, "Cannot issue mailbox in interrupt mode 0x%x\n",
+ *((uint8_t *)mbp->mb));
+ goto error_out;
+ }
+
+ if (mbm->mcurrent != NULL) {
+ /* Queue mbox cmd, if another mbox cmd is active */
+ if (mbp->mb_cbfn == NULL) {
+ rv = -EBUSY;
+ csio_dbg(hw, "Couldnt own Mailbox %x op:0x%x\n",
+ hw->pfn, *((uint8_t *)mbp->mb));
+
+ goto error_out;
+ } else {
+ list_add_tail(&mbp->list, &mbm->req_q);
+ CSIO_INC_STATS(mbm, n_activeq);
+
+ return 0;
+ }
+ }
+
+ /* Now get ownership of mailbox */
+ owner = MBOWNER_GET(csio_rd_reg32(hw, ctl_reg));
+
+ if (!csio_mb_is_host_owner(owner)) {
+
+ for (i = 0; (owner == CSIO_MBOWNER_NONE) && (i < 3); i++)
+ owner = MBOWNER_GET(csio_rd_reg32(hw, ctl_reg));
+ /*
+ * Mailbox unavailable. In immediate mode, fail the command.
+ * In other modes, enqueue the request.
+ */
+ if (!csio_mb_is_host_owner(owner)) {
+ if (mbp->mb_cbfn == NULL) {
+ rv = owner ? -EBUSY : -ETIMEDOUT;
+
+ csio_dbg(hw,
+ "Couldnt own Mailbox %x op:0x%x "
+ "owner:%x\n",
+ hw->pfn, *((uint8_t *)mbp->mb), owner);
+ goto error_out;
+ } else {
+ if (mbm->mcurrent == NULL) {
+ csio_err(hw,
+ "Couldnt own Mailbox %x "
+ "op:0x%x owner:%x\n",
+ hw->pfn, *((uint8_t *)mbp->mb),
+ owner);
+ csio_err(hw,
+ "No outstanding driver"
+ " mailbox as well\n");
+ goto error_out;
+ }
+ }
+ }
+ }
+
+ /* Mailbox is available, copy mailbox data into it */
+ for (i = 0; i < size; i += 8) {
+ csio_wr_reg64(hw, be64_to_cpu(*cmd), data_reg + i);
+ cmd++;
+ }
+
+ CSIO_DUMP_MB(hw, hw->pfn, data_reg);
+
+ /* Start completion timers in non-immediate modes and notify FW */
+ if (mbp->mb_cbfn != NULL) {
+ mbm->mcurrent = mbp;
+ mod_timer(&mbm->timer, jiffies + msecs_to_jiffies(mbp->tmo));
+ csio_wr_reg32(hw, MBMSGVALID | MBINTREQ |
+ MBOWNER(CSIO_MBOWNER_FW), ctl_reg);
+ } else
+ csio_wr_reg32(hw, MBMSGVALID | MBOWNER(CSIO_MBOWNER_FW),
+ ctl_reg);
+
+ /* Flush posted writes */
+ csio_rd_reg32(hw, ctl_reg);
+ wmb();
+
+ CSIO_INC_STATS(mbm, n_req);
+
+ if (mbp->mb_cbfn)
+ return 0;
+
+ /* Poll for completion in immediate mode */
+ cmd = mbp->mb;
+
+ for (ii = 0; ii < mbp->tmo; ii += CSIO_MB_POLL_FREQ) {
+ mdelay(CSIO_MB_POLL_FREQ);
+
+ /* Check for response */
+ ctl = csio_rd_reg32(hw, ctl_reg);
+ if (csio_mb_is_host_owner(MBOWNER_GET(ctl))) {
+
+ if (!(ctl & MBMSGVALID)) {
+ csio_wr_reg32(hw, 0, ctl_reg);
+ continue;
+ }
+
+ CSIO_DUMP_MB(hw, hw->pfn, data_reg);
+
+ hdr = cpu_to_be64(csio_rd_reg64(hw, data_reg));
+ fw_hdr = (struct fw_cmd_hdr *)&hdr;
+
+ switch (FW_CMD_OP_GET(ntohl(fw_hdr->hi))) {
+ case FW_DEBUG_CMD:
+ csio_mb_debug_cmd_handler(hw);
+ continue;
+ }
+
+ /* Copy response */
+ for (i = 0; i < size; i += 8)
+ *cmd++ = cpu_to_be64(csio_rd_reg64
+ (hw, data_reg + i));
+ csio_wr_reg32(hw, 0, ctl_reg);
+
+ if (FW_CMD_RETVAL_GET(*(mbp->mb)))
+ CSIO_INC_STATS(mbm, n_err);
+
+ CSIO_INC_STATS(mbm, n_rsp);
+ return 0;
+ }
+ }
+
+ CSIO_INC_STATS(mbm, n_tmo);
+
+ csio_err(hw, "Mailbox %x op:0x%x timed out!\n",
+ hw->pfn, *((uint8_t *)cmd));
+
+ return -ETIMEDOUT;
+
+error_out:
+ CSIO_INC_STATS(mbm, n_err);
+ return rv;
+}
+
+/*
+ * csio_mb_completions - Completion handler for Mailbox commands
+ * @hw: The HW structure
+ * @cbfn_q: Completion queue.
+ *
+ */
+void
+csio_mb_completions(struct csio_hw *hw, struct list_head *cbfn_q)
+{
+ struct csio_mb *mbp;
+ struct csio_mbm *mbm = &hw->mbm;
+ enum fw_retval rv;
+
+ while (!list_empty(cbfn_q)) {
+ mbp = list_first_entry(cbfn_q, struct csio_mb, list);
+ list_del_init(&mbp->list);
+
+ rv = csio_mb_fw_retval(mbp);
+ if ((rv != FW_SUCCESS) && (rv != FW_HOSTERROR))
+ CSIO_INC_STATS(mbm, n_err);
+ else if (rv != FW_HOSTERROR)
+ CSIO_INC_STATS(mbm, n_rsp);
+
+ if (mbp->mb_cbfn)
+ mbp->mb_cbfn(hw, mbp);
+ }
+}
+
+static void
+csio_mb_portmod_changed(struct csio_hw *hw, uint8_t port_id)
+{
+ static char *mod_str[] = {
+ NULL, "LR", "SR", "ER", "TWINAX", "active TWINAX", "LRM"
+ };
+
+ struct csio_pport *port = &hw->pport[port_id];
+
+ if (port->mod_type == FW_PORT_MOD_TYPE_NONE)
+ csio_info(hw, "Port:%d - port module unplugged\n", port_id);
+ else if (port->mod_type < ARRAY_SIZE(mod_str))
+ csio_info(hw, "Port:%d - %s port module inserted\n", port_id,
+ mod_str[port->mod_type]);
+ else if (port->mod_type == FW_PORT_MOD_TYPE_NOTSUPPORTED)
+ csio_info(hw,
+ "Port:%d - unsupported optical port module "
+ "inserted\n", port_id);
+ else if (port->mod_type == FW_PORT_MOD_TYPE_UNKNOWN)
+ csio_info(hw,
+ "Port:%d - unknown port module inserted, forcing "
+ "TWINAX\n", port_id);
+ else if (port->mod_type == FW_PORT_MOD_TYPE_ERROR)
+ csio_info(hw, "Port:%d - transceiver module error\n", port_id);
+ else
+ csio_info(hw, "Port:%d - unknown module type %d inserted\n",
+ port_id, port->mod_type);
+}
+
+int
+csio_mb_fwevt_handler(struct csio_hw *hw, __be64 *cmd)
+{
+ uint8_t opcode = *(uint8_t *)cmd;
+ struct fw_port_cmd *pcmd;
+ uint8_t port_id;
+ uint32_t link_status;
+ uint16_t action;
+ uint8_t mod_type;
+
+ if (opcode == FW_PORT_CMD) {
+ pcmd = (struct fw_port_cmd *)cmd;
+ port_id = FW_PORT_CMD_PORTID_GET(
+ ntohl(pcmd->op_to_portid));
+ action = FW_PORT_CMD_ACTION_GET(
+ ntohl(pcmd->action_to_len16));
+ if (action != FW_PORT_ACTION_GET_PORT_INFO) {
+ csio_err(hw, "Unhandled FW_PORT_CMD action: %u\n",
+ action);
+ return -EINVAL;
+ }
+
+ link_status = ntohl(pcmd->u.info.lstatus_to_modtype);
+ mod_type = FW_PORT_CMD_MODTYPE_GET(link_status);
+
+ hw->pport[port_id].link_status =
+ FW_PORT_CMD_LSTATUS_GET(link_status);
+ hw->pport[port_id].link_speed =
+ FW_PORT_CMD_LSPEED_GET(link_status);
+
+ csio_info(hw, "Port:%x - LINK %s\n", port_id,
+ FW_PORT_CMD_LSTATUS_GET(link_status) ? "UP" : "DOWN");
+
+ if (mod_type != hw->pport[port_id].mod_type) {
+ hw->pport[port_id].mod_type = mod_type;
+ csio_mb_portmod_changed(hw, port_id);
+ }
+ } else if (opcode == FW_DEBUG_CMD) {
+ csio_mb_dump_fw_dbg(hw, cmd);
+ } else {
+ csio_dbg(hw, "Gen MB can't handle op:0x%x on evtq.\n", opcode);
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+/*
+ * csio_mb_isr_handler - Handle mailboxes related interrupts.
+ * @hw: The HW structure
+ *
+ * Called from the ISR to handle Mailbox related interrupts.
+ * HW Lock should be held across this call.
+ */
+int
+csio_mb_isr_handler(struct csio_hw *hw)
+{
+ struct csio_mbm *mbm = &hw->mbm;
+ struct csio_mb *mbp = mbm->mcurrent;
+ __be64 *cmd;
+ uint32_t ctl, cim_cause, pl_cause;
+ int i;
+ uint32_t ctl_reg = PF_REG(hw->pfn, CIM_PF_MAILBOX_CTRL);
+ uint32_t data_reg = PF_REG(hw->pfn, CIM_PF_MAILBOX_DATA);
+ int size;
+ __be64 hdr;
+ struct fw_cmd_hdr *fw_hdr;
+
+ pl_cause = csio_rd_reg32(hw, MYPF_REG(PL_PF_INT_CAUSE));
+ cim_cause = csio_rd_reg32(hw, MYPF_REG(CIM_PF_HOST_INT_CAUSE));
+
+ if (!(pl_cause & PFCIM) || !(cim_cause & MBMSGRDYINT)) {
+ CSIO_INC_STATS(hw, n_mbint_unexp);
+ return -EINVAL;
+ }
+
+ /*
+ * The cause registers below HAVE to be cleared in the SAME
+ * order as below: The low level cause register followed by
+ * the upper level cause register. In other words, CIM-cause
+ * first followed by PL-Cause next.
+ */
+ csio_wr_reg32(hw, MBMSGRDYINT, MYPF_REG(CIM_PF_HOST_INT_CAUSE));
+ csio_wr_reg32(hw, PFCIM, MYPF_REG(PL_PF_INT_CAUSE));
+
+ ctl = csio_rd_reg32(hw, ctl_reg);
+
+ if (csio_mb_is_host_owner(MBOWNER_GET(ctl))) {
+
+ CSIO_DUMP_MB(hw, hw->pfn, data_reg);
+
+ if (!(ctl & MBMSGVALID)) {
+ csio_warn(hw,
+ "Stray mailbox interrupt recvd,"
+ " mailbox data not valid\n");
+ csio_wr_reg32(hw, 0, ctl_reg);
+ /* Flush */
+ csio_rd_reg32(hw, ctl_reg);
+ return -EINVAL;
+ }
+
+ hdr = cpu_to_be64(csio_rd_reg64(hw, data_reg));
+ fw_hdr = (struct fw_cmd_hdr *)&hdr;
+
+ switch (FW_CMD_OP_GET(ntohl(fw_hdr->hi))) {
+ case FW_DEBUG_CMD:
+ csio_mb_debug_cmd_handler(hw);
+ return -EINVAL;
+#if 0
+ case FW_ERROR_CMD:
+ case FW_INITIALIZE_CMD: /* When we are not master */
+#endif
+ }
+
+ CSIO_ASSERT(mbp != NULL);
+
+ cmd = mbp->mb;
+ size = mbp->mb_size;
+ /* Get response */
+ for (i = 0; i < size; i += 8)
+ *cmd++ = cpu_to_be64(csio_rd_reg64
+ (hw, data_reg + i));
+
+ csio_wr_reg32(hw, 0, ctl_reg);
+ /* Flush */
+ csio_rd_reg32(hw, ctl_reg);
+
+ mbm->mcurrent = NULL;
+
+ /* Add completion to tail of cbfn queue */
+ list_add_tail(&mbp->list, &mbm->cbfn_q);
+ CSIO_INC_STATS(mbm, n_cbfnq);
+
+ /*
+ * Enqueue event to EventQ. Events processing happens
+ * in Event worker thread context
+ */
+ if (csio_enqueue_evt(hw, CSIO_EVT_MBX, mbp, sizeof(mbp)))
+ CSIO_INC_STATS(hw, n_evt_drop);
+
+ return 0;
+
+ } else {
+ /*
+ * We can get here if mailbox MSIX vector is shared,
+ * or in INTx case. Or a stray interrupt.
+ */
+ csio_dbg(hw, "Host not owner, no mailbox interrupt\n");
+ CSIO_INC_STATS(hw, n_int_stray);
+ return -EINVAL;
+ }
+}
+
+/*
+ * csio_mb_tmo_handler - Timeout handler
+ * @hw: The HW structure
+ *
+ */
+struct csio_mb *
+csio_mb_tmo_handler(struct csio_hw *hw)
+{
+ struct csio_mbm *mbm = &hw->mbm;
+ struct csio_mb *mbp = mbm->mcurrent;
+ struct fw_cmd_hdr *fw_hdr;
+
+ /*
+ * Could be a race b/w the completion handler and the timer
+ * and the completion handler won that race.
+ */
+ if (mbp == NULL) {
+ CSIO_DB_ASSERT(0);
+ return NULL;
+ }
+
+ fw_hdr = (struct fw_cmd_hdr *)(mbp->mb);
+
+ csio_dbg(hw, "Mailbox num:%x op:0x%x timed out\n", hw->pfn,
+ FW_CMD_OP_GET(ntohl(fw_hdr->hi)));
+
+ mbm->mcurrent = NULL;
+ CSIO_INC_STATS(mbm, n_tmo);
+ fw_hdr->lo = htonl(FW_CMD_RETVAL(FW_ETIMEDOUT));
+
+ return mbp;
+}
+
+/*
+ * csio_mb_cancel_all - Cancel all waiting commands.
+ * @hw: The HW structure
+ * @cbfn_q: The callback queue.
+ *
+ * Caller should hold hw lock across this call.
+ */
+void
+csio_mb_cancel_all(struct csio_hw *hw, struct list_head *cbfn_q)
+{
+ struct csio_mb *mbp;
+ struct csio_mbm *mbm = &hw->mbm;
+ struct fw_cmd_hdr *hdr;
+ struct list_head *tmp;
+
+ if (mbm->mcurrent) {
+ mbp = mbm->mcurrent;
+
+ /* Stop mailbox completion timer */
+ del_timer_sync(&mbm->timer);
+
+ /* Add completion to tail of cbfn queue */
+ list_add_tail(&mbp->list, cbfn_q);
+ mbm->mcurrent = NULL;
+ }
+
+ if (!list_empty(&mbm->req_q)) {
+ list_splice_tail_init(&mbm->req_q, cbfn_q);
+ mbm->stats.n_activeq = 0;
+ }
+
+ if (!list_empty(&mbm->cbfn_q)) {
+ list_splice_tail_init(&mbm->cbfn_q, cbfn_q);
+ mbm->stats.n_cbfnq = 0;
+ }
+
+ if (list_empty(cbfn_q))
+ return;
+
+ list_for_each(tmp, cbfn_q) {
+ mbp = (struct csio_mb *)tmp;
+ hdr = (struct fw_cmd_hdr *)(mbp->mb);
+
+ csio_dbg(hw, "Cancelling pending mailbox num %x op:%x\n",
+ hw->pfn, FW_CMD_OP_GET(ntohl(hdr->hi)));
+
+ CSIO_INC_STATS(mbm, n_cancel);
+ hdr->lo = htonl(FW_CMD_RETVAL(FW_HOSTERROR));
+ }
+}
+
+/*
+ * csio_mbm_init - Initialize Mailbox module
+ * @mbm: Mailbox module
+ * @hw: The HW structure
+ * @timer: Timing function for interrupting mailboxes
+ *
+ * Initialize timer and the request/response queues.
+ */
+int
+csio_mbm_init(struct csio_mbm *mbm, struct csio_hw *hw,
+ void (*timer_fn)(uintptr_t))
+{
+ struct timer_list *timer = &mbm->timer;
+
+ init_timer(timer);
+ timer->function = timer_fn;
+ timer->data = (unsigned long)hw;
+
+ INIT_LIST_HEAD(&mbm->req_q);
+ INIT_LIST_HEAD(&mbm->cbfn_q);
+ csio_set_mb_intr_idx(mbm, -1);
+
+ return 0;
+}
+
+/*
+ * csio_mbm_exit - Uninitialize mailbox module
+ * @mbm: Mailbox module
+ *
+ * Stop timer.
+ */
+void
+csio_mbm_exit(struct csio_mbm *mbm)
+{
+ del_timer_sync(&mbm->timer);
+
+ CSIO_DB_ASSERT(mbm->mcurrent == NULL);
+ CSIO_DB_ASSERT(list_empty(&mbm->req_q));
+ CSIO_DB_ASSERT(list_empty(&mbm->cbfn_q));
+}
--
1.7.1
^ permalink raw reply related
* [V3 PATCH 2/9] csiostor: Chelsio FCoE offload driver submission (sources part 2).
From: Naresh Kumar Inna @ 2012-09-11 14:39 UTC (permalink / raw)
To: JBottomley, linux-scsi, dm, leedom; +Cc: netdev, naresh, chethan
In-Reply-To: <1347374347-3852-1-git-send-email-naresh@chelsio.com>
This patch contains code for driver initialization, driver resource
allocation and the Work Request module functionality. Driver initialization
includes module entry/exit points, registration with PCI, FC transport and
SCSI mid layer subsystems. The Work Request module provides services for
allocation of DMA queues, posting Work Requests on them and processing
completions.
Signed-off-by: Naresh Kumar Inna <naresh@chelsio.com>
---
V2: Removed module parameters.
V3: Corrected comment.
drivers/scsi/csiostor/csio_init.c | 1272 +++++++++++++++++++++++++++++
drivers/scsi/csiostor/csio_wr.c | 1632 +++++++++++++++++++++++++++++++++++++
2 files changed, 2904 insertions(+), 0 deletions(-)
create mode 100644 drivers/scsi/csiostor/csio_init.c
create mode 100644 drivers/scsi/csiostor/csio_wr.c
diff --git a/drivers/scsi/csiostor/csio_init.c b/drivers/scsi/csiostor/csio_init.c
new file mode 100644
index 0000000..1864250
--- /dev/null
+++ b/drivers/scsi/csiostor/csio_init.c
@@ -0,0 +1,1272 @@
+/*
+ * This file is part of the Chelsio FCoE driver for Linux.
+ *
+ * Copyright (c) 2008-2012 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses. You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ * Redistribution and use in source and binary forms, with or
+ * without modification, are permitted provided that the following
+ * conditions are met:
+ *
+ * - Redistributions of source code must retain the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer in the documentation and/or other materials
+ * provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/pci.h>
+#include <linux/aer.h>
+#include <linux/mm.h>
+#include <linux/notifier.h>
+#include <linux/kdebug.h>
+#include <linux/version.h>
+#include <linux/seq_file.h>
+#include <linux/debugfs.h>
+#include <linux/string.h>
+#include <linux/export.h>
+
+#include "csio_init.h"
+#include "csio_defs.h"
+
+#define CSIO_MIN_MEMPOOL_SZ 64
+
+static struct dentry *csio_debugfs_root;
+
+static struct scsi_transport_template *csio_fcoe_transport;
+static struct scsi_transport_template *csio_fcoe_transport_vport;
+
+/*
+ * debugfs support
+ */
+static int
+csio_mem_open(struct inode *inode, struct file *file)
+{
+ file->private_data = inode->i_private;
+ return 0;
+}
+
+static ssize_t
+csio_mem_read(struct file *file, char __user *buf, size_t count, loff_t *ppos)
+{
+ loff_t pos = *ppos;
+ loff_t avail = file->f_path.dentry->d_inode->i_size;
+ unsigned int mem = (uintptr_t)file->private_data & 3;
+ struct csio_hw *hw = file->private_data - mem;
+
+ if (pos < 0)
+ return -EINVAL;
+ if (pos >= avail)
+ return 0;
+ if (count > avail - pos)
+ count = avail - pos;
+
+ while (count) {
+ size_t len;
+ int ret, ofst;
+ __be32 data[16];
+
+ if (mem == MEM_MC)
+ ret = csio_hw_mc_read(hw, pos, data, NULL);
+ else
+ ret = csio_hw_edc_read(hw, mem, pos, data, NULL);
+ if (ret)
+ return ret;
+
+ ofst = pos % sizeof(data);
+ len = min(count, sizeof(data) - ofst);
+ if (copy_to_user(buf, (u8 *)data + ofst, len))
+ return -EFAULT;
+
+ buf += len;
+ pos += len;
+ count -= len;
+ }
+ count = pos - *ppos;
+ *ppos = pos;
+ return count;
+}
+
+static const struct file_operations csio_mem_debugfs_fops = {
+ .owner = THIS_MODULE,
+ .open = csio_mem_open,
+ .read = csio_mem_read,
+ .llseek = default_llseek,
+};
+
+static void __devinit
+csio_add_debugfs_mem(struct csio_hw *hw, const char *name,
+ unsigned int idx, unsigned int size_mb)
+{
+ struct dentry *de;
+
+ de = debugfs_create_file(name, S_IRUSR, hw->debugfs_root,
+ (void *)hw + idx, &csio_mem_debugfs_fops);
+ if (de && de->d_inode)
+ de->d_inode->i_size = size_mb << 20;
+}
+
+static int __devinit
+csio_setup_debugfs(struct csio_hw *hw)
+{
+ int i;
+
+ if (IS_ERR_OR_NULL(hw->debugfs_root))
+ return -1;
+
+ i = csio_rd_reg32(hw, MA_TARGET_MEM_ENABLE);
+ if (i & EDRAM0_ENABLE)
+ csio_add_debugfs_mem(hw, "edc0", MEM_EDC0, 5);
+ if (i & EDRAM1_ENABLE)
+ csio_add_debugfs_mem(hw, "edc1", MEM_EDC1, 5);
+ if (i & EXT_MEM_ENABLE)
+ csio_add_debugfs_mem(hw, "mc", MEM_MC,
+ EXT_MEM_SIZE_GET(csio_rd_reg32(hw, MA_EXT_MEMORY_BAR)));
+ return 0;
+}
+
+/*
+ * csio_dfs_create - Creates and sets up per-hw debugfs.
+ *
+ */
+static int
+csio_dfs_create(struct csio_hw *hw)
+{
+ if (csio_debugfs_root) {
+ hw->debugfs_root = debugfs_create_dir(pci_name(hw->pdev),
+ csio_debugfs_root);
+ csio_setup_debugfs(hw);
+ }
+
+ return 0;
+}
+
+/*
+ * csio_dfs_destroy - Destroys per-hw debugfs.
+ */
+static int
+csio_dfs_destroy(struct csio_hw *hw)
+{
+ if (hw->debugfs_root)
+ debugfs_remove_recursive(hw->debugfs_root);
+
+ return 0;
+}
+
+/*
+ * csio_dfs_init - Debug filesystem initialization for the module.
+ *
+ */
+static int
+csio_dfs_init(void)
+{
+ csio_debugfs_root = debugfs_create_dir(KBUILD_MODNAME, NULL);
+ if (!csio_debugfs_root)
+ pr_warn("Could not create debugfs entry, continuing\n");
+
+ return 0;
+}
+
+/*
+ * csio_dfs_exit - debugfs cleanup for the module.
+ */
+static void
+csio_dfs_exit(void)
+{
+ debugfs_remove(csio_debugfs_root);
+}
+
+/*
+ * csio_pci_init - PCI initialization.
+ * @pdev: PCI device.
+ * @bars: Bitmask of bars to be requested.
+ *
+ * Initializes the PCI function by enabling MMIO, setting bus
+ * mastership and setting DMA mask.
+ */
+static int
+csio_pci_init(struct pci_dev *pdev, int *bars)
+{
+ int rv = -ENODEV;
+
+ *bars = pci_select_bars(pdev, IORESOURCE_MEM);
+
+ if (pci_enable_device_mem(pdev))
+ goto err;
+
+ if (pci_request_selected_regions(pdev, *bars, KBUILD_MODNAME))
+ goto err_disable_device;
+
+ pci_set_master(pdev);
+ pci_try_set_mwi(pdev);
+
+ if (!pci_set_dma_mask(pdev, DMA_BIT_MASK(64))) {
+ pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(64));
+ } else if (!pci_set_dma_mask(pdev, DMA_BIT_MASK(32))) {
+ pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(32));
+ } else {
+ dev_err(&pdev->dev, "No suitable DMA available.\n");
+ goto err_release_regions;
+ }
+
+ return 0;
+
+err_release_regions:
+ pci_release_selected_regions(pdev, *bars);
+err_disable_device:
+ pci_disable_device(pdev);
+err:
+ return rv;
+
+}
+
+/*
+ * csio_pci_exit - PCI unitialization.
+ * @pdev: PCI device.
+ * @bars: Bars to be released.
+ *
+ */
+static void
+csio_pci_exit(struct pci_dev *pdev, int *bars)
+{
+ pci_release_selected_regions(pdev, *bars);
+ pci_disable_device(pdev);
+}
+
+/*
+ * csio_hw_init_workers - Initialize the HW module's worker threads.
+ * @hw: HW module.
+ *
+ */
+static void
+csio_hw_init_workers(struct csio_hw *hw)
+{
+ INIT_WORK(&hw->evtq_work, csio_evtq_worker);
+}
+
+static void
+csio_hw_exit_workers(struct csio_hw *hw)
+{
+ cancel_work_sync(&hw->evtq_work);
+ flush_scheduled_work();
+}
+
+static int
+csio_create_queues(struct csio_hw *hw)
+{
+ int i, j;
+ struct csio_mgmtm *mgmtm = csio_hw_to_mgmtm(hw);
+ int rv;
+ struct csio_scsi_cpu_info *info;
+
+ if (hw->flags & CSIO_HWF_Q_FW_ALLOCED)
+ return 0;
+
+ if (hw->intr_mode != CSIO_IM_MSIX) {
+ rv = csio_wr_iq_create(hw, NULL, hw->intr_iq_idx,
+ 0, hw->pport[0].portid, false, NULL);
+ if (rv != 0) {
+ csio_err(hw, " Forward Interrupt IQ failed!: %d\n", rv);
+ return rv;
+ }
+ }
+
+ /* FW event queue */
+ rv = csio_wr_iq_create(hw, NULL, hw->fwevt_iq_idx,
+ csio_get_fwevt_intr_idx(hw),
+ hw->pport[0].portid, true, NULL);
+ if (rv != 0) {
+ csio_err(hw, "FW event IQ config failed!: %d\n", rv);
+ return rv;
+ }
+
+ /* Create mgmt queue */
+ rv = csio_wr_eq_create(hw, NULL, mgmtm->eq_idx,
+ mgmtm->iq_idx, hw->pport[0].portid, NULL);
+
+ if (rv != 0) {
+ csio_err(hw, "Mgmt EQ create failed!: %d\n", rv);
+ goto err;
+ }
+
+ /* Create SCSI queues */
+ for (i = 0; i < hw->num_pports; i++) {
+ info = &hw->scsi_cpu_info[i];
+
+ for (j = 0; j < info->max_cpus; j++) {
+ struct csio_scsi_qset *sqset = &hw->sqset[i][j];
+
+ rv = csio_wr_iq_create(hw, NULL, sqset->iq_idx,
+ sqset->intr_idx, i, false, NULL);
+ if (rv != 0) {
+ csio_err(hw,
+ "SCSI module IQ config failed [%d][%d]:%d\n",
+ i, j, rv);
+ goto err;
+ }
+ rv = csio_wr_eq_create(hw, NULL, sqset->eq_idx,
+ sqset->iq_idx, i, NULL);
+ if (rv != 0) {
+ csio_err(hw,
+ "SCSI module EQ config failed [%d][%d]:%d\n",
+ i, j, rv);
+ goto err;
+ }
+ } /* for all CPUs */
+ } /* For all ports */
+
+ hw->flags |= CSIO_HWF_Q_FW_ALLOCED;
+ return 0;
+err:
+ csio_wr_destroy_queues(hw, true);
+ return -EINVAL;
+}
+
+/*
+ * csio_config_queues - Configure the DMA queues.
+ * @hw: HW module.
+ *
+ * Allocates memory for queues are registers them with FW.
+ */
+int
+csio_config_queues(struct csio_hw *hw)
+{
+ int i, j, idx, k = 0;
+ int rv;
+ struct csio_scsi_qset *sqset;
+ struct csio_mgmtm *mgmtm = csio_hw_to_mgmtm(hw);
+ struct csio_scsi_qset *orig;
+ struct csio_scsi_cpu_info *info;
+
+ if (hw->flags & CSIO_HWF_Q_MEM_ALLOCED)
+ return csio_create_queues(hw);
+
+ /* Calculate number of SCSI queues for MSIX we would like */
+ hw->num_scsi_msix_cpus = num_online_cpus();
+ hw->num_sqsets = num_online_cpus() * hw->num_pports;
+
+ if (hw->num_sqsets > CSIO_MAX_SCSI_QSETS) {
+ hw->num_sqsets = CSIO_MAX_SCSI_QSETS;
+ hw->num_scsi_msix_cpus = CSIO_MAX_SCSI_CPU;
+ }
+
+ /* Initialize max_cpus, may get reduced during msix allocations */
+ for (i = 0; i < hw->num_pports; i++)
+ hw->scsi_cpu_info[i].max_cpus = hw->num_scsi_msix_cpus;
+
+ csio_dbg(hw, "nsqsets:%d scpus:%d\n",
+ hw->num_sqsets, hw->num_scsi_msix_cpus);
+
+ csio_intr_enable(hw);
+
+ if (hw->intr_mode != CSIO_IM_MSIX) {
+
+ /* Allocate Forward interrupt iq. */
+ hw->intr_iq_idx = csio_wr_alloc_q(hw, CSIO_INTR_IQSIZE,
+ CSIO_INTR_WRSIZE, CSIO_INGRESS,
+ (void *)hw, 0, 0, NULL);
+ if (hw->intr_iq_idx == -1) {
+ csio_err(hw,
+ "Forward interrupt queue creation failed\n");
+ goto intr_disable;
+ }
+ }
+
+ /* Allocate the FW evt queue */
+ hw->fwevt_iq_idx = csio_wr_alloc_q(hw, CSIO_FWEVT_IQSIZE,
+ CSIO_FWEVT_WRSIZE,
+ CSIO_INGRESS, (void *)hw,
+ CSIO_FWEVT_FLBUFS, 0,
+ csio_fwevt_intx_handler);
+ if (hw->fwevt_iq_idx == -1) {
+ csio_err(hw, "FW evt queue creation failed\n");
+ goto intr_disable;
+ }
+
+ /* Allocate the mgmt queue */
+ mgmtm->eq_idx = csio_wr_alloc_q(hw, CSIO_MGMT_EQSIZE,
+ CSIO_MGMT_EQ_WRSIZE,
+ CSIO_EGRESS, (void *)hw, 0, 0, NULL);
+ if (mgmtm->eq_idx == -1) {
+ csio_err(hw, "Failed to alloc egress queue for mgmt module\n");
+ goto intr_disable;
+ }
+
+ /* Use FW IQ for MGMT req completion */
+ mgmtm->iq_idx = hw->fwevt_iq_idx;
+
+ /* Allocate SCSI queues */
+ for (i = 0; i < hw->num_pports; i++) {
+ info = &hw->scsi_cpu_info[i];
+
+ for (j = 0; j < hw->num_scsi_msix_cpus; j++) {
+ sqset = &hw->sqset[i][j];
+
+ if (j >= info->max_cpus) {
+ k = j % info->max_cpus;
+ orig = &hw->sqset[i][k];
+ sqset->eq_idx = orig->eq_idx;
+ sqset->iq_idx = orig->iq_idx;
+ continue;
+ }
+
+ idx = csio_wr_alloc_q(hw, csio_scsi_eqsize, 0,
+ CSIO_EGRESS, (void *)hw, 0, 0,
+ NULL);
+ if (idx == -1) {
+ csio_err(hw, "EQ creation failed for idx:%d\n",
+ idx);
+ goto intr_disable;
+ }
+
+ sqset->eq_idx = idx;
+
+ idx = csio_wr_alloc_q(hw, CSIO_SCSI_IQSIZE,
+ CSIO_SCSI_IQ_WRSZ, CSIO_INGRESS,
+ (void *)hw, 0, 0,
+ csio_scsi_intx_handler);
+ if (idx == -1) {
+ csio_err(hw, "IQ creation failed for idx:%d\n",
+ idx);
+ goto intr_disable;
+ }
+ sqset->iq_idx = idx;
+ } /* for all CPUs */
+ } /* For all ports */
+
+ hw->flags |= CSIO_HWF_Q_MEM_ALLOCED;
+
+ rv = csio_create_queues(hw);
+ if (rv != 0)
+ goto intr_disable;
+
+ /*
+ * Now request IRQs for the vectors. In the event of a failure,
+ * cleanup is handled internally by this function.
+ */
+ rv = csio_request_irqs(hw);
+ if (rv != 0)
+ return -EINVAL;
+
+ return 0;
+
+intr_disable:
+ csio_intr_disable(hw, false);
+
+ return -EINVAL;
+}
+
+static int
+csio_resource_alloc(struct csio_hw *hw)
+{
+ struct csio_wrm *wrm = csio_hw_to_wrm(hw);
+ int rv = -ENOMEM;
+
+ wrm->num_q = ((CSIO_MAX_SCSI_QSETS * 2) + CSIO_HW_NIQ +
+ CSIO_HW_NEQ + CSIO_HW_NFLQ + CSIO_HW_NINTXQ);
+
+ hw->mb_mempool = mempool_create_kmalloc_pool(CSIO_MIN_MEMPOOL_SZ,
+ sizeof(struct csio_mb));
+ if (!hw->mb_mempool)
+ goto err;
+
+ hw->rnode_mempool = mempool_create_kmalloc_pool(CSIO_MIN_MEMPOOL_SZ,
+ sizeof(struct csio_rnode));
+ if (!hw->rnode_mempool)
+ goto err_free_mb_mempool;
+
+ hw->scsi_pci_pool = pci_pool_create("csio_scsi_pci_pool", hw->pdev,
+ CSIO_SCSI_RSP_LEN, 8, 0);
+ if (!hw->scsi_pci_pool)
+ goto err_free_rn_pool;
+
+ return 0;
+
+err_free_rn_pool:
+ mempool_destroy(hw->rnode_mempool);
+ hw->rnode_mempool = NULL;
+err_free_mb_mempool:
+ mempool_destroy(hw->mb_mempool);
+ hw->mb_mempool = NULL;
+err:
+ return rv;
+}
+
+static void
+csio_resource_free(struct csio_hw *hw)
+{
+ pci_pool_destroy(hw->scsi_pci_pool);
+ hw->scsi_pci_pool = NULL;
+ mempool_destroy(hw->rnode_mempool);
+ hw->rnode_mempool = NULL;
+ mempool_destroy(hw->mb_mempool);
+ hw->mb_mempool = NULL;
+}
+
+/*
+ * csio_hw_alloc - Allocate and initialize the HW module.
+ * @pdev: PCI device.
+ *
+ * Allocates HW structure, DMA, memory resources, maps BARS to
+ * host memory and initializes HW module.
+ */
+static struct csio_hw * __devinit
+csio_hw_alloc(struct pci_dev *pdev)
+{
+ struct csio_hw *hw;
+
+ hw = kzalloc(sizeof(struct csio_hw), GFP_KERNEL);
+ if (!hw)
+ goto err;
+
+ hw->pdev = pdev;
+ strncpy(hw->drv_version, CSIO_DRV_VERSION, 32);
+
+ /* memory pool/DMA pool allocation */
+ if (csio_resource_alloc(hw))
+ goto err_free_hw;
+
+ /* Get the start address of registers from BAR 0 */
+ hw->regstart = ioremap_nocache(pci_resource_start(pdev, 0),
+ pci_resource_len(pdev, 0));
+ if (!hw->regstart) {
+ csio_err(hw, "Could not map BAR 0, regstart = %p\n",
+ hw->regstart);
+ goto err_resource_free;
+ }
+
+ csio_hw_init_workers(hw);
+
+ if (csio_hw_init(hw))
+ goto err_unmap_bar;
+
+ csio_dfs_create(hw);
+
+ csio_dbg(hw, "hw:%p\n", hw);
+
+ return hw;
+
+err_unmap_bar:
+ csio_hw_exit_workers(hw);
+ iounmap(hw->regstart);
+err_resource_free:
+ csio_resource_free(hw);
+err_free_hw:
+ kfree(hw);
+err:
+ return NULL;
+}
+
+/*
+ * csio_hw_free - Uninitialize and free the HW module.
+ * @hw: The HW module
+ *
+ * Disable interrupts, uninit the HW module, free resources, free hw.
+ */
+static void
+csio_hw_free(struct csio_hw *hw)
+{
+ csio_intr_disable(hw, true);
+ csio_hw_exit_workers(hw);
+ csio_hw_exit(hw);
+ iounmap(hw->regstart);
+ csio_dfs_destroy(hw);
+ csio_resource_free(hw);
+ kfree(hw);
+}
+
+/**
+ * csio_shost_init - Create and initialize the lnode module.
+ * @hw: The HW module.
+ * @dev: The device associated with this invocation.
+ * @probe: Called from probe context or not?
+ * @os_pln: Parent lnode if any.
+ *
+ * Allocates lnode structure via scsi_host_alloc, initializes
+ * shost, initializes lnode module and registers with SCSI ML
+ * via scsi_host_add. This function is shared between physical and
+ * virtual node ports.
+ */
+struct csio_lnode *
+csio_shost_init(struct csio_hw *hw, struct device *dev,
+ bool probe, struct csio_lnode *pln)
+{
+ struct Scsi_Host *shost = NULL;
+ struct csio_lnode *ln;
+
+ csio_fcoe_shost_template.cmd_per_lun = csio_lun_qdepth;
+ csio_fcoe_shost_vport_template.cmd_per_lun = csio_lun_qdepth;
+
+ /*
+ * hw->pdev is the physical port's PCI dev structure,
+ * which will be different from the NPIV dev structure.
+ */
+ if (dev == &hw->pdev->dev)
+ shost = scsi_host_alloc(
+ &csio_fcoe_shost_template,
+ sizeof(struct csio_lnode));
+ else
+ shost = scsi_host_alloc(
+ &csio_fcoe_shost_vport_template,
+ sizeof(struct csio_lnode));
+
+ if (!shost)
+ goto err;
+
+ ln = shost_priv(shost);
+ memset(ln, 0, sizeof(struct csio_lnode));
+
+ /* Link common lnode to this lnode */
+ ln->dev_num = (shost->host_no << 16);
+
+ shost->can_queue = CSIO_MAX_QUEUE;
+ shost->this_id = -1;
+ shost->unique_id = shost->host_no;
+ shost->max_cmd_len = 16; /* Max CDB length supported */
+ shost->max_id = min_t(uint32_t, csio_fcoe_rnodes,
+ hw->fres_info.max_ssns);
+ shost->max_lun = CSIO_MAX_LUN;
+ if (dev == &hw->pdev->dev)
+ shost->transportt = csio_fcoe_transport;
+ else
+ shost->transportt = csio_fcoe_transport_vport;
+
+ /* root lnode */
+ if (!hw->rln)
+ hw->rln = ln;
+
+ /* Other initialization here: Common, Transport specific */
+ if (csio_lnode_init(ln, hw, pln))
+ goto err_shost_put;
+
+ if (scsi_add_host(shost, dev))
+ goto err_lnode_exit;
+
+ return ln;
+
+err_lnode_exit:
+ csio_lnode_exit(ln);
+err_shost_put:
+ scsi_host_put(shost);
+err:
+ return NULL;
+}
+
+/**
+ * csio_shost_exit - De-instantiate the shost.
+ * @ln: The lnode module corresponding to the shost.
+ *
+ */
+void
+csio_shost_exit(struct csio_lnode *ln)
+{
+ struct Scsi_Host *shost = csio_ln_to_shost(ln);
+ struct csio_hw *hw = csio_lnode_to_hw(ln);
+
+ /* Inform transport */
+ fc_remove_host(shost);
+
+ /* Inform SCSI ML */
+ scsi_remove_host(shost);
+
+ /* Flush all the events, so that any rnode removal events
+ * already queued are all handled, before we remove the lnode.
+ */
+ spin_lock_irq(&hw->lock);
+ csio_evtq_flush(hw);
+ spin_unlock_irq(&hw->lock);
+
+ csio_lnode_exit(ln);
+ scsi_host_put(shost);
+}
+
+struct csio_lnode *
+csio_lnode_alloc(struct csio_hw *hw)
+{
+ return csio_shost_init(hw, &hw->pdev->dev, false, NULL);
+}
+
+void
+csio_lnodes_block_request(struct csio_hw *hw)
+{
+ struct Scsi_Host *shost;
+ struct csio_lnode *sln;
+ struct csio_lnode *ln;
+ struct list_head *cur_ln, *cur_cln;
+ struct csio_lnode **lnode_list;
+ int cur_cnt = 0, ii;
+
+ lnode_list = kzalloc((sizeof(struct csio_lnode *) * hw->num_lns),
+ GFP_KERNEL);
+ if (!lnode_list) {
+ csio_err(hw, "Failed to allocate lnodes_list");
+ return;
+ }
+
+ spin_lock_irq(&hw->lock);
+ /* Traverse sibling lnodes */
+ list_for_each(cur_ln, &hw->sln_head) {
+ sln = (struct csio_lnode *) cur_ln;
+ lnode_list[cur_cnt++] = sln;
+
+ /* Traverse children lnodes */
+ list_for_each(cur_cln, &sln->cln_head)
+ lnode_list[cur_cnt++] = (struct csio_lnode *) cur_cln;
+ }
+ spin_unlock_irq(&hw->lock);
+
+ for (ii = 0; ii < cur_cnt; ii++) {
+ csio_dbg(hw, "Blocking IOs on lnode: %p\n", lnode_list[ii]);
+ ln = lnode_list[ii];
+ shost = csio_ln_to_shost(ln);
+ scsi_block_requests(shost);
+
+ }
+ kfree(lnode_list);
+}
+
+void
+csio_lnodes_unblock_request(struct csio_hw *hw)
+{
+ struct csio_lnode *ln;
+ struct Scsi_Host *shost;
+ struct csio_lnode *sln;
+ struct list_head *cur_ln, *cur_cln;
+ struct csio_lnode **lnode_list;
+ int cur_cnt = 0, ii;
+
+ lnode_list = kzalloc((sizeof(struct csio_lnode *) * hw->num_lns),
+ GFP_KERNEL);
+ if (!lnode_list) {
+ csio_err(hw, "Failed to allocate lnodes_list");
+ return;
+ }
+
+ spin_lock_irq(&hw->lock);
+ /* Traverse sibling lnodes */
+ list_for_each(cur_ln, &hw->sln_head) {
+ sln = (struct csio_lnode *) cur_ln;
+ lnode_list[cur_cnt++] = sln;
+
+ /* Traverse children lnodes */
+ list_for_each(cur_cln, &sln->cln_head)
+ lnode_list[cur_cnt++] = (struct csio_lnode *) cur_cln;
+ }
+ spin_unlock_irq(&hw->lock);
+
+ for (ii = 0; ii < cur_cnt; ii++) {
+ csio_dbg(hw, "unblocking IOs on lnode: %p\n", lnode_list[ii]);
+ ln = lnode_list[ii];
+ shost = csio_ln_to_shost(ln);
+ scsi_unblock_requests(shost);
+ }
+ kfree(lnode_list);
+}
+
+void
+csio_lnodes_block_by_port(struct csio_hw *hw, uint8_t portid)
+{
+ struct csio_lnode *ln;
+ struct Scsi_Host *shost;
+ struct csio_lnode *sln;
+ struct list_head *cur_ln, *cur_cln;
+ struct csio_lnode **lnode_list;
+ int cur_cnt = 0, ii;
+
+ lnode_list = kzalloc((sizeof(struct csio_lnode *) * hw->num_lns),
+ GFP_KERNEL);
+ if (!lnode_list) {
+ csio_err(hw, "Failed to allocate lnodes_list");
+ return;
+ }
+
+ spin_lock_irq(&hw->lock);
+ /* Traverse sibling lnodes */
+ list_for_each(cur_ln, &hw->sln_head) {
+ sln = (struct csio_lnode *) cur_ln;
+ if (sln->portid != portid)
+ continue;
+
+ lnode_list[cur_cnt++] = sln;
+
+ /* Traverse children lnodes */
+ list_for_each(cur_cln, &sln->cln_head)
+ lnode_list[cur_cnt++] = (struct csio_lnode *) cur_cln;
+ }
+ spin_unlock_irq(&hw->lock);
+
+ for (ii = 0; ii < cur_cnt; ii++) {
+ csio_dbg(hw, "Blocking IOs on lnode: %p\n", lnode_list[ii]);
+ ln = lnode_list[ii];
+ shost = csio_ln_to_shost(ln);
+ scsi_block_requests(shost);
+ }
+ kfree(lnode_list);
+}
+
+void
+csio_lnodes_unblock_by_port(struct csio_hw *hw, uint8_t portid)
+{
+ struct csio_lnode *ln;
+ struct Scsi_Host *shost;
+ struct csio_lnode *sln;
+ struct list_head *cur_ln, *cur_cln;
+ struct csio_lnode **lnode_list;
+ int cur_cnt = 0, ii;
+
+ lnode_list = kzalloc((sizeof(struct csio_lnode *) * hw->num_lns),
+ GFP_KERNEL);
+ if (!lnode_list) {
+ csio_err(hw, "Failed to allocate lnodes_list");
+ return;
+ }
+
+ spin_lock_irq(&hw->lock);
+ /* Traverse sibling lnodes */
+ list_for_each(cur_ln, &hw->sln_head) {
+ sln = (struct csio_lnode *) cur_ln;
+ if (sln->portid != portid)
+ continue;
+ lnode_list[cur_cnt++] = sln;
+
+ /* Traverse children lnodes */
+ list_for_each(cur_cln, &sln->cln_head)
+ lnode_list[cur_cnt++] = (struct csio_lnode *) cur_cln;
+ }
+ spin_unlock_irq(&hw->lock);
+
+ for (ii = 0; ii < cur_cnt; ii++) {
+ csio_dbg(hw, "unblocking IOs on lnode: %p\n", lnode_list[ii]);
+ ln = lnode_list[ii];
+ shost = csio_ln_to_shost(ln);
+ scsi_unblock_requests(shost);
+ }
+ kfree(lnode_list);
+}
+
+void
+csio_lnodes_exit(struct csio_hw *hw, bool npiv)
+{
+ struct csio_lnode *sln;
+ struct csio_lnode *ln;
+ struct list_head *cur_ln, *cur_cln;
+ struct csio_lnode **lnode_list;
+ int cur_cnt = 0, ii;
+
+ lnode_list = kzalloc((sizeof(struct csio_lnode *) * hw->num_lns),
+ GFP_KERNEL);
+ if (!lnode_list) {
+ csio_err(hw, "lnodes_exit: Failed to allocate lnodes_list.\n");
+ return;
+ }
+
+ /* Get all child lnodes(NPIV ports) */
+ spin_lock_irq(&hw->lock);
+ list_for_each(cur_ln, &hw->sln_head) {
+ sln = (struct csio_lnode *) cur_ln;
+
+ /* Traverse children lnodes */
+ list_for_each(cur_cln, &sln->cln_head)
+ lnode_list[cur_cnt++] = (struct csio_lnode *) cur_cln;
+ }
+ spin_unlock_irq(&hw->lock);
+
+ /* Delete NPIV lnodes */
+ for (ii = 0; ii < cur_cnt; ii++) {
+ csio_dbg(hw, "Deleting child lnode: %p\n", lnode_list[ii]);
+ ln = lnode_list[ii];
+ fc_vport_terminate(ln->fc_vport);
+ }
+
+ /* Delete only npiv lnodes */
+ if (npiv)
+ goto free_lnodes;
+
+ cur_cnt = 0;
+ /* Get all physical lnodes */
+ spin_lock_irq(&hw->lock);
+ /* Traverse sibling lnodes */
+ list_for_each(cur_ln, &hw->sln_head) {
+ sln = (struct csio_lnode *) cur_ln;
+ lnode_list[cur_cnt++] = sln;
+ }
+ spin_unlock_irq(&hw->lock);
+
+ /* Delete physical lnodes */
+ for (ii = 0; ii < cur_cnt; ii++) {
+ csio_dbg(hw, "Deleting parent lnode: %p\n", lnode_list[ii]);
+ csio_shost_exit(lnode_list[ii]);
+ }
+
+free_lnodes:
+ kfree(lnode_list);
+}
+
+/*
+ * csio_lnode_init_post: Set lnode attributes after starting HW.
+ * @ln: lnode.
+ *
+ */
+static void
+csio_lnode_init_post(struct csio_lnode *ln)
+{
+ struct Scsi_Host *shost = csio_ln_to_shost(ln);
+
+ csio_fchost_attr_init(ln);
+
+ scsi_scan_host(shost);
+}
+
+/*
+ * csio_probe_one - Instantiate this function.
+ * @pdev: PCI device
+ * @id: Device ID
+ *
+ * This is the .probe() callback of the driver. This function:
+ * - Initializes the PCI function by enabling MMIO, setting bus
+ * mastership and setting DMA mask.
+ * - Allocates HW structure, DMA, memory resources, maps BARS to
+ * host memory and initializes HW module.
+ * - Allocates lnode structure via scsi_host_alloc, initializes
+ * shost, initialized lnode module and registers with SCSI ML
+ * via scsi_host_add.
+ * - Enables interrupts, and starts the chip by kicking off the
+ * HW state machine.
+ * - Once hardware is ready, initiated scan of the host via
+ * scsi_scan_host.
+ */
+static int __devinit
+csio_probe_one(struct pci_dev *pdev, const struct pci_device_id *id)
+{
+ int rv;
+ int bars;
+ int i;
+ struct csio_hw *hw;
+ struct csio_lnode *ln;
+
+ rv = csio_pci_init(pdev, &bars);
+ if (rv)
+ goto err;
+
+ hw = csio_hw_alloc(pdev);
+ if (!hw) {
+ rv = -ENODEV;
+ goto err_pci_exit;
+ }
+
+ pci_set_drvdata(pdev, hw);
+
+ if (csio_hw_start(hw) != 0) {
+ dev_err(&pdev->dev,
+ "Failed to start FW, continuing in debug mode.\n");
+ return 0;
+ }
+
+ sprintf(hw->fwrev_str, "%u.%u.%u.%u\n",
+ FW_HDR_FW_VER_MAJOR_GET(hw->fwrev),
+ FW_HDR_FW_VER_MINOR_GET(hw->fwrev),
+ FW_HDR_FW_VER_MICRO_GET(hw->fwrev),
+ FW_HDR_FW_VER_BUILD_GET(hw->fwrev));
+
+ for (i = 0; i < hw->num_pports; i++) {
+ ln = csio_shost_init(hw, &pdev->dev, true, NULL);
+ if (!ln) {
+ rv = -ENODEV;
+ break;
+ }
+ /* Initialize portid */
+ ln->portid = hw->pport[i].portid;
+
+ spin_lock_irq(&hw->lock);
+ if (csio_lnode_start(ln) != 0)
+ rv = -ENODEV;
+ spin_unlock_irq(&hw->lock);
+
+ if (rv)
+ break;
+
+ csio_lnode_init_post(ln);
+ }
+
+ if (rv)
+ goto err_lnode_exit;
+
+ return 0;
+
+err_lnode_exit:
+ csio_lnodes_block_request(hw);
+ spin_lock_irq(&hw->lock);
+ csio_hw_stop(hw);
+ spin_unlock_irq(&hw->lock);
+ csio_lnodes_unblock_request(hw);
+ pci_set_drvdata(hw->pdev, NULL);
+ csio_lnodes_exit(hw, 0);
+ csio_hw_free(hw);
+err_pci_exit:
+ csio_pci_exit(pdev, &bars);
+err:
+ dev_err(&pdev->dev, "probe of device failed: %d\n", rv);
+ return rv;
+}
+
+/*
+ * csio_remove_one - Remove one instance of the driver at this PCI function.
+ * @pdev: PCI device
+ *
+ * Used during hotplug operation.
+ */
+static void __devexit
+csio_remove_one(struct pci_dev *pdev)
+{
+ struct csio_hw *hw = pci_get_drvdata(pdev);
+ int bars = pci_select_bars(pdev, IORESOURCE_MEM);
+
+ csio_lnodes_block_request(hw);
+ spin_lock_irq(&hw->lock);
+
+ /* Stops lnode, Rnode s/m
+ * Quiesce IOs.
+ * All sessions with remote ports are unregistered.
+ */
+ csio_hw_stop(hw);
+ spin_unlock_irq(&hw->lock);
+ csio_lnodes_unblock_request(hw);
+
+ csio_lnodes_exit(hw, 0);
+ csio_hw_free(hw);
+ pci_set_drvdata(pdev, NULL);
+ csio_pci_exit(pdev, &bars);
+}
+
+/*
+ * csio_pci_error_detected - PCI error was detected
+ * @pdev: PCI device
+ *
+ */
+static pci_ers_result_t
+csio_pci_error_detected(struct pci_dev *pdev, pci_channel_state_t state)
+{
+ struct csio_hw *hw = pci_get_drvdata(pdev);
+
+ csio_lnodes_block_request(hw);
+ spin_lock_irq(&hw->lock);
+
+ /* Post PCI error detected evt to HW s/m
+ * HW s/m handles this evt by quiescing IOs, unregisters rports
+ * and finally takes the device to offline.
+ */
+ csio_post_event(&hw->sm, CSIO_HWE_PCIERR_DETECTED);
+ spin_unlock_irq(&hw->lock);
+ csio_lnodes_unblock_request(hw);
+ csio_lnodes_exit(hw, 0);
+ csio_intr_disable(hw, true);
+ pci_disable_device(pdev);
+ return state == pci_channel_io_perm_failure ?
+ PCI_ERS_RESULT_DISCONNECT : PCI_ERS_RESULT_NEED_RESET;
+}
+
+/*
+ * csio_pci_slot_reset - PCI slot has been reset.
+ * @pdev: PCI device
+ *
+ */
+static pci_ers_result_t
+csio_pci_slot_reset(struct pci_dev *pdev)
+{
+ struct csio_hw *hw = pci_get_drvdata(pdev);
+
+ if (pci_enable_device(pdev)) {
+ dev_err(&pdev->dev, "cannot re-enable device in slot reset\n");
+ return PCI_ERS_RESULT_DISCONNECT;
+ }
+
+ pci_set_master(pdev);
+ pci_restore_state(pdev);
+ pci_save_state(pdev);
+ pci_cleanup_aer_uncorrect_error_status(pdev);
+
+ /* Bring HW s/m to ready state.
+ * but don't resume IOs.
+ */
+ spin_lock_irq(&hw->lock);
+ csio_post_event(&hw->sm, CSIO_HWE_PCIERR_SLOT_RESET);
+ if (!csio_is_hw_ready(hw)) {
+ spin_unlock_irq(&hw->lock);
+ dev_err(&pdev->dev, "Can't initialize HW when in slot reset\n");
+ return PCI_ERS_RESULT_DISCONNECT;
+ }
+ spin_unlock_irq(&hw->lock);
+ return PCI_ERS_RESULT_RECOVERED;
+}
+
+/*
+ * csio_pci_resume - Resume normal operations
+ * @pdev: PCI device
+ *
+ */
+static void
+csio_pci_resume(struct pci_dev *pdev)
+{
+ struct csio_hw *hw = pci_get_drvdata(pdev);
+ struct csio_lnode *ln;
+ int rv = 0;
+ int i;
+
+ /* Bring the LINK UP and Resume IO */
+
+ for (i = 0; i < hw->num_pports; i++) {
+ ln = csio_shost_init(hw, &pdev->dev, true, NULL);
+ if (!ln) {
+ rv = -ENODEV;
+ break;
+ }
+ /* Initialize portid */
+ ln->portid = hw->pport[i].portid;
+
+ spin_lock_irq(&hw->lock);
+ if (csio_lnode_start(ln) != 0)
+ rv = -ENODEV;
+ spin_unlock_irq(&hw->lock);
+
+ if (rv)
+ break;
+
+ csio_lnode_init_post(ln);
+ }
+
+ if (rv)
+ goto err_resume_exit;
+
+ return;
+
+err_resume_exit:
+ csio_lnodes_block_request(hw);
+ spin_lock_irq(&hw->lock);
+ csio_hw_stop(hw);
+ spin_unlock_irq(&hw->lock);
+ csio_lnodes_unblock_request(hw);
+ csio_lnodes_exit(hw, 0);
+ csio_hw_free(hw);
+ dev_err(&pdev->dev, "resume of device failed: %d\n", rv);
+}
+
+static struct pci_error_handlers csio_err_handler = {
+ .error_detected = csio_pci_error_detected,
+ .slot_reset = csio_pci_slot_reset,
+ .resume = csio_pci_resume,
+};
+
+static DEFINE_PCI_DEVICE_TABLE(csio_pci_tbl) = {
+ CSIO_DEVICE(CSIO_DEVID_T440DBG_FCOE, 0), /* T440DBG FCOE */
+ CSIO_DEVICE(CSIO_DEVID_T420CR_FCOE, 0), /* T420CR FCOE */
+ CSIO_DEVICE(CSIO_DEVID_T422CR_FCOE, 0), /* T422CR FCOE */
+ CSIO_DEVICE(CSIO_DEVID_T440CR_FCOE, 0), /* T440CR FCOE */
+ CSIO_DEVICE(CSIO_DEVID_T420BCH_FCOE, 0), /* T420BCH FCOE */
+ CSIO_DEVICE(CSIO_DEVID_T440BCH_FCOE, 0), /* T440BCH FCOE */
+ CSIO_DEVICE(CSIO_DEVID_T440CH_FCOE, 0), /* T440CH FCOE */
+ CSIO_DEVICE(CSIO_DEVID_T420SO_FCOE, 0), /* T420SO FCOE */
+ CSIO_DEVICE(CSIO_DEVID_T420CX_FCOE, 0), /* T420CX FCOE */
+ CSIO_DEVICE(CSIO_DEVID_T420BT_FCOE, 0), /* T420BT FCOE */
+ CSIO_DEVICE(CSIO_DEVID_T404BT_FCOE, 0), /* T404BT FCOE */
+ CSIO_DEVICE(CSIO_DEVID_B420_FCOE, 0), /* B420 FCOE */
+ CSIO_DEVICE(CSIO_DEVID_B404_FCOE, 0), /* B404 FCOE */
+ CSIO_DEVICE(CSIO_DEVID_T480CR_FCOE, 0), /* T480 CR FCOE */
+ CSIO_DEVICE(CSIO_DEVID_T440LPCR_FCOE, 0), /* T440 LP-CR FCOE */
+ CSIO_DEVICE(CSIO_DEVID_PE10K, 0), /* PE10K FCOE */
+ CSIO_DEVICE(CSIO_DEVID_PE10K_PF1, 0), /* PE10K FCOE on PF1 */
+ { 0, 0, 0, 0, 0, 0, 0 }
+};
+
+
+static struct pci_driver csio_pci_driver = {
+ .name = KBUILD_MODNAME,
+ .driver = {
+ .owner = THIS_MODULE,
+ },
+ .id_table = csio_pci_tbl,
+ .probe = csio_probe_one,
+ .remove = csio_remove_one,
+ .err_handler = &csio_err_handler,
+};
+
+/*
+ * csio_init - Chelsio storage driver initialization function.
+ *
+ */
+static int __init
+csio_init(void)
+{
+ int rv = -ENOMEM;
+
+ pr_info("%s %s\n", CSIO_DRV_DESC, CSIO_DRV_VERSION);
+
+ csio_dfs_init();
+
+ csio_fcoe_transport = fc_attach_transport(&csio_fc_transport_funcs);
+ if (!csio_fcoe_transport)
+ goto err;
+
+ csio_fcoe_transport_vport =
+ fc_attach_transport(&csio_fc_transport_vport_funcs);
+ if (!csio_fcoe_transport_vport)
+ goto err_vport;
+
+ rv = pci_register_driver(&csio_pci_driver);
+ if (rv)
+ goto err_pci;
+
+ return 0;
+
+err_pci:
+ fc_release_transport(csio_fcoe_transport_vport);
+err_vport:
+ fc_release_transport(csio_fcoe_transport);
+err:
+ csio_dfs_exit();
+ return rv;
+}
+
+/*
+ * csio_exit - Chelsio storage driver uninitialization .
+ *
+ * Function that gets called in the unload path.
+ */
+static void __exit
+csio_exit(void)
+{
+ pci_unregister_driver(&csio_pci_driver);
+ csio_dfs_exit();
+ fc_release_transport(csio_fcoe_transport_vport);
+ fc_release_transport(csio_fcoe_transport);
+}
+
+module_init(csio_init);
+module_exit(csio_exit);
+MODULE_AUTHOR(CSIO_DRV_AUTHOR);
+MODULE_DESCRIPTION(CSIO_DRV_DESC);
+MODULE_LICENSE(CSIO_DRV_LICENSE);
+MODULE_DEVICE_TABLE(pci, csio_pci_tbl);
+MODULE_VERSION(CSIO_DRV_VERSION);
+MODULE_FIRMWARE(CSIO_FW_FNAME);
diff --git a/drivers/scsi/csiostor/csio_wr.c b/drivers/scsi/csiostor/csio_wr.c
new file mode 100644
index 0000000..329c6df
--- /dev/null
+++ b/drivers/scsi/csiostor/csio_wr.c
@@ -0,0 +1,1632 @@
+/*
+ * This file is part of the Chelsio FCoE driver for Linux.
+ *
+ * Copyright (c) 2008-2012 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses. You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ * Redistribution and use in source and binary forms, with or
+ * without modification, are permitted provided that the following
+ * conditions are met:
+ *
+ * - Redistributions of source code must retain the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer in the documentation and/or other materials
+ * provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include <linux/kernel.h>
+#include <linux/string.h>
+#include <linux/compiler.h>
+#include <linux/slab.h>
+#include <asm/page.h>
+#include <linux/cache.h>
+
+#include "csio_hw.h"
+#include "csio_wr.h"
+#include "csio_mb.h"
+#include "csio_defs.h"
+
+int csio_intr_coalesce_cnt; /* value:SGE_INGRESS_RX_THRESHOLD[0] */
+static int csio_sge_thresh_reg; /* SGE_INGRESS_RX_THRESHOLD[0] */
+
+int csio_intr_coalesce_time = 10; /* value:SGE_TIMER_VALUE_1 */
+static int csio_sge_timer_reg = 1;
+
+#define CSIO_SET_FLBUF_SIZE(_hw, _reg, _val) \
+ csio_wr_reg32((_hw), (_val), SGE_FL_BUFFER_SIZE##_reg)
+
+static void
+csio_get_flbuf_size(struct csio_hw *hw, struct csio_sge *sge, uint32_t reg)
+{
+ sge->sge_fl_buf_size[reg] = csio_rd_reg32(hw, SGE_FL_BUFFER_SIZE0 +
+ reg * sizeof(uint32_t));
+}
+
+/* Free list buffer size */
+static inline uint32_t
+csio_wr_fl_bufsz(struct csio_sge *sge, struct csio_dma_buf *buf)
+{
+ return sge->sge_fl_buf_size[buf->paddr & 0xF];
+}
+
+/* Size of the egress queue status page */
+static inline uint32_t
+csio_wr_qstat_pgsz(struct csio_hw *hw)
+{
+ return (hw->wrm.sge.sge_control & EGRSTATUSPAGESIZE(1)) ? 128 : 64;
+}
+
+/* Ring freelist doorbell */
+static inline void
+csio_wr_ring_fldb(struct csio_hw *hw, struct csio_q *flq)
+{
+ /*
+ * Ring the doorbell only when we have atleast CSIO_QCREDIT_SZ
+ * number of bytes in the freelist queue. This translates to atleast
+ * 8 freelist buffer pointers (since each pointer is 8 bytes).
+ */
+ if (flq->inc_idx >= 8) {
+ csio_wr_reg32(hw, DBPRIO(1) | QID(flq->un.fl.flid) |
+ PIDX(flq->inc_idx / 8),
+ MYPF_REG(SGE_PF_KDOORBELL));
+ flq->inc_idx &= 7;
+ }
+}
+
+/* Write a 0 cidx increment value to enable SGE interrupts for this queue */
+static void
+csio_wr_sge_intr_enable(struct csio_hw *hw, uint16_t iqid)
+{
+ csio_wr_reg32(hw, CIDXINC(0) |
+ INGRESSQID(iqid) |
+ TIMERREG(X_TIMERREG_RESTART_COUNTER),
+ MYPF_REG(SGE_PF_GTS));
+}
+
+/*
+ * csio_wr_fill_fl - Populate the FL buffers of a FL queue.
+ * @hw: HW module.
+ * @flq: Freelist queue.
+ *
+ * Fill up freelist buffer entries with buffers of size specified
+ * in the size register.
+ *
+ */
+static int
+csio_wr_fill_fl(struct csio_hw *hw, struct csio_q *flq)
+{
+ struct csio_wrm *wrm = csio_hw_to_wrm(hw);
+ struct csio_sge *sge = &wrm->sge;
+ __be64 *d = (__be64 *)(flq->vstart);
+ struct csio_dma_buf *buf = &flq->un.fl.bufs[0];
+ uint64_t paddr;
+ int sreg = flq->un.fl.sreg;
+ int n = flq->credits;
+
+ while (n--) {
+ buf->len = sge->sge_fl_buf_size[sreg];
+ buf->vaddr = pci_alloc_consistent(hw->pdev, buf->len,
+ &buf->paddr);
+ if (!buf->vaddr) {
+ csio_err(hw, "Could only fill %d buffers!\n", n + 1);
+ return -ENOMEM;
+ }
+
+ paddr = buf->paddr | (sreg & 0xF);
+
+ *d++ = cpu_to_be64(paddr);
+ buf++;
+ }
+
+ return 0;
+}
+
+/*
+ * csio_wr_update_fl -
+ * @hw: HW module.
+ * @flq: Freelist queue.
+ *
+ *
+ */
+static inline void
+csio_wr_update_fl(struct csio_hw *hw, struct csio_q *flq, uint16_t n)
+{
+
+ flq->inc_idx += n;
+ flq->pidx += n;
+ if (unlikely(flq->pidx >= flq->credits))
+ flq->pidx -= (uint16_t)flq->credits;
+
+ CSIO_INC_STATS(flq, n_flq_refill);
+}
+
+/*
+ * csio_wr_alloc_q - Allocate a WR queue and initialize it.
+ * @hw: HW module
+ * @qsize: Size of the queue in bytes
+ * @wrsize: Since of WR in this queue, if fixed.
+ * @type: Type of queue (Ingress/Egress/Freelist)
+ * @owner: Module that owns this queue.
+ * @nflb: Number of freelist buffers for FL.
+ * @sreg: What is the FL buffer size register?
+ * @iq_int_handler: Ingress queue handler in INTx mode.
+ *
+ * This function allocates and sets up a queue for the caller
+ * of size qsize, aligned at the required boundary. This is subject to
+ * be free entries being available in the queue array. If one is found,
+ * it is initialized with the allocated queue, marked as being used (owner),
+ * and a handle returned to the caller in form of the queue's index
+ * into the q_arr array.
+ * If user has indicated a freelist (by specifying nflb > 0), create
+ * another queue (with its own index into q_arr) for the freelist. Allocate
+ * memory for DMA buffer metadata (vaddr, len etc). Save off the freelist
+ * idx in the ingress queue's flq.idx. This is how a Freelist is associated
+ * with its owning ingress queue.
+ */
+int
+csio_wr_alloc_q(struct csio_hw *hw, uint32_t qsize, uint32_t wrsize,
+ uint16_t type, void *owner, uint32_t nflb, int sreg,
+ iq_handler_t iq_intx_handler)
+{
+ struct csio_wrm *wrm = csio_hw_to_wrm(hw);
+ struct csio_q *q, *flq;
+ int free_idx = wrm->free_qidx;
+ int ret_idx = free_idx;
+ uint32_t qsz;
+ int flq_idx;
+
+ if (free_idx >= wrm->num_q) {
+ csio_err(hw, "No more free queues.\n");
+ return -1;
+ }
+
+ switch (type) {
+ case CSIO_EGRESS:
+ qsz = ALIGN(qsize, CSIO_QCREDIT_SZ) + csio_wr_qstat_pgsz(hw);
+ break;
+ case CSIO_INGRESS:
+ switch (wrsize) {
+ case 16:
+ case 32:
+ case 64:
+ case 128:
+ break;
+ default:
+ csio_err(hw, "Invalid Ingress queue WR size:%d\n",
+ wrsize);
+ return -1;
+ }
+
+ /*
+ * Number of elements must be a multiple of 16
+ * So this includes status page size
+ */
+ qsz = ALIGN(qsize/wrsize, 16) * wrsize;
+
+ break;
+ case CSIO_FREELIST:
+ qsz = ALIGN(qsize/wrsize, 8) * wrsize + csio_wr_qstat_pgsz(hw);
+ break;
+ default:
+ csio_err(hw, "Invalid queue type: 0x%x\n", type);
+ return -1;
+ }
+
+ q = wrm->q_arr[free_idx];
+
+ q->vstart = pci_alloc_consistent(hw->pdev, qsz, &q->pstart);
+ if (!q->vstart) {
+ csio_err(hw,
+ "Failed to allocate DMA memory for "
+ "queue at id: %d size: %d\n", free_idx, qsize);
+ return -1;
+ }
+
+ /*
+ * We need to zero out the contents, importantly for ingress,
+ * since we start with a generatiom bit of 1 for ingress.
+ */
+ memset(q->vstart, 0, qsz);
+
+ q->type = type;
+ q->owner = owner;
+ q->pidx = q->cidx = q->inc_idx = 0;
+ q->size = qsz;
+ q->wr_sz = wrsize; /* If using fixed size WRs */
+
+ wrm->free_qidx++;
+
+ if (type == CSIO_INGRESS) {
+ /* Since queue area is set to zero */
+ q->un.iq.genbit = 1;
+
+ /*
+ * Ingress queue status page size is always the size of
+ * the ingress queue entry.
+ */
+ q->credits = (qsz - q->wr_sz) / q->wr_sz;
+ q->vwrap = (void *)((uintptr_t)(q->vstart) + qsz
+ - q->wr_sz);
+
+ /* Allocate memory for FL if requested */
+ if (nflb > 0) {
+ flq_idx = csio_wr_alloc_q(hw, nflb * sizeof(__be64),
+ sizeof(__be64), CSIO_FREELIST,
+ owner, 0, sreg, NULL);
+ if (flq_idx == -1) {
+ csio_err(hw,
+ "Failed to allocate FL queue"
+ " for IQ idx:%d\n", free_idx);
+ return -1;
+ }
+
+ /* Associate the new FL with the Ingress quue */
+ q->un.iq.flq_idx = flq_idx;
+
+ flq = wrm->q_arr[q->un.iq.flq_idx];
+ flq->un.fl.bufs = kzalloc(flq->credits *
+ sizeof(struct csio_dma_buf),
+ GFP_KERNEL);
+ if (!flq->un.fl.bufs) {
+ csio_err(hw,
+ "Failed to allocate FL queue bufs"
+ " for IQ idx:%d\n", free_idx);
+ return -1;
+ }
+
+ flq->un.fl.packen = 0;
+ flq->un.fl.offset = 0;
+ flq->un.fl.sreg = sreg;
+
+ /* Fill up the free list buffers */
+ if (csio_wr_fill_fl(hw, flq))
+ return -1;
+
+ /*
+ * Make sure in a FLQ, atleast 1 credit (8 FL buffers)
+ * remains unpopulated,otherwise HW thinks
+ * FLQ is empty.
+ */
+ flq->pidx = flq->inc_idx = flq->credits - 8;
+ } else {
+ q->un.iq.flq_idx = -1;
+ }
+
+ /* Associate the IQ INTx handler. */
+ q->un.iq.iq_intx_handler = iq_intx_handler;
+
+ csio_q_iqid(hw, ret_idx) = CSIO_MAX_QID;
+
+ } else if (type == CSIO_EGRESS) {
+ q->credits = (qsz - csio_wr_qstat_pgsz(hw)) / CSIO_QCREDIT_SZ;
+ q->vwrap = (void *)((uintptr_t)(q->vstart) + qsz
+ - csio_wr_qstat_pgsz(hw));
+ csio_q_eqid(hw, ret_idx) = CSIO_MAX_QID;
+ } else { /* Freelist */
+ q->credits = (qsz - csio_wr_qstat_pgsz(hw)) / sizeof(__be64);
+ q->vwrap = (void *)((uintptr_t)(q->vstart) + qsz
+ - csio_wr_qstat_pgsz(hw));
+ csio_q_flid(hw, ret_idx) = CSIO_MAX_QID;
+ }
+
+ return ret_idx;
+}
+
+/*
+ * csio_wr_iq_create_rsp - Response handler for IQ creation.
+ * @hw: The HW module.
+ * @mbp: Mailbox.
+ * @iq_idx: Ingress queue that got created.
+ *
+ * Handle FW_IQ_CMD mailbox completion. Save off the assigned IQ/FL ids.
+ */
+static int
+csio_wr_iq_create_rsp(struct csio_hw *hw, struct csio_mb *mbp, int iq_idx)
+{
+ struct csio_iq_params iqp;
+ enum fw_retval retval;
+ uint32_t iq_id;
+ int flq_idx;
+
+ memset(&iqp, 0, sizeof(struct csio_iq_params));
+
+ csio_mb_iq_alloc_write_rsp(hw, mbp, &retval, &iqp);
+
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "IQ cmd returned 0x%x!\n", retval);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ csio_q_iqid(hw, iq_idx) = iqp.iqid;
+ csio_q_physiqid(hw, iq_idx) = iqp.physiqid;
+ csio_q_pidx(hw, iq_idx) = csio_q_cidx(hw, iq_idx) = 0;
+ csio_q_inc_idx(hw, iq_idx) = 0;
+
+ /* Actual iq-id. */
+ iq_id = iqp.iqid - hw->wrm.fw_iq_start;
+
+ /* Set the iq-id to iq map table. */
+ if (iq_id >= CSIO_MAX_IQ) {
+ csio_err(hw,
+ "Exceeding MAX_IQ(%d) supported!"
+ " iqid:%d rel_iqid:%d FW iq_start:%d\n",
+ CSIO_MAX_IQ, iq_id, iqp.iqid, hw->wrm.fw_iq_start);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+ csio_q_set_intr_map(hw, iq_idx, iq_id);
+
+ /*
+ * During FW_IQ_CMD, FW sets interrupt_sent bit to 1 in the SGE
+ * ingress context of this queue. This will block interrupts to
+ * this queue until the next GTS write. Therefore, we do a
+ * 0-cidx increment GTS write for this queue just to clear the
+ * interrupt_sent bit. This will re-enable interrupts to this
+ * queue.
+ */
+ csio_wr_sge_intr_enable(hw, iqp.physiqid);
+
+ flq_idx = csio_q_iq_flq_idx(hw, iq_idx);
+ if (flq_idx != -1) {
+ struct csio_q *flq = hw->wrm.q_arr[flq_idx];
+
+ csio_q_flid(hw, flq_idx) = iqp.fl0id;
+ csio_q_cidx(hw, flq_idx) = 0;
+ csio_q_pidx(hw, flq_idx) = csio_q_credits(hw, flq_idx) - 8;
+ csio_q_inc_idx(hw, flq_idx) = csio_q_credits(hw, flq_idx) - 8;
+
+ /* Now update SGE about the buffers allocated during init */
+ csio_wr_ring_fldb(hw, flq);
+ }
+
+ mempool_free(mbp, hw->mb_mempool);
+
+ return 0;
+}
+
+/*
+ * csio_wr_iq_create - Configure an Ingress queue with FW.
+ * @hw: The HW module.
+ * @priv: Private data object.
+ * @iq_idx: Ingress queue index in the WR module.
+ * @vec: MSIX vector.
+ * @portid: PCIE Channel to be associated with this queue.
+ * @async: Is this a FW asynchronous message handling queue?
+ * @cbfn: Completion callback.
+ *
+ * This API configures an ingress queue with FW by issuing a FW_IQ_CMD mailbox
+ * with alloc/write bits set.
+ */
+int
+csio_wr_iq_create(struct csio_hw *hw, void *priv, int iq_idx,
+ uint32_t vec, uint8_t portid, bool async,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct csio_mb *mbp;
+ struct csio_iq_params iqp;
+ int flq_idx;
+
+ memset(&iqp, 0, sizeof(struct csio_iq_params));
+ csio_q_portid(hw, iq_idx) = portid;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ csio_err(hw, "IQ command out of memory!\n");
+ return -ENOMEM;
+ }
+
+ switch (hw->intr_mode) {
+ case CSIO_IM_INTX:
+ case CSIO_IM_MSI:
+ /* For interrupt forwarding queue only */
+ if (hw->intr_iq_idx == iq_idx)
+ iqp.iqandst = X_INTERRUPTDESTINATION_PCIE;
+ else
+ iqp.iqandst = X_INTERRUPTDESTINATION_IQ;
+ iqp.iqandstindex =
+ csio_q_physiqid(hw, hw->intr_iq_idx);
+ break;
+ case CSIO_IM_MSIX:
+ iqp.iqandst = X_INTERRUPTDESTINATION_PCIE;
+ iqp.iqandstindex = (uint16_t)vec;
+ break;
+ case CSIO_IM_NONE:
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ /* Pass in the ingress queue cmd parameters */
+ iqp.pfn = hw->pfn;
+ iqp.vfn = 0;
+ iqp.iq_start = 1;
+ iqp.viid = 0;
+ iqp.type = FW_IQ_TYPE_FL_INT_CAP;
+ iqp.iqasynch = async;
+ if (csio_intr_coalesce_cnt)
+ iqp.iqanus = X_UPDATESCHEDULING_COUNTER_OPTTIMER;
+ else
+ iqp.iqanus = X_UPDATESCHEDULING_TIMER;
+ iqp.iqanud = X_UPDATEDELIVERY_INTERRUPT;
+ iqp.iqpciech = portid;
+ iqp.iqintcntthresh = (uint8_t)csio_sge_thresh_reg;
+
+ switch (csio_q_wr_sz(hw, iq_idx)) {
+ case 16:
+ iqp.iqesize = 0; break;
+ case 32:
+ iqp.iqesize = 1; break;
+ case 64:
+ iqp.iqesize = 2; break;
+ case 128:
+ iqp.iqesize = 3; break;
+ }
+
+ iqp.iqsize = csio_q_size(hw, iq_idx) /
+ csio_q_wr_sz(hw, iq_idx);
+ iqp.iqaddr = csio_q_pstart(hw, iq_idx);
+
+ flq_idx = csio_q_iq_flq_idx(hw, iq_idx);
+ if (flq_idx != -1) {
+ struct csio_q *flq = hw->wrm.q_arr[flq_idx];
+
+ iqp.fl0paden = 1;
+ iqp.fl0packen = flq->un.fl.packen ? 1 : 0;
+ iqp.fl0fbmin = X_FETCHBURSTMIN_64B;
+ iqp.fl0fbmax = X_FETCHBURSTMAX_512B;
+ iqp.fl0size = csio_q_size(hw, flq_idx) / CSIO_QCREDIT_SZ;
+ iqp.fl0addr = csio_q_pstart(hw, flq_idx);
+ }
+
+ csio_mb_iq_alloc_write(hw, mbp, priv, CSIO_MB_DEFAULT_TMO, &iqp, cbfn);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of IQ cmd failed!\n");
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ if (cbfn != NULL)
+ return 0;
+
+ return csio_wr_iq_create_rsp(hw, mbp, iq_idx);
+}
+
+/*
+ * csio_wr_eq_create_rsp - Response handler for EQ creation.
+ * @hw: The HW module.
+ * @mbp: Mailbox.
+ * @eq_idx: Egress queue that got created.
+ *
+ * Handle FW_EQ_OFLD_CMD mailbox completion. Save off the assigned EQ ids.
+ */
+static int
+csio_wr_eq_cfg_rsp(struct csio_hw *hw, struct csio_mb *mbp, int eq_idx)
+{
+ struct csio_eq_params eqp;
+ enum fw_retval retval;
+
+ memset(&eqp, 0, sizeof(struct csio_eq_params));
+
+ csio_mb_eq_ofld_alloc_write_rsp(hw, mbp, &retval, &eqp);
+
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "EQ OFLD cmd returned 0x%x!\n", retval);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ csio_q_eqid(hw, eq_idx) = (uint16_t)eqp.eqid;
+ csio_q_physeqid(hw, eq_idx) = (uint16_t)eqp.physeqid;
+ csio_q_pidx(hw, eq_idx) = csio_q_cidx(hw, eq_idx) = 0;
+ csio_q_inc_idx(hw, eq_idx) = 0;
+
+ mempool_free(mbp, hw->mb_mempool);
+
+ return 0;
+}
+
+/*
+ * csio_wr_eq_create - Configure an Egress queue with FW.
+ * @hw: HW module.
+ * @priv: Private data.
+ * @eq_idx: Egress queue index in the WR module.
+ * @iq_idx: Associated ingress queue index.
+ * @cbfn: Completion callback.
+ *
+ * This API configures a offload egress queue with FW by issuing a
+ * FW_EQ_OFLD_CMD (with alloc + write ) mailbox.
+ */
+int
+csio_wr_eq_create(struct csio_hw *hw, void *priv, int eq_idx,
+ int iq_idx, uint8_t portid,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ struct csio_mb *mbp;
+ struct csio_eq_params eqp;
+
+ memset(&eqp, 0, sizeof(struct csio_eq_params));
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ csio_err(hw, "EQ command out of memory!\n");
+ return -ENOMEM;
+ }
+
+ eqp.pfn = hw->pfn;
+ eqp.vfn = 0;
+ eqp.eqstart = 1;
+ eqp.hostfcmode = X_HOSTFCMODE_STATUS_PAGE;
+ eqp.iqid = csio_q_iqid(hw, iq_idx);
+ eqp.fbmin = X_FETCHBURSTMIN_64B;
+ eqp.fbmax = X_FETCHBURSTMAX_512B;
+ eqp.cidxfthresh = 0;
+ eqp.pciechn = portid;
+ eqp.eqsize = csio_q_size(hw, eq_idx) / CSIO_QCREDIT_SZ;
+ eqp.eqaddr = csio_q_pstart(hw, eq_idx);
+
+ csio_mb_eq_ofld_alloc_write(hw, mbp, priv, CSIO_MB_DEFAULT_TMO,
+ &eqp, cbfn);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of EQ OFLD cmd failed!\n");
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ if (cbfn != NULL)
+ return 0;
+
+ return csio_wr_eq_cfg_rsp(hw, mbp, eq_idx);
+}
+
+/*
+ * csio_wr_iq_destroy_rsp - Response handler for IQ removal.
+ * @hw: The HW module.
+ * @mbp: Mailbox.
+ * @iq_idx: Ingress queue that was freed.
+ *
+ * Handle FW_IQ_CMD (free) mailbox completion.
+ */
+static int
+csio_wr_iq_destroy_rsp(struct csio_hw *hw, struct csio_mb *mbp, int iq_idx)
+{
+ enum fw_retval retval = csio_mb_fw_retval(mbp);
+ int rv = 0;
+
+ if (retval != FW_SUCCESS)
+ rv = -EINVAL;
+
+ mempool_free(mbp, hw->mb_mempool);
+
+ return rv;
+}
+
+/*
+ * csio_wr_iq_destroy - Free an ingress queue.
+ * @hw: The HW module.
+ * @priv: Private data object.
+ * @iq_idx: Ingress queue index to destroy
+ * @cbfn: Completion callback.
+ *
+ * This API frees an ingress queue by issuing the FW_IQ_CMD
+ * with the free bit set.
+ */
+static int
+csio_wr_iq_destroy(struct csio_hw *hw, void *priv, int iq_idx,
+ void (*cbfn)(struct csio_hw *, struct csio_mb *))
+{
+ int rv = 0;
+ struct csio_mb *mbp;
+ struct csio_iq_params iqp;
+ int flq_idx;
+
+ memset(&iqp, 0, sizeof(struct csio_iq_params));
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp)
+ return -ENOMEM;
+
+ iqp.pfn = hw->pfn;
+ iqp.vfn = 0;
+ iqp.iqid = csio_q_iqid(hw, iq_idx);
+ iqp.type = FW_IQ_TYPE_FL_INT_CAP;
+
+ flq_idx = csio_q_iq_flq_idx(hw, iq_idx);
+ if (flq_idx != -1)
+ iqp.fl0id = csio_q_flid(hw, flq_idx);
+ else
+ iqp.fl0id = 0xFFFF;
+
+ iqp.fl1id = 0xFFFF;
+
+ csio_mb_iq_free(hw, mbp, priv, CSIO_MB_DEFAULT_TMO, &iqp, cbfn);
+
+ rv = csio_mb_issue(hw, mbp);
+ if (rv != 0) {
+ mempool_free(mbp, hw->mb_mempool);
+ return rv;
+ }
+
+ if (cbfn != NULL)
+ return 0;
+
+ return csio_wr_iq_destroy_rsp(hw, mbp, iq_idx);
+}
+
+/*
+ * csio_wr_eq_destroy_rsp - Response handler for OFLD EQ creation.
+ * @hw: The HW module.
+ * @mbp: Mailbox.
+ * @eq_idx: Egress queue that was freed.
+ *
+ * Handle FW_OFLD_EQ_CMD (free) mailbox completion.
+ */
+static int
+csio_wr_eq_destroy_rsp(struct csio_hw *hw, struct csio_mb *mbp, int eq_idx)
+{
+ enum fw_retval retval = csio_mb_fw_retval(mbp);
+ int rv = 0;
+
+ if (retval != FW_SUCCESS)
+ rv = -EINVAL;
+
+ mempool_free(mbp, hw->mb_mempool);
+
+ return rv;
+}
+
+/*
+ * csio_wr_eq_destroy - Free an Egress queue.
+ * @hw: The HW module.
+ * @priv: Private data object.
+ * @eq_idx: Egress queue index to destroy
+ * @cbfn: Completion callback.
+ *
+ * This API frees an Egress queue by issuing the FW_EQ_OFLD_CMD
+ * with the free bit set.
+ */
+static int
+csio_wr_eq_destroy(struct csio_hw *hw, void *priv, int eq_idx,
+ void (*cbfn) (struct csio_hw *, struct csio_mb *))
+{
+ int rv = 0;
+ struct csio_mb *mbp;
+ struct csio_eq_params eqp;
+
+ memset(&eqp, 0, sizeof(struct csio_eq_params));
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp)
+ return -ENOMEM;
+
+ eqp.pfn = hw->pfn;
+ eqp.vfn = 0;
+ eqp.eqid = csio_q_eqid(hw, eq_idx);
+
+ csio_mb_eq_ofld_free(hw, mbp, priv, CSIO_MB_DEFAULT_TMO, &eqp, cbfn);
+
+ rv = csio_mb_issue(hw, mbp);
+ if (rv != 0) {
+ mempool_free(mbp, hw->mb_mempool);
+ return rv;
+ }
+
+ if (cbfn != NULL)
+ return 0;
+
+ return csio_wr_eq_destroy_rsp(hw, mbp, eq_idx);
+}
+
+/*
+ * csio_wr_cleanup_eq_stpg - Cleanup Egress queue status page
+ * @hw: HW module
+ * @qidx: Egress queue index
+ *
+ * Cleanup the Egress queue status page.
+ */
+static void
+csio_wr_cleanup_eq_stpg(struct csio_hw *hw, int qidx)
+{
+ struct csio_q *q = csio_hw_to_wrm(hw)->q_arr[qidx];
+ struct csio_qstatus_page *stp = (struct csio_qstatus_page *)q->vwrap;
+
+ memset(stp, 0, sizeof(*stp));
+}
+
+/*
+ * csio_wr_cleanup_iq_ftr - Cleanup Footer entries in IQ
+ * @hw: HW module
+ * @qidx: Ingress queue index
+ *
+ * Cleanup the footer entries in the given ingress queue,
+ * set to 1 the internal copy of genbit.
+ */
+static void
+csio_wr_cleanup_iq_ftr(struct csio_hw *hw, int qidx)
+{
+ struct csio_wrm *wrm = csio_hw_to_wrm(hw);
+ struct csio_q *q = wrm->q_arr[qidx];
+ void *wr;
+ struct csio_iqwr_footer *ftr;
+ uint32_t i = 0;
+
+ /* set to 1 since we are just about zero out genbit */
+ q->un.iq.genbit = 1;
+
+ for (i = 0; i < q->credits; i++) {
+ /* Get the WR */
+ wr = (void *)((uintptr_t)q->vstart +
+ (i * q->wr_sz));
+ /* Get the footer */
+ ftr = (struct csio_iqwr_footer *)((uintptr_t)wr +
+ (q->wr_sz - sizeof(*ftr)));
+ /* Zero out footer */
+ memset(ftr, 0, sizeof(*ftr));
+ }
+}
+
+int
+csio_wr_destroy_queues(struct csio_hw *hw, bool cmd)
+{
+ int i, flq_idx;
+ struct csio_q *q;
+ struct csio_wrm *wrm = csio_hw_to_wrm(hw);
+ int rv;
+
+ for (i = 0; i < wrm->free_qidx; i++) {
+ q = wrm->q_arr[i];
+
+ switch (q->type) {
+ case CSIO_EGRESS:
+ if (csio_q_eqid(hw, i) != CSIO_MAX_QID) {
+ csio_wr_cleanup_eq_stpg(hw, i);
+ if (!cmd) {
+ csio_q_eqid(hw, i) = CSIO_MAX_QID;
+ continue;
+ }
+
+ rv = csio_wr_eq_destroy(hw, NULL, i, NULL);
+ if ((rv == -EBUSY) || (rv == -ETIMEDOUT))
+ cmd = false;
+
+ csio_q_eqid(hw, i) = CSIO_MAX_QID;
+ }
+ case CSIO_INGRESS:
+ if (csio_q_iqid(hw, i) != CSIO_MAX_QID) {
+ csio_wr_cleanup_iq_ftr(hw, i);
+ if (!cmd) {
+ csio_q_iqid(hw, i) = CSIO_MAX_QID;
+ flq_idx = csio_q_iq_flq_idx(hw, i);
+ if (flq_idx != -1)
+ csio_q_flid(hw, flq_idx) =
+ CSIO_MAX_QID;
+ continue;
+ }
+
+ rv = csio_wr_iq_destroy(hw, NULL, i, NULL);
+ if ((rv == -EBUSY) || (rv == -ETIMEDOUT))
+ cmd = false;
+
+ csio_q_iqid(hw, i) = CSIO_MAX_QID;
+ flq_idx = csio_q_iq_flq_idx(hw, i);
+ if (flq_idx != -1)
+ csio_q_flid(hw, flq_idx) = CSIO_MAX_QID;
+ }
+ default:
+ break;
+ }
+ }
+
+ hw->flags &= ~CSIO_HWF_Q_FW_ALLOCED;
+
+ return 0;
+}
+
+/*
+ * csio_wr_get - Get requested size of WR entry/entries from queue.
+ * @hw: HW module.
+ * @qidx: Index of queue.
+ * @size: Cumulative size of Work request(s).
+ * @wrp: Work request pair.
+ *
+ * If requested credits are available, return the start address of the
+ * work request in the work request pair. Set pidx accordingly and
+ * return.
+ *
+ * NOTE about WR pair:
+ * ==================
+ * A WR can start towards the end of a queue, and then continue at the
+ * beginning, since the queue is considered to be circular. This will
+ * require a pair of address/size to be passed back to the caller -
+ * hence Work request pair format.
+ */
+int
+csio_wr_get(struct csio_hw *hw, int qidx, uint32_t size,
+ struct csio_wr_pair *wrp)
+{
+ struct csio_wrm *wrm = csio_hw_to_wrm(hw);
+ struct csio_q *q = wrm->q_arr[qidx];
+ void *cwr = (void *)((uintptr_t)(q->vstart) +
+ (q->pidx * CSIO_QCREDIT_SZ));
+ struct csio_qstatus_page *stp = (struct csio_qstatus_page *)q->vwrap;
+ uint16_t cidx = q->cidx = ntohs(stp->cidx);
+ uint16_t pidx = q->pidx;
+ uint32_t req_sz = ALIGN(size, CSIO_QCREDIT_SZ);
+ int req_credits = req_sz / CSIO_QCREDIT_SZ;
+ int credits;
+
+ CSIO_DB_ASSERT(q->owner != NULL);
+ CSIO_DB_ASSERT((qidx >= 0) && (qidx < wrm->free_qidx));
+ CSIO_DB_ASSERT(cidx <= q->credits);
+
+ /* Calculate credits */
+ if (pidx > cidx) {
+ credits = q->credits - (pidx - cidx) - 1;
+ } else if (cidx > pidx) {
+ credits = cidx - pidx - 1;
+ } else {
+ /* cidx == pidx, empty queue */
+ credits = q->credits;
+ CSIO_INC_STATS(q, n_qempty);
+ }
+
+ /*
+ * Check if we have enough credits.
+ * credits = 1 implies queue is full.
+ */
+ if (!credits || (req_credits > credits)) {
+ CSIO_INC_STATS(q, n_qfull);
+ return -EBUSY;
+ }
+
+ /*
+ * If we are here, we have enough credits to satisfy the
+ * request. Check if we are near the end of q, and if WR spills over.
+ * If it does, use the first addr/size to cover the queue until
+ * the end. Fit the remainder portion of the request at the top
+ * of queue and return it in the second addr/len. Set pidx
+ * accordingly.
+ */
+ if (unlikely(((uintptr_t)cwr + req_sz) > (uintptr_t)(q->vwrap))) {
+ wrp->addr1 = cwr;
+ wrp->size1 = (uint32_t)((uintptr_t)q->vwrap - (uintptr_t)cwr);
+ wrp->addr2 = q->vstart;
+ wrp->size2 = req_sz - wrp->size1;
+ q->pidx = (uint16_t)(ALIGN(wrp->size2, CSIO_QCREDIT_SZ) /
+ CSIO_QCREDIT_SZ);
+ CSIO_INC_STATS(q, n_qwrap);
+ CSIO_INC_STATS(q, n_eq_wr_split);
+ } else {
+ wrp->addr1 = cwr;
+ wrp->size1 = req_sz;
+ wrp->addr2 = NULL;
+ wrp->size2 = 0;
+ q->pidx += (uint16_t)req_credits;
+
+ /* We are the end of queue, roll back pidx to top of queue */
+ if (unlikely(q->pidx == q->credits)) {
+ q->pidx = 0;
+ CSIO_INC_STATS(q, n_qwrap);
+ }
+ }
+
+ q->inc_idx = (uint16_t)req_credits;
+
+ CSIO_INC_STATS(q, n_tot_reqs);
+
+ return 0;
+}
+
+/*
+ * csio_wr_copy_to_wrp - Copies given data into WR.
+ * @data_buf - Data buffer
+ * @wrp - Work request pair.
+ * @wr_off - Work request offset.
+ * @data_len - Data length.
+ *
+ * Copies the given data in Work Request. Work request pair(wrp) specifies
+ * address information of Work request.
+ * Returns: none
+ */
+void
+csio_wr_copy_to_wrp(void *data_buf, struct csio_wr_pair *wrp,
+ uint32_t wr_off, uint32_t data_len)
+{
+ uint32_t nbytes;
+
+ /* Number of space available in buffer addr1 of WRP */
+ nbytes = ((wrp->size1 - wr_off) >= data_len) ?
+ data_len : (wrp->size1 - wr_off);
+
+ memcpy((uint8_t *) wrp->addr1 + wr_off, data_buf, nbytes);
+ data_len -= nbytes;
+
+ /* Write the remaining data from the begining of circular buffer */
+ if (data_len) {
+ CSIO_DB_ASSERT(data_len <= wrp->size2);
+ CSIO_DB_ASSERT(wrp->addr2 != NULL);
+ memcpy(wrp->addr2, (uint8_t *) data_buf + nbytes, data_len);
+ }
+}
+
+/*
+ * csio_wr_issue - Notify chip of Work request.
+ * @hw: HW module.
+ * @qidx: Index of queue.
+ * @prio: 0: Low priority, 1: High priority
+ *
+ * Rings the SGE Doorbell by writing the current producer index of the passed
+ * in queue into the register.
+ *
+ */
+int
+csio_wr_issue(struct csio_hw *hw, int qidx, bool prio)
+{
+ struct csio_wrm *wrm = csio_hw_to_wrm(hw);
+ struct csio_q *q = wrm->q_arr[qidx];
+
+ CSIO_DB_ASSERT((qidx >= 0) && (qidx < wrm->free_qidx));
+
+ wmb();
+ /* Ring SGE Doorbell writing q->pidx into it */
+ csio_wr_reg32(hw, DBPRIO(prio) | QID(q->un.eq.physeqid) |
+ PIDX(q->inc_idx), MYPF_REG(SGE_PF_KDOORBELL));
+ q->inc_idx = 0;
+
+ return 0;
+}
+
+static inline uint32_t
+csio_wr_avail_qcredits(struct csio_q *q)
+{
+ if (q->pidx > q->cidx)
+ return q->pidx - q->cidx;
+ else if (q->cidx > q->pidx)
+ return q->credits - (q->cidx - q->pidx);
+ else
+ return 0; /* cidx == pidx, empty queue */
+}
+
+/*
+ * csio_wr_inval_flq_buf - Invalidate a free list buffer entry.
+ * @hw: HW module.
+ * @flq: The freelist queue.
+ *
+ * Invalidate the driver's version of a freelist buffer entry,
+ * without freeing the associated the DMA memory. The entry
+ * to be invalidated is picked up from the current Free list
+ * queue cidx.
+ *
+ */
+static inline void
+csio_wr_inval_flq_buf(struct csio_hw *hw, struct csio_q *flq)
+{
+ flq->cidx++;
+ if (flq->cidx == flq->credits) {
+ flq->cidx = 0;
+ CSIO_INC_STATS(flq, n_qwrap);
+ }
+}
+
+/*
+ * csio_wr_process_fl - Process a freelist completion.
+ * @hw: HW module.
+ * @q: The ingress queue attached to the Freelist.
+ * @wr: The freelist completion WR in the ingress queue.
+ * @len_to_qid: The lower 32-bits of the first flit of the RSP footer
+ * @iq_handler: Caller's handler for this completion.
+ * @priv: Private pointer of caller
+ *
+ */
+static inline void
+csio_wr_process_fl(struct csio_hw *hw, struct csio_q *q,
+ void *wr, uint32_t len_to_qid,
+ void (*iq_handler)(struct csio_hw *, void *,
+ uint32_t, struct csio_fl_dma_buf *,
+ void *),
+ void *priv)
+{
+ struct csio_wrm *wrm = csio_hw_to_wrm(hw);
+ struct csio_sge *sge = &wrm->sge;
+ struct csio_fl_dma_buf flb;
+ struct csio_dma_buf *buf, *fbuf;
+ uint32_t bufsz, len, lastlen = 0;
+ int n;
+ struct csio_q *flq = hw->wrm.q_arr[q->un.iq.flq_idx];
+
+ CSIO_DB_ASSERT(flq != NULL);
+
+ len = len_to_qid;
+
+ if (len & IQWRF_NEWBUF) {
+ if (flq->un.fl.offset > 0) {
+ csio_wr_inval_flq_buf(hw, flq);
+ flq->un.fl.offset = 0;
+ }
+ len = IQWRF_LEN_GET(len);
+ }
+
+ CSIO_DB_ASSERT(len != 0);
+
+ flb.totlen = len;
+
+ /* Consume all freelist buffers used for len bytes */
+ for (n = 0, fbuf = flb.flbufs; ; n++, fbuf++) {
+ buf = &flq->un.fl.bufs[flq->cidx];
+ bufsz = csio_wr_fl_bufsz(sge, buf);
+
+ fbuf->paddr = buf->paddr;
+ fbuf->vaddr = buf->vaddr;
+
+ flb.offset = flq->un.fl.offset;
+ lastlen = min(bufsz, len);
+ fbuf->len = lastlen;
+
+ len -= lastlen;
+ if (!len)
+ break;
+ csio_wr_inval_flq_buf(hw, flq);
+ }
+
+ flb.defer_free = flq->un.fl.packen ? 0 : 1;
+
+ iq_handler(hw, wr, q->wr_sz - sizeof(struct csio_iqwr_footer),
+ &flb, priv);
+
+ if (flq->un.fl.packen)
+ flq->un.fl.offset += ALIGN(lastlen, sge->csio_fl_align);
+ else
+ csio_wr_inval_flq_buf(hw, flq);
+
+}
+
+/*
+ * csio_is_new_iqwr - Is this a new Ingress queue entry ?
+ * @q: Ingress quueue.
+ * @ftr: Ingress queue WR SGE footer.
+ *
+ * The entry is new if our generation bit matches the corresponding
+ * bit in the footer of the current WR.
+ */
+static inline bool
+csio_is_new_iqwr(struct csio_q *q, struct csio_iqwr_footer *ftr)
+{
+ return (q->un.iq.genbit == (ftr->u.type_gen >> IQWRF_GEN_SHIFT));
+}
+
+/*
+ * csio_wr_process_iq - Process elements in Ingress queue.
+ * @hw: HW pointer
+ * @qidx: Index of queue
+ * @iq_handler: Handler for this queue
+ * @priv: Caller's private pointer
+ *
+ * This routine walks through every entry of the ingress queue, calling
+ * the provided iq_handler with the entry, until the generation bit
+ * flips.
+ */
+int
+csio_wr_process_iq(struct csio_hw *hw, struct csio_q *q,
+ void (*iq_handler)(struct csio_hw *, void *,
+ uint32_t, struct csio_fl_dma_buf *,
+ void *),
+ void *priv)
+{
+ struct csio_wrm *wrm = csio_hw_to_wrm(hw);
+ void *wr = (void *)((uintptr_t)q->vstart + (q->cidx * q->wr_sz));
+ struct csio_iqwr_footer *ftr;
+ uint32_t wr_type, fw_qid, qid;
+ struct csio_q *q_completed;
+ struct csio_q *flq = csio_iq_has_fl(q) ?
+ wrm->q_arr[q->un.iq.flq_idx] : NULL;
+ int rv = 0;
+
+ /* Get the footer */
+ ftr = (struct csio_iqwr_footer *)((uintptr_t)wr +
+ (q->wr_sz - sizeof(*ftr)));
+
+ /*
+ * When q wrapped around last time, driver should have inverted
+ * ic.genbit as well.
+ */
+ while (csio_is_new_iqwr(q, ftr)) {
+
+ CSIO_DB_ASSERT(((uintptr_t)wr + q->wr_sz) <=
+ (uintptr_t)q->vwrap);
+ rmb();
+ wr_type = IQWRF_TYPE_GET(ftr->u.type_gen);
+
+ switch (wr_type) {
+ case X_RSPD_TYPE_CPL:
+ /* Subtract footer from WR len */
+ iq_handler(hw, wr, q->wr_sz - sizeof(*ftr), NULL, priv);
+ break;
+ case X_RSPD_TYPE_FLBUF:
+ csio_wr_process_fl(hw, q, wr,
+ ntohl(ftr->pldbuflen_qid),
+ iq_handler, priv);
+ break;
+ case X_RSPD_TYPE_INTR:
+ fw_qid = ntohl(ftr->pldbuflen_qid);
+ qid = fw_qid - wrm->fw_iq_start;
+ q_completed = hw->wrm.intr_map[qid];
+
+ if (unlikely(qid ==
+ csio_q_physiqid(hw, hw->intr_iq_idx))) {
+ /*
+ * We are already in the Forward Interrupt
+ * Interrupt Queue Service! Do-not service
+ * again!
+ *
+ */
+ } else {
+ CSIO_DB_ASSERT(q_completed);
+ CSIO_DB_ASSERT(
+ q_completed->un.iq.iq_intx_handler);
+
+ /* Call the queue handler. */
+ q_completed->un.iq.iq_intx_handler(hw, NULL,
+ 0, NULL, (void *)q_completed);
+ }
+ break;
+ default:
+ csio_warn(hw, "Unknown resp type 0x%x received\n",
+ wr_type);
+ CSIO_INC_STATS(q, n_rsp_unknown);
+ break;
+ }
+
+ /*
+ * Ingress *always* has fixed size WR entries. Therefore,
+ * there should always be complete WRs towards the end of
+ * queue.
+ */
+ if (((uintptr_t)wr + q->wr_sz) == (uintptr_t)q->vwrap) {
+
+ /* Roll over to start of queue */
+ q->cidx = 0;
+ wr = q->vstart;
+
+ /* Toggle genbit */
+ q->un.iq.genbit ^= 0x1;
+
+ CSIO_INC_STATS(q, n_qwrap);
+ } else {
+ q->cidx++;
+ wr = (void *)((uintptr_t)(q->vstart) +
+ (q->cidx * q->wr_sz));
+ }
+
+ ftr = (struct csio_iqwr_footer *)((uintptr_t)wr +
+ (q->wr_sz - sizeof(*ftr)));
+ q->inc_idx++;
+
+ } /* while (q->un.iq.genbit == hdr->genbit) */
+
+ /*
+ * We need to re-arm SGE interrupts in case we got a stray interrupt,
+ * especially in msix mode. With INTx, this may be a common occurence.
+ */
+ if (unlikely(!q->inc_idx)) {
+ CSIO_INC_STATS(q, n_stray_comp);
+ rv = -EINVAL;
+ goto restart;
+ }
+
+ /* Replenish free list buffers if pending falls below low water mark */
+ if (flq) {
+ uint32_t avail = csio_wr_avail_qcredits(flq);
+ if (avail <= 16) {
+ /* Make sure in FLQ, atleast 1 credit (8 FL buffers)
+ * remains unpopulated otherwise HW thinks
+ * FLQ is empty.
+ */
+ csio_wr_update_fl(hw, flq, (flq->credits - 8) - avail);
+ csio_wr_ring_fldb(hw, flq);
+ }
+ }
+
+restart:
+ /* Now inform SGE about our incremental index value */
+ csio_wr_reg32(hw, CIDXINC(q->inc_idx) |
+ INGRESSQID(q->un.iq.physiqid) |
+ TIMERREG(csio_sge_timer_reg),
+ MYPF_REG(SGE_PF_GTS));
+ q->stats.n_tot_rsps += q->inc_idx;
+
+ q->inc_idx = 0;
+
+ return rv;
+}
+
+int
+csio_wr_process_iq_idx(struct csio_hw *hw, int qidx,
+ void (*iq_handler)(struct csio_hw *, void *,
+ uint32_t, struct csio_fl_dma_buf *,
+ void *),
+ void *priv)
+{
+ struct csio_wrm *wrm = csio_hw_to_wrm(hw);
+ struct csio_q *iq = wrm->q_arr[qidx];
+
+ return csio_wr_process_iq(hw, iq, iq_handler, priv);
+}
+
+static int
+csio_closest_timer(struct csio_sge *s, int time)
+{
+ int i, delta, match = 0, min_delta = INT_MAX;
+
+ for (i = 0; i < ARRAY_SIZE(s->timer_val); i++) {
+ delta = time - s->timer_val[i];
+ if (delta < 0)
+ delta = -delta;
+ if (delta < min_delta) {
+ min_delta = delta;
+ match = i;
+ }
+ }
+ return match;
+}
+
+static int
+csio_closest_thresh(struct csio_sge *s, int cnt)
+{
+ int i, delta, match = 0, min_delta = INT_MAX;
+
+ for (i = 0; i < ARRAY_SIZE(s->counter_val); i++) {
+ delta = cnt - s->counter_val[i];
+ if (delta < 0)
+ delta = -delta;
+ if (delta < min_delta) {
+ min_delta = delta;
+ match = i;
+ }
+ }
+ return match;
+}
+
+static void
+csio_wr_fixup_host_params(struct csio_hw *hw)
+{
+ struct csio_wrm *wrm = csio_hw_to_wrm(hw);
+ struct csio_sge *sge = &wrm->sge;
+ uint32_t clsz = L1_CACHE_BYTES;
+ uint32_t s_hps = PAGE_SHIFT - 10;
+ uint32_t ingpad = 0;
+ uint32_t stat_len = clsz > 64 ? 128 : 64;
+
+ csio_wr_reg32(hw, HOSTPAGESIZEPF0(s_hps) | HOSTPAGESIZEPF1(s_hps) |
+ HOSTPAGESIZEPF2(s_hps) | HOSTPAGESIZEPF3(s_hps) |
+ HOSTPAGESIZEPF4(s_hps) | HOSTPAGESIZEPF5(s_hps) |
+ HOSTPAGESIZEPF6(s_hps) | HOSTPAGESIZEPF7(s_hps),
+ SGE_HOST_PAGE_SIZE);
+
+ sge->csio_fl_align = clsz < 32 ? 32 : clsz;
+ ingpad = ilog2(sge->csio_fl_align) - 5;
+
+ csio_set_reg_field(hw, SGE_CONTROL, INGPADBOUNDARY_MASK |
+ EGRSTATUSPAGESIZE(1),
+ INGPADBOUNDARY(ingpad) |
+ EGRSTATUSPAGESIZE(stat_len != 64));
+
+ /* FL BUFFER SIZE#0 is Page size i,e already aligned to cache line */
+ csio_wr_reg32(hw, PAGE_SIZE, SGE_FL_BUFFER_SIZE0);
+ csio_wr_reg32(hw,
+ (csio_rd_reg32(hw, SGE_FL_BUFFER_SIZE2) +
+ sge->csio_fl_align - 1) & ~(sge->csio_fl_align - 1),
+ SGE_FL_BUFFER_SIZE2);
+ csio_wr_reg32(hw,
+ (csio_rd_reg32(hw, SGE_FL_BUFFER_SIZE3) +
+ sge->csio_fl_align - 1) & ~(sge->csio_fl_align - 1),
+ SGE_FL_BUFFER_SIZE3);
+
+ csio_wr_reg32(hw, HPZ0(PAGE_SHIFT - 12), ULP_RX_TDDP_PSZ);
+
+ /* default value of rx_dma_offset of the NIC driver */
+ csio_set_reg_field(hw, SGE_CONTROL, PKTSHIFT_MASK,
+ PKTSHIFT(CSIO_SGE_RX_DMA_OFFSET));
+}
+
+static void
+csio_init_intr_coalesce_parms(struct csio_hw *hw)
+{
+ struct csio_wrm *wrm = csio_hw_to_wrm(hw);
+ struct csio_sge *sge = &wrm->sge;
+
+ csio_sge_thresh_reg = csio_closest_thresh(sge, csio_intr_coalesce_cnt);
+ if (csio_intr_coalesce_cnt) {
+ csio_sge_thresh_reg = 0;
+ csio_sge_timer_reg = X_TIMERREG_RESTART_COUNTER;
+ return;
+ }
+
+ csio_sge_timer_reg = csio_closest_timer(sge, csio_intr_coalesce_time);
+}
+
+/*
+ * csio_wr_get_sge - Get SGE register values.
+ * @hw: HW module.
+ *
+ * Used by non-master functions and by master-functions relying on config file.
+ */
+static void
+csio_wr_get_sge(struct csio_hw *hw)
+{
+ struct csio_wrm *wrm = csio_hw_to_wrm(hw);
+ struct csio_sge *sge = &wrm->sge;
+ uint32_t ingpad;
+ int i;
+ u32 timer_value_0_and_1, timer_value_2_and_3, timer_value_4_and_5;
+ u32 ingress_rx_threshold;
+
+ sge->sge_control = csio_rd_reg32(hw, SGE_CONTROL);
+
+ ingpad = INGPADBOUNDARY_GET(sge->sge_control);
+
+ switch (ingpad) {
+ case X_INGPCIEBOUNDARY_32B:
+ sge->csio_fl_align = 32; break;
+ case X_INGPCIEBOUNDARY_64B:
+ sge->csio_fl_align = 64; break;
+ case X_INGPCIEBOUNDARY_128B:
+ sge->csio_fl_align = 128; break;
+ case X_INGPCIEBOUNDARY_256B:
+ sge->csio_fl_align = 256; break;
+ case X_INGPCIEBOUNDARY_512B:
+ sge->csio_fl_align = 512; break;
+ case X_INGPCIEBOUNDARY_1024B:
+ sge->csio_fl_align = 1024; break;
+ case X_INGPCIEBOUNDARY_2048B:
+ sge->csio_fl_align = 2048; break;
+ case X_INGPCIEBOUNDARY_4096B:
+ sge->csio_fl_align = 4096; break;
+ }
+
+ for (i = 0; i < CSIO_SGE_FL_SIZE_REGS; i++)
+ csio_get_flbuf_size(hw, sge, i);
+
+ timer_value_0_and_1 = csio_rd_reg32(hw, SGE_TIMER_VALUE_0_AND_1);
+ timer_value_2_and_3 = csio_rd_reg32(hw, SGE_TIMER_VALUE_2_AND_3);
+ timer_value_4_and_5 = csio_rd_reg32(hw, SGE_TIMER_VALUE_4_AND_5);
+
+ sge->timer_val[0] = (uint16_t)csio_core_ticks_to_us(hw,
+ TIMERVALUE0_GET(timer_value_0_and_1));
+ sge->timer_val[1] = (uint16_t)csio_core_ticks_to_us(hw,
+ TIMERVALUE1_GET(timer_value_0_and_1));
+ sge->timer_val[2] = (uint16_t)csio_core_ticks_to_us(hw,
+ TIMERVALUE2_GET(timer_value_2_and_3));
+ sge->timer_val[3] = (uint16_t)csio_core_ticks_to_us(hw,
+ TIMERVALUE3_GET(timer_value_2_and_3));
+ sge->timer_val[4] = (uint16_t)csio_core_ticks_to_us(hw,
+ TIMERVALUE4_GET(timer_value_4_and_5));
+ sge->timer_val[5] = (uint16_t)csio_core_ticks_to_us(hw,
+ TIMERVALUE5_GET(timer_value_4_and_5));
+
+ ingress_rx_threshold = csio_rd_reg32(hw, SGE_INGRESS_RX_THRESHOLD);
+ sge->counter_val[0] = THRESHOLD_0_GET(ingress_rx_threshold);
+ sge->counter_val[1] = THRESHOLD_1_GET(ingress_rx_threshold);
+ sge->counter_val[2] = THRESHOLD_2_GET(ingress_rx_threshold);
+ sge->counter_val[3] = THRESHOLD_3_GET(ingress_rx_threshold);
+
+ csio_init_intr_coalesce_parms(hw);
+}
+
+/*
+ * csio_wr_set_sge - Initialize SGE registers
+ * @hw: HW module.
+ *
+ * Used by Master function to initialize SGE registers in the absence
+ * of a config file.
+ */
+static void
+csio_wr_set_sge(struct csio_hw *hw)
+{
+ struct csio_wrm *wrm = csio_hw_to_wrm(hw);
+ struct csio_sge *sge = &wrm->sge;
+ int i;
+
+ /*
+ * Set up our basic SGE mode to deliver CPL messages to our Ingress
+ * Queue and Packet Date to the Free List.
+ */
+ csio_set_reg_field(hw, SGE_CONTROL, RXPKTCPLMODE, RXPKTCPLMODE);
+
+ sge->sge_control = csio_rd_reg32(hw, SGE_CONTROL);
+
+ /* sge->csio_fl_align is set up by csio_wr_fixup_host_params(). */
+
+ /*
+ * Set up to drop DOORBELL writes when the DOORBELL FIFO overflows
+ * and generate an interrupt when this occurs so we can recover.
+ */
+ csio_set_reg_field(hw, SGE_DBFIFO_STATUS,
+ HP_INT_THRESH(HP_INT_THRESH_MASK) |
+ LP_INT_THRESH(LP_INT_THRESH_MASK),
+ HP_INT_THRESH(CSIO_SGE_DBFIFO_INT_THRESH) |
+ LP_INT_THRESH(CSIO_SGE_DBFIFO_INT_THRESH));
+ csio_set_reg_field(hw, SGE_DOORBELL_CONTROL, ENABLE_DROP,
+ ENABLE_DROP);
+
+ /* SGE_FL_BUFFER_SIZE0 is set up by csio_wr_fixup_host_params(). */
+
+ CSIO_SET_FLBUF_SIZE(hw, 1, CSIO_SGE_FLBUF_SIZE1);
+ CSIO_SET_FLBUF_SIZE(hw, 2, CSIO_SGE_FLBUF_SIZE2);
+ CSIO_SET_FLBUF_SIZE(hw, 3, CSIO_SGE_FLBUF_SIZE3);
+ CSIO_SET_FLBUF_SIZE(hw, 4, CSIO_SGE_FLBUF_SIZE4);
+ CSIO_SET_FLBUF_SIZE(hw, 5, CSIO_SGE_FLBUF_SIZE5);
+ CSIO_SET_FLBUF_SIZE(hw, 6, CSIO_SGE_FLBUF_SIZE6);
+ CSIO_SET_FLBUF_SIZE(hw, 7, CSIO_SGE_FLBUF_SIZE7);
+ CSIO_SET_FLBUF_SIZE(hw, 8, CSIO_SGE_FLBUF_SIZE8);
+
+ for (i = 0; i < CSIO_SGE_FL_SIZE_REGS; i++)
+ csio_get_flbuf_size(hw, sge, i);
+
+ /* Initialize interrupt coalescing attributes */
+ sge->timer_val[0] = CSIO_SGE_TIMER_VAL_0;
+ sge->timer_val[1] = CSIO_SGE_TIMER_VAL_1;
+ sge->timer_val[2] = CSIO_SGE_TIMER_VAL_2;
+ sge->timer_val[3] = CSIO_SGE_TIMER_VAL_3;
+ sge->timer_val[4] = CSIO_SGE_TIMER_VAL_4;
+ sge->timer_val[5] = CSIO_SGE_TIMER_VAL_5;
+
+ sge->counter_val[0] = CSIO_SGE_INT_CNT_VAL_0;
+ sge->counter_val[1] = CSIO_SGE_INT_CNT_VAL_1;
+ sge->counter_val[2] = CSIO_SGE_INT_CNT_VAL_2;
+ sge->counter_val[3] = CSIO_SGE_INT_CNT_VAL_3;
+
+ csio_wr_reg32(hw, THRESHOLD_0(sge->counter_val[0]) |
+ THRESHOLD_1(sge->counter_val[1]) |
+ THRESHOLD_2(sge->counter_val[2]) |
+ THRESHOLD_3(sge->counter_val[3]),
+ SGE_INGRESS_RX_THRESHOLD);
+
+ csio_wr_reg32(hw,
+ TIMERVALUE0(csio_us_to_core_ticks(hw, sge->timer_val[0])) |
+ TIMERVALUE1(csio_us_to_core_ticks(hw, sge->timer_val[1])),
+ SGE_TIMER_VALUE_0_AND_1);
+
+ csio_wr_reg32(hw,
+ TIMERVALUE2(csio_us_to_core_ticks(hw, sge->timer_val[2])) |
+ TIMERVALUE3(csio_us_to_core_ticks(hw, sge->timer_val[3])),
+ SGE_TIMER_VALUE_2_AND_3);
+
+ csio_wr_reg32(hw,
+ TIMERVALUE4(csio_us_to_core_ticks(hw, sge->timer_val[4])) |
+ TIMERVALUE5(csio_us_to_core_ticks(hw, sge->timer_val[5])),
+ SGE_TIMER_VALUE_4_AND_5);
+
+ csio_init_intr_coalesce_parms(hw);
+}
+
+void
+csio_wr_sge_init(struct csio_hw *hw)
+{
+ /*
+ * If we are master:
+ * - If we plan to use the config file, we need to fixup some
+ * host specific registers, and read the rest of the SGE
+ * configuration.
+ * - If we dont plan to use the config file, we need to initialize
+ * SGE entirely, including fixing the host specific registers.
+ * If we arent the master, we are only allowed to read and work off of
+ * the already initialized SGE values.
+ *
+ * Therefore, before calling this function, we assume that the master-
+ * ship of the card, and whether to use config file or not, have
+ * already been decided. In other words, CSIO_HWF_USING_SOFT_PARAMS and
+ * CSIO_HWF_MASTER should be set/unset.
+ */
+ if (csio_is_hw_master(hw)) {
+ csio_wr_fixup_host_params(hw);
+
+ if (hw->flags & CSIO_HWF_USING_SOFT_PARAMS)
+ csio_wr_get_sge(hw);
+ else
+ csio_wr_set_sge(hw);
+ } else
+ csio_wr_get_sge(hw);
+}
+
+/*
+ * csio_wrm_init - Initialize Work request module.
+ * @wrm: WR module
+ * @hw: HW pointer
+ *
+ * Allocates memory for an array of queue pointers starting at q_arr.
+ */
+int
+csio_wrm_init(struct csio_wrm *wrm, struct csio_hw *hw)
+{
+ int i;
+
+ if (!wrm->num_q) {
+ csio_err(hw, "Num queues is not set\n");
+ return -EINVAL;
+ }
+
+ wrm->q_arr = kzalloc(sizeof(struct csio_q *) * wrm->num_q, GFP_KERNEL);
+ if (!wrm->q_arr)
+ goto err;
+
+ for (i = 0; i < wrm->num_q; i++) {
+ wrm->q_arr[i] = kzalloc(sizeof(struct csio_q), GFP_KERNEL);
+ if (!wrm->q_arr[i]) {
+ while (--i >= 0)
+ kfree(wrm->q_arr[i]);
+ goto err_free_arr;
+ }
+ }
+ wrm->free_qidx = 0;
+
+ return 0;
+
+err_free_arr:
+ kfree(wrm->q_arr);
+err:
+ return -ENOMEM;
+}
+
+/*
+ * csio_wrm_exit - Initialize Work request module.
+ * @wrm: WR module
+ * @hw: HW module
+ *
+ * Uninitialize WR module. Free q_arr and pointers in it.
+ * We have the additional job of freeing the DMA memory associated
+ * with the queues.
+ */
+void
+csio_wrm_exit(struct csio_wrm *wrm, struct csio_hw *hw)
+{
+ int i;
+ uint32_t j;
+ struct csio_q *q;
+ struct csio_dma_buf *buf;
+
+ for (i = 0; i < wrm->num_q; i++) {
+ q = wrm->q_arr[i];
+
+ if (wrm->free_qidx && (i < wrm->free_qidx)) {
+ if (q->type == CSIO_FREELIST) {
+ if (!q->un.fl.bufs)
+ continue;
+ for (j = 0; j < q->credits; j++) {
+ buf = &q->un.fl.bufs[j];
+ if (!buf->vaddr)
+ continue;
+ pci_free_consistent(hw->pdev, buf->len,
+ buf->vaddr,
+ buf->paddr);
+ }
+ kfree(q->un.fl.bufs);
+ }
+ pci_free_consistent(hw->pdev, q->size,
+ q->vstart, q->pstart);
+ }
+ kfree(q);
+ }
+
+ hw->flags &= ~CSIO_HWF_Q_MEM_ALLOCED;
+
+ kfree(wrm->q_arr);
+}
--
1.7.1
^ permalink raw reply related
* [V3 PATCH 1/9] csiostor: Chelsio FCoE offload driver submission (sources part 1).
From: Naresh Kumar Inna @ 2012-09-11 14:38 UTC (permalink / raw)
To: JBottomley, linux-scsi, dm, leedom; +Cc: netdev, naresh, chethan
In-Reply-To: <1347374347-3852-1-git-send-email-naresh@chelsio.com>
This patch contains the hardware interfacing functionality (chip/firmware
initialization/setup) and error handling. It also has slow path event
handling functionality, the Makefile and Kconfig changes.
Signed-off-by: Naresh Kumar Inna <naresh@chelsio.com>
---
V3: Added const qualifier to adapter descriptor table.
drivers/scsi/Kconfig | 1 +
drivers/scsi/Makefile | 1 +
drivers/scsi/csiostor/Kconfig | 20 +
drivers/scsi/csiostor/Makefile | 11 +
drivers/scsi/csiostor/csio_hw.c | 4396 +++++++++++++++++++++++++++++++++++++++
5 files changed, 4429 insertions(+), 0 deletions(-)
create mode 100644 drivers/scsi/csiostor/Kconfig
create mode 100644 drivers/scsi/csiostor/Makefile
create mode 100644 drivers/scsi/csiostor/csio_hw.c
diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig
index 74bf1aa..af7a3e7 100644
--- a/drivers/scsi/Kconfig
+++ b/drivers/scsi/Kconfig
@@ -1812,6 +1812,7 @@ config SCSI_VIRTIO
This is the virtual HBA driver for virtio. If the kernel will
be used in a virtual machine, say Y or M.
+source "drivers/scsi/csiostor/Kconfig"
endif # SCSI_LOWLEVEL
diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile
index 888f73a..8739aa7 100644
--- a/drivers/scsi/Makefile
+++ b/drivers/scsi/Makefile
@@ -90,6 +90,7 @@ obj-$(CONFIG_SCSI_QLA_FC) += qla2xxx/
obj-$(CONFIG_SCSI_QLA_ISCSI) += libiscsi.o qla4xxx/
obj-$(CONFIG_SCSI_LPFC) += lpfc/
obj-$(CONFIG_SCSI_BFA_FC) += bfa/
+obj-$(CONFIG_SCSI_CHELSIO_FCOE) += csiostor/
obj-$(CONFIG_SCSI_PAS16) += pas16.o
obj-$(CONFIG_SCSI_T128) += t128.o
obj-$(CONFIG_SCSI_DMX3191D) += dmx3191d.o
diff --git a/drivers/scsi/csiostor/Kconfig b/drivers/scsi/csiostor/Kconfig
new file mode 100644
index 0000000..c2acf02
--- /dev/null
+++ b/drivers/scsi/csiostor/Kconfig
@@ -0,0 +1,20 @@
+config SCSI_CHELSIO_FCOE
+ tristate "Chelsio Communications FCoE support"
+ depends on PCI && SCSI
+ select SCSI_FC_ATTRS
+ select FW_LOADER
+ help
+ This driver supports FCoE Offload functionality over
+ Chelsio T4-based 10Gb Converged Network Adapters.
+
+ For general information about Chelsio and our products, visit
+ our website at <http://www.chelsio.com>.
+
+ For customer support, please visit our customer support page at
+ <http://www.chelsio.com/support.html>.
+
+ Please send feedback to <linux-bugs@chelsio.com>.
+
+ To compile this driver as a module choose M here; the module
+ will be called csiostor.
+
diff --git a/drivers/scsi/csiostor/Makefile b/drivers/scsi/csiostor/Makefile
new file mode 100644
index 0000000..b581966
--- /dev/null
+++ b/drivers/scsi/csiostor/Makefile
@@ -0,0 +1,11 @@
+#
+## Chelsio FCoE driver
+#
+##
+
+ccflags-y += -I$(srctree)/drivers/net/ethernet/chelsio/cxgb4
+
+obj-$(CONFIG_SCSI_CHELSIO_FCOE) += csiostor.o
+
+csiostor-objs := csio_attr.o csio_init.o csio_lnode.o csio_scsi.o \
+ csio_hw.o csio_isr.o csio_mb.o csio_rnode.o csio_wr.o
diff --git a/drivers/scsi/csiostor/csio_hw.c b/drivers/scsi/csiostor/csio_hw.c
new file mode 100644
index 0000000..990dae9
--- /dev/null
+++ b/drivers/scsi/csiostor/csio_hw.c
@@ -0,0 +1,4396 @@
+/*
+ * This file is part of the Chelsio FCoE driver for Linux.
+ *
+ * Copyright (c) 2008-2012 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses. You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ * Redistribution and use in source and binary forms, with or
+ * without modification, are permitted provided that the following
+ * conditions are met:
+ *
+ * - Redistributions of source code must retain the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer in the documentation and/or other materials
+ * provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include <linux/pci.h>
+#include <linux/pci_regs.h>
+#include <linux/firmware.h>
+#include <linux/stddef.h>
+#include <linux/delay.h>
+#include <linux/string.h>
+#include <linux/compiler.h>
+#include <linux/jiffies.h>
+#include <linux/kernel.h>
+#include <linux/log2.h>
+
+#include "csio_hw.h"
+#include "csio_lnode.h"
+#include "csio_rnode.h"
+
+int csio_force_master;
+int csio_dbg_level = 0xFEFF;
+unsigned int csio_port_mask = 0xf;
+
+/* Default FW event queue entries. */
+static uint32_t csio_evtq_sz = CSIO_EVTQ_SIZE;
+
+/* Default MSI param level */
+int csio_msi = 2;
+
+/* FCoE function instances */
+static int dev_num;
+
+/* FCoE Adapter types & its description */
+static const struct csio_adap_desc csio_fcoe_adapters[] = {
+ {"T440-Dbg 10G", "Chelsio T440-Dbg 10G [FCoE]"},
+ {"T420-CR 10G", "Chelsio T420-CR 10G [FCoE]"},
+ {"T422-CR 10G/1G", "Chelsio T422-CR 10G/1G [FCoE]"},
+ {"T440-CR 10G", "Chelsio T440-CR 10G [FCoE]"},
+ {"T420-BCH 10G", "Chelsio T420-BCH 10G [FCoE]"},
+ {"T440-BCH 10G", "Chelsio T440-BCH 10G [FCoE]"},
+ {"T440-CH 10G", "Chelsio T440-CH 10G [FCoE]"},
+ {"T420-SO 10G", "Chelsio T420-SO 10G [FCoE]"},
+ {"T420-CX4 10G", "Chelsio T420-CX4 10G [FCoE]"},
+ {"T420-BT 10G", "Chelsio T420-BT 10G [FCoE]"},
+ {"T404-BT 1G", "Chelsio T404-BT 1G [FCoE]"},
+ {"B420-SR 10G", "Chelsio B420-SR 10G [FCoE]"},
+ {"B404-BT 1G", "Chelsio B404-BT 1G [FCoE]"},
+ {"T480-CR 10G", "Chelsio T480-CR 10G [FCoE]"},
+ {"T440-LP-CR 10G", "Chelsio T440-LP-CR 10G [FCoE]"},
+ {"T4 FPGA", "Chelsio T4 FPGA [FCoE]"}
+};
+
+static void csio_mgmtm_cleanup(struct csio_mgmtm *);
+static void csio_hw_mbm_cleanup(struct csio_hw *);
+
+/* State machine forward declarations */
+static void csio_hws_uninit(struct csio_hw *, enum csio_hw_ev);
+static void csio_hws_configuring(struct csio_hw *, enum csio_hw_ev);
+static void csio_hws_initializing(struct csio_hw *, enum csio_hw_ev);
+static void csio_hws_ready(struct csio_hw *, enum csio_hw_ev);
+static void csio_hws_quiescing(struct csio_hw *, enum csio_hw_ev);
+static void csio_hws_quiesced(struct csio_hw *, enum csio_hw_ev);
+static void csio_hws_resetting(struct csio_hw *, enum csio_hw_ev);
+static void csio_hws_removing(struct csio_hw *, enum csio_hw_ev);
+static void csio_hws_pcierr(struct csio_hw *, enum csio_hw_ev);
+
+static void csio_hw_initialize(struct csio_hw *hw);
+static void csio_evtq_stop(struct csio_hw *hw);
+static void csio_evtq_start(struct csio_hw *hw);
+
+int csio_is_hw_ready(struct csio_hw *hw)
+{
+ return csio_match_state(hw, csio_hws_ready);
+}
+
+int csio_is_hw_removing(struct csio_hw *hw)
+{
+ return csio_match_state(hw, csio_hws_removing);
+}
+
+
+/*
+ * csio_hw_wait_op_done_val - wait until an operation is completed
+ * @hw: the HW module
+ * @reg: the register to check for completion
+ * @mask: a single-bit field within @reg that indicates completion
+ * @polarity: the value of the field when the operation is completed
+ * @attempts: number of check iterations
+ * @delay: delay in usecs between iterations
+ * @valp: where to store the value of the register at completion time
+ *
+ * Wait until an operation is completed by checking a bit in a register
+ * up to @attempts times. If @valp is not NULL the value of the register
+ * at the time it indicated completion is stored there. Returns 0 if the
+ * operation completes and -EAGAIN otherwise.
+ */
+static int
+csio_hw_wait_op_done_val(struct csio_hw *hw, int reg, uint32_t mask,
+ int polarity, int attempts, int delay, uint32_t *valp)
+{
+ uint32_t val;
+ while (1) {
+ val = csio_rd_reg32(hw, reg);
+
+ if (!!(val & mask) == polarity) {
+ if (valp)
+ *valp = val;
+ return 0;
+ }
+
+ if (--attempts == 0)
+ return -EAGAIN;
+ if (delay)
+ udelay(delay);
+ }
+}
+
+void
+csio_set_reg_field(struct csio_hw *hw, uint32_t reg, uint32_t mask,
+ uint32_t value)
+{
+ uint32_t val = csio_rd_reg32(hw, reg) & ~mask;
+
+ csio_wr_reg32(hw, val | value, reg);
+ /* Flush */
+ csio_rd_reg32(hw, reg);
+
+}
+
+/*
+ * csio_hw_mc_read - read from MC through backdoor accesses
+ * @hw: the hw module
+ * @addr: address of first byte requested
+ * @data: 64 bytes of data containing the requested address
+ * @ecc: where to store the corresponding 64-bit ECC word
+ *
+ * Read 64 bytes of data from MC starting at a 64-byte-aligned address
+ * that covers the requested address @addr. If @parity is not %NULL it
+ * is assigned the 64-bit ECC word for the read data.
+ */
+int
+csio_hw_mc_read(struct csio_hw *hw, uint32_t addr, uint32_t *data,
+ uint64_t *ecc)
+{
+ int i;
+
+ if (csio_rd_reg32(hw, MC_BIST_CMD) & START_BIST)
+ return -EBUSY;
+ csio_wr_reg32(hw, addr & ~0x3fU, MC_BIST_CMD_ADDR);
+ csio_wr_reg32(hw, 64, MC_BIST_CMD_LEN);
+ csio_wr_reg32(hw, 0xc, MC_BIST_DATA_PATTERN);
+ csio_wr_reg32(hw, BIST_OPCODE(1) | START_BIST | BIST_CMD_GAP(1),
+ MC_BIST_CMD);
+ i = csio_hw_wait_op_done_val(hw, MC_BIST_CMD, START_BIST,
+ 0, 10, 1, NULL);
+ if (i)
+ return i;
+
+#define MC_DATA(i) MC_BIST_STATUS_REG(MC_BIST_STATUS_RDATA, i)
+
+ for (i = 15; i >= 0; i--)
+ *data++ = htonl(csio_rd_reg32(hw, MC_DATA(i)));
+ if (ecc)
+ *ecc = csio_rd_reg64(hw, MC_DATA(16));
+#undef MC_DATA
+ return 0;
+}
+
+/*
+ * csio_hw_edc_read - read from EDC through backdoor accesses
+ * @hw: the hw module
+ * @idx: which EDC to access
+ * @addr: address of first byte requested
+ * @data: 64 bytes of data containing the requested address
+ * @ecc: where to store the corresponding 64-bit ECC word
+ *
+ * Read 64 bytes of data from EDC starting at a 64-byte-aligned address
+ * that covers the requested address @addr. If @parity is not %NULL it
+ * is assigned the 64-bit ECC word for the read data.
+ */
+int
+csio_hw_edc_read(struct csio_hw *hw, int idx, uint32_t addr, uint32_t *data,
+ uint64_t *ecc)
+{
+ int i;
+
+ idx *= EDC_STRIDE;
+ if (csio_rd_reg32(hw, EDC_BIST_CMD + idx) & START_BIST)
+ return -EBUSY;
+ csio_wr_reg32(hw, addr & ~0x3fU, EDC_BIST_CMD_ADDR + idx);
+ csio_wr_reg32(hw, 64, EDC_BIST_CMD_LEN + idx);
+ csio_wr_reg32(hw, 0xc, EDC_BIST_DATA_PATTERN + idx);
+ csio_wr_reg32(hw, BIST_OPCODE(1) | BIST_CMD_GAP(1) | START_BIST,
+ EDC_BIST_CMD + idx);
+ i = csio_hw_wait_op_done_val(hw, EDC_BIST_CMD + idx, START_BIST,
+ 0, 10, 1, NULL);
+ if (i)
+ return i;
+
+#define EDC_DATA(i) (EDC_BIST_STATUS_REG(EDC_BIST_STATUS_RDATA, i) + idx)
+
+ for (i = 15; i >= 0; i--)
+ *data++ = htonl(csio_rd_reg32(hw, EDC_DATA(i)));
+ if (ecc)
+ *ecc = csio_rd_reg64(hw, EDC_DATA(16));
+#undef EDC_DATA
+ return 0;
+}
+
+/*
+ * csio_mem_win_rw - read/write memory through PCIE memory window
+ * @hw: the adapter
+ * @addr: address of first byte requested
+ * @data: MEMWIN0_APERTURE bytes of data containing the requested address
+ * @dir: direction of transfer 1 => read, 0 => write
+ *
+ * Read/write MEMWIN0_APERTURE bytes of data from MC starting at a
+ * MEMWIN0_APERTURE-byte-aligned address that covers the requested
+ * address @addr.
+ */
+static int
+csio_mem_win_rw(struct csio_hw *hw, u32 addr, __be32 *data, int dir)
+{
+ int i;
+
+ /*
+ * Setup offset into PCIE memory window. Address must be a
+ * MEMWIN0_APERTURE-byte-aligned address. (Read back MA register to
+ * ensure that changes propagate before we attempt to use the new
+ * values.)
+ */
+ csio_wr_reg32(hw, addr & ~(MEMWIN0_APERTURE - 1),
+ PCIE_MEM_ACCESS_OFFSET);
+ csio_rd_reg32(hw, PCIE_MEM_ACCESS_OFFSET);
+
+ /* Collecting data 4 bytes at a time upto MEMWIN0_APERTURE */
+ for (i = 0; i < MEMWIN0_APERTURE; i = i + sizeof(__be32)) {
+ if (dir)
+ *data++ = csio_rd_reg32(hw, (MEMWIN0_BASE + i));
+ else
+ csio_wr_reg32(hw, *data++, (MEMWIN0_BASE + i));
+ }
+
+ return 0;
+}
+
+/*
+ * csio_memory_rw - read/write EDC 0, EDC 1 or MC via PCIE memory window
+ * @hw: the csio_hw
+ * @mtype: memory type: MEM_EDC0, MEM_EDC1 or MEM_MC
+ * @addr: address within indicated memory type
+ * @len: amount of memory to transfer
+ * @buf: host memory buffer
+ * @dir: direction of transfer 1 => read, 0 => write
+ *
+ * Reads/writes an [almost] arbitrary memory region in the firmware: the
+ * firmware memory address, length and host buffer must be aligned on
+ * 32-bit boudaries. The memory is transferred as a raw byte sequence
+ * from/to the firmware's memory. If this memory contains data
+ * structures which contain multi-byte integers, it's the callers
+ * responsibility to perform appropriate byte order conversions.
+ */
+static int
+csio_memory_rw(struct csio_hw *hw, int mtype, u32 addr, u32 len,
+ uint32_t *buf, int dir)
+{
+ uint32_t pos, start, end, offset, memoffset;
+ int ret;
+ __be32 *data;
+
+ /*
+ * Argument sanity checks ...
+ */
+ if ((addr & 0x3) || (len & 0x3))
+ return -EINVAL;
+
+ data = kzalloc(MEMWIN0_APERTURE, GFP_KERNEL);
+ if (!data)
+ return -ENOMEM;
+
+ /* Offset into the region of memory which is being accessed
+ * MEM_EDC0 = 0
+ * MEM_EDC1 = 1
+ * MEM_MC = 2
+ */
+ memoffset = (mtype * (5 * 1024 * 1024));
+
+ /* Determine the PCIE_MEM_ACCESS_OFFSET */
+ addr = addr + memoffset;
+
+ /*
+ * The underlaying EDC/MC read routines read MEMWIN0_APERTURE bytes
+ * at a time so we need to round down the start and round up the end.
+ * We'll start copying out of the first line at (addr - start) a word
+ * at a time.
+ */
+ start = addr & ~(MEMWIN0_APERTURE-1);
+ end = (addr + len + MEMWIN0_APERTURE-1) & ~(MEMWIN0_APERTURE-1);
+ offset = (addr - start)/sizeof(__be32);
+
+ for (pos = start; pos < end; pos += MEMWIN0_APERTURE, offset = 0) {
+ /*
+ * If we're writing, copy the data from the caller's memory
+ * buffer
+ */
+ if (!dir) {
+ /*
+ * If we're doing a partial write, then we need to do
+ * a read-modify-write ...
+ */
+ if (offset || len < MEMWIN0_APERTURE) {
+ ret = csio_mem_win_rw(hw, pos, data, 1);
+ if (ret) {
+ kfree(data);
+ return ret;
+ }
+ }
+ while (offset < (MEMWIN0_APERTURE/sizeof(__be32)) &&
+ len > 0) {
+ data[offset++] = *buf++;
+ len -= sizeof(__be32);
+ }
+ }
+
+ /*
+ * Transfer a block of memory and bail if there's an error.
+ */
+ ret = csio_mem_win_rw(hw, pos, data, dir);
+ if (ret) {
+ kfree(data);
+ return ret;
+ }
+
+ /*
+ * If we're reading, copy the data into the caller's memory
+ * buffer.
+ */
+ if (dir)
+ while (offset < (MEMWIN0_APERTURE/sizeof(__be32)) &&
+ len > 0) {
+ *buf++ = data[offset++];
+ len -= sizeof(__be32);
+ }
+ }
+
+ kfree(data);
+
+ return 0;
+}
+
+static int
+csio_memory_write(struct csio_hw *hw, int mtype, u32 addr, u32 len, __be32 *buf)
+{
+ return csio_memory_rw(hw, mtype, addr, len, buf, 0);
+}
+
+/*
+ * EEPROM reads take a few tens of us while writes can take a bit over 5 ms.
+ */
+#define EEPROM_MAX_RD_POLL 40
+#define EEPROM_MAX_WR_POLL 6
+#define EEPROM_STAT_ADDR 0x7bfc
+#define VPD_BASE 0x400
+#define VPD_BASE_OLD 0
+#define VPD_LEN 512
+#define VPD_INFO_FLD_HDR_SIZE 3
+
+/*
+ * csio_hw_seeprom_read - read a serial EEPROM location
+ * @hw: hw to read
+ * @addr: EEPROM virtual address
+ * @data: where to store the read data
+ *
+ * Read a 32-bit word from a location in serial EEPROM using the card's PCI
+ * VPD capability. Note that this function must be called with a virtual
+ * address.
+ */
+static int
+csio_hw_seeprom_read(struct csio_hw *hw, uint32_t addr, uint32_t *data)
+{
+ uint16_t val = 0;
+ int attempts = EEPROM_MAX_RD_POLL;
+ uint32_t base = hw->params.pci.vpd_cap_addr;
+
+ if (addr >= EEPROMVSIZE || (addr & 3))
+ return -EINVAL;
+
+ pci_write_config_word(hw->pdev, base + PCI_VPD_ADDR, (uint16_t)addr);
+
+ do {
+ udelay(10);
+ pci_read_config_word(hw->pdev, base + PCI_VPD_ADDR, &val);
+ } while (!(val & PCI_VPD_ADDR_F) && --attempts);
+
+ if (!(val & PCI_VPD_ADDR_F)) {
+ csio_err(hw, "reading EEPROM address 0x%x failed\n", addr);
+ return -EINVAL;
+ }
+
+ pci_read_config_dword(hw->pdev, base + PCI_VPD_DATA, data);
+ *data = le32_to_cpu(*data);
+ return 0;
+}
+
+/*
+ * Partial EEPROM Vital Product Data structure. Includes only the ID and
+ * VPD-R sections.
+ */
+struct t4_vpd_hdr {
+ u8 id_tag;
+ u8 id_len[2];
+ u8 id_data[ID_LEN];
+ u8 vpdr_tag;
+ u8 vpdr_len[2];
+};
+
+/*
+ * csio_hw_get_vpd_keyword_val - Locates an information field keyword in
+ * the VPD
+ * @v: Pointer to buffered vpd data structure
+ * @kw: The keyword to search for
+ *
+ * Returns the value of the information field keyword or
+ * -EINVAL otherwise.
+ */
+static int
+csio_hw_get_vpd_keyword_val(const struct t4_vpd_hdr *v, const char *kw)
+{
+ int32_t i;
+ int32_t offset , len;
+ const uint8_t *buf = &v->id_tag;
+ const uint8_t *vpdr_len = &v->vpdr_tag;
+ offset = sizeof(struct t4_vpd_hdr);
+ len = (uint16_t)vpdr_len[1] + ((uint16_t)vpdr_len[2] << 8);
+
+ if (len + sizeof(struct t4_vpd_hdr) > VPD_LEN)
+ return -EINVAL;
+
+ for (i = offset; (i + VPD_INFO_FLD_HDR_SIZE) <= (offset + len);) {
+ if (memcmp(buf + i , kw, 2) == 0) {
+ i += VPD_INFO_FLD_HDR_SIZE;
+ return i;
+ }
+
+ i += VPD_INFO_FLD_HDR_SIZE + buf[i+2];
+ }
+
+ return -EINVAL;
+}
+
+static int
+csio_pci_capability(struct pci_dev *pdev, int cap, int *pos)
+{
+ *pos = pci_find_capability(pdev, cap);
+ if (*pos)
+ return 0;
+
+ return -1;
+}
+
+/*
+ * csio_hw_get_vpd_params - read VPD parameters from VPD EEPROM
+ * @hw: HW module
+ * @p: where to store the parameters
+ *
+ * Reads card parameters stored in VPD EEPROM.
+ */
+static int
+csio_hw_get_vpd_params(struct csio_hw *hw, struct csio_vpd *p)
+{
+ int i, ret, ec, sn, addr;
+ uint8_t *vpd, csum;
+ const struct t4_vpd_hdr *v;
+ /* To get around compilation warning from strstrip */
+ char *s;
+
+ if (csio_is_valid_vpd(hw))
+ return 0;
+
+ ret = csio_pci_capability(hw->pdev, PCI_CAP_ID_VPD,
+ &hw->params.pci.vpd_cap_addr);
+ if (ret)
+ return -EINVAL;
+
+ vpd = kzalloc(VPD_LEN, GFP_ATOMIC);
+ if (vpd == NULL)
+ return -ENOMEM;
+
+ /*
+ * Card information normally starts at VPD_BASE but early cards had
+ * it at 0.
+ */
+ ret = csio_hw_seeprom_read(hw, VPD_BASE, (uint32_t *)(vpd));
+ addr = *vpd == 0x82 ? VPD_BASE : VPD_BASE_OLD;
+
+ for (i = 0; i < VPD_LEN; i += 4) {
+ ret = csio_hw_seeprom_read(hw, addr + i, (uint32_t *)(vpd + i));
+ if (ret) {
+ kfree(vpd);
+ return ret;
+ }
+ }
+
+ /* Reset the VPD flag! */
+ hw->flags &= (~CSIO_HWF_VPD_VALID);
+
+ v = (const struct t4_vpd_hdr *)vpd;
+
+#define FIND_VPD_KW(var, name) do { \
+ var = csio_hw_get_vpd_keyword_val(v, name); \
+ if (var < 0) { \
+ csio_err(hw, "missing VPD keyword " name "\n"); \
+ kfree(vpd); \
+ return -EINVAL; \
+ } \
+} while (0)
+
+ FIND_VPD_KW(i, "RV");
+ for (csum = 0; i >= 0; i--)
+ csum += vpd[i];
+
+ if (csum) {
+ csio_err(hw, "corrupted VPD EEPROM, actual csum %u\n", csum);
+ kfree(vpd);
+ return -EINVAL;
+ }
+ FIND_VPD_KW(ec, "EC");
+ FIND_VPD_KW(sn, "SN");
+#undef FIND_VPD_KW
+
+ memcpy(p->id, v->id_data, ID_LEN);
+ s = strstrip(p->id);
+ memcpy(p->ec, vpd + ec, EC_LEN);
+ s = strstrip(p->ec);
+ i = vpd[sn - VPD_INFO_FLD_HDR_SIZE + 2];
+ memcpy(p->sn, vpd + sn, min(i, SERNUM_LEN));
+ s = strstrip(p->sn);
+
+ csio_valid_vpd_copied(hw);
+
+ kfree(vpd);
+ return 0;
+}
+
+/*
+ * csio_hw_sf1_read - read data from the serial flash
+ * @hw: the HW module
+ * @byte_cnt: number of bytes to read
+ * @cont: whether another operation will be chained
+ * @lock: whether to lock SF for PL access only
+ * @valp: where to store the read data
+ *
+ * Reads up to 4 bytes of data from the serial flash. The location of
+ * the read needs to be specified prior to calling this by issuing the
+ * appropriate commands to the serial flash.
+ */
+static int
+csio_hw_sf1_read(struct csio_hw *hw, uint32_t byte_cnt, int32_t cont,
+ int32_t lock, uint32_t *valp)
+{
+ int ret;
+
+ if (!byte_cnt || byte_cnt > 4)
+ return -EINVAL;
+ if (csio_rd_reg32(hw, SF_OP) & SF_BUSY)
+ return -EBUSY;
+
+ cont = cont ? SF_CONT : 0;
+ lock = lock ? SF_LOCK : 0;
+
+ csio_wr_reg32(hw, lock | cont | BYTECNT(byte_cnt - 1), SF_OP);
+ ret = csio_hw_wait_op_done_val(hw, SF_OP, SF_BUSY, 0, SF_ATTEMPTS,
+ 10, NULL);
+ if (!ret)
+ *valp = csio_rd_reg32(hw, SF_DATA);
+ return ret;
+}
+
+/*
+ * csio_hw_sf1_write - write data to the serial flash
+ * @hw: the HW module
+ * @byte_cnt: number of bytes to write
+ * @cont: whether another operation will be chained
+ * @lock: whether to lock SF for PL access only
+ * @val: value to write
+ *
+ * Writes up to 4 bytes of data to the serial flash. The location of
+ * the write needs to be specified prior to calling this by issuing the
+ * appropriate commands to the serial flash.
+ */
+static int
+csio_hw_sf1_write(struct csio_hw *hw, uint32_t byte_cnt, uint32_t cont,
+ int32_t lock, uint32_t val)
+{
+ if (!byte_cnt || byte_cnt > 4)
+ return -EINVAL;
+ if (csio_rd_reg32(hw, SF_OP) & SF_BUSY)
+ return -EBUSY;
+
+ cont = cont ? SF_CONT : 0;
+ lock = lock ? SF_LOCK : 0;
+
+ csio_wr_reg32(hw, val, SF_DATA);
+ csio_wr_reg32(hw, cont | BYTECNT(byte_cnt - 1) | OP_WR | lock, SF_OP);
+
+ return csio_hw_wait_op_done_val(hw, SF_OP, SF_BUSY, 0, SF_ATTEMPTS,
+ 10, NULL);
+}
+
+/*
+ * csio_hw_flash_wait_op - wait for a flash operation to complete
+ * @hw: the HW module
+ * @attempts: max number of polls of the status register
+ * @delay: delay between polls in ms
+ *
+ * Wait for a flash operation to complete by polling the status register.
+ */
+static int
+csio_hw_flash_wait_op(struct csio_hw *hw, int32_t attempts, int32_t delay)
+{
+ int ret;
+ uint32_t status;
+
+ while (1) {
+ ret = csio_hw_sf1_write(hw, 1, 1, 1, SF_RD_STATUS);
+ if (ret != 0)
+ return ret;
+
+ ret = csio_hw_sf1_read(hw, 1, 0, 1, &status);
+ if (ret != 0)
+ return ret;
+
+ if (!(status & 1))
+ return 0;
+ if (--attempts == 0)
+ return -EAGAIN;
+ if (delay)
+ msleep(delay);
+ }
+}
+
+/*
+ * csio_hw_read_flash - read words from serial flash
+ * @hw: the HW module
+ * @addr: the start address for the read
+ * @nwords: how many 32-bit words to read
+ * @data: where to store the read data
+ * @byte_oriented: whether to store data as bytes or as words
+ *
+ * Read the specified number of 32-bit words from the serial flash.
+ * If @byte_oriented is set the read data is stored as a byte array
+ * (i.e., big-endian), otherwise as 32-bit words in the platform's
+ * natural endianess.
+ */
+static int
+csio_hw_read_flash(struct csio_hw *hw, uint32_t addr, uint32_t nwords,
+ uint32_t *data, int32_t byte_oriented)
+{
+ int ret;
+
+ if (addr + nwords * sizeof(uint32_t) > hw->params.sf_size || (addr & 3))
+ return -EINVAL;
+
+ addr = swab32(addr) | SF_RD_DATA_FAST;
+
+ ret = csio_hw_sf1_write(hw, 4, 1, 0, addr);
+ if (ret != 0)
+ return ret;
+
+ ret = csio_hw_sf1_read(hw, 1, 1, 0, data);
+ if (ret != 0)
+ return ret;
+
+ for ( ; nwords; nwords--, data++) {
+ ret = csio_hw_sf1_read(hw, 4, nwords > 1, nwords == 1, data);
+ if (nwords == 1)
+ csio_wr_reg32(hw, 0, SF_OP); /* unlock SF */
+ if (ret)
+ return ret;
+ if (byte_oriented)
+ *data = htonl(*data);
+ }
+ return 0;
+}
+
+/*
+ * csio_hw_write_flash - write up to a page of data to the serial flash
+ * @hw: the hw
+ * @addr: the start address to write
+ * @n: length of data to write in bytes
+ * @data: the data to write
+ *
+ * Writes up to a page of data (256 bytes) to the serial flash starting
+ * at the given address. All the data must be written to the same page.
+ */
+static int
+csio_hw_write_flash(struct csio_hw *hw, uint32_t addr,
+ uint32_t n, const uint8_t *data)
+{
+ int ret = -EINVAL;
+ uint32_t buf[64];
+ uint32_t i, c, left, val, offset = addr & 0xff;
+
+ if (addr >= hw->params.sf_size || offset + n > SF_PAGE_SIZE)
+ return -EINVAL;
+
+ val = swab32(addr) | SF_PROG_PAGE;
+
+ ret = csio_hw_sf1_write(hw, 1, 0, 1, SF_WR_ENABLE);
+ if (ret != 0)
+ goto unlock;
+
+ ret = csio_hw_sf1_write(hw, 4, 1, 1, val);
+ if (ret != 0)
+ goto unlock;
+
+ for (left = n; left; left -= c) {
+ c = min(left, 4U);
+ for (val = 0, i = 0; i < c; ++i)
+ val = (val << 8) + *data++;
+
+ ret = csio_hw_sf1_write(hw, c, c != left, 1, val);
+ if (ret)
+ goto unlock;
+ }
+ ret = csio_hw_flash_wait_op(hw, 8, 1);
+ if (ret)
+ goto unlock;
+
+ csio_wr_reg32(hw, 0, SF_OP); /* unlock SF */
+
+ /* Read the page to verify the write succeeded */
+ ret = csio_hw_read_flash(hw, addr & ~0xff, ARRAY_SIZE(buf), buf, 1);
+ if (ret)
+ return ret;
+
+ if (memcmp(data - n, (uint8_t *)buf + offset, n)) {
+ csio_err(hw,
+ "failed to correctly write the flash page at %#x\n",
+ addr);
+ return -EINVAL;
+ }
+
+ return 0;
+
+unlock:
+ csio_wr_reg32(hw, 0, SF_OP); /* unlock SF */
+ return ret;
+}
+
+/*
+ * csio_hw_flash_erase_sectors - erase a range of flash sectors
+ * @hw: the HW module
+ * @start: the first sector to erase
+ * @end: the last sector to erase
+ *
+ * Erases the sectors in the given inclusive range.
+ */
+static int
+csio_hw_flash_erase_sectors(struct csio_hw *hw, int32_t start, int32_t end)
+{
+ int ret = 0;
+
+ while (start <= end) {
+
+ ret = csio_hw_sf1_write(hw, 1, 0, 1, SF_WR_ENABLE);
+ if (ret != 0)
+ goto out;
+
+ ret = csio_hw_sf1_write(hw, 4, 0, 1,
+ SF_ERASE_SECTOR | (start << 8));
+ if (ret != 0)
+ goto out;
+
+ ret = csio_hw_flash_wait_op(hw, 14, 500);
+ if (ret != 0)
+ goto out;
+
+ start++;
+ }
+out:
+ if (ret)
+ csio_err(hw, "erase of flash sector %d failed, error %d\n",
+ start, ret);
+ csio_wr_reg32(hw, 0, SF_OP); /* unlock SF */
+ return 0;
+}
+
+/*
+ * csio_hw_flash_cfg_addr - return the address of the flash
+ * configuration file
+ * @hw: the HW module
+ *
+ * Return the address within the flash where the Firmware Configuration
+ * File is stored.
+ */
+static unsigned int
+csio_hw_flash_cfg_addr(struct csio_hw *hw)
+{
+ if (hw->params.sf_size == 0x100000)
+ return FPGA_FLASH_CFG_OFFSET;
+ else
+ return FLASH_CFG_OFFSET;
+}
+
+static void
+csio_hw_print_fw_version(struct csio_hw *hw, char *str)
+{
+ csio_info(hw, "%s: %u.%u.%u.%u\n", str,
+ FW_HDR_FW_VER_MAJOR_GET(hw->fwrev),
+ FW_HDR_FW_VER_MINOR_GET(hw->fwrev),
+ FW_HDR_FW_VER_MICRO_GET(hw->fwrev),
+ FW_HDR_FW_VER_BUILD_GET(hw->fwrev));
+}
+
+/*
+ * csio_hw_get_fw_version - read the firmware version
+ * @hw: HW module
+ * @vers: where to place the version
+ *
+ * Reads the FW version from flash.
+ */
+static int
+csio_hw_get_fw_version(struct csio_hw *hw, uint32_t *vers)
+{
+ return csio_hw_read_flash(hw, FW_IMG_START +
+ offsetof(struct fw_hdr, fw_ver), 1,
+ vers, 0);
+}
+
+/*
+ * csio_hw_get_tp_version - read the TP microcode version
+ * @hw: HW module
+ * @vers: where to place the version
+ *
+ * Reads the TP microcode version from flash.
+ */
+static int
+csio_hw_get_tp_version(struct csio_hw *hw, u32 *vers)
+{
+ return csio_hw_read_flash(hw, FLASH_FW_START +
+ offsetof(struct fw_hdr, tp_microcode_ver), 1,
+ vers, 0);
+}
+
+/*
+ * csio_hw_check_fw_version - check if the FW is compatible with
+ * this driver
+ * @hw: HW module
+ *
+ * Checks if an adapter's FW is compatible with the driver. Returns 0
+ * if there's exact match, a negative error if the version could not be
+ * read or there's a major/minor version mismatch/minor.
+ */
+static int
+csio_hw_check_fw_version(struct csio_hw *hw)
+{
+ int ret, major, minor, micro;
+
+ ret = csio_hw_get_fw_version(hw, &hw->fwrev);
+ if (!ret)
+ ret = csio_hw_get_tp_version(hw, &hw->tp_vers);
+ if (ret)
+ return ret;
+
+ major = FW_HDR_FW_VER_MAJOR_GET(hw->fwrev);
+ minor = FW_HDR_FW_VER_MINOR_GET(hw->fwrev);
+ micro = FW_HDR_FW_VER_MICRO_GET(hw->fwrev);
+
+ if (major != FW_VERSION_MAJOR) { /* major mismatch - fail */
+ csio_err(hw, "card FW has major version %u, driver wants %u\n",
+ major, FW_VERSION_MAJOR);
+ return -EINVAL;
+ }
+
+ if (minor == FW_VERSION_MINOR && micro == FW_VERSION_MICRO)
+ return 0; /* perfect match */
+
+ /* Minor/micro version mismatch */
+ return -EINVAL;
+}
+
+/*
+ * csio_hw_fw_dload - download firmware.
+ * @hw: HW module
+ * @fw_data: firmware image to write.
+ * @size: image size
+ *
+ * Write the supplied firmware image to the card's serial flash.
+ */
+static int
+csio_hw_fw_dload(struct csio_hw *hw, uint8_t *fw_data, uint32_t size)
+{
+ uint32_t csum;
+ int32_t addr;
+ int ret;
+ uint32_t i;
+ uint8_t first_page[SF_PAGE_SIZE];
+ const uint32_t *p = (const uint32_t *)fw_data;
+ struct fw_hdr *hdr = (struct fw_hdr *)fw_data;
+ uint32_t sf_sec_size;
+
+ if ((!hw->params.sf_size) || (!hw->params.sf_nsec)) {
+ csio_err(hw, "Serial Flash data invalid\n");
+ return -EINVAL;
+ }
+
+ if (!size) {
+ csio_err(hw, "FW image has no data\n");
+ return -EINVAL;
+ }
+
+ if (size & 511) {
+ csio_err(hw, "FW image size not multiple of 512 bytes\n");
+ return -EINVAL;
+ }
+
+ if (ntohs(hdr->len512) * 512 != size) {
+ csio_err(hw, "FW image size differs from size in FW header\n");
+ return -EINVAL;
+ }
+
+ if (size > FW_MAX_SIZE) {
+ csio_err(hw, "FW image too large, max is %u bytes\n",
+ FW_MAX_SIZE);
+ return -EINVAL;
+ }
+
+ for (csum = 0, i = 0; i < size / sizeof(csum); i++)
+ csum += ntohl(p[i]);
+
+ if (csum != 0xffffffff) {
+ csio_err(hw, "corrupted firmware image, checksum %#x\n", csum);
+ return -EINVAL;
+ }
+
+ sf_sec_size = hw->params.sf_size / hw->params.sf_nsec;
+ i = DIV_ROUND_UP(size, sf_sec_size); /* # of sectors spanned */
+
+ csio_dbg(hw, "Erasing sectors... start:%d end:%d\n",
+ FW_START_SEC, FW_START_SEC + i - 1);
+
+ ret = csio_hw_flash_erase_sectors(hw, FW_START_SEC,
+ FW_START_SEC + i - 1);
+ if (ret) {
+ csio_err(hw, "Flash Erase failed\n");
+ goto out;
+ }
+
+ /*
+ * We write the correct version at the end so the driver can see a bad
+ * version if the FW write fails. Start by writing a copy of the
+ * first page with a bad version.
+ */
+ memcpy(first_page, fw_data, SF_PAGE_SIZE);
+ ((struct fw_hdr *)first_page)->fw_ver = htonl(0xffffffff);
+ ret = csio_hw_write_flash(hw, FW_IMG_START, SF_PAGE_SIZE, first_page);
+ if (ret)
+ goto out;
+
+ csio_dbg(hw, "Writing Flash .. start:%d end:%d\n",
+ FW_IMG_START, FW_IMG_START + size);
+
+ addr = FW_IMG_START;
+ for (size -= SF_PAGE_SIZE; size; size -= SF_PAGE_SIZE) {
+ addr += SF_PAGE_SIZE;
+ fw_data += SF_PAGE_SIZE;
+ ret = csio_hw_write_flash(hw, addr, SF_PAGE_SIZE, fw_data);
+ if (ret)
+ goto out;
+ }
+
+ ret = csio_hw_write_flash(hw,
+ FW_IMG_START +
+ offsetof(struct fw_hdr, fw_ver),
+ sizeof(hdr->fw_ver),
+ (const uint8_t *)&hdr->fw_ver);
+
+out:
+ if (ret)
+ csio_err(hw, "firmware download failed, error %d\n", ret);
+ return ret;
+}
+
+static int
+csio_hw_get_flash_params(struct csio_hw *hw)
+{
+ int ret;
+ uint32_t info = 0;
+
+ ret = csio_hw_sf1_write(hw, 1, 1, 0, SF_RD_ID);
+ if (!ret)
+ ret = csio_hw_sf1_read(hw, 3, 0, 1, &info);
+ csio_wr_reg32(hw, 0, SF_OP); /* unlock SF */
+ if (ret != 0)
+ return ret;
+
+ if ((info & 0xff) != 0x20) /* not a Numonix flash */
+ return -EINVAL;
+ info >>= 16; /* log2 of size */
+ if (info >= 0x14 && info < 0x18)
+ hw->params.sf_nsec = 1 << (info - 16);
+ else if (info == 0x18)
+ hw->params.sf_nsec = 64;
+ else
+ return -EINVAL;
+ hw->params.sf_size = 1 << info;
+
+ return 0;
+}
+
+static void
+csio_set_pcie_completion_timeout(struct csio_hw *hw, u8 range)
+{
+ uint16_t val;
+ uint32_t pcie_cap;
+
+ if (!csio_pci_capability(hw->pdev, PCI_CAP_ID_EXP, &pcie_cap)) {
+ pci_read_config_word(hw->pdev,
+ pcie_cap + PCI_EXP_DEVCTL2, &val);
+ val &= 0xfff0;
+ val |= range ;
+ pci_write_config_word(hw->pdev,
+ pcie_cap + PCI_EXP_DEVCTL2, val);
+ }
+}
+
+
+/*
+ * Return the specified PCI-E Configuration Space register from our Physical
+ * Function. We try first via a Firmware LDST Command since we prefer to let
+ * the firmware own all of these registers, but if that fails we go for it
+ * directly ourselves.
+ */
+static uint32_t
+csio_read_pcie_cfg4(struct csio_hw *hw, int reg)
+{
+ u32 val = 0;
+ struct csio_mb *mbp;
+ int rv;
+ struct fw_ldst_cmd *ldst_cmd;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ pci_read_config_dword(hw->pdev, reg, &val);
+ return val;
+ }
+
+ csio_mb_ldst(hw, mbp, CSIO_MB_DEFAULT_TMO, reg);
+
+ rv = csio_mb_issue(hw, mbp);
+
+ /*
+ * If the LDST Command suucceeded, exctract the returned register
+ * value. Otherwise read it directly ourself.
+ */
+ if (rv == 0) {
+ ldst_cmd = (struct fw_ldst_cmd *)(mbp->mb);
+ val = ntohl(ldst_cmd->u.pcie.data[0]);
+ } else
+ pci_read_config_dword(hw->pdev, reg, &val);
+
+ mempool_free(mbp, hw->mb_mempool);
+
+ return val;
+} /* csio_read_pcie_cfg4 */
+
+static int
+csio_hw_set_mem_win(struct csio_hw *hw)
+{
+ u32 bar0;
+
+ /*
+ * Truncation intentional: we only read the bottom 32-bits of the
+ * 64-bit BAR0/BAR1 ... We use the hardware backdoor mechanism to
+ * read BAR0 instead of using pci_resource_start() because we could be
+ * operating from within a Virtual Machine which is trapping our
+ * accesses to our Configuration Space and we need to set up the PCI-E
+ * Memory Window decoders with the actual addresses which will be
+ * coming across the PCI-E link.
+ */
+ bar0 = csio_read_pcie_cfg4(hw, PCI_BASE_ADDRESS_0);
+ bar0 &= PCI_BASE_ADDRESS_MEM_MASK;
+
+ /*
+ * Set up memory window for accessing adapter memory ranges. (Read
+ * back MA register to ensure that changes propagate before we attempt
+ * to use the new values.)
+ */
+ csio_wr_reg32(hw, (bar0 + MEMWIN0_BASE) | BIR(0) |
+ WINDOW(ilog2(MEMWIN0_APERTURE) - 10),
+ PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN, 0));
+ csio_wr_reg32(hw, (bar0 + MEMWIN1_BASE) | BIR(0) |
+ WINDOW(ilog2(MEMWIN1_APERTURE) - 10),
+ PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN, 1));
+ csio_wr_reg32(hw, (bar0 + MEMWIN2_BASE) | BIR(0) |
+ WINDOW(ilog2(MEMWIN2_APERTURE) - 10),
+ PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN, 2));
+ csio_rd_reg32(hw, PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN, 2));
+ return 0;
+} /* csio_hw_set_mem_win */
+
+
+
+/*****************************************************************************/
+/* HW State machine assists */
+/*****************************************************************************/
+
+static int
+csio_hw_dev_ready(struct csio_hw *hw)
+{
+ uint32_t reg;
+ int cnt = 6;
+
+ while (((reg = csio_rd_reg32(hw, PL_WHOAMI)) == 0xFFFFFFFF) &&
+ (--cnt != 0))
+ mdelay(100);
+
+ if ((cnt == 0) && (((int32_t)(SOURCEPF_GET(reg)) < 0) ||
+ (SOURCEPF_GET(reg) >= CSIO_MAX_PFN))) {
+ csio_err(hw, "PL_WHOAMI returned 0x%x, cnt:%d\n", reg, cnt);
+ return -EIO;
+ }
+
+ hw->pfn = SOURCEPF_GET(reg);
+
+ return 0;
+}
+
+/*
+ * csio_do_hello - Perform the HELLO FW Mailbox command and process response.
+ * @hw: HW module
+ * @state: Device state
+ *
+ * FW_HELLO_CMD has to be polled for completion.
+ */
+static int
+csio_do_hello(struct csio_hw *hw, enum csio_dev_state *state)
+{
+ struct csio_mb *mbp;
+ int rv = 0;
+ enum csio_dev_master master;
+ enum fw_retval retval;
+ uint8_t mpfn;
+ char state_str[16];
+ int retries = FW_CMD_HELLO_RETRIES;
+
+ memset(state_str, 0, sizeof(state_str));
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ rv = -ENOMEM;
+ CSIO_INC_STATS(hw, n_err_nomem);
+ goto out;
+ }
+
+ master = csio_force_master ? CSIO_MASTER_MUST : CSIO_MASTER_MAY;
+
+retry:
+ csio_mb_hello(hw, mbp, CSIO_MB_DEFAULT_TMO, hw->pfn,
+ hw->pfn, master, NULL);
+
+ rv = csio_mb_issue(hw, mbp);
+ if (rv) {
+ csio_err(hw, "failed to issue HELLO cmd. ret:%d.\n", rv);
+ goto out_free_mb;
+ }
+
+ csio_mb_process_hello_rsp(hw, mbp, &retval, state, &mpfn);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "HELLO cmd failed with ret: %d\n", retval);
+ rv = -EINVAL;
+ goto out_free_mb;
+ }
+
+ /* Firmware has designated us to be master */
+ if (hw->pfn == mpfn) {
+ hw->flags |= CSIO_HWF_MASTER;
+ } else if (*state == CSIO_DEV_STATE_UNINIT) {
+ /*
+ * If we're not the Master PF then we need to wait around for
+ * the Master PF Driver to finish setting up the adapter.
+ *
+ * Note that we also do this wait if we're a non-Master-capable
+ * PF and there is no current Master PF; a Master PF may show up
+ * momentarily and we wouldn't want to fail pointlessly. (This
+ * can happen when an OS loads lots of different drivers rapidly
+ * at the same time). In this case, the Master PF returned by
+ * the firmware will be PCIE_FW_MASTER_MASK so the test below
+ * will work ...
+ */
+
+ int waiting = FW_CMD_HELLO_TIMEOUT;
+
+ /*
+ * Wait for the firmware to either indicate an error or
+ * initialized state. If we see either of these we bail out
+ * and report the issue to the caller. If we exhaust the
+ * "hello timeout" and we haven't exhausted our retries, try
+ * again. Otherwise bail with a timeout error.
+ */
+ for (;;) {
+ uint32_t pcie_fw;
+
+ msleep(50);
+ waiting -= 50;
+
+ /*
+ * If neither Error nor Initialialized are indicated
+ * by the firmware keep waiting till we exaust our
+ * timeout ... and then retry if we haven't exhausted
+ * our retries ...
+ */
+ pcie_fw = csio_rd_reg32(hw, PCIE_FW);
+ if (!(pcie_fw & (PCIE_FW_ERR|PCIE_FW_INIT))) {
+ if (waiting <= 0) {
+ if (retries-- > 0)
+ goto retry;
+
+ rv = -ETIMEDOUT;
+ break;
+ }
+ continue;
+ }
+
+ /*
+ * We either have an Error or Initialized condition
+ * report errors preferentially.
+ */
+ if (state) {
+ if (pcie_fw & PCIE_FW_ERR) {
+ *state = CSIO_DEV_STATE_ERR;
+ rv = -ETIMEDOUT;
+ } else if (pcie_fw & PCIE_FW_INIT)
+ *state = CSIO_DEV_STATE_INIT;
+ }
+
+ /*
+ * If we arrived before a Master PF was selected and
+ * there's not a valid Master PF, grab its identity
+ * for our caller.
+ */
+ if (mpfn == PCIE_FW_MASTER_MASK &&
+ (pcie_fw & PCIE_FW_MASTER_VLD))
+ mpfn = PCIE_FW_MASTER_GET(pcie_fw);
+ break;
+ }
+ hw->flags &= ~CSIO_HWF_MASTER;
+ }
+
+ switch (*state) {
+ case CSIO_DEV_STATE_UNINIT:
+ strcpy(state_str, "Initializing");
+ break;
+ case CSIO_DEV_STATE_INIT:
+ strcpy(state_str, "Initialized");
+ break;
+ case CSIO_DEV_STATE_ERR:
+ strcpy(state_str, "Error");
+ break;
+ default:
+ strcpy(state_str, "Unknown");
+ break;
+ }
+
+ if (hw->pfn == mpfn)
+ csio_info(hw, "PF: %d, Coming up as MASTER, HW state: %s\n",
+ hw->pfn, state_str);
+ else
+ csio_info(hw,
+ "PF: %d, Coming up as SLAVE, Master PF: %d, HW state: %s\n",
+ hw->pfn, mpfn, state_str);
+
+out_free_mb:
+ mempool_free(mbp, hw->mb_mempool);
+out:
+ return rv;
+}
+
+/*
+ * csio_do_bye - Perform the BYE FW Mailbox command and process response.
+ * @hw: HW module
+ *
+ */
+static int
+csio_do_bye(struct csio_hw *hw)
+{
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ csio_mb_bye(hw, mbp, CSIO_MB_DEFAULT_TMO, NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of BYE command failed\n");
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ retval = csio_mb_fw_retval(mbp);
+ if (retval != FW_SUCCESS) {
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ mempool_free(mbp, hw->mb_mempool);
+
+ return 0;
+}
+
+/*
+ * csio_do_reset- Perform the device reset.
+ * @hw: HW module
+ * @fw_rst: FW reset
+ *
+ * If fw_rst is set, issues FW reset mbox cmd otherwise
+ * does PIO reset.
+ * Performs reset of the function.
+ */
+static int
+csio_do_reset(struct csio_hw *hw, bool fw_rst)
+{
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+
+ if (!fw_rst) {
+ /* PIO reset */
+ csio_wr_reg32(hw, PIORSTMODE | PIORST, PL_RST);
+ mdelay(2000);
+ return 0;
+ }
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ csio_mb_reset(hw, mbp, CSIO_MB_DEFAULT_TMO,
+ PIORSTMODE | PIORST, 0, NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of RESET command failed.n");
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ retval = csio_mb_fw_retval(mbp);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "RESET cmd failed with ret:0x%x.\n", retval);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ mempool_free(mbp, hw->mb_mempool);
+
+ return 0;
+}
+
+static int
+csio_hw_validate_caps(struct csio_hw *hw, struct csio_mb *mbp)
+{
+ struct fw_caps_config_cmd *rsp = (struct fw_caps_config_cmd *)mbp->mb;
+ uint16_t caps;
+
+ caps = ntohs(rsp->fcoecaps);
+
+ if (!(caps & FW_CAPS_CONFIG_FCOE_INITIATOR)) {
+ csio_err(hw, "No FCoE Initiator capability in the firmware.\n");
+ return -EINVAL;
+ }
+
+ if (!(caps & FW_CAPS_CONFIG_FCOE_CTRL_OFLD)) {
+ csio_err(hw, "No FCoE Control Offload capability\n");
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+/*
+ * csio_hw_fw_halt - issue a reset/halt to FW and put uP into RESET
+ * @hw: the HW module
+ * @mbox: mailbox to use for the FW RESET command (if desired)
+ * @force: force uP into RESET even if FW RESET command fails
+ *
+ * Issues a RESET command to firmware (if desired) with a HALT indication
+ * and then puts the microprocessor into RESET state. The RESET command
+ * will only be issued if a legitimate mailbox is provided (mbox <=
+ * PCIE_FW_MASTER_MASK).
+ *
+ * This is generally used in order for the host to safely manipulate the
+ * adapter without fear of conflicting with whatever the firmware might
+ * be doing. The only way out of this state is to RESTART the firmware
+ * ...
+ */
+static int
+csio_hw_fw_halt(struct csio_hw *hw, uint32_t mbox, int32_t force)
+{
+ enum fw_retval retval = 0;
+
+ /*
+ * If a legitimate mailbox is provided, issue a RESET command
+ * with a HALT indication.
+ */
+ if (mbox <= PCIE_FW_MASTER_MASK) {
+ struct csio_mb *mbp;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ csio_mb_reset(hw, mbp, CSIO_MB_DEFAULT_TMO,
+ PIORSTMODE | PIORST, FW_RESET_CMD_HALT,
+ NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of RESET command failed!\n");
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ retval = csio_mb_fw_retval(mbp);
+ mempool_free(mbp, hw->mb_mempool);
+ }
+
+ /*
+ * Normally we won't complete the operation if the firmware RESET
+ * command fails but if our caller insists we'll go ahead and put the
+ * uP into RESET. This can be useful if the firmware is hung or even
+ * missing ... We'll have to take the risk of putting the uP into
+ * RESET without the cooperation of firmware in that case.
+ *
+ * We also force the firmware's HALT flag to be on in case we bypassed
+ * the firmware RESET command above or we're dealing with old firmware
+ * which doesn't have the HALT capability. This will serve as a flag
+ * for the incoming firmware to know that it's coming out of a HALT
+ * rather than a RESET ... if it's new enough to understand that ...
+ */
+ if (retval == 0 || force) {
+ csio_set_reg_field(hw, CIM_BOOT_CFG, UPCRST, UPCRST);
+ csio_set_reg_field(hw, PCIE_FW, PCIE_FW_HALT, PCIE_FW_HALT);
+ }
+
+ /*
+ * And we always return the result of the firmware RESET command
+ * even when we force the uP into RESET ...
+ */
+ return retval ? -EINVAL : 0;
+}
+
+/*
+ * csio_hw_fw_restart - restart the firmware by taking the uP out of RESET
+ * @hw: the HW module
+ * @reset: if we want to do a RESET to restart things
+ *
+ * Restart firmware previously halted by csio_hw_fw_halt(). On successful
+ * return the previous PF Master remains as the new PF Master and there
+ * is no need to issue a new HELLO command, etc.
+ *
+ * We do this in two ways:
+ *
+ * 1. If we're dealing with newer firmware we'll simply want to take
+ * the chip's microprocessor out of RESET. This will cause the
+ * firmware to start up from its start vector. And then we'll loop
+ * until the firmware indicates it's started again (PCIE_FW.HALT
+ * reset to 0) or we timeout.
+ *
+ * 2. If we're dealing with older firmware then we'll need to RESET
+ * the chip since older firmware won't recognize the PCIE_FW.HALT
+ * flag and automatically RESET itself on startup.
+ */
+static int
+csio_hw_fw_restart(struct csio_hw *hw, uint32_t mbox, int32_t reset)
+{
+ if (reset) {
+ /*
+ * Since we're directing the RESET instead of the firmware
+ * doing it automatically, we need to clear the PCIE_FW.HALT
+ * bit.
+ */
+ csio_set_reg_field(hw, PCIE_FW, PCIE_FW_HALT, 0);
+
+ /*
+ * If we've been given a valid mailbox, first try to get the
+ * firmware to do the RESET. If that works, great and we can
+ * return success. Otherwise, if we haven't been given a
+ * valid mailbox or the RESET command failed, fall back to
+ * hitting the chip with a hammer.
+ */
+ if (mbox <= PCIE_FW_MASTER_MASK) {
+ csio_set_reg_field(hw, CIM_BOOT_CFG, UPCRST, 0);
+ msleep(100);
+ if (csio_do_reset(hw, true) == 0)
+ return 0;
+ }
+
+ csio_wr_reg32(hw, PIORSTMODE | PIORST, PL_RST);
+ msleep(2000);
+ } else {
+ int ms;
+
+ csio_set_reg_field(hw, CIM_BOOT_CFG, UPCRST, 0);
+ for (ms = 0; ms < FW_CMD_MAX_TIMEOUT; ) {
+ if (!(csio_rd_reg32(hw, PCIE_FW) & PCIE_FW_HALT))
+ return 0;
+ msleep(100);
+ ms += 100;
+ }
+ return -ETIMEDOUT;
+ }
+ return 0;
+}
+
+/*
+ * csio_hw_fw_upgrade - perform all of the steps necessary to upgrade FW
+ * @hw: the HW module
+ * @mbox: mailbox to use for the FW RESET command (if desired)
+ * @fw_data: the firmware image to write
+ * @size: image size
+ * @force: force upgrade even if firmware doesn't cooperate
+ *
+ * Perform all of the steps necessary for upgrading an adapter's
+ * firmware image. Normally this requires the cooperation of the
+ * existing firmware in order to halt all existing activities
+ * but if an invalid mailbox token is passed in we skip that step
+ * (though we'll still put the adapter microprocessor into RESET in
+ * that case).
+ *
+ * On successful return the new firmware will have been loaded and
+ * the adapter will have been fully RESET losing all previous setup
+ * state. On unsuccessful return the adapter may be completely hosed ...
+ * positive errno indicates that the adapter is ~probably~ intact, a
+ * negative errno indicates that things are looking bad ...
+ */
+static int
+csio_hw_fw_upgrade(struct csio_hw *hw, uint32_t mbox,
+ const u8 *fw_data, uint32_t size, int32_t force)
+{
+ const struct fw_hdr *fw_hdr = (const struct fw_hdr *)fw_data;
+ int reset, ret;
+
+ ret = csio_hw_fw_halt(hw, mbox, force);
+ if (ret != 0 && !force)
+ return ret;
+
+ ret = csio_hw_fw_dload(hw, (uint8_t *) fw_data, size);
+ if (ret != 0)
+ return ret;
+
+ /*
+ * Older versions of the firmware don't understand the new
+ * PCIE_FW.HALT flag and so won't know to perform a RESET when they
+ * restart. So for newly loaded older firmware we'll have to do the
+ * RESET for it so it starts up on a clean slate. We can tell if
+ * the newly loaded firmware will handle this right by checking
+ * its header flags to see if it advertises the capability.
+ */
+ reset = ((ntohl(fw_hdr->flags) & FW_HDR_FLAGS_RESET_HALT) == 0);
+ return csio_hw_fw_restart(hw, mbox, reset);
+}
+
+
+/*
+ * csio_hw_fw_config_file - setup an adapter via a Configuration File
+ * @hw: the HW module
+ * @mbox: mailbox to use for the FW command
+ * @mtype: the memory type where the Configuration File is located
+ * @maddr: the memory address where the Configuration File is located
+ * @finiver: return value for CF [fini] version
+ * @finicsum: return value for CF [fini] checksum
+ * @cfcsum: return value for CF computed checksum
+ *
+ * Issue a command to get the firmware to process the Configuration
+ * File located at the specified mtype/maddress. If the Configuration
+ * File is processed successfully and return value pointers are
+ * provided, the Configuration File "[fini] section version and
+ * checksum values will be returned along with the computed checksum.
+ * It's up to the caller to decide how it wants to respond to the
+ * checksums not matching but it recommended that a prominant warning
+ * be emitted in order to help people rapidly identify changed or
+ * corrupted Configuration Files.
+ *
+ * Also note that it's possible to modify things like "niccaps",
+ * "toecaps",etc. between processing the Configuration File and telling
+ * the firmware to use the new configuration. Callers which want to
+ * do this will need to "hand-roll" their own CAPS_CONFIGS commands for
+ * Configuration Files if they want to do this.
+ */
+static int
+csio_hw_fw_config_file(struct csio_hw *hw,
+ unsigned int mtype, unsigned int maddr,
+ uint32_t *finiver, uint32_t *finicsum, uint32_t *cfcsum)
+{
+ struct csio_mb *mbp;
+ struct fw_caps_config_cmd *caps_cmd;
+ int rv = -EINVAL;
+ enum fw_retval ret;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+ /*
+ * Tell the firmware to process the indicated Configuration File.
+ * If there are no errors and the caller has provided return value
+ * pointers for the [fini] section version, checksum and computed
+ * checksum, pass those back to the caller.
+ */
+ caps_cmd = (struct fw_caps_config_cmd *)(mbp->mb);
+ CSIO_INIT_MBP(mbp, caps_cmd, CSIO_MB_DEFAULT_TMO, hw, NULL, 1);
+ caps_cmd->op_to_write =
+ htonl(FW_CMD_OP(FW_CAPS_CONFIG_CMD) |
+ FW_CMD_REQUEST |
+ FW_CMD_READ);
+ caps_cmd->cfvalid_to_len16 =
+ htonl(FW_CAPS_CONFIG_CMD_CFVALID |
+ FW_CAPS_CONFIG_CMD_MEMTYPE_CF(mtype) |
+ FW_CAPS_CONFIG_CMD_MEMADDR64K_CF(maddr >> 16) |
+ FW_LEN16(*caps_cmd));
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of FW_CAPS_CONFIG_CMD failed!\n");
+ goto out;
+ }
+
+ ret = csio_mb_fw_retval(mbp);
+ if (ret != FW_SUCCESS) {
+ csio_dbg(hw, "FW_CAPS_CONFIG_CMD returned %d!\n", rv);
+ goto out;
+ }
+
+ if (finiver)
+ *finiver = ntohl(caps_cmd->finiver);
+ if (finicsum)
+ *finicsum = ntohl(caps_cmd->finicsum);
+ if (cfcsum)
+ *cfcsum = ntohl(caps_cmd->cfcsum);
+
+ /* Validate device capabilities */
+ if (csio_hw_validate_caps(hw, mbp)) {
+ rv = -ENOENT;
+ goto out;
+ }
+
+ /*
+ * And now tell the firmware to use the configuration we just loaded.
+ */
+ caps_cmd->op_to_write =
+ htonl(FW_CMD_OP(FW_CAPS_CONFIG_CMD) |
+ FW_CMD_REQUEST |
+ FW_CMD_WRITE);
+ caps_cmd->cfvalid_to_len16 = htonl(FW_LEN16(*caps_cmd));
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of FW_CAPS_CONFIG_CMD failed!\n");
+ goto out;
+ }
+
+ ret = csio_mb_fw_retval(mbp);
+ if (ret != FW_SUCCESS) {
+ csio_dbg(hw, "FW_CAPS_CONFIG_CMD returned %d!\n", rv);
+ goto out;
+ }
+
+ rv = 0;
+out:
+ mempool_free(mbp, hw->mb_mempool);
+ return rv;
+}
+
+/*
+ * csio_get_device_params - Get device parameters.
+ * @hw: HW module
+ *
+ */
+static int
+csio_get_device_params(struct csio_hw *hw)
+{
+ struct csio_wrm *wrm = csio_hw_to_wrm(hw);
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+ u32 param[6];
+ int i, j = 0;
+
+ /* Initialize portids to -1 */
+ for (i = 0; i < CSIO_MAX_PPORTS; i++)
+ hw->pport[i].portid = -1;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ /* Get port vec information. */
+ param[0] = FW_PARAM_DEV(PORTVEC);
+
+ /* Get Core clock. */
+ param[1] = FW_PARAM_DEV(CCLK);
+
+ /* Get EQ id start and end. */
+ param[2] = FW_PARAM_PFVF(EQ_START);
+ param[3] = FW_PARAM_PFVF(EQ_END);
+
+ /* Get IQ id start and end. */
+ param[4] = FW_PARAM_PFVF(IQFLINT_START);
+ param[5] = FW_PARAM_PFVF(IQFLINT_END);
+
+ csio_mb_params(hw, mbp, CSIO_MB_DEFAULT_TMO, hw->pfn, 0,
+ ARRAY_SIZE(param), param, NULL, false, NULL);
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of FW_PARAMS_CMD(read) failed!\n");
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ csio_mb_process_read_params_rsp(hw, mbp, &retval,
+ ARRAY_SIZE(param), param);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_PARAMS_CMD(read) failed with ret:0x%x!\n",
+ retval);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ /* cache the information. */
+ hw->port_vec = param[0];
+ hw->vpd.cclk = param[1];
+ wrm->fw_eq_start = param[2];
+ wrm->fw_iq_start = param[4];
+
+ /* Using FW configured max iqs & eqs */
+ if ((hw->flags & CSIO_HWF_USING_SOFT_PARAMS) ||
+ !csio_is_hw_master(hw)) {
+ hw->cfg_niq = param[5] - param[4] + 1;
+ hw->cfg_neq = param[3] - param[2] + 1;
+ csio_dbg(hw, "Using fwconfig max niqs %d neqs %d\n",
+ hw->cfg_niq, hw->cfg_neq);
+ }
+
+ hw->port_vec &= csio_port_mask;
+
+ hw->num_pports = hweight32(hw->port_vec);
+
+ csio_dbg(hw, "Port vector: 0x%x, #ports: %d\n",
+ hw->port_vec, hw->num_pports);
+
+ for (i = 0; i < hw->num_pports; i++) {
+ while ((hw->port_vec & (1 << j)) == 0)
+ j++;
+ hw->pport[i].portid = j++;
+ csio_dbg(hw, "Found Port:%d\n", hw->pport[i].portid);
+ }
+ mempool_free(mbp, hw->mb_mempool);
+
+ return 0;
+}
+
+
+/*
+ * csio_config_device_caps - Get and set device capabilities.
+ * @hw: HW module
+ *
+ */
+static int
+csio_config_device_caps(struct csio_hw *hw)
+{
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+ int rv = -EINVAL;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ /* Get device capabilities */
+ csio_mb_caps_config(hw, mbp, CSIO_MB_DEFAULT_TMO, 0, 0, 0, 0, NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of FW_CAPS_CONFIG_CMD(r) failed!\n");
+ goto out;
+ }
+
+ retval = csio_mb_fw_retval(mbp);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_CAPS_CONFIG_CMD(r) returned %d!\n", retval);
+ goto out;
+ }
+
+ /* Validate device capabilities */
+ if (csio_hw_validate_caps(hw, mbp))
+ goto out;
+
+ /* Don't config device capabilities if already configured */
+ if (hw->fw_state == CSIO_DEV_STATE_INIT) {
+ rv = 0;
+ goto out;
+ }
+
+ /* Write back desired device capabilities */
+ csio_mb_caps_config(hw, mbp, CSIO_MB_DEFAULT_TMO, true, true,
+ false, true, NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of FW_CAPS_CONFIG_CMD(w) failed!\n");
+ goto out;
+ }
+
+ retval = csio_mb_fw_retval(mbp);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_CAPS_CONFIG_CMD(w) returned %d!\n", retval);
+ goto out;
+ }
+
+ rv = 0;
+out:
+ mempool_free(mbp, hw->mb_mempool);
+ return rv;
+}
+
+static int
+csio_config_global_rss(struct csio_hw *hw)
+{
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ csio_rss_glb_config(hw, mbp, CSIO_MB_DEFAULT_TMO,
+ FW_RSS_GLB_CONFIG_CMD_MODE_BASICVIRTUAL,
+ FW_RSS_GLB_CONFIG_CMD_TNLMAPEN |
+ FW_RSS_GLB_CONFIG_CMD_HASHTOEPLITZ |
+ FW_RSS_GLB_CONFIG_CMD_TNLALLLKP,
+ NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of FW_RSS_GLB_CONFIG_CMD failed!\n");
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ retval = csio_mb_fw_retval(mbp);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_RSS_GLB_CONFIG_CMD returned 0x%x!\n", retval);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ mempool_free(mbp, hw->mb_mempool);
+
+ return 0;
+}
+
+/*
+ * csio_config_pfvf - Configure Physical/Virtual functions settings.
+ * @hw: HW module
+ *
+ */
+static int
+csio_config_pfvf(struct csio_hw *hw)
+{
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ /*
+ * For now, allow all PFs to access to all ports using a pmask
+ * value of 0xF (M_FW_PFVF_CMD_PMASK). Once we have VFs, we will
+ * need to provide access based on some rule.
+ */
+ csio_mb_pfvf(hw, mbp, CSIO_MB_DEFAULT_TMO, hw->pfn, 0, CSIO_NEQ,
+ CSIO_NETH_CTRL, CSIO_NIQ_FLINT, 0, 0, CSIO_NVI, CSIO_CMASK,
+ CSIO_PMASK, CSIO_NEXACTF, CSIO_R_CAPS, CSIO_WX_CAPS, NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of FW_PFVF_CMD failed!\n");
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ retval = csio_mb_fw_retval(mbp);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_PFVF_CMD returned 0x%x!\n", retval);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ mempool_free(mbp, hw->mb_mempool);
+
+ return 0;
+}
+
+/*
+ * csio_enable_ports - Bring up all available ports.
+ * @hw: HW module.
+ *
+ */
+static int
+csio_enable_ports(struct csio_hw *hw)
+{
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+ uint8_t portid;
+ int i;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ for (i = 0; i < hw->num_pports; i++) {
+ portid = hw->pport[i].portid;
+
+ /* Read PORT information */
+ csio_mb_port(hw, mbp, CSIO_MB_DEFAULT_TMO, portid,
+ false, 0, 0, NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "failed to issue FW_PORT_CMD(r) port:%d\n",
+ portid);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ csio_mb_process_read_port_rsp(hw, mbp, &retval,
+ &hw->pport[i].pcap);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_PORT_CMD(r) port:%d failed: 0x%x\n",
+ portid, retval);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ /* Write back PORT information */
+ csio_mb_port(hw, mbp, CSIO_MB_DEFAULT_TMO, portid, true,
+ (PAUSE_RX | PAUSE_TX), hw->pport[i].pcap, NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "failed to issue FW_PORT_CMD(w) port:%d\n",
+ portid);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ retval = csio_mb_fw_retval(mbp);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_PORT_CMD(w) port:%d failed :0x%x\n",
+ portid, retval);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ } /* For all ports */
+
+ mempool_free(mbp, hw->mb_mempool);
+
+ return 0;
+}
+
+/*
+ * csio_get_fcoe_resinfo - Read fcoe fw resource info.
+ * @hw: HW module
+ * Issued with lock held.
+ */
+static int
+csio_get_fcoe_resinfo(struct csio_hw *hw)
+{
+ struct csio_fcoe_res_info *res_info = &hw->fres_info;
+ struct fw_fcoe_res_info_cmd *rsp;
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ /* Get FCoE FW resource information */
+ csio_fcoe_read_res_info_init_mb(hw, mbp, CSIO_MB_DEFAULT_TMO, NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "failed to issue FW_FCOE_RES_INFO_CMD\n");
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ rsp = (struct fw_fcoe_res_info_cmd *)(mbp->mb);
+ retval = FW_CMD_RETVAL_GET(ntohl(rsp->retval_len16));
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_FCOE_RES_INFO_CMD failed with ret x%x\n",
+ retval);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ res_info->e_d_tov = ntohs(rsp->e_d_tov);
+ res_info->r_a_tov_seq = ntohs(rsp->r_a_tov_seq);
+ res_info->r_a_tov_els = ntohs(rsp->r_a_tov_els);
+ res_info->r_r_tov = ntohs(rsp->r_r_tov);
+ res_info->max_xchgs = ntohl(rsp->max_xchgs);
+ res_info->max_ssns = ntohl(rsp->max_ssns);
+ res_info->used_xchgs = ntohl(rsp->used_xchgs);
+ res_info->used_ssns = ntohl(rsp->used_ssns);
+ res_info->max_fcfs = ntohl(rsp->max_fcfs);
+ res_info->max_vnps = ntohl(rsp->max_vnps);
+ res_info->used_fcfs = ntohl(rsp->used_fcfs);
+ res_info->used_vnps = ntohl(rsp->used_vnps);
+
+ csio_dbg(hw, "max ssns:%d max xchgs:%d\n", res_info->max_ssns,
+ res_info->max_xchgs);
+ mempool_free(mbp, hw->mb_mempool);
+
+ return 0;
+}
+
+static int
+csio_hw_check_fwconfig(struct csio_hw *hw, u32 *param)
+{
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+ u32 _param[1];
+
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp) {
+ CSIO_INC_STATS(hw, n_err_nomem);
+ return -ENOMEM;
+ }
+
+ /*
+ * Find out whether we're dealing with a version of
+ * the firmware which has configuration file support.
+ */
+ _param[0] = (FW_PARAMS_MNEM(FW_PARAMS_MNEM_DEV) |
+ FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_DEV_CF));
+
+ csio_mb_params(hw, mbp, CSIO_MB_DEFAULT_TMO, hw->pfn, 0,
+ ARRAY_SIZE(_param), _param, NULL, false, NULL);
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of FW_PARAMS_CMD(read) failed!\n");
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ csio_mb_process_read_params_rsp(hw, mbp, &retval,
+ ARRAY_SIZE(_param), _param);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_PARAMS_CMD(read) failed with ret:0x%x!\n",
+ retval);
+ mempool_free(mbp, hw->mb_mempool);
+ return -EINVAL;
+ }
+
+ mempool_free(mbp, hw->mb_mempool);
+ *param = _param[0];
+
+ return 0;
+}
+
+static int
+csio_hw_flash_config(struct csio_hw *hw, u32 *fw_cfg_param, char *path)
+{
+ int ret = 0;
+ const struct firmware *cf;
+ struct pci_dev *pci_dev = hw->pdev;
+ struct device *dev = &pci_dev->dev;
+
+ unsigned int mtype = 0, maddr = 0;
+ uint32_t *cfg_data;
+ int value_to_add = 0;
+
+ if (request_firmware(&cf, CSIO_CF_FNAME, dev) < 0) {
+ csio_err(hw, "could not find config file " CSIO_CF_FNAME
+ ",err: %d\n", ret);
+ return -ENOENT;
+ }
+
+ if (cf->size%4 != 0)
+ value_to_add = 4 - (cf->size % 4);
+
+ cfg_data = kzalloc(cf->size+value_to_add, GFP_KERNEL);
+ if (cfg_data == NULL)
+ return -ENOMEM;
+
+ memcpy((void *)cfg_data, (const void *)cf->data, cf->size);
+
+ if (csio_hw_check_fwconfig(hw, fw_cfg_param) != 0)
+ return -EINVAL;
+
+ mtype = FW_PARAMS_PARAM_Y_GET(*fw_cfg_param);
+ maddr = FW_PARAMS_PARAM_Z_GET(*fw_cfg_param) << 16;
+
+ ret = csio_memory_write(hw, mtype, maddr,
+ cf->size + value_to_add, cfg_data);
+ if (ret == 0) {
+ csio_info(hw, "config file upgraded to " CSIO_CF_FNAME "\n");
+ strncpy(path, "/lib/firmware/" CSIO_CF_FNAME, 64);
+ }
+
+ kfree(cfg_data);
+ release_firmware(cf);
+
+ return ret;
+}
+
+/*
+ * HW initialization: contact FW, obtain config, perform basic init.
+ *
+ * If the firmware we're dealing with has Configuration File support, then
+ * we use that to perform all configuration -- either using the configuration
+ * file stored in flash on the adapter or using a filesystem-local file
+ * if available.
+ *
+ * If we don't have configuration file support in the firmware, then we'll
+ * have to set things up the old fashioned way with hard-coded register
+ * writes and firmware commands ...
+ */
+
+/*
+ * Attempt to initialize the HW via a Firmware Configuration File.
+ */
+static int
+csio_hw_use_fwconfig(struct csio_hw *hw, int reset, u32 *fw_cfg_param)
+{
+ unsigned int mtype, maddr;
+ int rv;
+ uint32_t finiver, finicsum, cfcsum;
+ int using_flash;
+ char path[64];
+
+ /*
+ * Reset device if necessary
+ */
+ if (reset) {
+ rv = csio_do_reset(hw, true);
+ if (rv != 0)
+ goto bye;
+ }
+
+ /*
+ * If we have a configuration file in host ,
+ * then use that. Otherwise, use the configuration file stored
+ * in the HW flash ...
+ */
+ spin_unlock_irq(&hw->lock);
+ rv = csio_hw_flash_config(hw, fw_cfg_param, path);
+ spin_lock_irq(&hw->lock);
+ if (rv != 0) {
+ if (rv == -ENOENT) {
+ /*
+ * config file was not found. Use default
+ * config file from flash.
+ */
+ mtype = FW_MEMTYPE_CF_FLASH;
+ maddr = csio_hw_flash_cfg_addr(hw);
+ using_flash = 1;
+ } else {
+ /*
+ * we revert back to the hardwired config if
+ * flashing failed.
+ */
+ goto bye;
+ }
+ } else {
+ mtype = FW_PARAMS_PARAM_Y_GET(*fw_cfg_param);
+ maddr = FW_PARAMS_PARAM_Z_GET(*fw_cfg_param) << 16;
+ using_flash = 0;
+ }
+
+ hw->cfg_store = (uint8_t)mtype;
+
+ /*
+ * Issue a Capability Configuration command to the firmware to get it
+ * to parse the Configuration File.
+ */
+ rv = csio_hw_fw_config_file(hw, mtype, maddr, &finiver,
+ &finicsum, &cfcsum);
+ if (rv != 0)
+ goto bye;
+
+ hw->cfg_finiver = finiver;
+ hw->cfg_finicsum = finicsum;
+ hw->cfg_cfcsum = cfcsum;
+ hw->cfg_csum_status = true;
+
+ if (finicsum != cfcsum) {
+ csio_warn(hw,
+ "Config File checksum mismatch: csum=%#x, computed=%#x\n",
+ finicsum, cfcsum);
+
+ hw->cfg_csum_status = false;
+ }
+
+ /*
+ * Note that we're operating with parameters
+ * not supplied by the driver, rather than from hard-wired
+ * initialization constants buried in the driver.
+ */
+ hw->flags |= CSIO_HWF_USING_SOFT_PARAMS;
+
+ /* device parameters */
+ rv = csio_get_device_params(hw);
+ if (rv != 0)
+ goto bye;
+
+ /* Configure SGE */
+ csio_wr_sge_init(hw);
+
+ /*
+ * And finally tell the firmware to initialize itself using the
+ * parameters from the Configuration File.
+ */
+ /* Post event to notify completion of configuration */
+ csio_post_event(&hw->sm, CSIO_HWE_INIT);
+
+ csio_info(hw,
+ "Firmware Configuration File %s, version %#x, computed checksum %#x\n",
+ (using_flash ? "in device FLASH" : path), finiver, cfcsum);
+
+ return 0;
+
+ /*
+ * Something bad happened. Return the error ...
+ */
+bye:
+ hw->flags &= ~CSIO_HWF_USING_SOFT_PARAMS;
+ csio_dbg(hw, "Configuration file error %d\n", rv);
+ return rv;
+}
+
+/*
+ * Attempt to initialize the adapter via hard-coded, driver supplied
+ * parameters ...
+ */
+static int
+csio_hw_no_fwconfig(struct csio_hw *hw, int reset)
+{
+ int rv;
+ /*
+ * Reset device if necessary
+ */
+ if (reset) {
+ rv = csio_do_reset(hw, true);
+ if (rv != 0)
+ goto out;
+ }
+
+ /* Get and set device capabilities */
+ rv = csio_config_device_caps(hw);
+ if (rv != 0)
+ goto out;
+
+ /* Config Global RSS command */
+ rv = csio_config_global_rss(hw);
+ if (rv != 0)
+ goto out;
+
+ /* Configure PF/VF capabilities of device */
+ rv = csio_config_pfvf(hw);
+ if (rv != 0)
+ goto out;
+
+ /* device parameters */
+ rv = csio_get_device_params(hw);
+ if (rv != 0)
+ goto out;
+
+ /* Configure SGE */
+ csio_wr_sge_init(hw);
+
+ /* Post event to notify completion of configuration */
+ csio_post_event(&hw->sm, CSIO_HWE_INIT);
+
+out:
+ return rv;
+}
+
+/*
+ * Returns -EINVAL if attempts to flash the firmware failed
+ * else returns 0,
+ * if flashing was not attempted because the card had the
+ * latest firmware ECANCELED is returned
+ */
+static int
+csio_hw_flash_fw(struct csio_hw *hw)
+{
+ int ret = -ECANCELED;
+ const struct firmware *fw;
+ const struct fw_hdr *hdr;
+ u32 fw_ver;
+ struct pci_dev *pci_dev = hw->pdev;
+ struct device *dev = &pci_dev->dev ;
+
+ if (request_firmware(&fw, CSIO_FW_FNAME, dev) < 0) {
+ csio_err(hw, "could not find firmware image " CSIO_FW_FNAME
+ ",err: %d\n", ret);
+ return -EINVAL;
+ }
+
+ hdr = (const struct fw_hdr *)fw->data;
+ fw_ver = ntohl(hdr->fw_ver);
+ if (FW_HDR_FW_VER_MAJOR_GET(fw_ver) != FW_VERSION_MAJOR)
+ return -EINVAL; /* wrong major version, won't do */
+
+ /*
+ * If the flash FW is unusable or we found something newer, load it.
+ */
+ if (FW_HDR_FW_VER_MAJOR_GET(hw->fwrev) != FW_VERSION_MAJOR ||
+ fw_ver > hw->fwrev) {
+ ret = csio_hw_fw_upgrade(hw, hw->pfn, fw->data, fw->size,
+ /*force=*/false);
+ if (!ret)
+ csio_info(hw, "firmware upgraded to version %pI4 from "
+ CSIO_FW_FNAME "\n", &hdr->fw_ver);
+ else
+ csio_err(hw, "firmware upgrade failed! err=%d\n", ret);
+ }
+
+ release_firmware(fw);
+
+ return ret;
+}
+
+
+/*
+ * csio_hw_configure - Configure HW
+ * @hw - HW module
+ *
+ */
+static void
+csio_hw_configure(struct csio_hw *hw)
+{
+ int reset = 1;
+ int rv;
+ u32 param[1];
+
+ rv = csio_hw_dev_ready(hw);
+ if (rv != 0) {
+ CSIO_INC_STATS(hw, n_err_fatal);
+ csio_post_event(&hw->sm, CSIO_HWE_FATAL);
+ goto out;
+ }
+
+ /* HW version */
+ hw->chip_ver = (char)csio_rd_reg32(hw, PL_REV);
+
+ /* Needed for FW download */
+ rv = csio_hw_get_flash_params(hw);
+ if (rv != 0) {
+ csio_err(hw, "Failed to get serial flash params rv:%d\n", rv);
+ csio_post_event(&hw->sm, CSIO_HWE_FATAL);
+ goto out;
+ }
+
+ /* Set pci completion timeout value to 4 seconds. */
+ csio_set_pcie_completion_timeout(hw, 0xd);
+
+ csio_hw_set_mem_win(hw);
+
+ rv = csio_hw_get_fw_version(hw, &hw->fwrev);
+ if (rv != 0)
+ goto out;
+
+ csio_hw_print_fw_version(hw, "Firmware revision");
+
+ rv = csio_do_hello(hw, &hw->fw_state);
+ if (rv != 0) {
+ CSIO_INC_STATS(hw, n_err_fatal);
+ csio_post_event(&hw->sm, CSIO_HWE_FATAL);
+ goto out;
+ }
+
+ /* Read vpd */
+ rv = csio_hw_get_vpd_params(hw, &hw->vpd);
+ if (rv != 0)
+ goto out;
+
+ if (csio_is_hw_master(hw) && hw->fw_state != CSIO_DEV_STATE_INIT) {
+ rv = csio_hw_check_fw_version(hw);
+ if (rv == -EINVAL) {
+
+ /* Do firmware update */
+ spin_unlock_irq(&hw->lock);
+ rv = csio_hw_flash_fw(hw);
+ spin_lock_irq(&hw->lock);
+
+ if (rv == 0) {
+ reset = 0;
+ /*
+ * Note that the chip was reset as part of the
+ * firmware upgrade so we don't reset it again
+ * below and grab the new firmware version.
+ */
+ rv = csio_hw_check_fw_version(hw);
+ }
+ }
+ /*
+ * If the firmware doesn't support Configuration
+ * Files, use the old Driver-based, hard-wired
+ * initialization. Otherwise, try using the
+ * Configuration File support and fall back to the
+ * Driver-based initialization if there's no
+ * Configuration File found.
+ */
+ if (csio_hw_check_fwconfig(hw, param) == 0) {
+ rv = csio_hw_use_fwconfig(hw, reset, param);
+ if (rv == -ENOENT)
+ goto out;
+ if (rv != 0) {
+ csio_info(hw,
+ "No Configuration File present "
+ "on adapter. Using hard-wired "
+ "configuration parameters.\n");
+ rv = csio_hw_no_fwconfig(hw, reset);
+ }
+ } else {
+ rv = csio_hw_no_fwconfig(hw, reset);
+ }
+
+ if (rv != 0)
+ goto out;
+
+ } else {
+ if (hw->fw_state == CSIO_DEV_STATE_INIT) {
+
+ /* device parameters */
+ rv = csio_get_device_params(hw);
+ if (rv != 0)
+ goto out;
+
+ /* Get device capabilities */
+ rv = csio_config_device_caps(hw);
+ if (rv != 0)
+ goto out;
+
+ /* Configure SGE */
+ csio_wr_sge_init(hw);
+
+ /* Post event to notify completion of configuration */
+ csio_post_event(&hw->sm, CSIO_HWE_INIT);
+ goto out;
+ }
+ } /* if not master */
+
+out:
+ return;
+}
+
+/*
+ * csio_hw_initialize - Initialize HW
+ * @hw - HW module
+ *
+ */
+static void
+csio_hw_initialize(struct csio_hw *hw)
+{
+ struct csio_mb *mbp;
+ enum fw_retval retval;
+ int rv;
+ int i;
+
+ if (csio_is_hw_master(hw) && hw->fw_state != CSIO_DEV_STATE_INIT) {
+ mbp = mempool_alloc(hw->mb_mempool, GFP_ATOMIC);
+ if (!mbp)
+ goto out;
+
+ csio_mb_initialize(hw, mbp, CSIO_MB_DEFAULT_TMO, NULL);
+
+ if (csio_mb_issue(hw, mbp)) {
+ csio_err(hw, "Issue of FW_INITIALIZE_CMD failed!\n");
+ goto free_and_out;
+ }
+
+ retval = csio_mb_fw_retval(mbp);
+ if (retval != FW_SUCCESS) {
+ csio_err(hw, "FW_INITIALIZE_CMD returned 0x%x!\n",
+ retval);
+ goto free_and_out;
+ }
+
+ mempool_free(mbp, hw->mb_mempool);
+ }
+
+ rv = csio_get_fcoe_resinfo(hw);
+ if (rv != 0) {
+ csio_err(hw, "Failed to read fcoe resource info: %d\n", rv);
+ goto out;
+ }
+
+ spin_unlock_irq(&hw->lock);
+ rv = csio_config_queues(hw);
+ spin_lock_irq(&hw->lock);
+
+ if (rv != 0) {
+ csio_err(hw, "Config of queues failed!: %d\n", rv);
+ goto out;
+ }
+
+ for (i = 0; i < hw->num_pports; i++)
+ hw->pport[i].mod_type = FW_PORT_MOD_TYPE_NA;
+
+ if (csio_is_hw_master(hw) && hw->fw_state != CSIO_DEV_STATE_INIT) {
+ rv = csio_enable_ports(hw);
+ if (rv != 0) {
+ csio_err(hw, "Failed to enable ports: %d\n", rv);
+ goto out;
+ }
+ }
+
+ csio_post_event(&hw->sm, CSIO_HWE_INIT_DONE);
+ return;
+
+free_and_out:
+ mempool_free(mbp, hw->mb_mempool);
+out:
+ return;
+}
+
+#define PF_INTR_MASK (PFSW | PFCIM)
+
+/*
+ * csio_hw_intr_enable - Enable HW interrupts
+ * @hw: Pointer to HW module.
+ *
+ * Enable interrupts in HW registers.
+ */
+static void
+csio_hw_intr_enable(struct csio_hw *hw)
+{
+ uint16_t vec = (uint16_t)csio_get_mb_intr_idx(csio_hw_to_mbm(hw));
+ uint32_t pf = SOURCEPF_GET(csio_rd_reg32(hw, PL_WHOAMI));
+ uint32_t pl = csio_rd_reg32(hw, PL_INT_ENABLE);
+
+ /*
+ * Set aivec for MSI/MSIX. PCIE_PF_CFG.INTXType is set up
+ * by FW, so do nothing for INTX.
+ */
+ if (hw->intr_mode == CSIO_IM_MSIX)
+ csio_set_reg_field(hw, MYPF_REG(PCIE_PF_CFG),
+ AIVEC(AIVEC_MASK), vec);
+ else if (hw->intr_mode == CSIO_IM_MSI)
+ csio_set_reg_field(hw, MYPF_REG(PCIE_PF_CFG),
+ AIVEC(AIVEC_MASK), 0);
+
+ csio_wr_reg32(hw, PF_INTR_MASK, MYPF_REG(PL_PF_INT_ENABLE));
+
+ /* Turn on MB interrupts - this will internally flush PIO as well */
+ csio_mb_intr_enable(hw);
+
+ /* These are common registers - only a master can modify them */
+ if (csio_is_hw_master(hw)) {
+ /*
+ * Disable the Serial FLASH interrupt, if enabled!
+ */
+ pl &= (~SF);
+ csio_wr_reg32(hw, pl, PL_INT_ENABLE);
+
+ csio_wr_reg32(hw, ERR_CPL_EXCEED_IQE_SIZE |
+ EGRESS_SIZE_ERR | ERR_INVALID_CIDX_INC |
+ ERR_CPL_OPCODE_0 | ERR_DROPPED_DB |
+ ERR_DATA_CPL_ON_HIGH_QID1 |
+ ERR_DATA_CPL_ON_HIGH_QID0 | ERR_BAD_DB_PIDX3 |
+ ERR_BAD_DB_PIDX2 | ERR_BAD_DB_PIDX1 |
+ ERR_BAD_DB_PIDX0 | ERR_ING_CTXT_PRIO |
+ ERR_EGR_CTXT_PRIO | INGRESS_SIZE_ERR,
+ SGE_INT_ENABLE3);
+ csio_set_reg_field(hw, PL_INT_MAP0, 0, 1 << pf);
+ }
+
+ hw->flags |= CSIO_HWF_HW_INTR_ENABLED;
+
+}
+
+/*
+ * csio_hw_intr_disable - Disable HW interrupts
+ * @hw: Pointer to HW module.
+ *
+ * Turn off Mailbox and PCI_PF_CFG interrupts.
+ */
+void
+csio_hw_intr_disable(struct csio_hw *hw)
+{
+ uint32_t pf = SOURCEPF_GET(csio_rd_reg32(hw, PL_WHOAMI));
+
+ if (!(hw->flags & CSIO_HWF_HW_INTR_ENABLED))
+ return;
+
+ hw->flags &= ~CSIO_HWF_HW_INTR_ENABLED;
+
+ csio_wr_reg32(hw, 0, MYPF_REG(PL_PF_INT_ENABLE));
+ if (csio_is_hw_master(hw))
+ csio_set_reg_field(hw, PL_INT_MAP0, 1 << pf, 0);
+
+ /* Turn off MB interrupts */
+ csio_mb_intr_disable(hw);
+
+}
+
+static void
+csio_hw_fatal_err(struct csio_hw *hw)
+{
+ csio_set_reg_field(hw, SGE_CONTROL, GLOBALENABLE, 0);
+ csio_hw_intr_disable(hw);
+
+ /* Do not reset HW, we may need FW state for debugging */
+ csio_fatal(hw, "HW Fatal error encountered!\n");
+}
+
+/*****************************************************************************/
+/* START: HW SM */
+/*****************************************************************************/
+/*
+ * csio_hws_uninit - Uninit state
+ * @hw - HW module
+ * @evt - Event
+ *
+ */
+static void
+csio_hws_uninit(struct csio_hw *hw, enum csio_hw_ev evt)
+{
+ hw->prev_evt = hw->cur_evt;
+ hw->cur_evt = evt;
+ CSIO_INC_STATS(hw, n_evt_sm[evt]);
+
+ switch (evt) {
+ case CSIO_HWE_CFG:
+ csio_set_state(&hw->sm, csio_hws_configuring);
+ csio_hw_configure(hw);
+ break;
+
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+ }
+}
+
+/*
+ * csio_hws_configuring - Configuring state
+ * @hw - HW module
+ * @evt - Event
+ *
+ */
+static void
+csio_hws_configuring(struct csio_hw *hw, enum csio_hw_ev evt)
+{
+ hw->prev_evt = hw->cur_evt;
+ hw->cur_evt = evt;
+ CSIO_INC_STATS(hw, n_evt_sm[evt]);
+
+ switch (evt) {
+ case CSIO_HWE_INIT:
+ csio_set_state(&hw->sm, csio_hws_initializing);
+ csio_hw_initialize(hw);
+ break;
+
+ case CSIO_HWE_INIT_DONE:
+ csio_set_state(&hw->sm, csio_hws_ready);
+ /* Fan out event to all lnode SMs */
+ csio_notify_lnodes(hw, CSIO_LN_NOTIFY_HWREADY);
+ break;
+
+ case CSIO_HWE_FATAL:
+ csio_set_state(&hw->sm, csio_hws_uninit);
+ break;
+
+ case CSIO_HWE_PCI_REMOVE:
+ csio_do_bye(hw);
+ break;
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+ }
+}
+
+/*
+ * csio_hws_initializing - Initialiazing state
+ * @hw - HW module
+ * @evt - Event
+ *
+ */
+static void
+csio_hws_initializing(struct csio_hw *hw, enum csio_hw_ev evt)
+{
+ hw->prev_evt = hw->cur_evt;
+ hw->cur_evt = evt;
+ CSIO_INC_STATS(hw, n_evt_sm[evt]);
+
+ switch (evt) {
+ case CSIO_HWE_INIT_DONE:
+ csio_set_state(&hw->sm, csio_hws_ready);
+
+ /* Fan out event to all lnode SMs */
+ csio_notify_lnodes(hw, CSIO_LN_NOTIFY_HWREADY);
+
+ /* Enable interrupts */
+ csio_hw_intr_enable(hw);
+ break;
+
+ case CSIO_HWE_FATAL:
+ csio_set_state(&hw->sm, csio_hws_uninit);
+ break;
+
+ case CSIO_HWE_PCI_REMOVE:
+ csio_do_bye(hw);
+ break;
+
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+ }
+}
+
+/*
+ * csio_hws_ready - Ready state
+ * @hw - HW module
+ * @evt - Event
+ *
+ */
+static void
+csio_hws_ready(struct csio_hw *hw, enum csio_hw_ev evt)
+{
+ /* Remember the event */
+ hw->evtflag = evt;
+
+ hw->prev_evt = hw->cur_evt;
+ hw->cur_evt = evt;
+ CSIO_INC_STATS(hw, n_evt_sm[evt]);
+
+ switch (evt) {
+ case CSIO_HWE_HBA_RESET:
+ case CSIO_HWE_FW_DLOAD:
+ case CSIO_HWE_SUSPEND:
+ case CSIO_HWE_PCI_REMOVE:
+ case CSIO_HWE_PCIERR_DETECTED:
+ csio_set_state(&hw->sm, csio_hws_quiescing);
+ /* cleanup all outstanding cmds */
+ if (evt == CSIO_HWE_HBA_RESET ||
+ evt == CSIO_HWE_PCIERR_DETECTED)
+ csio_scsim_cleanup_io(csio_hw_to_scsim(hw), false);
+ else
+ csio_scsim_cleanup_io(csio_hw_to_scsim(hw), true);
+
+ csio_hw_intr_disable(hw);
+ csio_hw_mbm_cleanup(hw);
+ csio_evtq_stop(hw);
+ csio_notify_lnodes(hw, CSIO_LN_NOTIFY_HWSTOP);
+ csio_evtq_flush(hw);
+ csio_mgmtm_cleanup(csio_hw_to_mgmtm(hw));
+ csio_post_event(&hw->sm, CSIO_HWE_QUIESCED);
+ break;
+
+ case CSIO_HWE_FATAL:
+ csio_set_state(&hw->sm, csio_hws_uninit);
+ break;
+
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+ }
+}
+
+/*
+ * csio_hws_quiescing - Quiescing state
+ * @hw - HW module
+ * @evt - Event
+ *
+ */
+static void
+csio_hws_quiescing(struct csio_hw *hw, enum csio_hw_ev evt)
+{
+ hw->prev_evt = hw->cur_evt;
+ hw->cur_evt = evt;
+ CSIO_INC_STATS(hw, n_evt_sm[evt]);
+
+ switch (evt) {
+ case CSIO_HWE_QUIESCED:
+ switch (hw->evtflag) {
+ case CSIO_HWE_FW_DLOAD:
+ csio_set_state(&hw->sm, csio_hws_resetting);
+ /* Download firmware */
+ /* Fall through */
+
+ case CSIO_HWE_HBA_RESET:
+ csio_set_state(&hw->sm, csio_hws_resetting);
+ /* Start reset of the HBA */
+ csio_notify_lnodes(hw, CSIO_LN_NOTIFY_HWRESET);
+ csio_wr_destroy_queues(hw, false);
+ csio_do_reset(hw, false);
+ csio_post_event(&hw->sm, CSIO_HWE_HBA_RESET_DONE);
+ break;
+
+ case CSIO_HWE_PCI_REMOVE:
+ csio_set_state(&hw->sm, csio_hws_removing);
+ csio_notify_lnodes(hw, CSIO_LN_NOTIFY_HWREMOVE);
+ csio_wr_destroy_queues(hw, true);
+ /* Now send the bye command */
+ csio_do_bye(hw);
+ break;
+
+ case CSIO_HWE_SUSPEND:
+ csio_set_state(&hw->sm, csio_hws_quiesced);
+ break;
+
+ case CSIO_HWE_PCIERR_DETECTED:
+ csio_set_state(&hw->sm, csio_hws_pcierr);
+ csio_wr_destroy_queues(hw, false);
+ break;
+
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+
+ }
+ break;
+
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+ }
+}
+
+/*
+ * csio_hws_quiesced - Quiesced state
+ * @hw - HW module
+ * @evt - Event
+ *
+ */
+static void
+csio_hws_quiesced(struct csio_hw *hw, enum csio_hw_ev evt)
+{
+ hw->prev_evt = hw->cur_evt;
+ hw->cur_evt = evt;
+ CSIO_INC_STATS(hw, n_evt_sm[evt]);
+
+ switch (evt) {
+ case CSIO_HWE_RESUME:
+ csio_set_state(&hw->sm, csio_hws_configuring);
+ csio_hw_configure(hw);
+ break;
+
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+ }
+}
+
+/*
+ * csio_hws_resetting - HW Resetting state
+ * @hw - HW module
+ * @evt - Event
+ *
+ */
+static void
+csio_hws_resetting(struct csio_hw *hw, enum csio_hw_ev evt)
+{
+ hw->prev_evt = hw->cur_evt;
+ hw->cur_evt = evt;
+ CSIO_INC_STATS(hw, n_evt_sm[evt]);
+
+ switch (evt) {
+ case CSIO_HWE_HBA_RESET_DONE:
+ csio_evtq_start(hw);
+ csio_set_state(&hw->sm, csio_hws_configuring);
+ csio_hw_configure(hw);
+ break;
+
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+ }
+}
+
+/*
+ * csio_hws_removing - PCI Hotplug removing state
+ * @hw - HW module
+ * @evt - Event
+ *
+ */
+static void
+csio_hws_removing(struct csio_hw *hw, enum csio_hw_ev evt)
+{
+ hw->prev_evt = hw->cur_evt;
+ hw->cur_evt = evt;
+ CSIO_INC_STATS(hw, n_evt_sm[evt]);
+
+ switch (evt) {
+ case CSIO_HWE_HBA_RESET:
+ if (!csio_is_hw_master(hw))
+ break;
+ /*
+ * The BYE should have alerady been issued, so we cant
+ * use the mailbox interface. Hence we use the PL_RST
+ * register directly.
+ */
+ csio_err(hw, "Resetting HW and waiting 2 seconds...\n");
+ csio_wr_reg32(hw, PIORSTMODE | PIORST, PL_RST);
+ mdelay(2000);
+ break;
+
+ /* Should never receive any new events */
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+
+ }
+}
+
+/*
+ * csio_hws_pcierr - PCI Error state
+ * @hw - HW module
+ * @evt - Event
+ *
+ */
+static void
+csio_hws_pcierr(struct csio_hw *hw, enum csio_hw_ev evt)
+{
+ hw->prev_evt = hw->cur_evt;
+ hw->cur_evt = evt;
+ CSIO_INC_STATS(hw, n_evt_sm[evt]);
+
+ switch (evt) {
+ case CSIO_HWE_PCIERR_SLOT_RESET:
+ csio_evtq_start(hw);
+ csio_set_state(&hw->sm, csio_hws_configuring);
+ csio_hw_configure(hw);
+ break;
+
+ default:
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+ }
+}
+
+/*****************************************************************************/
+/* END: HW SM */
+/*****************************************************************************/
+
+/* Slow path handlers */
+struct intr_info {
+ unsigned int mask; /* bits to check in interrupt status */
+ const char *msg; /* message to print or NULL */
+ short stat_idx; /* stat counter to increment or -1 */
+ unsigned short fatal; /* whether the condition reported is fatal */
+};
+
+/*
+ * csio_handle_intr_status - table driven interrupt handler
+ * @hw: HW instance
+ * @reg: the interrupt status register to process
+ * @acts: table of interrupt actions
+ *
+ * A table driven interrupt handler that applies a set of masks to an
+ * interrupt status word and performs the corresponding actions if the
+ * interrupts described by the mask have occured. The actions include
+ * optionally emitting a warning or alert message. The table is terminated
+ * by an entry specifying mask 0. Returns the number of fatal interrupt
+ * conditions.
+ */
+static int
+csio_handle_intr_status(struct csio_hw *hw, unsigned int reg,
+ const struct intr_info *acts)
+{
+ int fatal = 0;
+ unsigned int mask = 0;
+ unsigned int status = csio_rd_reg32(hw, reg);
+
+ for ( ; acts->mask; ++acts) {
+ if (!(status & acts->mask))
+ continue;
+ if (acts->fatal) {
+ fatal++;
+ csio_fatal(hw, "Fatal %s (0x%x)\n",
+ acts->msg, status & acts->mask);
+ } else if (acts->msg)
+ csio_info(hw, "%s (0x%x)\n",
+ acts->msg, status & acts->mask);
+ mask |= acts->mask;
+ }
+ status &= mask;
+ if (status) /* clear processed interrupts */
+ csio_wr_reg32(hw, status, reg);
+ return fatal;
+}
+
+/*
+ * Interrupt handler for the PCIE module.
+ */
+static void
+csio_pcie_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info sysbus_intr_info[] = {
+ { RNPP, "RXNP array parity error", -1, 1 },
+ { RPCP, "RXPC array parity error", -1, 1 },
+ { RCIP, "RXCIF array parity error", -1, 1 },
+ { RCCP, "Rx completions control array parity error", -1, 1 },
+ { RFTP, "RXFT array parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+ static struct intr_info pcie_port_intr_info[] = {
+ { TPCP, "TXPC array parity error", -1, 1 },
+ { TNPP, "TXNP array parity error", -1, 1 },
+ { TFTP, "TXFT array parity error", -1, 1 },
+ { TCAP, "TXCA array parity error", -1, 1 },
+ { TCIP, "TXCIF array parity error", -1, 1 },
+ { RCAP, "RXCA array parity error", -1, 1 },
+ { OTDD, "outbound request TLP discarded", -1, 1 },
+ { RDPE, "Rx data parity error", -1, 1 },
+ { TDUE, "Tx uncorrectable data error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+ static struct intr_info pcie_intr_info[] = {
+ { MSIADDRLPERR, "MSI AddrL parity error", -1, 1 },
+ { MSIADDRHPERR, "MSI AddrH parity error", -1, 1 },
+ { MSIDATAPERR, "MSI data parity error", -1, 1 },
+ { MSIXADDRLPERR, "MSI-X AddrL parity error", -1, 1 },
+ { MSIXADDRHPERR, "MSI-X AddrH parity error", -1, 1 },
+ { MSIXDATAPERR, "MSI-X data parity error", -1, 1 },
+ { MSIXDIPERR, "MSI-X DI parity error", -1, 1 },
+ { PIOCPLPERR, "PCI PIO completion FIFO parity error", -1, 1 },
+ { PIOREQPERR, "PCI PIO request FIFO parity error", -1, 1 },
+ { TARTAGPERR, "PCI PCI target tag FIFO parity error", -1, 1 },
+ { CCNTPERR, "PCI CMD channel count parity error", -1, 1 },
+ { CREQPERR, "PCI CMD channel request parity error", -1, 1 },
+ { CRSPPERR, "PCI CMD channel response parity error", -1, 1 },
+ { DCNTPERR, "PCI DMA channel count parity error", -1, 1 },
+ { DREQPERR, "PCI DMA channel request parity error", -1, 1 },
+ { DRSPPERR, "PCI DMA channel response parity error", -1, 1 },
+ { HCNTPERR, "PCI HMA channel count parity error", -1, 1 },
+ { HREQPERR, "PCI HMA channel request parity error", -1, 1 },
+ { HRSPPERR, "PCI HMA channel response parity error", -1, 1 },
+ { CFGSNPPERR, "PCI config snoop FIFO parity error", -1, 1 },
+ { FIDPERR, "PCI FID parity error", -1, 1 },
+ { INTXCLRPERR, "PCI INTx clear parity error", -1, 1 },
+ { MATAGPERR, "PCI MA tag parity error", -1, 1 },
+ { PIOTAGPERR, "PCI PIO tag parity error", -1, 1 },
+ { RXCPLPERR, "PCI Rx completion parity error", -1, 1 },
+ { RXWRPERR, "PCI Rx write parity error", -1, 1 },
+ { RPLPERR, "PCI replay buffer parity error", -1, 1 },
+ { PCIESINT, "PCI core secondary fault", -1, 1 },
+ { PCIEPINT, "PCI core primary fault", -1, 1 },
+ { UNXSPLCPLERR, "PCI unexpected split completion error", -1,
+ 0 },
+ { 0, NULL, 0, 0 }
+ };
+
+ int fat;
+
+ fat = csio_handle_intr_status(hw,
+ PCIE_CORE_UTL_SYSTEM_BUS_AGENT_STATUS,
+ sysbus_intr_info) +
+ csio_handle_intr_status(hw,
+ PCIE_CORE_UTL_PCI_EXPRESS_PORT_STATUS,
+ pcie_port_intr_info) +
+ csio_handle_intr_status(hw, PCIE_INT_CAUSE, pcie_intr_info);
+ if (fat)
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * TP interrupt handler.
+ */
+static void csio_tp_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info tp_intr_info[] = {
+ { 0x3fffffff, "TP parity error", -1, 1 },
+ { FLMTXFLSTEMPTY, "TP out of Tx pages", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, TP_INT_CAUSE, tp_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * SGE interrupt handler.
+ */
+static void csio_sge_intr_handler(struct csio_hw *hw)
+{
+ uint64_t v;
+
+ static struct intr_info sge_intr_info[] = {
+ { ERR_CPL_EXCEED_IQE_SIZE,
+ "SGE received CPL exceeding IQE size", -1, 1 },
+ { ERR_INVALID_CIDX_INC,
+ "SGE GTS CIDX increment too large", -1, 0 },
+ { ERR_CPL_OPCODE_0, "SGE received 0-length CPL", -1, 0 },
+ { ERR_DROPPED_DB, "SGE doorbell dropped", -1, 0 },
+ { ERR_DATA_CPL_ON_HIGH_QID1 | ERR_DATA_CPL_ON_HIGH_QID0,
+ "SGE IQID > 1023 received CPL for FL", -1, 0 },
+ { ERR_BAD_DB_PIDX3, "SGE DBP 3 pidx increment too large", -1,
+ 0 },
+ { ERR_BAD_DB_PIDX2, "SGE DBP 2 pidx increment too large", -1,
+ 0 },
+ { ERR_BAD_DB_PIDX1, "SGE DBP 1 pidx increment too large", -1,
+ 0 },
+ { ERR_BAD_DB_PIDX0, "SGE DBP 0 pidx increment too large", -1,
+ 0 },
+ { ERR_ING_CTXT_PRIO,
+ "SGE too many priority ingress contexts", -1, 0 },
+ { ERR_EGR_CTXT_PRIO,
+ "SGE too many priority egress contexts", -1, 0 },
+ { INGRESS_SIZE_ERR, "SGE illegal ingress QID", -1, 0 },
+ { EGRESS_SIZE_ERR, "SGE illegal egress QID", -1, 0 },
+ { 0, NULL, 0, 0 }
+ };
+
+ v = (uint64_t)csio_rd_reg32(hw, SGE_INT_CAUSE1) |
+ ((uint64_t)csio_rd_reg32(hw, SGE_INT_CAUSE2) << 32);
+ if (v) {
+ csio_fatal(hw, "SGE parity error (%#llx)\n",
+ (unsigned long long)v);
+ csio_wr_reg32(hw, (uint32_t)(v & 0xFFFFFFFF),
+ SGE_INT_CAUSE1);
+ csio_wr_reg32(hw, (uint32_t)(v >> 32), SGE_INT_CAUSE2);
+ }
+
+ v |= csio_handle_intr_status(hw, SGE_INT_CAUSE3, sge_intr_info);
+
+ if (csio_handle_intr_status(hw, SGE_INT_CAUSE3, sge_intr_info) ||
+ v != 0)
+ csio_hw_fatal_err(hw);
+}
+
+#define CIM_OBQ_INTR (OBQULP0PARERR | OBQULP1PARERR | OBQULP2PARERR |\
+ OBQULP3PARERR | OBQSGEPARERR | OBQNCSIPARERR)
+#define CIM_IBQ_INTR (IBQTP0PARERR | IBQTP1PARERR | IBQULPPARERR |\
+ IBQSGEHIPARERR | IBQSGELOPARERR | IBQNCSIPARERR)
+
+/*
+ * CIM interrupt handler.
+ */
+static void csio_cim_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info cim_intr_info[] = {
+ { PREFDROPINT, "CIM control register prefetch drop", -1, 1 },
+ { CIM_OBQ_INTR, "CIM OBQ parity error", -1, 1 },
+ { CIM_IBQ_INTR, "CIM IBQ parity error", -1, 1 },
+ { MBUPPARERR, "CIM mailbox uP parity error", -1, 1 },
+ { MBHOSTPARERR, "CIM mailbox host parity error", -1, 1 },
+ { TIEQINPARERRINT, "CIM TIEQ outgoing parity error", -1, 1 },
+ { TIEQOUTPARERRINT, "CIM TIEQ incoming parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+ static struct intr_info cim_upintr_info[] = {
+ { RSVDSPACEINT, "CIM reserved space access", -1, 1 },
+ { ILLTRANSINT, "CIM illegal transaction", -1, 1 },
+ { ILLWRINT, "CIM illegal write", -1, 1 },
+ { ILLRDINT, "CIM illegal read", -1, 1 },
+ { ILLRDBEINT, "CIM illegal read BE", -1, 1 },
+ { ILLWRBEINT, "CIM illegal write BE", -1, 1 },
+ { SGLRDBOOTINT, "CIM single read from boot space", -1, 1 },
+ { SGLWRBOOTINT, "CIM single write to boot space", -1, 1 },
+ { BLKWRBOOTINT, "CIM block write to boot space", -1, 1 },
+ { SGLRDFLASHINT, "CIM single read from flash space", -1, 1 },
+ { SGLWRFLASHINT, "CIM single write to flash space", -1, 1 },
+ { BLKWRFLASHINT, "CIM block write to flash space", -1, 1 },
+ { SGLRDEEPROMINT, "CIM single EEPROM read", -1, 1 },
+ { SGLWREEPROMINT, "CIM single EEPROM write", -1, 1 },
+ { BLKRDEEPROMINT, "CIM block EEPROM read", -1, 1 },
+ { BLKWREEPROMINT, "CIM block EEPROM write", -1, 1 },
+ { SGLRDCTLINT , "CIM single read from CTL space", -1, 1 },
+ { SGLWRCTLINT , "CIM single write to CTL space", -1, 1 },
+ { BLKRDCTLINT , "CIM block read from CTL space", -1, 1 },
+ { BLKWRCTLINT , "CIM block write to CTL space", -1, 1 },
+ { SGLRDPLINT , "CIM single read from PL space", -1, 1 },
+ { SGLWRPLINT , "CIM single write to PL space", -1, 1 },
+ { BLKRDPLINT , "CIM block read from PL space", -1, 1 },
+ { BLKWRPLINT , "CIM block write to PL space", -1, 1 },
+ { REQOVRLOOKUPINT , "CIM request FIFO overwrite", -1, 1 },
+ { RSPOVRLOOKUPINT , "CIM response FIFO overwrite", -1, 1 },
+ { TIMEOUTINT , "CIM PIF timeout", -1, 1 },
+ { TIMEOUTMAINT , "CIM PIF MA timeout", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ int fat;
+
+ fat = csio_handle_intr_status(hw, CIM_HOST_INT_CAUSE,
+ cim_intr_info) +
+ csio_handle_intr_status(hw, CIM_HOST_UPACC_INT_CAUSE,
+ cim_upintr_info);
+ if (fat)
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * ULP RX interrupt handler.
+ */
+static void csio_ulprx_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info ulprx_intr_info[] = {
+ { 0x1800000, "ULPRX context error", -1, 1 },
+ { 0x7fffff, "ULPRX parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, ULP_RX_INT_CAUSE, ulprx_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * ULP TX interrupt handler.
+ */
+static void csio_ulptx_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info ulptx_intr_info[] = {
+ { PBL_BOUND_ERR_CH3, "ULPTX channel 3 PBL out of bounds", -1,
+ 0 },
+ { PBL_BOUND_ERR_CH2, "ULPTX channel 2 PBL out of bounds", -1,
+ 0 },
+ { PBL_BOUND_ERR_CH1, "ULPTX channel 1 PBL out of bounds", -1,
+ 0 },
+ { PBL_BOUND_ERR_CH0, "ULPTX channel 0 PBL out of bounds", -1,
+ 0 },
+ { 0xfffffff, "ULPTX parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, ULP_TX_INT_CAUSE, ulptx_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * PM TX interrupt handler.
+ */
+static void csio_pmtx_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info pmtx_intr_info[] = {
+ { PCMD_LEN_OVFL0, "PMTX channel 0 pcmd too large", -1, 1 },
+ { PCMD_LEN_OVFL1, "PMTX channel 1 pcmd too large", -1, 1 },
+ { PCMD_LEN_OVFL2, "PMTX channel 2 pcmd too large", -1, 1 },
+ { ZERO_C_CMD_ERROR, "PMTX 0-length pcmd", -1, 1 },
+ { 0xffffff0, "PMTX framing error", -1, 1 },
+ { OESPI_PAR_ERROR, "PMTX oespi parity error", -1, 1 },
+ { DB_OPTIONS_PAR_ERROR, "PMTX db_options parity error", -1,
+ 1 },
+ { ICSPI_PAR_ERROR, "PMTX icspi parity error", -1, 1 },
+ { C_PCMD_PAR_ERROR, "PMTX c_pcmd parity error", -1, 1},
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, PM_TX_INT_CAUSE, pmtx_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * PM RX interrupt handler.
+ */
+static void csio_pmrx_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info pmrx_intr_info[] = {
+ { ZERO_E_CMD_ERROR, "PMRX 0-length pcmd", -1, 1 },
+ { 0x3ffff0, "PMRX framing error", -1, 1 },
+ { OCSPI_PAR_ERROR, "PMRX ocspi parity error", -1, 1 },
+ { DB_OPTIONS_PAR_ERROR, "PMRX db_options parity error", -1,
+ 1 },
+ { IESPI_PAR_ERROR, "PMRX iespi parity error", -1, 1 },
+ { E_PCMD_PAR_ERROR, "PMRX e_pcmd parity error", -1, 1},
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, PM_RX_INT_CAUSE, pmrx_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * CPL switch interrupt handler.
+ */
+static void csio_cplsw_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info cplsw_intr_info[] = {
+ { CIM_OP_MAP_PERR, "CPLSW CIM op_map parity error", -1, 1 },
+ { CIM_OVFL_ERROR, "CPLSW CIM overflow", -1, 1 },
+ { TP_FRAMING_ERROR, "CPLSW TP framing error", -1, 1 },
+ { SGE_FRAMING_ERROR, "CPLSW SGE framing error", -1, 1 },
+ { CIM_FRAMING_ERROR, "CPLSW CIM framing error", -1, 1 },
+ { ZERO_SWITCH_ERROR, "CPLSW no-switch error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, CPL_INTR_CAUSE, cplsw_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * LE interrupt handler.
+ */
+static void csio_le_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info le_intr_info[] = {
+ { LIPMISS, "LE LIP miss", -1, 0 },
+ { LIP0, "LE 0 LIP error", -1, 0 },
+ { PARITYERR, "LE parity error", -1, 1 },
+ { UNKNOWNCMD, "LE unknown command", -1, 1 },
+ { REQQPARERR, "LE request queue parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, LE_DB_INT_CAUSE, le_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * MPS interrupt handler.
+ */
+static void csio_mps_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info mps_rx_intr_info[] = {
+ { 0xffffff, "MPS Rx parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+ static struct intr_info mps_tx_intr_info[] = {
+ { TPFIFO, "MPS Tx TP FIFO parity error", -1, 1 },
+ { NCSIFIFO, "MPS Tx NC-SI FIFO parity error", -1, 1 },
+ { TXDATAFIFO, "MPS Tx data FIFO parity error", -1, 1 },
+ { TXDESCFIFO, "MPS Tx desc FIFO parity error", -1, 1 },
+ { BUBBLE, "MPS Tx underflow", -1, 1 },
+ { SECNTERR, "MPS Tx SOP/EOP error", -1, 1 },
+ { FRMERR, "MPS Tx framing error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+ static struct intr_info mps_trc_intr_info[] = {
+ { FILTMEM, "MPS TRC filter parity error", -1, 1 },
+ { PKTFIFO, "MPS TRC packet FIFO parity error", -1, 1 },
+ { MISCPERR, "MPS TRC misc parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+ static struct intr_info mps_stat_sram_intr_info[] = {
+ { 0x1fffff, "MPS statistics SRAM parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+ static struct intr_info mps_stat_tx_intr_info[] = {
+ { 0xfffff, "MPS statistics Tx FIFO parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+ static struct intr_info mps_stat_rx_intr_info[] = {
+ { 0xffffff, "MPS statistics Rx FIFO parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+ static struct intr_info mps_cls_intr_info[] = {
+ { MATCHSRAM, "MPS match SRAM parity error", -1, 1 },
+ { MATCHTCAM, "MPS match TCAM parity error", -1, 1 },
+ { HASHSRAM, "MPS hash SRAM parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ int fat;
+
+ fat = csio_handle_intr_status(hw, MPS_RX_PERR_INT_CAUSE,
+ mps_rx_intr_info) +
+ csio_handle_intr_status(hw, MPS_TX_INT_CAUSE,
+ mps_tx_intr_info) +
+ csio_handle_intr_status(hw, MPS_TRC_INT_CAUSE,
+ mps_trc_intr_info) +
+ csio_handle_intr_status(hw, MPS_STAT_PERR_INT_CAUSE_SRAM,
+ mps_stat_sram_intr_info) +
+ csio_handle_intr_status(hw, MPS_STAT_PERR_INT_CAUSE_TX_FIFO,
+ mps_stat_tx_intr_info) +
+ csio_handle_intr_status(hw, MPS_STAT_PERR_INT_CAUSE_RX_FIFO,
+ mps_stat_rx_intr_info) +
+ csio_handle_intr_status(hw, MPS_CLS_INT_CAUSE,
+ mps_cls_intr_info);
+
+ csio_wr_reg32(hw, 0, MPS_INT_CAUSE);
+ csio_rd_reg32(hw, MPS_INT_CAUSE); /* flush */
+ if (fat)
+ csio_hw_fatal_err(hw);
+}
+
+#define MEM_INT_MASK (PERR_INT_CAUSE | ECC_CE_INT_CAUSE | ECC_UE_INT_CAUSE)
+
+/*
+ * EDC/MC interrupt handler.
+ */
+static void csio_mem_intr_handler(struct csio_hw *hw, int idx)
+{
+ static const char name[3][5] = { "EDC0", "EDC1", "MC" };
+
+ unsigned int addr, cnt_addr, v;
+
+ if (idx <= MEM_EDC1) {
+ addr = EDC_REG(EDC_INT_CAUSE, idx);
+ cnt_addr = EDC_REG(EDC_ECC_STATUS, idx);
+ } else {
+ addr = MC_INT_CAUSE;
+ cnt_addr = MC_ECC_STATUS;
+ }
+
+ v = csio_rd_reg32(hw, addr) & MEM_INT_MASK;
+ if (v & PERR_INT_CAUSE)
+ csio_fatal(hw, "%s FIFO parity error\n", name[idx]);
+ if (v & ECC_CE_INT_CAUSE) {
+ uint32_t cnt = ECC_CECNT_GET(csio_rd_reg32(hw, cnt_addr));
+
+ csio_wr_reg32(hw, ECC_CECNT_MASK, cnt_addr);
+ csio_warn(hw, "%u %s correctable ECC data error%s\n",
+ cnt, name[idx], cnt > 1 ? "s" : "");
+ }
+ if (v & ECC_UE_INT_CAUSE)
+ csio_fatal(hw, "%s uncorrectable ECC data error\n", name[idx]);
+
+ csio_wr_reg32(hw, v, addr);
+ if (v & (PERR_INT_CAUSE | ECC_UE_INT_CAUSE))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * MA interrupt handler.
+ */
+static void csio_ma_intr_handler(struct csio_hw *hw)
+{
+ uint32_t v, status = csio_rd_reg32(hw, MA_INT_CAUSE);
+
+ if (status & MEM_PERR_INT_CAUSE)
+ csio_fatal(hw, "MA parity error, parity status %#x\n",
+ csio_rd_reg32(hw, MA_PARITY_ERROR_STATUS));
+ if (status & MEM_WRAP_INT_CAUSE) {
+ v = csio_rd_reg32(hw, MA_INT_WRAP_STATUS);
+ csio_fatal(hw,
+ "MA address wrap-around error by client %u to address %#x\n",
+ MEM_WRAP_CLIENT_NUM_GET(v), MEM_WRAP_ADDRESS_GET(v) << 4);
+ }
+ csio_wr_reg32(hw, status, MA_INT_CAUSE);
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * SMB interrupt handler.
+ */
+static void csio_smb_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info smb_intr_info[] = {
+ { MSTTXFIFOPARINT, "SMB master Tx FIFO parity error", -1, 1 },
+ { MSTRXFIFOPARINT, "SMB master Rx FIFO parity error", -1, 1 },
+ { SLVFIFOPARINT, "SMB slave FIFO parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, SMB_INT_CAUSE, smb_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * NC-SI interrupt handler.
+ */
+static void csio_ncsi_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info ncsi_intr_info[] = {
+ { CIM_DM_PRTY_ERR, "NC-SI CIM parity error", -1, 1 },
+ { MPS_DM_PRTY_ERR, "NC-SI MPS parity error", -1, 1 },
+ { TXFIFO_PRTY_ERR, "NC-SI Tx FIFO parity error", -1, 1 },
+ { RXFIFO_PRTY_ERR, "NC-SI Rx FIFO parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, NCSI_INT_CAUSE, ncsi_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * XGMAC interrupt handler.
+ */
+static void csio_xgmac_intr_handler(struct csio_hw *hw, int port)
+{
+ uint32_t v = csio_rd_reg32(hw, PORT_REG(port, XGMAC_PORT_INT_CAUSE));
+
+ v &= TXFIFO_PRTY_ERR | RXFIFO_PRTY_ERR;
+ if (!v)
+ return;
+
+ if (v & TXFIFO_PRTY_ERR)
+ csio_fatal(hw, "XGMAC %d Tx FIFO parity error\n", port);
+ if (v & RXFIFO_PRTY_ERR)
+ csio_fatal(hw, "XGMAC %d Rx FIFO parity error\n", port);
+ csio_wr_reg32(hw, v, PORT_REG(port, XGMAC_PORT_INT_CAUSE));
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * PL interrupt handler.
+ */
+static void csio_pl_intr_handler(struct csio_hw *hw)
+{
+ static struct intr_info pl_intr_info[] = {
+ { FATALPERR, "T4 fatal parity error", -1, 1 },
+ { PERRVFID, "PL VFID_MAP parity error", -1, 1 },
+ { 0, NULL, 0, 0 }
+ };
+
+ if (csio_handle_intr_status(hw, PL_PL_INT_CAUSE, pl_intr_info))
+ csio_hw_fatal_err(hw);
+}
+
+/*
+ * csio_hw_slow_intr_handler - control path interrupt handler
+ * @hw: HW module
+ *
+ * Interrupt handler for non-data global interrupt events, e.g., errors.
+ * The designation 'slow' is because it involves register reads, while
+ * data interrupts typically don't involve any MMIOs.
+ */
+int
+csio_hw_slow_intr_handler(struct csio_hw *hw)
+{
+ uint32_t cause = csio_rd_reg32(hw, PL_INT_CAUSE);
+
+ if (!(cause & CSIO_GLBL_INTR_MASK)) {
+ CSIO_INC_STATS(hw, n_plint_unexp);
+ return 0;
+ }
+
+ csio_dbg(hw, "Slow interrupt! cause: 0x%x\n", cause);
+
+ CSIO_INC_STATS(hw, n_plint_cnt);
+
+ if (cause & CIM)
+ csio_cim_intr_handler(hw);
+
+ if (cause & MPS)
+ csio_mps_intr_handler(hw);
+
+ if (cause & NCSI)
+ csio_ncsi_intr_handler(hw);
+
+ if (cause & PL)
+ csio_pl_intr_handler(hw);
+
+ if (cause & SMB)
+ csio_smb_intr_handler(hw);
+
+ if (cause & XGMAC0)
+ csio_xgmac_intr_handler(hw, 0);
+
+ if (cause & XGMAC1)
+ csio_xgmac_intr_handler(hw, 1);
+
+ if (cause & XGMAC_KR0)
+ csio_xgmac_intr_handler(hw, 2);
+
+ if (cause & XGMAC_KR1)
+ csio_xgmac_intr_handler(hw, 3);
+
+ if (cause & PCIE)
+ csio_pcie_intr_handler(hw);
+
+ if (cause & MC)
+ csio_mem_intr_handler(hw, MEM_MC);
+
+ if (cause & EDC0)
+ csio_mem_intr_handler(hw, MEM_EDC0);
+
+ if (cause & EDC1)
+ csio_mem_intr_handler(hw, MEM_EDC1);
+
+ if (cause & LE)
+ csio_le_intr_handler(hw);
+
+ if (cause & TP)
+ csio_tp_intr_handler(hw);
+
+ if (cause & MA)
+ csio_ma_intr_handler(hw);
+
+ if (cause & PM_TX)
+ csio_pmtx_intr_handler(hw);
+
+ if (cause & PM_RX)
+ csio_pmrx_intr_handler(hw);
+
+ if (cause & ULP_RX)
+ csio_ulprx_intr_handler(hw);
+
+ if (cause & CPL_SWITCH)
+ csio_cplsw_intr_handler(hw);
+
+ if (cause & SGE)
+ csio_sge_intr_handler(hw);
+
+ if (cause & ULP_TX)
+ csio_ulptx_intr_handler(hw);
+
+ /* Clear the interrupts just processed for which we are the master. */
+ csio_wr_reg32(hw, cause & CSIO_GLBL_INTR_MASK, PL_INT_CAUSE);
+ csio_rd_reg32(hw, PL_INT_CAUSE); /* flush */
+
+ return 1;
+}
+
+/*****************************************************************************
+ * HW <--> mailbox interfacing routines.
+ ****************************************************************************/
+/*
+ * csio_mberr_worker - Worker thread (dpc) for mailbox/error completions
+ *
+ * @data: Private data pointer.
+ *
+ * Called from worker thread context.
+ */
+static void
+csio_mberr_worker(void *data)
+{
+ struct csio_hw *hw = (struct csio_hw *)data;
+ struct csio_mbm *mbm = &hw->mbm;
+ LIST_HEAD(cbfn_q);
+ struct csio_mb *mbp_next;
+ int rv;
+
+ del_timer_sync(&mbm->timer);
+
+ spin_lock_irq(&hw->lock);
+ if (list_empty(&mbm->cbfn_q)) {
+ spin_unlock_irq(&hw->lock);
+ return;
+ }
+
+ list_splice_tail_init(&mbm->cbfn_q, &cbfn_q);
+ mbm->stats.n_cbfnq = 0;
+
+ /* Try to start waiting mailboxes */
+ if (!list_empty(&mbm->req_q)) {
+ mbp_next = list_first_entry(&mbm->req_q, struct csio_mb, list);
+ list_del_init(&mbp_next->list);
+
+ rv = csio_mb_issue(hw, mbp_next);
+ if (rv != 0)
+ list_add_tail(&mbp_next->list, &mbm->req_q);
+ else
+ CSIO_DEC_STATS(mbm, n_activeq);
+ }
+ spin_unlock_irq(&hw->lock);
+
+ /* Now callback completions */
+ csio_mb_completions(hw, &cbfn_q);
+}
+
+/*
+ * csio_hw_mb_timer - Top-level Mailbox timeout handler.
+ *
+ * @data: private data pointer
+ *
+ **/
+static void
+csio_hw_mb_timer(uintptr_t data)
+{
+ struct csio_hw *hw = (struct csio_hw *)data;
+ struct csio_mb *mbp = NULL;
+
+ spin_lock_irq(&hw->lock);
+ mbp = csio_mb_tmo_handler(hw);
+ spin_unlock_irq(&hw->lock);
+
+ /* Call back the function for the timed-out Mailbox */
+ if (mbp)
+ mbp->mb_cbfn(hw, mbp);
+
+}
+
+/*
+ * csio_hw_mbm_cleanup - Cleanup Mailbox module.
+ * @hw: HW module
+ *
+ * Called with lock held, should exit with lock held.
+ * Cancels outstanding mailboxes (waiting, in-flight) and gathers them
+ * into a local queue. Drops lock and calls the completions. Holds
+ * lock and returns.
+ */
+static void
+csio_hw_mbm_cleanup(struct csio_hw *hw)
+{
+ LIST_HEAD(cbfn_q);
+
+ csio_mb_cancel_all(hw, &cbfn_q);
+
+ spin_unlock_irq(&hw->lock);
+ csio_mb_completions(hw, &cbfn_q);
+ spin_lock_irq(&hw->lock);
+}
+
+/*****************************************************************************
+ * Event handling
+ ****************************************************************************/
+int
+csio_enqueue_evt(struct csio_hw *hw, enum csio_evt type, void *evt_msg,
+ uint16_t len)
+{
+ struct csio_evt_msg *evt_entry = NULL;
+
+ if (type >= CSIO_EVT_MAX)
+ return -EINVAL;
+
+ if (len > CSIO_EVT_MSG_SIZE)
+ return -EINVAL;
+
+ if (hw->flags & CSIO_HWF_FWEVT_STOP)
+ return -EINVAL;
+
+ if (list_empty(&hw->evt_free_q)) {
+ csio_err(hw, "Failed to alloc evt entry, msg type %d len %d\n",
+ type, len);
+ return -ENOMEM;
+ }
+
+ evt_entry = list_first_entry(&hw->evt_free_q,
+ struct csio_evt_msg, list);
+ list_del_init(&evt_entry->list);
+
+ /* copy event msg and queue the event */
+ evt_entry->type = type;
+ memcpy((void *)evt_entry->data, evt_msg, len);
+ list_add_tail(&evt_entry->list, &hw->evt_active_q);
+
+ CSIO_DEC_STATS(hw, n_evt_freeq);
+ CSIO_INC_STATS(hw, n_evt_activeq);
+
+ return 0;
+}
+
+static int
+csio_enqueue_evt_lock(struct csio_hw *hw, enum csio_evt type, void *evt_msg,
+ uint16_t len, bool msg_sg)
+{
+ struct csio_evt_msg *evt_entry = NULL;
+ struct csio_fl_dma_buf *fl_sg;
+ uint32_t off = 0;
+ unsigned long flags;
+ int n;
+
+ if (type >= CSIO_EVT_MAX)
+ return -EINVAL;
+
+ if (len > CSIO_EVT_MSG_SIZE)
+ return -EINVAL;
+
+ spin_lock_irqsave(&hw->lock, flags);
+ if (hw->flags & CSIO_HWF_FWEVT_STOP) {
+ spin_unlock_irqrestore(&hw->lock, flags);
+ return -EINVAL;
+ }
+
+ if (list_empty(&hw->evt_free_q)) {
+ csio_err(hw, "Failed to alloc evt entry, msg type %d len %d\n",
+ type, len);
+ spin_unlock_irqrestore(&hw->lock, flags);
+ return -ENOMEM;
+ }
+
+ evt_entry = list_first_entry(&hw->evt_free_q,
+ struct csio_evt_msg, list);
+ list_del_init(&evt_entry->list);
+
+ /* copy event msg and queue the event */
+ evt_entry->type = type;
+
+ /* If Payload in SG list*/
+ if (msg_sg) {
+ fl_sg = (struct csio_fl_dma_buf *) evt_msg;
+ for (n = 0; (n < CSIO_MAX_FLBUF_PER_IQWR && off < len); n++) {
+ memcpy((void *)((uintptr_t)evt_entry->data + off),
+ fl_sg->flbufs[n].vaddr,
+ fl_sg->flbufs[n].len);
+ off += fl_sg->flbufs[n].len;
+ }
+ } else
+ memcpy((void *)evt_entry->data, evt_msg, len);
+
+ list_add_tail(&evt_entry->list, &hw->evt_active_q);
+ spin_unlock_irqrestore(&hw->lock, flags);
+
+ CSIO_DEC_STATS(hw, n_evt_freeq);
+ CSIO_INC_STATS(hw, n_evt_activeq);
+
+ return 0;
+}
+
+static void
+csio_free_evt(struct csio_hw *hw, struct csio_evt_msg *evt_entry)
+{
+ if (evt_entry) {
+ spin_lock_irq(&hw->lock);
+ list_del_init(&evt_entry->list);
+ list_add_tail(&evt_entry->list, &hw->evt_free_q);
+ CSIO_DEC_STATS(hw, n_evt_activeq);
+ CSIO_INC_STATS(hw, n_evt_freeq);
+ spin_unlock_irq(&hw->lock);
+ }
+}
+
+void
+csio_evtq_flush(struct csio_hw *hw)
+{
+ uint32_t count;
+ count = 30;
+ while (hw->flags & CSIO_HWF_FWEVT_PENDING && count--) {
+ spin_unlock_irq(&hw->lock);
+ msleep(2000);
+ spin_lock_irq(&hw->lock);
+ }
+
+ CSIO_DB_ASSERT(!(hw->flags & CSIO_HWF_FWEVT_PENDING));
+}
+
+static void
+csio_evtq_stop(struct csio_hw *hw)
+{
+ hw->flags |= CSIO_HWF_FWEVT_STOP;
+}
+
+static void
+csio_evtq_start(struct csio_hw *hw)
+{
+ hw->flags &= ~CSIO_HWF_FWEVT_STOP;
+}
+
+static void
+csio_evtq_cleanup(struct csio_hw *hw)
+{
+ struct list_head *evt_entry, *next_entry;
+
+ /* Release outstanding events from activeq to freeq*/
+ if (!list_empty(&hw->evt_active_q))
+ list_splice_tail_init(&hw->evt_active_q, &hw->evt_free_q);
+
+ hw->stats.n_evt_activeq = 0;
+ hw->flags &= ~CSIO_HWF_FWEVT_PENDING;
+
+ /* Freeup event entry */
+ list_for_each_safe(evt_entry, next_entry, &hw->evt_free_q) {
+ kfree(evt_entry);
+ CSIO_DEC_STATS(hw, n_evt_freeq);
+ }
+
+ hw->stats.n_evt_freeq = 0;
+}
+
+
+static void
+csio_process_fwevtq_entry(struct csio_hw *hw, void *wr, uint32_t len,
+ struct csio_fl_dma_buf *flb, void *priv)
+{
+ __u8 op;
+ __be64 *data;
+ void *msg = NULL;
+ uint32_t msg_len = 0;
+ bool msg_sg = 0;
+
+ op = ((struct rss_header *) wr)->opcode;
+ if (op == CPL_FW6_PLD) {
+ CSIO_INC_STATS(hw, n_cpl_fw6_pld);
+ if (!flb || !flb->totlen) {
+ CSIO_INC_STATS(hw, n_cpl_unexp);
+ return;
+ }
+
+ msg = (void *) flb;
+ msg_len = flb->totlen;
+ msg_sg = 1;
+
+ data = (__be64 *) msg;
+ } else if (op == CPL_FW6_MSG || op == CPL_FW4_MSG) {
+
+ CSIO_INC_STATS(hw, n_cpl_fw6_msg);
+ /* skip RSS header */
+ msg = (void *)((uintptr_t)wr + sizeof(__be64));
+ msg_len = (op == CPL_FW6_MSG) ? sizeof(struct cpl_fw6_msg) :
+ sizeof(struct cpl_fw4_msg);
+
+ data = (__be64 *) msg;
+ } else {
+ csio_warn(hw, "unexpected CPL %#x on FW event queue\n", op);
+ CSIO_INC_STATS(hw, n_cpl_unexp);
+ return;
+ }
+
+ /*
+ * Enqueue event to EventQ. Events processing happens
+ * in Event worker thread context
+ */
+ if (csio_enqueue_evt_lock(hw, CSIO_EVT_FW, msg,
+ (uint16_t)msg_len, msg_sg))
+ CSIO_INC_STATS(hw, n_evt_drop);
+}
+
+void
+csio_evtq_worker(struct work_struct *work)
+{
+ struct csio_hw *hw = container_of(work, struct csio_hw, evtq_work);
+ struct list_head *evt_entry, *next_entry;
+ LIST_HEAD(evt_q);
+ struct csio_evt_msg *evt_msg;
+ struct cpl_fw6_msg *msg;
+ struct csio_rnode *rn;
+ int rv = 0;
+ uint8_t evtq_stop = 0;
+
+ csio_dbg(hw, "event worker thread active evts#%d\n",
+ hw->stats.n_evt_activeq);
+
+ spin_lock_irq(&hw->lock);
+ while (!list_empty(&hw->evt_active_q)) {
+ list_splice_tail_init(&hw->evt_active_q, &evt_q);
+ spin_unlock_irq(&hw->lock);
+
+ list_for_each_safe(evt_entry, next_entry, &evt_q) {
+ evt_msg = (struct csio_evt_msg *) evt_entry;
+
+ /* Drop events if queue is STOPPED */
+ spin_lock_irq(&hw->lock);
+ if (hw->flags & CSIO_HWF_FWEVT_STOP)
+ evtq_stop = 1;
+ spin_unlock_irq(&hw->lock);
+ if (evtq_stop) {
+ CSIO_INC_STATS(hw, n_evt_drop);
+ goto free_evt;
+ }
+
+ switch (evt_msg->type) {
+ case CSIO_EVT_FW:
+ msg = (struct cpl_fw6_msg *)(evt_msg->data);
+
+ if ((msg->opcode == CPL_FW6_MSG ||
+ msg->opcode == CPL_FW4_MSG) &&
+ !msg->type) {
+ rv = csio_mb_fwevt_handler(hw,
+ msg->data);
+ if (!rv)
+ break;
+ /* Handle any remaining fw events */
+ csio_fcoe_fwevt_handler(hw,
+ msg->opcode, msg->data);
+ } else if (msg->opcode == CPL_FW6_PLD) {
+
+ csio_fcoe_fwevt_handler(hw,
+ msg->opcode, msg->data);
+ } else {
+ csio_warn(hw,
+ "Unhandled FW msg op %x type %x\n",
+ msg->opcode, msg->type);
+ CSIO_INC_STATS(hw, n_evt_drop);
+ }
+ break;
+
+ case CSIO_EVT_MBX:
+ csio_mberr_worker(hw);
+ break;
+
+ case CSIO_EVT_DEV_LOSS:
+ memcpy(&rn, evt_msg->data, sizeof(rn));
+ csio_rnode_devloss_handler(rn);
+ break;
+
+ default:
+ csio_warn(hw, "Unhandled event %x on evtq\n",
+ evt_msg->type);
+ CSIO_INC_STATS(hw, n_evt_unexp);
+ break;
+ }
+free_evt:
+ csio_free_evt(hw, evt_msg);
+ }
+
+ spin_lock_irq(&hw->lock);
+ }
+ hw->flags &= ~CSIO_HWF_FWEVT_PENDING;
+ spin_unlock_irq(&hw->lock);
+}
+
+int
+csio_fwevtq_handler(struct csio_hw *hw)
+{
+ int rv;
+
+ if (csio_q_iqid(hw, hw->fwevt_iq_idx) == CSIO_MAX_QID) {
+ CSIO_INC_STATS(hw, n_int_stray);
+ return -EINVAL;
+ }
+
+ rv = csio_wr_process_iq_idx(hw, hw->fwevt_iq_idx,
+ csio_process_fwevtq_entry, NULL);
+ return rv;
+}
+
+/****************************************************************************
+ * Entry points
+ ****************************************************************************/
+
+/* Management module */
+/*
+ * csio_mgmt_req_lookup - Lookup the given IO req exist in Active Q.
+ * mgmt - mgmt module
+ * @io_req - io request
+ *
+ * Return - 0:if given IO Req exists in active Q.
+ * -EINVAL :if lookup fails.
+ */
+int
+csio_mgmt_req_lookup(struct csio_mgmtm *mgmtm, struct csio_ioreq *io_req)
+{
+ struct list_head *tmp;
+
+ /* Lookup ioreq in the ACTIVEQ */
+ list_for_each(tmp, &mgmtm->active_q) {
+ if (io_req == (struct csio_ioreq *)tmp)
+ return 0;
+ }
+ return -EINVAL;
+}
+
+#define ECM_MIN_TMO 1000 /* Minimum timeout value for req */
+
+/*
+ * csio_mgmts_tmo_handler - MGMT IO Timeout handler.
+ * @data - Event data.
+ *
+ * Return - none.
+ */
+static void
+csio_mgmt_tmo_handler(uintptr_t data)
+{
+ struct csio_mgmtm *mgmtm = (struct csio_mgmtm *) data;
+ struct list_head *tmp;
+ struct csio_ioreq *io_req;
+
+ csio_dbg(mgmtm->hw, "Mgmt timer invoked!\n");
+
+ spin_lock_irq(&mgmtm->hw->lock);
+
+ list_for_each(tmp, &mgmtm->active_q) {
+ io_req = (struct csio_ioreq *) tmp;
+ io_req->tmo -= min_t(uint32_t, io_req->tmo, ECM_MIN_TMO);
+
+ if (!io_req->tmo) {
+ /* Dequeue the request from retry Q. */
+ tmp = csio_list_prev(tmp);
+ list_del_init(&io_req->sm.sm_list);
+ if (io_req->io_cbfn) {
+ /* io_req will be freed by completion handler */
+ io_req->wr_status = -ETIMEDOUT;
+ io_req->io_cbfn(mgmtm->hw, io_req);
+ } else {
+ CSIO_DB_ASSERT(0);
+ }
+ }
+ }
+
+ /* If retry queue is not empty, re-arm timer */
+ if (!list_empty(&mgmtm->active_q))
+ mod_timer(&mgmtm->mgmt_timer,
+ jiffies + msecs_to_jiffies(ECM_MIN_TMO));
+ spin_unlock_irq(&mgmtm->hw->lock);
+}
+
+static void
+csio_mgmtm_cleanup(struct csio_mgmtm *mgmtm)
+{
+ struct csio_hw *hw = mgmtm->hw;
+ struct csio_ioreq *io_req;
+ struct list_head *tmp;
+ uint32_t count;
+
+ count = 30;
+ /* Wait for all outstanding req to complete gracefully */
+ while ((!list_empty(&mgmtm->active_q)) && count--) {
+ spin_unlock_irq(&hw->lock);
+ msleep(2000);
+ spin_lock_irq(&hw->lock);
+ }
+
+ /* release outstanding req from ACTIVEQ */
+ list_for_each(tmp, &mgmtm->active_q) {
+ io_req = (struct csio_ioreq *) tmp;
+ tmp = csio_list_prev(tmp);
+ list_del_init(&io_req->sm.sm_list);
+ mgmtm->stats.n_active--;
+ if (io_req->io_cbfn) {
+ /* io_req will be freed by completion handler */
+ io_req->wr_status = -ETIMEDOUT;
+ io_req->io_cbfn(mgmtm->hw, io_req);
+ }
+ }
+}
+
+/*
+ * csio_mgmt_init - Mgmt module init entry point
+ * @mgmtsm - mgmt module
+ * @hw - HW module
+ *
+ * Initialize mgmt timer, resource wait queue, active queue,
+ * completion q. Allocate Egress and Ingress
+ * WR queues and save off the queue index returned by the WR
+ * module for future use. Allocate and save off mgmt reqs in the
+ * mgmt_req_freelist for future use. Make sure their SM is initialized
+ * to uninit state.
+ * Returns: 0 - on success
+ * -ENOMEM - on error.
+ */
+static int
+csio_mgmtm_init(struct csio_mgmtm *mgmtm, struct csio_hw *hw)
+{
+ struct timer_list *timer = &mgmtm->mgmt_timer;
+
+ init_timer(timer);
+ timer->function = csio_mgmt_tmo_handler;
+ timer->data = (unsigned long)mgmtm;
+
+ INIT_LIST_HEAD(&mgmtm->active_q);
+ INIT_LIST_HEAD(&mgmtm->cbfn_q);
+
+ mgmtm->hw = hw;
+ /*mgmtm->iq_idx = hw->fwevt_iq_idx;*/
+
+ return 0;
+}
+
+/*
+ * csio_mgmtm_exit - MGMT module exit entry point
+ * @mgmtsm - mgmt module
+ *
+ * This function called during MGMT module uninit.
+ * Stop timers, free ioreqs allocated.
+ * Returns: None
+ *
+ */
+static void
+csio_mgmtm_exit(struct csio_mgmtm *mgmtm)
+{
+ del_timer_sync(&mgmtm->mgmt_timer);
+}
+
+
+/**
+ * csio_hw_start - Kicks off the HW State machine
+ * @hw: Pointer to HW module.
+ *
+ * It is assumed that the initialization is a synchronous operation.
+ * So when we return afer posting the event, the HW SM should be in
+ * the ready state, if there were no errors during init.
+ */
+int
+csio_hw_start(struct csio_hw *hw)
+{
+ spin_lock_irq(&hw->lock);
+ csio_post_event(&hw->sm, CSIO_HWE_CFG);
+ spin_unlock_irq(&hw->lock);
+
+ if (csio_is_hw_ready(hw))
+ return 0;
+ else
+ return -EINVAL;
+}
+
+int
+csio_hw_stop(struct csio_hw *hw)
+{
+ csio_post_event(&hw->sm, CSIO_HWE_PCI_REMOVE);
+
+ if (csio_is_hw_removing(hw))
+ return 0;
+ else
+ return -EINVAL;
+}
+
+/* Max reset retries */
+#define CSIO_MAX_RESET_RETRIES 3
+
+/**
+ * csio_hw_reset - Reset the hardware
+ * @hw: HW module.
+ *
+ * Caller should hold lock across this function.
+ */
+int
+csio_hw_reset(struct csio_hw *hw)
+{
+ if (!csio_is_hw_master(hw))
+ return -EPERM;
+
+ if (hw->rst_retries >= CSIO_MAX_RESET_RETRIES) {
+ csio_dbg(hw, "Max hw reset attempts reached..");
+ return -EINVAL;
+ }
+
+ hw->rst_retries++;
+ csio_post_event(&hw->sm, CSIO_HWE_HBA_RESET);
+
+ if (csio_is_hw_ready(hw)) {
+ hw->rst_retries = 0;
+ hw->stats.n_reset_start = jiffies_to_msecs(jiffies);
+ return 0;
+ } else
+ return -EINVAL;
+}
+
+/*
+ * csio_hw_get_device_id - Caches the Adapter's vendor & device id.
+ * @hw: HW module.
+ */
+static void
+csio_hw_get_device_id(struct csio_hw *hw)
+{
+ /* Is the adapter device id cached already ?*/
+ if (csio_is_dev_id_cached(hw))
+ return;
+
+ /* Get the PCI vendor & device id */
+ pci_read_config_word(hw->pdev, PCI_VENDOR_ID,
+ &hw->params.pci.vendor_id);
+ pci_read_config_word(hw->pdev, PCI_DEVICE_ID,
+ &hw->params.pci.device_id);
+
+ csio_dev_id_cached(hw);
+
+} /* csio_hw_get_device_id */
+
+/*
+ * csio_hw_set_description - Set the model, description of the hw.
+ * @hw: HW module.
+ * @ven_id: PCI Vendor ID
+ * @dev_id: PCI Device ID
+ */
+static void
+csio_hw_set_description(struct csio_hw *hw, uint16_t ven_id, uint16_t dev_id)
+{
+ uint32_t adap_type, prot_type;
+
+ if (ven_id == CSIO_VENDOR_ID) {
+ prot_type = (dev_id & CSIO_ASIC_DEVID_PROTO_MASK);
+ adap_type = (dev_id & CSIO_ASIC_DEVID_TYPE_MASK);
+
+ if (prot_type == CSIO_FPGA) {
+ memcpy(hw->model_desc,
+ csio_fcoe_adapters[13].description, 32);
+ } else if (prot_type == CSIO_T4_FCOE_ASIC) {
+ memcpy(hw->hw_ver,
+ csio_fcoe_adapters[adap_type].model_no, 16);
+ memcpy(hw->model_desc,
+ csio_fcoe_adapters[adap_type].description, 32);
+ } else {
+ char tempName[32] = "Chelsio FCoE Controller";
+ memcpy(hw->model_desc, tempName, 32);
+
+ CSIO_DB_ASSERT(0);
+ }
+ }
+} /* csio_hw_set_description */
+
+/**
+ * csio_hw_init - Initialize HW module.
+ * @hw: Pointer to HW module.
+ *
+ * Initialize the members of the HW module.
+ */
+int
+csio_hw_init(struct csio_hw *hw)
+{
+ int rv = -EINVAL;
+ uint32_t i;
+ uint16_t ven_id, dev_id;
+ struct csio_evt_msg *evt_entry;
+
+ INIT_LIST_HEAD(&hw->sm.sm_list);
+ csio_init_state(&hw->sm, csio_hws_uninit);
+ spin_lock_init(&hw->lock);
+ INIT_LIST_HEAD(&hw->sln_head);
+
+ /* Get the PCI vendor & device id */
+ csio_hw_get_device_id(hw);
+
+ strcpy(hw->name, CSIO_HW_NAME);
+
+ /* Set the model & its description */
+
+ ven_id = hw->params.pci.vendor_id;
+ dev_id = hw->params.pci.device_id;
+
+ csio_hw_set_description(hw, ven_id, dev_id);
+
+ /* Initialize default log level */
+ hw->params.log_level = (uint32_t) csio_dbg_level;
+
+ csio_set_fwevt_intr_idx(hw, -1);
+ csio_set_nondata_intr_idx(hw, -1);
+
+ /* Init all the modules: Mailbox, WorkRequest and Transport */
+ if (csio_mbm_init(csio_hw_to_mbm(hw), hw, csio_hw_mb_timer))
+ goto err;
+
+ rv = csio_wrm_init(csio_hw_to_wrm(hw), hw);
+ if (rv)
+ goto err_mbm_exit;
+
+ rv = csio_scsim_init(csio_hw_to_scsim(hw), hw);
+ if (rv)
+ goto err_wrm_exit;
+
+ rv = csio_mgmtm_init(csio_hw_to_mgmtm(hw), hw);
+ if (rv)
+ goto err_scsim_exit;
+ /* Pre-allocate evtq and initialize them */
+ INIT_LIST_HEAD(&hw->evt_active_q);
+ INIT_LIST_HEAD(&hw->evt_free_q);
+ for (i = 0; i < csio_evtq_sz; i++) {
+
+ evt_entry = kzalloc(sizeof(struct csio_evt_msg), GFP_KERNEL);
+ if (!evt_entry) {
+ csio_err(hw, "Failed to initialize eventq");
+ goto err_evtq_cleanup;
+ }
+
+ list_add_tail(&evt_entry->list, &hw->evt_free_q);
+ CSIO_INC_STATS(hw, n_evt_freeq);
+ }
+
+ hw->dev_num = dev_num;
+ dev_num++;
+
+ return 0;
+
+err_evtq_cleanup:
+ csio_evtq_cleanup(hw);
+ csio_mgmtm_exit(csio_hw_to_mgmtm(hw));
+err_scsim_exit:
+ csio_scsim_exit(csio_hw_to_scsim(hw));
+err_wrm_exit:
+ csio_wrm_exit(csio_hw_to_wrm(hw), hw);
+err_mbm_exit:
+ csio_mbm_exit(csio_hw_to_mbm(hw));
+err:
+ return rv;
+}
+
+/**
+ * csio_hw_exit - Un-initialize HW module.
+ * @hw: Pointer to HW module.
+ *
+ */
+void
+csio_hw_exit(struct csio_hw *hw)
+{
+ csio_evtq_cleanup(hw);
+ csio_mgmtm_exit(csio_hw_to_mgmtm(hw));
+ csio_scsim_exit(csio_hw_to_scsim(hw));
+ csio_wrm_exit(csio_hw_to_wrm(hw), hw);
+ csio_mbm_exit(csio_hw_to_mbm(hw));
+}
--
1.7.1
^ permalink raw reply related
* [V3 PATCH 0/9] csiostor: Chelsio FCoE offload driver submission
From: Naresh Kumar Inna @ 2012-09-11 14:38 UTC (permalink / raw)
To: JBottomley, linux-scsi, dm, leedom; +Cc: netdev, naresh, chethan
This is the initial submission of the Chelsio FCoE offload driver (csiostor)
to the upstream kernel. This driver currently supports FCoE offload
functionality over Chelsio T4-based 10Gb Converged Network Adapters.
The following patches contain the driver sources for csiostor driver and
updates to firmware/hardware header files shared between csiostor,
cxgb4 (Chelsio T4-based NIC driver) and cxgb4vf (Chelsio T4-based Virtual
Function NIC driver). The csiostor driver is dependent on these
header updates. These patches have been generated against scsi 'misc' branch.
csiostor is a low level SCSI driver that interfaces with PCI, SCSI midlayer and
FC transport subsystems. This driver claims the FCoE PCIe function on
Chelsio Converged Network Adapters. It relies on firmware events for slow path
operations like discovery, thereby offloading session management. The driver
programs firmware via Work Request interfaces for fast path I/O offload
features.
V3 patches address review comments from Ben Hutchings and Stephen Hemminger.
These changes are listed in the individual patch emails.
Here is the brief description of patches:
[V3 PATCH 1/9]: Hardware interface, Makefile and Kconfig changes.
[V3 PATCH 2/9]: Driver initialization and Work Request services.
[V3 PATCH 3/9]: FC transport interfaces and mailbox services.
[V3 PATCH 4/9]: Local and remote port state tracking functionality.
[V3 PATCH 5/9]: Interrupt handling and fast path I/O functionality.
[V3 PATCH 6/9]: Header files part 1.
[V3 PATCH 7/9]: Header files part 2.
[V3 PATCH 8/9]: Updates to header files shared between cxgb4, cxgb4vf and
csiostor.
[V3 PATCH 9/9]: Header file compatibility fixes to cxgb4vf.
Naresh Kumar Inna (9):
csiostor: Chelsio FCoE offload driver submission (sources part 1).
csiostor: Chelsio FCoE offload driver submission (sources part 2).
csiostor: Chelsio FCoE offload driver submission (sources part 3).
csiostor: Chelsio FCoE offload driver submission (sources part 4).
csiostor: Chelsio FCoE offload driver submission (sources part 5).
csiostor: Chelsio FCoE offload driver submission (headers part 1).
csiostor: Chelsio FCoE offload driver submission (headers part 2).
cxgb4: Chelsio FCoE offload driver submission (cxgb4 common header
updates).
cxgb4vf: Chelsio FCoE offload driver submission (header compatibility
fixes).
drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c | 2 +-
drivers/net/ethernet/chelsio/cxgb4/sge.c | 10 +-
drivers/net/ethernet/chelsio/cxgb4/t4_hw.c | 16 +-
drivers/net/ethernet/chelsio/cxgb4/t4_msg.h | 1 +
drivers/net/ethernet/chelsio/cxgb4/t4_regs.h | 69 +-
drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h | 104 +-
drivers/net/ethernet/chelsio/cxgb4vf/sge.c | 11 +-
drivers/scsi/Kconfig | 1 +
drivers/scsi/Makefile | 1 +
drivers/scsi/csiostor/Kconfig | 20 +
drivers/scsi/csiostor/Makefile | 11 +
drivers/scsi/csiostor/csio_attr.c | 809 +++++
drivers/scsi/csiostor/csio_defs.h | 108 +
drivers/scsi/csiostor/csio_hw.c | 4396 +++++++++++++++++++++++
drivers/scsi/csiostor/csio_hw.h | 666 ++++
drivers/scsi/csiostor/csio_init.c | 1272 +++++++
drivers/scsi/csiostor/csio_init.h | 158 +
drivers/scsi/csiostor/csio_isr.c | 624 ++++
drivers/scsi/csiostor/csio_lnode.c | 2148 +++++++++++
drivers/scsi/csiostor/csio_lnode.h | 255 ++
drivers/scsi/csiostor/csio_mb.c | 1769 +++++++++
drivers/scsi/csiostor/csio_mb.h | 278 ++
drivers/scsi/csiostor/csio_rnode.c | 889 +++++
drivers/scsi/csiostor/csio_rnode.h | 141 +
drivers/scsi/csiostor/csio_scsi.c | 2561 +++++++++++++
drivers/scsi/csiostor/csio_scsi.h | 342 ++
drivers/scsi/csiostor/csio_wr.c | 1632 +++++++++
drivers/scsi/csiostor/csio_wr.h | 512 +++
drivers/scsi/csiostor/t4fw_api_stor.h | 578 +++
29 files changed, 19347 insertions(+), 37 deletions(-)
create mode 100644 drivers/scsi/csiostor/Kconfig
create mode 100644 drivers/scsi/csiostor/Makefile
create mode 100644 drivers/scsi/csiostor/csio_attr.c
create mode 100644 drivers/scsi/csiostor/csio_defs.h
create mode 100644 drivers/scsi/csiostor/csio_hw.c
create mode 100644 drivers/scsi/csiostor/csio_hw.h
create mode 100644 drivers/scsi/csiostor/csio_init.c
create mode 100644 drivers/scsi/csiostor/csio_init.h
create mode 100644 drivers/scsi/csiostor/csio_isr.c
create mode 100644 drivers/scsi/csiostor/csio_lnode.c
create mode 100644 drivers/scsi/csiostor/csio_lnode.h
create mode 100644 drivers/scsi/csiostor/csio_mb.c
create mode 100644 drivers/scsi/csiostor/csio_mb.h
create mode 100644 drivers/scsi/csiostor/csio_rnode.c
create mode 100644 drivers/scsi/csiostor/csio_rnode.h
create mode 100644 drivers/scsi/csiostor/csio_scsi.c
create mode 100644 drivers/scsi/csiostor/csio_scsi.h
create mode 100644 drivers/scsi/csiostor/csio_wr.c
create mode 100644 drivers/scsi/csiostor/csio_wr.h
create mode 100644 drivers/scsi/csiostor/t4fw_api_stor.h
^ permalink raw reply
* [PATCH v2 1/2] iproute2: add libgenl files
From: Julian Anastasov @ 2012-09-11 9:04 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: netdev
In-Reply-To: <1347354274-3119-1-git-send-email-ja@ssi.bg>
Create libgenl.h and libgenl.c. They will contain
common code for GENL users such as ipl2tp, tcp_metrics, etc.
Signed-off-by: Julian Anastasov <ja@ssi.bg>
---
include/libgenl.h | 29 +++++++++++++++++++++++
lib/Makefile | 2 +-
lib/libgenl.c | 65 +++++++++++++++++++++++++++++++++++++++++++++++++++++
3 files changed, 95 insertions(+), 1 deletions(-)
create mode 100644 include/libgenl.h
create mode 100644 lib/libgenl.c
diff --git a/include/libgenl.h b/include/libgenl.h
new file mode 100644
index 0000000..7862292
--- /dev/null
+++ b/include/libgenl.h
@@ -0,0 +1,29 @@
+#ifndef __LIBGENL_H__
+#define __LIBGENL_H__
+
+#include "libnetlink.h"
+
+#define GENL_DEFINE_REQUEST(req, hdrsize, bufsiz) \
+struct { \
+ struct nlmsghdr n; \
+ struct genlmsghdr g; \
+ char buf[NLMSG_ALIGN(hdrsize) + (bufsiz)]; \
+} req
+
+#define GENL_INIT_REQUEST(req, family, hdrsize, ver, cmd_, flags) \
+ do { \
+ req.n = (struct nlmsghdr) { \
+ .nlmsg_type = (family), \
+ .nlmsg_flags = (flags), \
+ .nlmsg_len = NLMSG_LENGTH(GENL_HDRLEN + \
+ (hdrsize)), \
+ }; \
+ req.g = (struct genlmsghdr) { \
+ .cmd = (cmd_), \
+ .version = (ver), \
+ }; \
+ } while (0)
+
+extern int genl_resolve_family(struct rtnl_handle *grth, const char *family);
+
+#endif /* __LIBGENL_H__ */
diff --git a/lib/Makefile b/lib/Makefile
index da2f0fc..bfbe672 100644
--- a/lib/Makefile
+++ b/lib/Makefile
@@ -2,7 +2,7 @@ CFLAGS += -fPIC
UTILOBJ=utils.o rt_names.o ll_types.o ll_proto.o ll_addr.o inet_proto.o
-NLOBJ=ll_map.o libnetlink.o
+NLOBJ=libgenl.o ll_map.o libnetlink.o
all: libnetlink.a libutil.a
diff --git a/lib/libgenl.c b/lib/libgenl.c
new file mode 100644
index 0000000..e7ddf95
--- /dev/null
+++ b/lib/libgenl.c
@@ -0,0 +1,65 @@
+/*
+ * libgenl.c GENL library
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <unistd.h>
+
+#include <linux/genetlink.h>
+#include "libgenl.h"
+
+static int genl_parse_getfamily(struct nlmsghdr *nlh)
+{
+ struct rtattr *tb[CTRL_ATTR_MAX + 1];
+ struct genlmsghdr *ghdr = NLMSG_DATA(nlh);
+ int len = nlh->nlmsg_len;
+ struct rtattr *attrs;
+
+ if (nlh->nlmsg_type != GENL_ID_CTRL) {
+ fprintf(stderr, "Not a controller message, nlmsg_len=%d "
+ "nlmsg_type=0x%x\n", nlh->nlmsg_len, nlh->nlmsg_type);
+ return -1;
+ }
+
+ len -= NLMSG_LENGTH(GENL_HDRLEN);
+
+ if (len < 0) {
+ fprintf(stderr, "wrong controller message len %d\n", len);
+ return -1;
+ }
+
+ if (ghdr->cmd != CTRL_CMD_NEWFAMILY) {
+ fprintf(stderr, "Unknown controller command %d\n", ghdr->cmd);
+ return -1;
+ }
+
+ attrs = (struct rtattr *) ((char *) ghdr + GENL_HDRLEN);
+ parse_rtattr(tb, CTRL_ATTR_MAX, attrs, len);
+
+ if (tb[CTRL_ATTR_FAMILY_ID] == NULL) {
+ fprintf(stderr, "Missing family id TLV\n");
+ return -1;
+ }
+
+ return rta_getattr_u16(tb[CTRL_ATTR_FAMILY_ID]);
+}
+
+int genl_resolve_family(struct rtnl_handle *grth, const char *family)
+{
+ GENL_DEFINE_REQUEST(req, 0, 1024);
+
+ GENL_INIT_REQUEST(req, GENL_ID_CTRL, 0, 0, CTRL_CMD_GETFAMILY,
+ NLM_F_REQUEST);
+
+ addattr_l(&req.n, 1024, CTRL_ATTR_FAMILY_NAME,
+ family, strlen(family) + 1);
+
+ if (rtnl_talk(grth, &req.n, 0, 0, &req.n) < 0) {
+ fprintf(stderr, "Error talking to the kernel\n");
+ return -2;
+ }
+
+ return genl_parse_getfamily(&req.n);
+}
+
--
1.7.3.4
^ permalink raw reply related
* [PATCH v2 2/2] iproute2: use libgenl in ipl2tp
From: Julian Anastasov @ 2012-09-11 9:04 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: netdev
In-Reply-To: <1347354274-3119-1-git-send-email-ja@ssi.bg>
Use the common code from libgenl.c to parse family.
Signed-off-by: Julian Anastasov <ja@ssi.bg>
---
ip/ipl2tp.c | 64 +---------------------------------------------------------
1 files changed, 2 insertions(+), 62 deletions(-)
diff --git a/ip/ipl2tp.c b/ip/ipl2tp.c
index 1cbed8d..c7893c2 100644
--- a/ip/ipl2tp.c
+++ b/ip/ipl2tp.c
@@ -25,6 +25,7 @@
#include <linux/genetlink.h>
#include <linux/l2tp.h>
+#include "libgenl.h"
#include "utils.h"
#include "ip_common.h"
@@ -747,67 +748,6 @@ static int do_show(int argc, char **argv)
return 0;
}
-static int genl_parse_getfamily(struct nlmsghdr *nlh)
-{
- struct rtattr *tb[CTRL_ATTR_MAX + 1];
- struct genlmsghdr *ghdr = NLMSG_DATA(nlh);
- int len = nlh->nlmsg_len;
- struct rtattr *attrs;
-
- if (nlh->nlmsg_type != GENL_ID_CTRL) {
- fprintf(stderr, "Not a controller message, nlmsg_len=%d "
- "nlmsg_type=0x%x\n", nlh->nlmsg_len, nlh->nlmsg_type);
- return -1;
- }
-
- if (ghdr->cmd != CTRL_CMD_NEWFAMILY) {
- fprintf(stderr, "Unknown controller command %d\n", ghdr->cmd);
- return -1;
- }
-
- len -= NLMSG_LENGTH(GENL_HDRLEN);
-
- if (len < 0) {
- fprintf(stderr, "wrong controller message len %d\n", len);
- return -1;
- }
-
- attrs = (struct rtattr *) ((char *) ghdr + GENL_HDRLEN);
- parse_rtattr(tb, CTRL_ATTR_MAX, attrs, len);
-
- if (tb[CTRL_ATTR_FAMILY_ID] == NULL) {
- fprintf(stderr, "Missing family id TLV\n");
- return -1;
- }
-
- return rta_getattr_u16(tb[CTRL_ATTR_FAMILY_ID]);
-}
-
-int genl_ctrl_resolve_family(const char *family)
-{
- struct {
- struct nlmsghdr n;
- struct genlmsghdr g;
- char buf[1024];
- } req;
-
- memset(&req, 0, sizeof(req));
- req.n.nlmsg_len = NLMSG_LENGTH(GENL_HDRLEN);
- req.n.nlmsg_flags = NLM_F_REQUEST;
- req.n.nlmsg_type = GENL_ID_CTRL;
- req.g.cmd = CTRL_CMD_GETFAMILY;
-
- addattr_l(&req.n, 1024, CTRL_ATTR_FAMILY_NAME,
- family, strlen(family) + 1);
-
- if (rtnl_talk(&genl_rth, &req.n, 0, 0, &req.n) < 0) {
- fprintf(stderr, "Error talking to the kernel\n");
- return -2;
- }
-
- return genl_parse_getfamily(&req.n);
-}
-
int do_ipl2tp(int argc, char **argv)
{
if (genl_family < 0) {
@@ -816,7 +756,7 @@ int do_ipl2tp(int argc, char **argv)
exit(1);
}
- genl_family = genl_ctrl_resolve_family(L2TP_GENL_NAME);
+ genl_family = genl_resolve_family(&genl_rth, L2TP_GENL_NAME);
if (genl_family < 0)
exit(1);
}
--
1.7.3.4
^ permalink raw reply related
* [PATCH v2 0/2] iproute2: add GENL helpers
From: Julian Anastasov @ 2012-09-11 9:04 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: netdev
Put common code for GENL users in new files.
v2:
- remove memset for buffer in GENL_INIT_REQUEST and use
C99 style initializers
- rename funcs: libgenl_ to genl_
Julian Anastasov (2):
iproute2: add libgenl files
iproute2: use libgenl in ipl2tp
include/libgenl.h | 29 +++++++++++++++++++++++
ip/ipl2tp.c | 64 +--------------------------------------------------
lib/Makefile | 2 +-
lib/libgenl.c | 65 +++++++++++++++++++++++++++++++++++++++++++++++++++++
4 files changed, 97 insertions(+), 63 deletions(-)
create mode 100644 include/libgenl.h
create mode 100644 lib/libgenl.c
--
1.7.3.4
^ permalink raw reply
* Re: [PATCH 2/3] ipv6, route: remove BACKTRACK() macro
From: Cong Wang @ 2012-09-11 8:39 UTC (permalink / raw)
To: David Miller; +Cc: netdev
In-Reply-To: <20120910.214245.368123682481775366.davem@davemloft.net>
On Mon, 2012-09-10 at 21:42 -0400, David Miller wrote:
>
> Yes, I know this conflicts with cases like how we killed all of
> the netlink macros with embedded gotos.
>
> But this patch made things worse and added code duplication. Fix
> it without the code duplication side effect and it'll be fine.
Ok, will do.
Thanks, David!
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox