* net-next-2.6 [PATCH 0/3] dccp: Ack Vectors in circular buffer instead of array
From: Gerrit Renker @ 2010-11-11 6:07 UTC (permalink / raw)
To: davem; +Cc: dccp, netdev
In-Reply-To: <ack_vector_circular_buffer_dccp>
Dave,
please can you consider this set, which is part I of a two-part patch set
to fix issues in the DCCP Ack Vector implementation.
It is a self-contained set to move away from a linear array to a circular
buffer with overflow handling. Part II would then follow next week.
Patch #1: cleans up the old interface to prepare for the improved one.
Patch #2: also tidies up the old interface, by separating the internals
of Ack Vectors from the option-parsing code.
Patch #3: Completes the implementation of a circular Ack Vector buffer.
I have also placed this in into a fresh (today's) copy of net-next-2.6, on
git://eden-feed.erg.abdn.ac.uk/net-next-2.6 [subtree 'dccp']
The set has been tested for 3 years, and is fully bisectable.
---
net/dccp/ackvec.c | 251 ++++++++++++++++--------------------------------
net/dccp/ackvec.h | 115 +++++++++++++----------
net/dccp/ccids/ccid2.c | 13 +--
net/dccp/dccp.h | 11 ++-
net/dccp/input.c | 6 +-
net/dccp/options.c | 65 ++++++++++++-
6 files changed, 225 insertions(+), 236 deletions(-)
^ permalink raw reply
* [PATCH 2/3] dccp ccid-2: Separate internals of Ack Vectors from option-parsing code
From: Gerrit Renker @ 2010-11-11 6:07 UTC (permalink / raw)
To: davem; +Cc: dccp, netdev, Gerrit Renker
In-Reply-To: <1289455653-5463-2-git-send-email-gerrit@erg.abdn.ac.uk>
This patch
* separates Ack Vector housekeeping code from option-insertion code;
* shifts option-specific code from ackvec.c into options.c;
* introduces a dedicated routine to take care of the Ack Vector records;
* simplifies the dccp_ackvec_insert_avr() routine: the BUG_ON was redundant,
since the list is automatically arranged in descending order of ack_seqno.
Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
---
net/dccp/ackvec.h | 2 +-
net/dccp/ackvec.c | 100 +++++++++------------------------------------------
net/dccp/options.c | 60 +++++++++++++++++++++++++++++++
3 files changed, 79 insertions(+), 83 deletions(-)
--- a/net/dccp/ackvec.h
+++ b/net/dccp/ackvec.h
@@ -111,7 +111,7 @@ extern int dccp_ackvec_parse(struct sock *sk, const struct sk_buff *skb,
u64 *ackno, const u8 opt,
const u8 *value, const u8 len);
-extern int dccp_insert_option_ackvec(struct sock *sk, struct sk_buff *skb);
+extern int dccp_ackvec_update_records(struct dccp_ackvec *av, u64 seq, u8 sum);
static inline int dccp_ackvec_pending(const struct dccp_ackvec *av)
{
--- a/net/dccp/ackvec.c
+++ b/net/dccp/ackvec.c
@@ -52,99 +52,35 @@ void dccp_ackvec_free(struct dccp_ackvec *av)
}
}
-static void dccp_ackvec_insert_avr(struct dccp_ackvec *av,
- struct dccp_ackvec_record *avr)
-{
- /*
- * AVRs are sorted by seqno. Since we are sending them in order, we
- * just add the AVR at the head of the list.
- * -sorbo.
- */
- if (!list_empty(&av->av_records)) {
- const struct dccp_ackvec_record *head =
- list_entry(av->av_records.next,
- struct dccp_ackvec_record,
- avr_node);
- BUG_ON(before48(avr->avr_ack_seqno, head->avr_ack_seqno));
- }
-
- list_add(&avr->avr_node, &av->av_records);
-}
-
-int dccp_insert_option_ackvec(struct sock *sk, struct sk_buff *skb)
+/**
+ * dccp_ackvec_update_records - Record information about sent Ack Vectors
+ * @av: Ack Vector records to update
+ * @seqno: Sequence number of the packet carrying the Ack Vector just sent
+ * @nonce_sum: The sum of all buffer nonces contained in the Ack Vector
+ */
+int dccp_ackvec_update_records(struct dccp_ackvec *av, u64 seqno, u8 nonce_sum)
{
- struct dccp_sock *dp = dccp_sk(sk);
- struct dccp_ackvec *av = dp->dccps_hc_rx_ackvec;
- /* Figure out how many options do we need to represent the ackvec */
- const u8 nr_opts = DIV_ROUND_UP(av->av_vec_len, DCCP_SINGLE_OPT_MAXLEN);
- u16 len = av->av_vec_len + 2 * nr_opts;
- u8 i, nonce = 0;
- const unsigned char *tail, *from;
- unsigned char *to;
struct dccp_ackvec_record *avr;
- if (DCCP_SKB_CB(skb)->dccpd_opt_len + len > DCCP_MAX_OPT_LEN)
- return -1;
-
avr = kmem_cache_alloc(dccp_ackvec_record_slab, GFP_ATOMIC);
if (avr == NULL)
- return -1;
-
- DCCP_SKB_CB(skb)->dccpd_opt_len += len;
-
- to = skb_push(skb, len);
- len = av->av_vec_len;
- from = av->av_buf + av->av_buf_head;
- tail = av->av_buf + DCCPAV_MAX_ACKVEC_LEN;
-
- for (i = 0; i < nr_opts; ++i) {
- int copylen = len;
-
- if (len > DCCP_SINGLE_OPT_MAXLEN)
- copylen = DCCP_SINGLE_OPT_MAXLEN;
-
- /*
- * RFC 4340, 12.2: Encode the Nonce Echo for this Ack Vector via
- * its type; ack_nonce is the sum of all individual buf_nonce's.
- */
- nonce ^= av->av_buf_nonce[i];
-
- *to++ = DCCPO_ACK_VECTOR_0 + av->av_buf_nonce[i];
- *to++ = copylen + 2;
-
- /* Check if buf_head wraps */
- if (from + copylen > tail) {
- const u16 tailsize = tail - from;
-
- memcpy(to, from, tailsize);
- to += tailsize;
- len -= tailsize;
- copylen -= tailsize;
- from = av->av_buf;
- }
-
- memcpy(to, from, copylen);
- from += copylen;
- to += copylen;
- len -= copylen;
- }
+ return -ENOBUFS;
- /*
- * Each sent Ack Vector is recorded in the list, as per A.2 of RFC 4340.
- */
- avr->avr_ack_seqno = DCCP_SKB_CB(skb)->dccpd_seq;
+ avr->avr_ack_seqno = seqno;
avr->avr_ack_ptr = av->av_buf_head;
avr->avr_ack_ackno = av->av_buf_ackno;
- avr->avr_ack_nonce = nonce;
+ avr->avr_ack_nonce = nonce_sum;
avr->avr_ack_runlen = dccp_ackvec_runlen(av->av_buf + av->av_buf_head);
+ /*
+ * Since GSS is incremented for each packet, the list is automatically
+ * arranged in descending order of @ack_seqno.
+ */
+ list_add(&avr->avr_node, &av->av_records);
- dccp_ackvec_insert_avr(av, avr);
-
- dccp_pr_debug("%s ACK Vector 0, len=%d, ack_seqno=%llu, "
- "ack_ackno=%llu\n",
- dccp_role(sk), avr->avr_ack_runlen,
+ dccp_pr_debug("Added Vector, ack_seqno=%llu, ack_ackno=%llu (rl=%u)\n",
(unsigned long long)avr->avr_ack_seqno,
- (unsigned long long)avr->avr_ack_ackno);
+ (unsigned long long)avr->avr_ack_ackno,
+ avr->avr_ack_runlen);
return 0;
}
--- a/net/dccp/options.c
+++ b/net/dccp/options.c
@@ -425,6 +425,66 @@ static int dccp_insert_option_timestamp_echo(struct dccp_sock *dp,
return 0;
}
+static int dccp_insert_option_ackvec(struct sock *sk, struct sk_buff *skb)
+{
+ struct dccp_sock *dp = dccp_sk(sk);
+ struct dccp_ackvec *av = dp->dccps_hc_rx_ackvec;
+ /* Figure out how many options do we need to represent the ackvec */
+ const u8 nr_opts = DIV_ROUND_UP(av->av_vec_len, DCCP_SINGLE_OPT_MAXLEN);
+ u16 len = av->av_vec_len + 2 * nr_opts;
+ u8 i, nonce = 0;
+ const unsigned char *tail, *from;
+ unsigned char *to;
+
+ if (DCCP_SKB_CB(skb)->dccpd_opt_len + len > DCCP_MAX_OPT_LEN)
+ return -1;
+
+ DCCP_SKB_CB(skb)->dccpd_opt_len += len;
+
+ to = skb_push(skb, len);
+ len = av->av_vec_len;
+ from = av->av_buf + av->av_buf_head;
+ tail = av->av_buf + DCCPAV_MAX_ACKVEC_LEN;
+
+ for (i = 0; i < nr_opts; ++i) {
+ int copylen = len;
+
+ if (len > DCCP_SINGLE_OPT_MAXLEN)
+ copylen = DCCP_SINGLE_OPT_MAXLEN;
+
+ /*
+ * RFC 4340, 12.2: Encode the Nonce Echo for this Ack Vector via
+ * its type; ack_nonce is the sum of all individual buf_nonce's.
+ */
+ nonce ^= av->av_buf_nonce[i];
+
+ *to++ = DCCPO_ACK_VECTOR_0 + av->av_buf_nonce[i];
+ *to++ = copylen + 2;
+
+ /* Check if buf_head wraps */
+ if (from + copylen > tail) {
+ const u16 tailsize = tail - from;
+
+ memcpy(to, from, tailsize);
+ to += tailsize;
+ len -= tailsize;
+ copylen -= tailsize;
+ from = av->av_buf;
+ }
+
+ memcpy(to, from, copylen);
+ from += copylen;
+ to += copylen;
+ len -= copylen;
+ }
+ /*
+ * Each sent Ack Vector is recorded in the list, as per A.2 of RFC 4340.
+ */
+ if (dccp_ackvec_update_records(av, DCCP_SKB_CB(skb)->dccpd_seq, nonce))
+ return -ENOBUFS;
+ return 0;
+}
+
/**
* dccp_insert_option_mandatory - Mandatory option (5.8.2)
* Note that since we are using skb_push, this function needs to be called
^ permalink raw reply
* [PATCH] dlm: Handle application limited situations properly.
From: David Miller @ 2010-11-11 5:56 UTC (permalink / raw)
To: ccaulfie; +Cc: teigland, cluster-devel, netdev, linux-kernel
In the normal regime where an application uses non-blocking I/O
writes on a socket, they will handle -EAGAIN and use poll() to
wait for send space.
They don't actually sleep on the socket I/O write.
But kernel level RPC layers that do socket I/O operations directly
and key off of -EAGAIN on the write() to "try again later" don't
use poll(), they instead have their own sleeping mechanism and
rely upon ->sk_write_space() to trigger the wakeup.
So they do effectively sleep on the write(), but this mechanism
alone does not let the socket layers know what's going on.
Therefore they must emulate what would have happened, otherwise
TCP cannot possibly see that the connection is application window
size limited.
Handle this, therefore, like SUNRPC by setting SOCK_NOSPACE and
bumping the ->sk_write_count as needed when we hit the send buffer
limits.
This should make TCP send buffer size auto-tuning and the
->sk_write_space() callback invocations actually happen.
Signed-off-by: David S. Miller <davem@davemloft.net>
diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c
index 37a34c2..77720f8 100644
--- a/fs/dlm/lowcomms.c
+++ b/fs/dlm/lowcomms.c
@@ -108,6 +108,7 @@ struct connection {
#define CF_INIT_PENDING 4
#define CF_IS_OTHERCON 5
#define CF_CLOSE 6
+#define CF_APP_LIMITED 7
struct list_head writequeue; /* List of outgoing writequeue_entries */
spinlock_t writequeue_lock;
int (*rx_action) (struct connection *); /* What to do when active */
@@ -295,7 +296,17 @@ static void lowcomms_write_space(struct sock *sk)
{
struct connection *con = sock2con(sk);
- if (con && !test_and_set_bit(CF_WRITE_PENDING, &con->flags))
+ if (!con)
+ return;
+
+ clear_bit(SOCK_NOSPACE, &con->sock->flags);
+
+ if (test_and_clear_bit(CF_APP_LIMITED, &con->flags)) {
+ con->sock->sk->sk_write_pending--;
+ clear_bit(SOCK_ASYNC_NOSPACE, &con->sock->flags);
+ }
+
+ if (!test_and_set_bit(CF_WRITE_PENDING, &con->flags))
queue_work(send_workqueue, &con->swork);
}
@@ -1319,6 +1330,15 @@ static void send_to_sock(struct connection *con)
ret = kernel_sendpage(con->sock, e->page, offset, len,
msg_flags);
if (ret == -EAGAIN || ret == 0) {
+ if (ret == -EAGAIN &&
+ test_bit(SOCK_ASYNC_NOSPACE, &con->sock->flags) &&
+ !test_and_set_bit(CF_APP_LIMITED, &con->flags)) {
+ /* Notify TCP that we're limited by the
+ * application window size.
+ */
+ set_bit(SOCK_NOSPACE, &con->sock->flags);
+ con->sock->sk->sk_write_pending++;
+ }
cond_resched();
goto out;
}
^ permalink raw reply related
* Re: possible kernel oops from user MSS
From: Shan Wei @ 2010-11-11 5:37 UTC (permalink / raw)
To: David Miller; +Cc: schen, netdev
In-Reply-To: <20101110.213313.71108940.davem@davemloft.net>
David Miller wrote, at 11/11/2010 01:33 PM:
> From: Shan Wei <shanwei@cn.fujitsu.com>
> Date: Thu, 11 Nov 2010 13:15:01 +0800
>
>> In tcp_connect_init(), if tcp_header_len includes TCPOLEN_TSTAMP_ALIGNED(12 bytes)
>> and TCPOLEN_MD5SIG_ALIGNED(20 bytes).
>
> If you knew this, why didn't you mention it in your initial report? :-/
Firstly reported by Steve Chen, not me. :-)
I just review your patch.
> I'll make the minimum 64 or something like that.
Welcome.
--
Best Regards
-----
Shan Wei
^ permalink raw reply
* Re: possible kernel oops from user MSS
From: David Miller @ 2010-11-11 5:36 UTC (permalink / raw)
To: shanwei; +Cc: schen, netdev
In-Reply-To: <20101110.213313.71108940.davem@davemloft.net>
From: David Miller <davem@davemloft.net>
Date: Wed, 10 Nov 2010 21:33:13 -0800 (PST)
> I'll make the minimum 64 or something like that.
Here is the patch I will use:
--------------------
tcp: Increase TCP_MAXSEG socket option minimum.
As noted by Steve Chen, since commit
f5fff5dc8a7a3f395b0525c02ba92c95d42b7390 ("tcp: advertise MSS
requested by user") we can end up with a situation where
tcp_select_initial_window() does a divide by a zero (or
even negative) mss value.
The problem is that sometimes we effectively subtract
TCPOLEN_TSTAMP_ALIGNED and/or TCPOLEN_MD5SIG_ALIGNED from the mss.
Fix this by increasing the minimum from 8 to 64.
Reported-by: Steve Chen <schen@mvista.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
net/ipv4/tcp.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 245603c..0814199 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -2246,7 +2246,7 @@ static int do_tcp_setsockopt(struct sock *sk, int level,
/* Values greater than interface MTU won't take effect. However
* at the point when this call is done we typically don't yet
* know which interface is going to be used */
- if (val < 8 || val > MAX_TCP_WINDOW) {
+ if (val < 64 || val > MAX_TCP_WINDOW) {
err = -EINVAL;
break;
}
--
1.7.3.2
^ permalink raw reply related
* Re: possible kernel oops from user MSS
From: David Miller @ 2010-11-11 5:33 UTC (permalink / raw)
To: shanwei; +Cc: schen, netdev
In-Reply-To: <4CDB7BD5.6030204@cn.fujitsu.com>
From: Shan Wei <shanwei@cn.fujitsu.com>
Date: Thu, 11 Nov 2010 13:15:01 +0800
> In tcp_connect_init(), if tcp_header_len includes TCPOLEN_TSTAMP_ALIGNED(12 bytes)
> and TCPOLEN_MD5SIG_ALIGNED(20 bytes).
If you knew this, why didn't you mention it in your initial report? :-/
I'll make the minimum 64 or something like that.
^ permalink raw reply
* Re: possible kernel oops from user MSS
From: Shan Wei @ 2010-11-11 5:15 UTC (permalink / raw)
To: David Miller; +Cc: schen, netdev
In-Reply-To: <20101110.124119.102563803.davem@davemloft.net>
David Miller wrote, at 11/11/2010 04:41 AM:
> From: Steve Chen <schen@mvista.com>
> Date: Wed, 10 Nov 2010 07:24:51 -0600
>
>> With commit f5fff5dc8a7a3f395b0525c02ba92c95d42b7390, a user program
>> can pass in TCP_MAXSEG of 12 (or TCPOLEN_TSTAMP_ALIGNED), and cause
>> kernel oops with division by 0
>> in tcp_select_initial_window. One way to prevent it is to change the
>> minimum value for TCP_MAXSEG in do_tcp_setsockopt from 8 to some value
>> over 12. Two questions.
>>
>> 1. Is this the right solution?
>> 2. If it is, what is a good minimum value?
>
> Thanks Steve, I'll fix this like so:
>
> --------------------
> tcp: Increase TCP_MAXSEG socket option minimum.
>
> As noted by Steve Chen, since commit
> f5fff5dc8a7a3f395b0525c02ba92c95d42b7390 ("tcp: advertise MSS
> requested by user") we can end up with a situation where
> tcp_select_initial_window() does a divide by a zero (or
> even negative) mss value.
>
> The problem is that sometimes we subtract TCPOLEN_TSTAMP_ALIGNED
> from the mss.
>
> Fix this by increasing the minimum from 8 to 8 plus the value
> of TCPOLEN_TSTATMP_ALIGNED.
In tcp_connect_init(), if tcp_header_len includes TCPOLEN_TSTAMP_ALIGNED(12 bytes)
and TCPOLEN_MD5SIG_ALIGNED(20 bytes).
This fix is still not perfect.
The minimum value of TCP_MAXSEG is 20 tytes, tcp_select_initial_window() still be
called with negative mss value.
--
Best Regards
-----
Shan Wei
> Reported-by: Steve Chen <schen@mvista.com>
> Signed-off-by: David S. Miller <davem@davemloft.net>
> ---
> net/ipv4/tcp.c | 2 +-
> 1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
> index 245603c..6b0eb4d 100644
> --- a/net/ipv4/tcp.c
> +++ b/net/ipv4/tcp.c
> @@ -2246,7 +2246,7 @@ static int do_tcp_setsockopt(struct sock *sk, int level,
> /* Values greater than interface MTU won't take effect. However
> * at the point when this call is done we typically don't yet
> * know which interface is going to be used */
> - if (val < 8 || val > MAX_TCP_WINDOW) {
> + if (val < TCPOLEN_TSTAMP_ALIGNED + 8 || val > MAX_TCP_WINDOW) {
> err = -EINVAL;
> break;
> }
^ permalink raw reply
* [PATCH] drivers/isdn/hisax: Add printf format/argument verification and fix fallout
From: Joe Perches @ 2010-11-11 4:54 UTC (permalink / raw)
To: linux-kernel; +Cc: Karsten Keil, netdev
Add __attribute__((format... to several functins
Make formats and arguments match.
Signed-off-by: Joe Perches <joe@perches.com>
---
drivers/isdn/hisax/avm_pci.c | 2 +-
drivers/isdn/hisax/callc.c | 4 ++--
drivers/isdn/hisax/hfc_2bds0.c | 4 ++--
drivers/isdn/hisax/hfc_2bs0.c | 2 +-
drivers/isdn/hisax/hfc_pci.c | 4 ++--
drivers/isdn/hisax/hfc_sx.c | 6 +++---
drivers/isdn/hisax/hisax.h | 2 ++
drivers/isdn/hisax/ipacx.c | 2 +-
drivers/isdn/hisax/isar.c | 6 +++---
drivers/isdn/hisax/isdnl1.h | 1 +
drivers/isdn/hisax/isdnl3.c | 2 +-
drivers/isdn/hisax/netjet.c | 10 +++++-----
drivers/isdn/hisax/st5481_d.c | 6 ++++--
13 files changed, 28 insertions(+), 23 deletions(-)
diff --git a/drivers/isdn/hisax/avm_pci.c b/drivers/isdn/hisax/avm_pci.c
index fcf4ed1..0e66af1 100644
--- a/drivers/isdn/hisax/avm_pci.c
+++ b/drivers/isdn/hisax/avm_pci.c
@@ -314,7 +314,7 @@ hdlc_fill_fifo(struct BCState *bcs)
bcs->hw.hdlc.ctrl.sr.cmd |= HDLC_CMD_XME;
}
if ((cs->debug & L1_DEB_HSCX) && !(cs->debug & L1_DEB_HSCX_FIFO))
- debugl1(cs, "hdlc_fill_fifo %d/%ld", count, bcs->tx_skb->len);
+ debugl1(cs, "hdlc_fill_fifo %d/%u", count, bcs->tx_skb->len);
p = bcs->tx_skb->data;
ptr = (u_int *)p;
skb_pull(bcs->tx_skb, count);
diff --git a/drivers/isdn/hisax/callc.c b/drivers/isdn/hisax/callc.c
index f150330..37e685e 100644
--- a/drivers/isdn/hisax/callc.c
+++ b/drivers/isdn/hisax/callc.c
@@ -65,7 +65,7 @@ hisax_findcard(int driverid)
return (struct IsdnCardState *) 0;
}
-static void
+static __attribute__((format(printf, 3, 4))) void
link_debug(struct Channel *chanp, int direction, char *fmt, ...)
{
va_list args;
@@ -1068,7 +1068,7 @@ init_d_st(struct Channel *chanp)
return 0;
}
-static void
+static __attribute__((format(printf, 2, 3))) void
callc_debug(struct FsmInst *fi, char *fmt, ...)
{
va_list args;
diff --git a/drivers/isdn/hisax/hfc_2bds0.c b/drivers/isdn/hisax/hfc_2bds0.c
index 7250f56..a16459a 100644
--- a/drivers/isdn/hisax/hfc_2bds0.c
+++ b/drivers/isdn/hisax/hfc_2bds0.c
@@ -292,7 +292,7 @@ hfc_fill_fifo(struct BCState *bcs)
}
count = GetFreeFifoBytes_B(bcs);
if (cs->debug & L1_DEB_HSCX)
- debugl1(cs, "hfc_fill_fifo %d count(%ld/%d),%lx",
+ debugl1(cs, "hfc_fill_fifo %d count(%u/%d),%lx",
bcs->channel, bcs->tx_skb->len,
count, current->state);
if (count < bcs->tx_skb->len) {
@@ -719,7 +719,7 @@ hfc_fill_dfifo(struct IsdnCardState *cs)
}
count = GetFreeFifoBytes_D(cs);
if (cs->debug & L1_DEB_ISAC)
- debugl1(cs, "hfc_fill_Dfifo count(%ld/%d)",
+ debugl1(cs, "hfc_fill_Dfifo count(%u/%d)",
cs->tx_skb->len, count);
if (count < cs->tx_skb->len) {
if (cs->debug & L1_DEB_ISAC)
diff --git a/drivers/isdn/hisax/hfc_2bs0.c b/drivers/isdn/hisax/hfc_2bs0.c
index b1f6481..626f85d 100644
--- a/drivers/isdn/hisax/hfc_2bs0.c
+++ b/drivers/isdn/hisax/hfc_2bs0.c
@@ -282,7 +282,7 @@ hfc_fill_fifo(struct BCState *bcs)
count += cs->hw.hfc.fifosize;
} /* L1_MODE_TRANS */
if (cs->debug & L1_DEB_HSCX)
- debugl1(cs, "hfc_fill_fifo %d count(%ld/%d)",
+ debugl1(cs, "hfc_fill_fifo %d count(%u/%d)",
bcs->channel, bcs->tx_skb->len,
count);
if (count < bcs->tx_skb->len) {
diff --git a/drivers/isdn/hisax/hfc_pci.c b/drivers/isdn/hisax/hfc_pci.c
index 917cc84..3147020 100644
--- a/drivers/isdn/hisax/hfc_pci.c
+++ b/drivers/isdn/hisax/hfc_pci.c
@@ -550,7 +550,7 @@ hfcpci_fill_dfifo(struct IsdnCardState *cs)
count += D_FIFO_SIZE; /* count now contains available bytes */
if (cs->debug & L1_DEB_ISAC)
- debugl1(cs, "hfcpci_fill_Dfifo count(%ld/%d)",
+ debugl1(cs, "hfcpci_fill_Dfifo count(%u/%d)",
cs->tx_skb->len, count);
if (count < cs->tx_skb->len) {
if (cs->debug & L1_DEB_ISAC)
@@ -681,7 +681,7 @@ hfcpci_fill_fifo(struct BCState *bcs)
count += B_FIFO_SIZE; /* count now contains available bytes */
if (cs->debug & L1_DEB_HSCX)
- debugl1(cs, "hfcpci_fill_fifo %d count(%ld/%d),%lx",
+ debugl1(cs, "hfcpci_fill_fifo %d count(%u/%d),%lx",
bcs->channel, bcs->tx_skb->len,
count, current->state);
diff --git a/drivers/isdn/hisax/hfc_sx.c b/drivers/isdn/hisax/hfc_sx.c
index 5aa138e..1235b71 100644
--- a/drivers/isdn/hisax/hfc_sx.c
+++ b/drivers/isdn/hisax/hfc_sx.c
@@ -179,7 +179,7 @@ write_fifo(struct IsdnCardState *cs, struct sk_buff *skb, u_char fifo, int trans
count += fifo_size; /* count now contains available bytes */
if (cs->debug & L1_DEB_ISAC_FIFO)
- debugl1(cs, "hfcsx_write_fifo %d count(%ld/%d)",
+ debugl1(cs, "hfcsx_write_fifo %d count(%u/%d)",
fifo, skb->len, count);
if (count < skb->len) {
if (cs->debug & L1_DEB_ISAC_FIFO)
@@ -265,7 +265,7 @@ read_fifo(struct IsdnCardState *cs, u_char fifo, int trans_max)
count++;
if (cs->debug & L1_DEB_ISAC_FIFO)
- debugl1(cs, "hfcsx_read_fifo %d count %ld)",
+ debugl1(cs, "hfcsx_read_fifo %d count %u)",
fifo, count);
if ((count > fifo_size) || (count < 4)) {
@@ -986,7 +986,7 @@ HFCSX_l1hw(struct PStack *st, int pr, void *arg)
default:
spin_unlock_irqrestore(&cs->lock, flags);
if (cs->debug & L1_DEB_WARN)
- debugl1(cs, "hfcsx_l1hw loop invalid %4lx", arg);
+ debugl1(cs, "hfcsx_l1hw loop invalid %4lx", (unsigned long)arg);
return;
}
cs->hw.hfcsx.trm |= 0x80; /* enable IOM-loop */
diff --git a/drivers/isdn/hisax/hisax.h b/drivers/isdn/hisax/hisax.h
index 32ab392..de1c669 100644
--- a/drivers/isdn/hisax/hisax.h
+++ b/drivers/isdn/hisax/hisax.h
@@ -1286,7 +1286,9 @@ int jiftime(char *s, long mark);
int HiSax_command(isdn_ctrl * ic);
int HiSax_writebuf_skb(int id, int chan, int ack, struct sk_buff *skb);
+__attribute__((format(printf, 3, 4)))
void HiSax_putstatus(struct IsdnCardState *cs, char *head, char *fmt, ...);
+__attribute__((format(printf, 3, 0)))
void VHiSax_putstatus(struct IsdnCardState *cs, char *head, char *fmt, va_list args);
void HiSax_reportcard(int cardnr, int sel);
int QuickHex(char *txt, u_char * p, int cnt);
diff --git a/drivers/isdn/hisax/ipacx.c b/drivers/isdn/hisax/ipacx.c
index 751b25f..3321041 100644
--- a/drivers/isdn/hisax/ipacx.c
+++ b/drivers/isdn/hisax/ipacx.c
@@ -717,7 +717,7 @@ bch_mode(struct BCState *bcs, int mode, int bc)
bc = bc ? 1 : 0; // in case bc is greater than 1
if (cs->debug & L1_DEB_HSCX)
- debugl1(cs, "mode_bch() switch B-% mode %d chan %d", hscx, mode, bc);
+ debugl1(cs, "mode_bch() switch B-%d mode %d chan %d", hscx, mode, bc);
bcs->mode = mode;
bcs->channel = bc;
diff --git a/drivers/isdn/hisax/isar.c b/drivers/isdn/hisax/isar.c
index 2e72227..1be4552 100644
--- a/drivers/isdn/hisax/isar.c
+++ b/drivers/isdn/hisax/isar.c
@@ -953,7 +953,7 @@ isar_pump_statev_modem(struct BCState *bcs, u_char devt) {
break;
case PSEV_GSTN_CLR:
if (cs->debug & L1_DEB_HSCX)
- debugl1(cs, "pump stev GSTN CLEAR", devt);
+ debugl1(cs, "pump stev GSTN CLEAR");
break;
default:
if (cs->debug & L1_DEB_HSCX)
@@ -1268,7 +1268,7 @@ isar_int_main(struct IsdnCardState *cs)
static void
ftimer_handler(struct BCState *bcs) {
if (bcs->cs->debug)
- debugl1(bcs->cs, "ftimer flags %04x",
+ debugl1(bcs->cs, "ftimer flags %04lx",
bcs->Flag);
test_and_clear_bit(BC_FLG_FTI_RUN, &bcs->Flag);
if (test_and_clear_bit(BC_FLG_LL_CONN, &bcs->Flag)) {
@@ -1748,7 +1748,7 @@ isar_auxcmd(struct IsdnCardState *cs, isdn_ctrl *ic) {
struct BCState *bcs;
if (cs->debug & L1_DEB_HSCX)
- debugl1(cs, "isar_auxcmd cmd/ch %x/%d", ic->command, ic->arg);
+ debugl1(cs, "isar_auxcmd cmd/ch %x/%ld", ic->command, ic->arg);
switch (ic->command) {
case (ISDN_CMD_FAXCMD):
bcs = cs->channel[ic->arg].bcs;
diff --git a/drivers/isdn/hisax/isdnl1.h b/drivers/isdn/hisax/isdnl1.h
index 172ad4c..425d861 100644
--- a/drivers/isdn/hisax/isdnl1.h
+++ b/drivers/isdn/hisax/isdnl1.h
@@ -21,6 +21,7 @@
#define B_XMTBUFREADY 1
#define B_ACKPENDING 2
+__attribute__((format(printf, 2, 3)))
void debugl1(struct IsdnCardState *cs, char *fmt, ...);
void DChannel_proc_xmt(struct IsdnCardState *cs);
void DChannel_proc_rcv(struct IsdnCardState *cs);
diff --git a/drivers/isdn/hisax/isdnl3.c b/drivers/isdn/hisax/isdnl3.c
index fd0b643..ad291f2 100644
--- a/drivers/isdn/hisax/isdnl3.c
+++ b/drivers/isdn/hisax/isdnl3.c
@@ -66,7 +66,7 @@ static char *strL3Event[] =
"EV_TIMEOUT",
};
-static void
+static __attribute__((format(printf, 2, 3))) void
l3m_debug(struct FsmInst *fi, char *fmt, ...)
{
va_list args;
diff --git a/drivers/isdn/hisax/netjet.c b/drivers/isdn/hisax/netjet.c
index 5d7f0f2..644891e 100644
--- a/drivers/isdn/hisax/netjet.c
+++ b/drivers/isdn/hisax/netjet.c
@@ -254,7 +254,7 @@ static int make_raw_data(struct BCState *bcs) {
val >>= 1;
}
if (bcs->cs->debug & L1_DEB_HSCX)
- debugl1(bcs->cs,"tiger make_raw: in %ld out %d.%d",
+ debugl1(bcs->cs,"tiger make_raw: in %u out %d.%d",
bcs->tx_skb->len, s_cnt, bitcnt);
if (bitcnt) {
while (8>bitcnt++) {
@@ -361,7 +361,7 @@ static int make_raw_data_56k(struct BCState *bcs) {
val >>= 1;
}
if (bcs->cs->debug & L1_DEB_HSCX)
- debugl1(bcs->cs,"tiger make_raw_56k: in %ld out %d.%d",
+ debugl1(bcs->cs,"tiger make_raw_56k: in %u out %d.%d",
bcs->tx_skb->len, s_cnt, bitcnt);
if (bitcnt) {
while (8>bitcnt++) {
@@ -612,7 +612,7 @@ void netjet_fill_dma(struct BCState *bcs)
if (!bcs->tx_skb)
return;
if (bcs->cs->debug & L1_DEB_HSCX)
- debugl1(bcs->cs,"tiger fill_dma1: c%d %4x", bcs->channel,
+ debugl1(bcs->cs,"tiger fill_dma1: c%d %4lx", bcs->channel,
bcs->Flag);
if (test_and_set_bit(BC_FLG_BUSY, &bcs->Flag))
return;
@@ -625,7 +625,7 @@ void netjet_fill_dma(struct BCState *bcs)
return;
};
if (bcs->cs->debug & L1_DEB_HSCX)
- debugl1(bcs->cs,"tiger fill_dma2: c%d %4x", bcs->channel,
+ debugl1(bcs->cs,"tiger fill_dma2: c%d %4lx", bcs->channel,
bcs->Flag);
if (test_and_clear_bit(BC_FLG_NOFRAME, &bcs->Flag)) {
write_raw(bcs, bcs->hw.tiger.sendp, bcs->hw.tiger.free);
@@ -667,7 +667,7 @@ void netjet_fill_dma(struct BCState *bcs)
write_raw(bcs, p, cnt);
}
if (bcs->cs->debug & L1_DEB_HSCX)
- debugl1(bcs->cs,"tiger fill_dma3: c%d %4x", bcs->channel,
+ debugl1(bcs->cs,"tiger fill_dma3: c%d %4lx", bcs->channel,
bcs->Flag);
}
diff --git a/drivers/isdn/hisax/st5481_d.c b/drivers/isdn/hisax/st5481_d.c
index b7876b1..4408263 100644
--- a/drivers/isdn/hisax/st5481_d.c
+++ b/drivers/isdn/hisax/st5481_d.c
@@ -167,7 +167,8 @@ static struct FsmNode L1FnList[] __initdata =
{ST_L1_F8, EV_IND_RSY, l1_ignore},
};
-static void l1m_debug(struct FsmInst *fi, char *fmt, ...)
+static __attribute__((format(printf, 2, 3)))
+void l1m_debug(struct FsmInst *fi, char *fmt, ...)
{
va_list args;
char buf[256];
@@ -269,7 +270,8 @@ static char *strDoutEvent[] =
"EV_DOUT_UNDERRUN",
};
-static void dout_debug(struct FsmInst *fi, char *fmt, ...)
+static __attribute__((format(printf, 2, 3)))
+void dout_debug(struct FsmInst *fi, char *fmt, ...)
{
va_list args;
char buf[256];
--
1.7.3.1.g432b3.dirty
^ permalink raw reply related
* Re: [PATCH 01/10] cxgb4vf: minor comment/symbolic name cleanup.
From: Joe Perches @ 2010-11-11 3:19 UTC (permalink / raw)
To: Casey Leedom; +Cc: netdev, davem
In-Reply-To: <1289441052-4925-1-git-send-email-leedom@chelsio.com>
On Wed, 2010-11-10 at 18:04 -0800, Casey Leedom wrote:
> Minor cleanup of comments and symbolic constant names for clarity.
> diff --git a/drivers/net/cxgb4vf/cxgb4vf_main.c b/drivers/net/cxgb4vf/cxgb4vf_main.c
> index 6de5e2e..f5259a1 100644
> --- a/drivers/net/cxgb4vf/cxgb4vf_main.c
> +++ b/drivers/net/cxgb4vf/cxgb4vf_main.c
> @@ -280,7 +280,7 @@ static void name_msix_vecs(struct adapter *adapter)
> const struct port_info *pi = netdev_priv(dev);
> int qs, msi;
>
> - for (qs = 0, msi = MSIX_NIQFLINT;
> + for (qs = 0, msi = MSIX_IQFLINT;
> qs < pi->nqsets;
> qs++, msi++) {
This for now fits on a single line.
> diff --git a/drivers/net/cxgb4vf/t4vf_hw.c b/drivers/net/cxgb4vf/t4vf_hw.c
> index ea1c123..2180181 100644
> --- a/drivers/net/cxgb4vf/t4vf_hw.c
> +++ b/drivers/net/cxgb4vf/t4vf_hw.c
> @@ -1257,7 +1257,7 @@ int t4vf_eth_eq_free(struct adapter *adapter, unsigned int eqid)
> */
> int t4vf_handle_fw_rpl(struct adapter *adapter, const __be64 *rpl)
> - struct fw_cmd_hdr *cmd_hdr = (struct fw_cmd_hdr *)rpl;
> + const struct fw_cmd_hdr *cmd_hdr = (const struct fw_cmd_hdr *)rpl;
> @@ -1265,7 +1265,7 @@ int t4vf_handle_fw_rpl(struct adapter *adapter, const __be64 *rpl)
> - const struct fw_port_cmd *port_cmd = (void *)rpl;
> + const struct fw_port_cmd *port_cmd = (const void *)rpl;
might be better to have a consistent casting style.
1st uses a direct cast, 2nd an implicit one.
^ permalink raw reply
* [PATCH 10/10] cxgb4vf: add call to Firmware to reset VF State.
From: Casey Leedom @ 2010-11-11 2:04 UTC (permalink / raw)
To: netdev; +Cc: davem, Casey Leedom
In-Reply-To: <450C6E98-D0D9-42A6-9F15-4DB98647DB04@chelsio.com>
Add call to Firmware to reset its VF State when we first attach to the VF.
Signed-off-by: Casey Leedom <leedom@chelsio.com>
---
drivers/net/cxgb4vf/cxgb4vf_main.c | 16 ++++++++++++++++
drivers/net/cxgb4vf/t4vf_common.h | 1 +
drivers/net/cxgb4vf/t4vf_hw.c | 19 +++++++++++++++++++
3 files changed, 36 insertions(+), 0 deletions(-)
diff --git a/drivers/net/cxgb4vf/cxgb4vf_main.c b/drivers/net/cxgb4vf/cxgb4vf_main.c
index 15b14ef..3456a9b 100644
--- a/drivers/net/cxgb4vf/cxgb4vf_main.c
+++ b/drivers/net/cxgb4vf/cxgb4vf_main.c
@@ -2076,6 +2076,22 @@ static int __devinit adap_init0(struct adapter *adapter)
}
/*
+ * Some environments do not properly handle PCIE FLRs -- e.g. in Linux
+ * 2.6.31 and later we can't call pci_reset_function() in order to
+ * issue an FLR because of a self- deadlock on the device semaphore.
+ * Meanwhile, the OS infrastructure doesn't issue FLRs in all the
+ * cases where they're needed -- for instance, some versions of KVM
+ * fail to reset "Assigned Devices" when the VM reboots. Therefore we
+ * use the firmware based reset in order to reset any per function
+ * state.
+ */
+ err = t4vf_fw_reset(adapter);
+ if (err < 0) {
+ dev_err(adapter->pdev_dev, "FW reset failed: err=%d\n", err);
+ return err;
+ }
+
+ /*
* Grab basic operational parameters. These will predominantly have
* been set up by the Physical Function Driver or will be hard coded
* into the adapter. We just have to live with them ... Note that
diff --git a/drivers/net/cxgb4vf/t4vf_common.h b/drivers/net/cxgb4vf/t4vf_common.h
index 659616b..7541a60 100644
--- a/drivers/net/cxgb4vf/t4vf_common.h
+++ b/drivers/net/cxgb4vf/t4vf_common.h
@@ -235,6 +235,7 @@ static inline int t4vf_wr_mbox_ns(struct adapter *adapter, const void *cmd,
int __devinit t4vf_wait_dev_ready(struct adapter *);
int __devinit t4vf_port_init(struct adapter *, int);
+int t4vf_fw_reset(struct adapter *);
int t4vf_query_params(struct adapter *, unsigned int, const u32 *, u32 *);
int t4vf_set_params(struct adapter *, unsigned int, const u32 *, const u32 *);
diff --git a/drivers/net/cxgb4vf/t4vf_hw.c b/drivers/net/cxgb4vf/t4vf_hw.c
index 2180181..591f161 100644
--- a/drivers/net/cxgb4vf/t4vf_hw.c
+++ b/drivers/net/cxgb4vf/t4vf_hw.c
@@ -326,6 +326,25 @@ int __devinit t4vf_port_init(struct adapter *adapter, int pidx)
}
/**
+ * t4vf_fw_reset - issue a reset to FW
+ * @adapter: the adapter
+ *
+ * Issues a reset command to FW. For a Physical Function this would
+ * result in the Firmware reseting all of its state. For a Virtual
+ * Function this just resets the state associated with the VF.
+ */
+int t4vf_fw_reset(struct adapter *adapter)
+{
+ struct fw_reset_cmd cmd;
+
+ memset(&cmd, 0, sizeof(cmd));
+ cmd.op_to_write = cpu_to_be32(FW_CMD_OP(FW_RESET_CMD) |
+ FW_CMD_WRITE);
+ cmd.retval_len16 = cpu_to_be32(FW_LEN16(cmd));
+ return t4vf_wr_mbox(adapter, &cmd, sizeof(cmd), NULL);
+}
+
+/**
* t4vf_query_params - query FW or device parameters
* @adapter: the adapter
* @nparams: the number of parameters
--
1.7.0.4
^ permalink raw reply related
* [PATCH 09/10] cxgb4vf: Mark "UDP [RSS Hash] Enable" as a 1-bit field.
From: Casey Leedom @ 2010-11-11 2:04 UTC (permalink / raw)
To: netdev; +Cc: davem, Casey Leedom
In-Reply-To: <450C6E98-D0D9-42A6-9F15-4DB98647DB04@chelsio.com>
Mark the UDP RSS Hash Enable field as 1-bit in length. Also clean up
formatting from previous changeset which changed the RSS 1-bit fields from
"int" to "unsigned int".
Signed-off-by: Casey Leedom <leedom@chelsio.com>
---
drivers/net/cxgb4vf/t4vf_common.h | 28 ++++++++++++++--------------
1 files changed, 14 insertions(+), 14 deletions(-)
diff --git a/drivers/net/cxgb4vf/t4vf_common.h b/drivers/net/cxgb4vf/t4vf_common.h
index 873cb7d..659616b 100644
--- a/drivers/net/cxgb4vf/t4vf_common.h
+++ b/drivers/net/cxgb4vf/t4vf_common.h
@@ -132,15 +132,15 @@ struct rss_params {
unsigned int mode; /* RSS mode */
union {
struct {
- unsigned int synmapen:1; /* SYN Map Enable */
- unsigned int syn4tupenipv6:1; /* enable hashing 4-tuple IPv6 SYNs */
- unsigned int syn2tupenipv6:1; /* enable hashing 2-tuple IPv6 SYNs */
- unsigned int syn4tupenipv4:1; /* enable hashing 4-tuple IPv4 SYNs */
- unsigned int syn2tupenipv4:1; /* enable hashing 2-tuple IPv4 SYNs */
- unsigned int ofdmapen:1; /* Offload Map Enable */
- unsigned int tnlmapen:1; /* Tunnel Map Enable */
- unsigned int tnlalllookup:1; /* Tunnel All Lookup */
- unsigned int hashtoeplitz:1; /* use Toeplitz hash */
+ uint synmapen:1; /* SYN Map Enable */
+ uint syn4tupenipv6:1; /* enable hashing 4-tuple IPv6 SYNs */
+ uint syn2tupenipv6:1; /* enable hashing 2-tuple IPv6 SYNs */
+ uint syn4tupenipv4:1; /* enable hashing 4-tuple IPv4 SYNs */
+ uint syn2tupenipv4:1; /* enable hashing 2-tuple IPv4 SYNs */
+ uint ofdmapen:1; /* Offload Map Enable */
+ uint tnlmapen:1; /* Tunnel Map Enable */
+ uint tnlalllookup:1; /* Tunnel All Lookup */
+ uint hashtoeplitz:1; /* use Toeplitz hash */
} basicvirtual;
} u;
};
@@ -151,11 +151,11 @@ struct rss_params {
union rss_vi_config {
struct {
u16 defaultq; /* Ingress Queue ID for !tnlalllookup */
- unsigned int ip6fourtupen:1; /* hash 4-tuple IPv6 ingress packets */
- unsigned int ip6twotupen:1; /* hash 2-tuple IPv6 ingress packets */
- unsigned int ip4fourtupen:1; /* hash 4-tuple IPv4 ingress packets */
- unsigned int ip4twotupen:1; /* hash 2-tuple IPv4 ingress packets */
- int udpen; /* hash 4-tuple UDP ingress packets */
+ uint ip6fourtupen:1; /* hash 4-tuple IPv6 ingress packets */
+ uint ip6twotupen:1; /* hash 2-tuple IPv6 ingress packets */
+ uint ip4fourtupen:1; /* hash 4-tuple IPv4 ingress packets */
+ uint ip4twotupen:1; /* hash 2-tuple IPv4 ingress packets */
+ uint udpen:1; /* hash 4-tuple UDP ingress packets */
} basicvirtual;
};
--
1.7.0.4
^ permalink raw reply related
* [PATCH 08/10] cxgb4vf: Advertise NETIF_F_TSO_ECN.
From: Casey Leedom @ 2010-11-11 2:04 UTC (permalink / raw)
To: netdev; +Cc: davem, Casey Leedom
In-Reply-To: <450C6E98-D0D9-42A6-9F15-4DB98647DB04@chelsio.com>
Advertise NETIF_F_TSO_ECN.
Signed-off-by: Casey Leedom <leedom@chelsio.com>
---
drivers/net/cxgb4vf/cxgb4vf_main.c | 11 ++++++++---
1 files changed, 8 insertions(+), 3 deletions(-)
diff --git a/drivers/net/cxgb4vf/cxgb4vf_main.c b/drivers/net/cxgb4vf/cxgb4vf_main.c
index 2d09e83..15b14ef 100644
--- a/drivers/net/cxgb4vf/cxgb4vf_main.c
+++ b/drivers/net/cxgb4vf/cxgb4vf_main.c
@@ -1536,14 +1536,19 @@ static void cxgb4vf_get_wol(struct net_device *dev,
}
/*
+ * TCP Segmentation Offload flags which we support.
+ */
+#define TSO_FLAGS (NETIF_F_TSO | NETIF_F_TSO6 | NETIF_F_TSO_ECN)
+
+/*
* Set TCP Segmentation Offloading feature capabilities.
*/
static int cxgb4vf_set_tso(struct net_device *dev, u32 tso)
{
if (tso)
- dev->features |= NETIF_F_TSO | NETIF_F_TSO6;
+ dev->features |= TSO_FLAGS;
else
- dev->features &= ~(NETIF_F_TSO | NETIF_F_TSO6);
+ dev->features &= ~TSO_FLAGS;
return 0;
}
@@ -2596,7 +2601,7 @@ static int __devinit cxgb4vf_pci_probe(struct pci_dev *pdev,
netif_carrier_off(netdev);
netdev->irq = pdev->irq;
- netdev->features = (NETIF_F_SG | NETIF_F_TSO | NETIF_F_TSO6 |
+ netdev->features = (NETIF_F_SG | TSO_FLAGS |
NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM |
NETIF_F_HW_VLAN_TX | NETIF_F_HW_VLAN_RX |
NETIF_F_GRO);
--
1.7.0.4
^ permalink raw reply related
* [PATCH 07/10] cxgb4vf: Fail open if link_start() fails.
From: Casey Leedom @ 2010-11-11 2:04 UTC (permalink / raw)
To: netdev; +Cc: davem, Casey Leedom
In-Reply-To: <450C6E98-D0D9-42A6-9F15-4DB98647DB04@chelsio.com>
Fail open if link_start() fails.
Signed-off-by: Casey Leedom <leedom@chelsio.com>
---
drivers/net/cxgb4vf/cxgb4vf_main.c | 4 +++-
1 files changed, 3 insertions(+), 1 deletions(-)
diff --git a/drivers/net/cxgb4vf/cxgb4vf_main.c b/drivers/net/cxgb4vf/cxgb4vf_main.c
index b5c0bff..2d09e83 100644
--- a/drivers/net/cxgb4vf/cxgb4vf_main.c
+++ b/drivers/net/cxgb4vf/cxgb4vf_main.c
@@ -753,7 +753,9 @@ static int cxgb4vf_open(struct net_device *dev)
if (err)
return err;
set_bit(pi->port_id, &adapter->open_device_map);
- link_start(dev);
+ err = link_start(dev);
+ if (err)
+ return err;
netif_tx_start_all_queues(dev);
return 0;
}
--
1.7.0.4
^ permalink raw reply related
* [PATCH 05/10] cxgb4vf: flesh out PCI Device ID Table ...
From: Casey Leedom @ 2010-11-11 2:04 UTC (permalink / raw)
To: netdev; +Cc: davem, Casey Leedom
In-Reply-To: <450C6E98-D0D9-42A6-9F15-4DB98647DB04@chelsio.com>
Add a bunch of T4 Device IDs for the VF Driver.
Signed-off-by: Casey Leedom <leedom@chelsio.com>
---
drivers/net/cxgb4vf/cxgb4vf_main.c | 8 ++++++++
1 files changed, 8 insertions(+), 0 deletions(-)
diff --git a/drivers/net/cxgb4vf/cxgb4vf_main.c b/drivers/net/cxgb4vf/cxgb4vf_main.c
index 9e9e4f6..75b85ca 100644
--- a/drivers/net/cxgb4vf/cxgb4vf_main.c
+++ b/drivers/net/cxgb4vf/cxgb4vf_main.c
@@ -2834,6 +2834,14 @@ static struct pci_device_id cxgb4vf_pci_tbl[] = {
CH_DEVICE(0x4800, 0), /* T440-dbg */
CH_DEVICE(0x4801, 0), /* T420-cr */
CH_DEVICE(0x4802, 0), /* T422-cr */
+ CH_DEVICE(0x4803, 0), /* T440-cr */
+ CH_DEVICE(0x4804, 0), /* T420-bch */
+ CH_DEVICE(0x4805, 0), /* T440-bch */
+ CH_DEVICE(0x4806, 0), /* T460-ch */
+ CH_DEVICE(0x4807, 0), /* T420-so */
+ CH_DEVICE(0x4808, 0), /* T420-cx */
+ CH_DEVICE(0x4809, 0), /* T420-bt */
+ CH_DEVICE(0x480a, 0), /* T404-bt */
{ 0, }
};
--
1.7.0.4
^ permalink raw reply related
* [PATCH 06/10] cxgb4vf: fix up "Section Mismatch" compiler warning.
From: Casey Leedom @ 2010-11-11 2:04 UTC (permalink / raw)
To: netdev; +Cc: davem, Casey Leedom
In-Reply-To: <450C6E98-D0D9-42A6-9F15-4DB98647DB04@chelsio.com>
Fix up "Section Mismatch" compiler warning and mark another routine as
__devinit.
Signed-off-by: Casey Leedom <leedom@chelsio.com>
---
drivers/net/cxgb4vf/cxgb4vf_main.c | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/cxgb4vf/cxgb4vf_main.c b/drivers/net/cxgb4vf/cxgb4vf_main.c
index 75b85ca..b5c0bff 100644
--- a/drivers/net/cxgb4vf/cxgb4vf_main.c
+++ b/drivers/net/cxgb4vf/cxgb4vf_main.c
@@ -2032,7 +2032,7 @@ static int __devinit setup_debugfs(struct adapter *adapter)
* Tear down the /sys/kernel/debug/cxgb4vf sub-nodes created above. We leave
* it to our caller to tear down the directory (debugfs_root).
*/
-static void __devexit cleanup_debugfs(struct adapter *adapter)
+static void cleanup_debugfs(struct adapter *adapter)
{
BUG_ON(adapter->debugfs_root == NULL);
@@ -2050,7 +2050,7 @@ static void __devexit cleanup_debugfs(struct adapter *adapter)
* adapter parameters we're going to be using and initialize basic adapter
* hardware support.
*/
-static int adap_init0(struct adapter *adapter)
+static int __devinit adap_init0(struct adapter *adapter)
{
struct vf_resources *vfres = &adapter->params.vfres;
struct sge_params *sge_params = &adapter->params.sge;
--
1.7.0.4
^ permalink raw reply related
* [PATCH 03/10] cxgb4vf: fix bug in Generic Receive Offload
From: Casey Leedom @ 2010-11-11 2:04 UTC (permalink / raw)
To: netdev; +Cc: davem, Casey Leedom
In-Reply-To: <450C6E98-D0D9-42A6-9F15-4DB98647DB04@chelsio.com>
Fix botch in Generic Receive Offload (the Packet Gather List Total length
field wasn't being initialized). Also add ethtool statistics for GRO.
Signed-off-by: Casey Leedom <leedom@chelsio.com>
---
drivers/net/cxgb4vf/cxgb4vf_main.c | 6 ++++++
drivers/net/cxgb4vf/sge.c | 1 +
2 files changed, 7 insertions(+), 0 deletions(-)
diff --git a/drivers/net/cxgb4vf/cxgb4vf_main.c b/drivers/net/cxgb4vf/cxgb4vf_main.c
index 224b36c..9e9e4f6 100644
--- a/drivers/net/cxgb4vf/cxgb4vf_main.c
+++ b/drivers/net/cxgb4vf/cxgb4vf_main.c
@@ -1346,6 +1346,8 @@ struct queue_port_stats {
u64 rx_csum;
u64 vlan_ex;
u64 vlan_ins;
+ u64 lro_pkts;
+ u64 lro_merged;
};
/*
@@ -1383,6 +1385,8 @@ static const char stats_strings[][ETH_GSTRING_LEN] = {
"RxCsumGood ",
"VLANextractions ",
"VLANinsertions ",
+ "GROPackets ",
+ "GROMerged ",
};
/*
@@ -1432,6 +1436,8 @@ static void collect_sge_port_stats(const struct adapter *adapter,
stats->rx_csum += rxq->stats.rx_cso;
stats->vlan_ex += rxq->stats.vlan_ex;
stats->vlan_ins += txq->vlan_ins;
+ stats->lro_pkts += rxq->stats.lro_pkts;
+ stats->lro_merged += rxq->stats.lro_merged;
}
}
diff --git a/drivers/net/cxgb4vf/sge.c b/drivers/net/cxgb4vf/sge.c
index d791adb..1767bda 100644
--- a/drivers/net/cxgb4vf/sge.c
+++ b/drivers/net/cxgb4vf/sge.c
@@ -1682,6 +1682,7 @@ int process_responses(struct sge_rspq *rspq, int budget)
}
len = RSPD_LEN(len);
}
+ gl.tot_len = len;
/*
* Gather packet fragments.
--
1.7.0.4
^ permalink raw reply related
* [PATCH 04/10] cxgb4vf: fix some errors in Gather List to skb conversion
From: Casey Leedom @ 2010-11-11 2:04 UTC (permalink / raw)
To: netdev; +Cc: davem, Casey Leedom
In-Reply-To: <450C6E98-D0D9-42A6-9F15-4DB98647DB04@chelsio.com>
There were some errors in the way that internal Gather Lists were being
translated into skb's. This also makes the VF Driver look more like the PF
Driver to facilitate easier comarison.
Signed-off-by: Casey Leedom <leedom@chelsio.com>
---
drivers/net/cxgb4vf/sge.c | 121 +++++++++++++++++++++++++++-----------------
1 files changed, 74 insertions(+), 47 deletions(-)
diff --git a/drivers/net/cxgb4vf/sge.c b/drivers/net/cxgb4vf/sge.c
index 1767bda..e0b3d1b 100644
--- a/drivers/net/cxgb4vf/sge.c
+++ b/drivers/net/cxgb4vf/sge.c
@@ -154,13 +154,14 @@ enum {
*/
RX_COPY_THRES = 256,
RX_PULL_LEN = 128,
-};
-/*
- * Can't define this in the above enum because PKTSHIFT isn't a constant in
- * the VF Driver ...
- */
-#define RX_PKT_PULL_LEN (RX_PULL_LEN + PKTSHIFT)
+ /*
+ * Main body length for sk_buffs used for RX Ethernet packets with
+ * fragments. Should be >= RX_PULL_LEN but possibly bigger to give
+ * pskb_may_pull() some room.
+ */
+ RX_SKB_LEN = 512,
+};
/*
* Software state per TX descriptor.
@@ -1355,6 +1356,67 @@ out_free:
}
/**
+ * t4vf_pktgl_to_skb - build an sk_buff from a packet gather list
+ * @gl: the gather list
+ * @skb_len: size of sk_buff main body if it carries fragments
+ * @pull_len: amount of data to move to the sk_buff's main body
+ *
+ * Builds an sk_buff from the given packet gather list. Returns the
+ * sk_buff or %NULL if sk_buff allocation failed.
+ */
+struct sk_buff *t4vf_pktgl_to_skb(const struct pkt_gl *gl,
+ unsigned int skb_len, unsigned int pull_len)
+{
+ struct sk_buff *skb;
+ struct skb_shared_info *ssi;
+
+ /*
+ * If the ingress packet is small enough, allocate an skb large enough
+ * for all of the data and copy it inline. Otherwise, allocate an skb
+ * with enough room to pull in the header and reference the rest of
+ * the data via the skb fragment list.
+ *
+ * Below we rely on RX_COPY_THRES being less than the smallest Rx
+ * buff! size, which is expected since buffers are at least
+ * PAGE_SIZEd. In this case packets up to RX_COPY_THRES have only one
+ * fragment.
+ */
+ if (gl->tot_len <= RX_COPY_THRES) {
+ /* small packets have only one fragment */
+ skb = alloc_skb(gl->tot_len, GFP_ATOMIC);
+ if (unlikely(!skb))
+ goto out;
+ __skb_put(skb, gl->tot_len);
+ skb_copy_to_linear_data(skb, gl->va, gl->tot_len);
+ } else {
+ skb = alloc_skb(skb_len, GFP_ATOMIC);
+ if (unlikely(!skb))
+ goto out;
+ __skb_put(skb, pull_len);
+ skb_copy_to_linear_data(skb, gl->va, pull_len);
+
+ ssi = skb_shinfo(skb);
+ ssi->frags[0].page = gl->frags[0].page;
+ ssi->frags[0].page_offset = gl->frags[0].page_offset + pull_len;
+ ssi->frags[0].size = gl->frags[0].size - pull_len;
+ if (gl->nfrags > 1)
+ memcpy(&ssi->frags[1], &gl->frags[1],
+ (gl->nfrags-1) * sizeof(skb_frag_t));
+ ssi->nr_frags = gl->nfrags;
+
+ skb->len = gl->tot_len;
+ skb->data_len = skb->len - pull_len;
+ skb->truesize += skb->data_len;
+
+ /* Get a reference for the last page, we don't own it */
+ get_page(gl->frags[gl->nfrags - 1].page);
+ }
+
+out:
+ return skb;
+}
+
+/**
* t4vf_pktgl_free - free a packet gather list
* @gl: the gather list
*
@@ -1463,10 +1525,8 @@ int t4vf_ethrx_handler(struct sge_rspq *rspq, const __be64 *rsp,
{
struct sk_buff *skb;
struct port_info *pi;
- struct skb_shared_info *ssi;
const struct cpl_rx_pkt *pkt = (void *)&rsp[1];
bool csum_ok = pkt->csum_calc && !pkt->err_vec;
- unsigned int len = be16_to_cpu(pkt->len);
struct sge_eth_rxq *rxq = container_of(rspq, struct sge_eth_rxq, rspq);
/*
@@ -1481,42 +1541,14 @@ int t4vf_ethrx_handler(struct sge_rspq *rspq, const __be64 *rsp,
}
/*
- * If the ingress packet is small enough, allocate an skb large enough
- * for all of the data and copy it inline. Otherwise, allocate an skb
- * with enough room to pull in the header and reference the rest of
- * the data via the skb fragment list.
+ * Convert the Packet Gather List into an skb.
*/
- if (len <= RX_COPY_THRES) {
- /* small packets have only one fragment */
- skb = alloc_skb(gl->frags[0].size, GFP_ATOMIC);
- if (!skb)
- goto nomem;
- __skb_put(skb, gl->frags[0].size);
- skb_copy_to_linear_data(skb, gl->va, gl->frags[0].size);
- } else {
- skb = alloc_skb(RX_PKT_PULL_LEN, GFP_ATOMIC);
- if (!skb)
- goto nomem;
- __skb_put(skb, RX_PKT_PULL_LEN);
- skb_copy_to_linear_data(skb, gl->va, RX_PKT_PULL_LEN);
-
- ssi = skb_shinfo(skb);
- ssi->frags[0].page = gl->frags[0].page;
- ssi->frags[0].page_offset = (gl->frags[0].page_offset +
- RX_PKT_PULL_LEN);
- ssi->frags[0].size = gl->frags[0].size - RX_PKT_PULL_LEN;
- if (gl->nfrags > 1)
- memcpy(&ssi->frags[1], &gl->frags[1],
- (gl->nfrags-1) * sizeof(skb_frag_t));
- ssi->nr_frags = gl->nfrags;
- skb->len = len + PKTSHIFT;
- skb->data_len = skb->len - RX_PKT_PULL_LEN;
- skb->truesize += skb->data_len;
-
- /* Get a reference for the last page, we don't own it */
- get_page(gl->frags[gl->nfrags - 1].page);
+ skb = t4vf_pktgl_to_skb(gl, RX_SKB_LEN, RX_PULL_LEN);
+ if (unlikely(!skb)) {
+ t4vf_pktgl_free(gl);
+ rxq->stats.rx_drops++;
+ return 0;
}
-
__skb_pull(skb, PKTSHIFT);
skb->protocol = eth_type_trans(skb, rspq->netdev);
skb_record_rx_queue(skb, rspq->idx);
@@ -1552,11 +1584,6 @@ int t4vf_ethrx_handler(struct sge_rspq *rspq, const __be64 *rsp,
netif_receive_skb(skb);
return 0;
-
-nomem:
- t4vf_pktgl_free(gl);
- rxq->stats.rx_drops++;
- return 0;
}
/**
--
1.7.0.4
^ permalink raw reply related
* [PATCH 01/10] cxgb4vf: minor comment/symbolic name cleanup.
From: Casey Leedom @ 2010-11-11 2:04 UTC (permalink / raw)
To: netdev; +Cc: davem, Casey Leedom
In-Reply-To: <450C6E98-D0D9-42A6-9F15-4DB98647DB04@chelsio.com>
Minor cleanup of comments and symbolic constant names for clarity.
Signed-off-by: Casey Leedom <leedom@chelsio.com>
---
drivers/net/cxgb4vf/adapter.h | 2 +-
drivers/net/cxgb4vf/cxgb4vf_main.c | 9 ++++-----
drivers/net/cxgb4vf/sge.c | 9 ++++++---
drivers/net/cxgb4vf/t4vf_hw.c | 4 ++--
4 files changed, 13 insertions(+), 11 deletions(-)
diff --git a/drivers/net/cxgb4vf/adapter.h b/drivers/net/cxgb4vf/adapter.h
index 8ea0196..4766b41 100644
--- a/drivers/net/cxgb4vf/adapter.h
+++ b/drivers/net/cxgb4vf/adapter.h
@@ -60,7 +60,7 @@ enum {
* MSI-X interrupt index usage.
*/
MSIX_FW = 0, /* MSI-X index for firmware Q */
- MSIX_NIQFLINT = 1, /* MSI-X index base for Ingress Qs */
+ MSIX_IQFLINT = 1, /* MSI-X index base for Ingress Qs */
MSIX_EXTRAS = 1,
MSIX_ENTRIES = MAX_ETH_QSETS + MSIX_EXTRAS,
diff --git a/drivers/net/cxgb4vf/cxgb4vf_main.c b/drivers/net/cxgb4vf/cxgb4vf_main.c
index 6de5e2e..f5259a1 100644
--- a/drivers/net/cxgb4vf/cxgb4vf_main.c
+++ b/drivers/net/cxgb4vf/cxgb4vf_main.c
@@ -280,7 +280,7 @@ static void name_msix_vecs(struct adapter *adapter)
const struct port_info *pi = netdev_priv(dev);
int qs, msi;
- for (qs = 0, msi = MSIX_NIQFLINT;
+ for (qs = 0, msi = MSIX_IQFLINT;
qs < pi->nqsets;
qs++, msi++) {
snprintf(adapter->msix_info[msi].desc, namelen,
@@ -309,7 +309,7 @@ static int request_msix_queue_irqs(struct adapter *adapter)
/*
* Ethernet queues.
*/
- msi = MSIX_NIQFLINT;
+ msi = MSIX_IQFLINT;
for_each_ethrxq(s, rxq) {
err = request_irq(adapter->msix_info[msi].vec,
t4vf_sge_intr_msix, 0,
@@ -337,7 +337,7 @@ static void free_msix_queue_irqs(struct adapter *adapter)
int rxq, msi;
free_irq(adapter->msix_info[MSIX_FW].vec, &s->fw_evtq);
- msi = MSIX_NIQFLINT;
+ msi = MSIX_IQFLINT;
for_each_ethrxq(s, rxq)
free_irq(adapter->msix_info[msi++].vec,
&s->ethrxq[rxq].rspq);
@@ -527,7 +527,7 @@ static int setup_sge_queues(struct adapter *adapter)
* brought up at which point lots of things get nailed down
* permanently ...
*/
- msix = MSIX_NIQFLINT;
+ msix = MSIX_IQFLINT;
for_each_port(adapter, pidx) {
struct net_device *dev = adapter->port[pidx];
struct port_info *pi = netdev_priv(dev);
@@ -2465,7 +2465,6 @@ static int __devinit cxgb4vf_pci_probe(struct pci_dev *pdev,
version_printed = 1;
}
-
/*
* Initialize generic PCI device state.
*/
diff --git a/drivers/net/cxgb4vf/sge.c b/drivers/net/cxgb4vf/sge.c
index f10864d..d791adb 100644
--- a/drivers/net/cxgb4vf/sge.c
+++ b/drivers/net/cxgb4vf/sge.c
@@ -1536,6 +1536,9 @@ int t4vf_ethrx_handler(struct sge_rspq *rspq, const __be64 *rsp,
} else
skb_checksum_none_assert(skb);
+ /*
+ * Deliver the packet to the stack.
+ */
if (unlikely(pkt->vlan_ex)) {
struct vlan_group *grp = pi->vlan_grp;
@@ -2115,7 +2118,7 @@ int t4vf_sge_alloc_rxq(struct adapter *adapter, struct sge_rspq *rspq,
/*
* Calculate the size of the hardware free list ring plus
- * status page (which the SGE will place at the end of the
+ * Status Page (which the SGE will place after the end of the
* free list ring) in Egress Queue Units.
*/
flsz = (fl->size / FL_PER_EQ_UNIT +
@@ -2212,8 +2215,8 @@ int t4vf_sge_alloc_eth_txq(struct adapter *adapter, struct sge_eth_txq *txq,
struct port_info *pi = netdev_priv(dev);
/*
- * Calculate the size of the hardware TX Queue (including the
- * status age on the end) in units of TX Descriptors.
+ * Calculate the size of the hardware TX Queue (including the Status
+ * Page on the end of the TX Queue) in units of TX Descriptors.
*/
nentries = txq->q.size + STAT_LEN / sizeof(struct tx_desc);
diff --git a/drivers/net/cxgb4vf/t4vf_hw.c b/drivers/net/cxgb4vf/t4vf_hw.c
index ea1c123..2180181 100644
--- a/drivers/net/cxgb4vf/t4vf_hw.c
+++ b/drivers/net/cxgb4vf/t4vf_hw.c
@@ -1257,7 +1257,7 @@ int t4vf_eth_eq_free(struct adapter *adapter, unsigned int eqid)
*/
int t4vf_handle_fw_rpl(struct adapter *adapter, const __be64 *rpl)
{
- struct fw_cmd_hdr *cmd_hdr = (struct fw_cmd_hdr *)rpl;
+ const struct fw_cmd_hdr *cmd_hdr = (const struct fw_cmd_hdr *)rpl;
u8 opcode = FW_CMD_OP_GET(be32_to_cpu(cmd_hdr->hi));
switch (opcode) {
@@ -1265,7 +1265,7 @@ int t4vf_handle_fw_rpl(struct adapter *adapter, const __be64 *rpl)
/*
* Link/module state change message.
*/
- const struct fw_port_cmd *port_cmd = (void *)rpl;
+ const struct fw_port_cmd *port_cmd = (const void *)rpl;
u32 word;
int action, port_id, link_ok, speed, fc, pidx;
--
1.7.0.4
^ permalink raw reply related
* [PATCH 02/10] cxgb4vf: don't implement trivial (and incorrect) ndo_select_queue()
From: Casey Leedom @ 2010-11-11 2:04 UTC (permalink / raw)
To: netdev; +Cc: davem, Casey Leedom
In-Reply-To: <450C6E98-D0D9-42A6-9F15-4DB98647DB04@chelsio.com>
Don't implement (struct net_device_ops *)->ndo_select_queue() with simple
call to skb_tx_hash(). This leads to non-persistent TX queue selection in
the Linux dev_pick_tx() routine for TCP connections.
Signed-off-by: Casey Leedom <leedom@chelsio.com>
---
drivers/net/cxgb4vf/cxgb4vf_main.c | 14 --------------
1 files changed, 0 insertions(+), 14 deletions(-)
diff --git a/drivers/net/cxgb4vf/cxgb4vf_main.c b/drivers/net/cxgb4vf/cxgb4vf_main.c
index f5259a1..224b36c 100644
--- a/drivers/net/cxgb4vf/cxgb4vf_main.c
+++ b/drivers/net/cxgb4vf/cxgb4vf_main.c
@@ -1103,18 +1103,6 @@ static int cxgb4vf_set_mac_addr(struct net_device *dev, void *_addr)
return 0;
}
-/*
- * Return a TX Queue on which to send the specified skb.
- */
-static u16 cxgb4vf_select_queue(struct net_device *dev, struct sk_buff *skb)
-{
- /*
- * XXX For now just use the default hash but we probably want to
- * XXX look at other possibilities ...
- */
- return skb_tx_hash(dev, skb);
-}
-
#ifdef CONFIG_NET_POLL_CONTROLLER
/*
* Poll all of our receive queues. This is called outside of normal interrupt
@@ -2417,7 +2405,6 @@ static const struct net_device_ops cxgb4vf_netdev_ops = {
.ndo_get_stats = cxgb4vf_get_stats,
.ndo_set_rx_mode = cxgb4vf_set_rxmode,
.ndo_set_mac_address = cxgb4vf_set_mac_addr,
- .ndo_select_queue = cxgb4vf_select_queue,
.ndo_validate_addr = eth_validate_addr,
.ndo_do_ioctl = cxgb4vf_do_ioctl,
.ndo_change_mtu = cxgb4vf_change_mtu,
@@ -2623,7 +2610,6 @@ static int __devinit cxgb4vf_pci_probe(struct pci_dev *pdev,
netdev->do_ioctl = cxgb4vf_do_ioctl;
netdev->change_mtu = cxgb4vf_change_mtu;
netdev->set_mac_address = cxgb4vf_set_mac_addr;
- netdev->select_queue = cxgb4vf_select_queue;
#ifdef CONFIG_NET_POLL_CONTROLLER
netdev->poll_controller = cxgb4vf_poll_controller;
#endif
--
1.7.0.4
^ permalink raw reply related
* [PATCH 00/10] cxgb4vf: a number of bug fixes and minor cleanup
From: Casey Leedom @ 2010-11-11 2:01 UTC (permalink / raw)
To: netdev; +Cc: David Miller
The following patch set includes a number of bug fixes and some minor cleanup for the cxgb4vf network driver. If there are any problems with formatting, etc. of the patch set please just reject the patch set and kick my butt — I'm still a novice at putting these together!
drivers/net/cxgb4vf/adapter.h | 2 +-
drivers/net/cxgb4vf/cxgb4vf_main.c | 72 +++++++++++++-------
drivers/net/cxgb4vf/sge.c | 131 ++++++++++++++++++++++--------------
drivers/net/cxgb4vf/t4vf_common.h | 29 ++++----
drivers/net/cxgb4vf/t4vf_hw.c | 23 ++++++-
5 files changed, 165 insertions(+), 92 deletions(-)
Casey
^ permalink raw reply
* [PATCH] virtio-net: init link state correctly
From: Rusty Russell @ 2010-11-11 0:45 UTC (permalink / raw)
To: netdev; +Cc: Michael S. Tsirkin, Jason Wang
From: Jason Wang <jasowang@redhat.com>
For device that supports VIRTIO_NET_F_STATUS, there's no need to
assume the link is up and we need to call nerif_carrier_off() before
querying device status, otherwise we may get wrong operstate after
diver was loaded because the link watch event was not fired as
expected.
For device that does not support VIRITO_NET_F_STATUS, we could not get
its status through virtnet_update_status() and what we can only do is
always assuming the link is up.
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
---
drivers/net/virtio_net.c | 12 +++++++++---
1 files changed, 9 insertions(+), 3 deletions(-)
diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
index bb6b67f..b6d4028 100644
--- a/drivers/net/virtio_net.c
+++ b/drivers/net/virtio_net.c
@@ -986,9 +986,15 @@ static int virtnet_probe(struct virtio_device *vdev)
goto unregister;
}
- vi->status = VIRTIO_NET_S_LINK_UP;
- virtnet_update_status(vi);
- netif_carrier_on(dev);
+ /* Assume link up if device can't report link status,
+ otherwise get link status from config. */
+ if (virtio_has_feature(vi->vdev, VIRTIO_NET_F_STATUS)) {
+ netif_carrier_off(dev);
+ virtnet_update_status(vi);
+ } else {
+ vi->status = VIRTIO_NET_S_LINK_UP;
+ netif_carrier_on(dev);
+ }
pr_debug("virtnet: registered device %s\n", dev->name);
return 0;
^ permalink raw reply related
* Re: [RESEND PATCH] virtio-net: init link state correctly
From: Rusty Russell @ 2010-11-11 0:45 UTC (permalink / raw)
To: Michael S. Tsirkin; +Cc: Jason Wang, netdev, linux-kernel, davem, markmc, kvm
In-Reply-To: <20101110150823.GA5622@redhat.com>
On Thu, 11 Nov 2010 01:38:23 am Michael S. Tsirkin wrote:
> On Mon, Nov 08, 2010 at 09:41:27AM +1030, Rusty Russell wrote:
> > On Fri, 5 Nov 2010 08:17:18 pm Jason Wang wrote:
> > > For device that supports VIRTIO_NET_F_STATUS, there's no need to
> > > assume the link is up and we need to call nerif_carrier_off() before
> > > querying device status, otherwise we may get wrong operstate after
> > > diver was loaded because the link watch event was not fired as
> > > expected.
> > >
> > > For device that does not support VIRITO_NET_F_STATUS, we could not get
> > > its status through virtnet_update_status() and what we can only do is
> > > always assuming the link is up.
> > >
> > > Acked-by: Michael S. Tsirkin <mst@redhat.com>
> > > Signed-off-by: Jason Wang <jasowang@redhat.com>
> >
> > Acked-by: Rusty Russell <rusty@rustcorp.com.au>
> >
> > Thanks!
> > Rusty.
>
> Rusty, just to verify: who shall be queueing this up?
Sorry, that was sloppy of me; was meant for DaveM to take.
Sending explicitly now...
Rusty.
^ permalink raw reply
* Re: [PATCH] macvlan: lockless tx path
From: Ben Greear @ 2010-11-10 23:46 UTC (permalink / raw)
To: Eric Dumazet; +Cc: netdev
In-Reply-To: <1289432184.17691.141.camel@edumazet-laptop>
On 11/10/2010 03:36 PM, Eric Dumazet wrote:
> Le mercredi 10 novembre 2010 à 14:53 -0800, Ben Greear a écrit :
>
>> I did similar, and then wrote extra code to detect a 64-bit kernel and if
>> so assume that the counters wrap at 64 bits so I didn't have to poll so
>> often to make sure I didn't miss a wrap for a 10G NIC. If instead one wraps at 33
>> bits and the other at 36, there is no way for me to deal with the wrap
>> properly w/out explicitly knowing about that 33 and 36.
>>
>
> How do you define 'wrap around' ? Maybe your definition is wrong.
Maybe so. My algorithm looks like:
// uint64 accum;
// uint32 old;
// uint32 new;
if (old > new) {
// This assumes counters wrap at 32 bits (ie, 0xFFFFFFFF).
accum += ((uint32)(0xFFFFFFFF) - old) + new;
}
else if (old < new) {
accum += new - old;
}
old = new;
...
Is there some way I can do this w/out the (0xFFFFFFFF - old),
and thus the assumption of 32-bit counters?
Thanks,
Ben
--
Ben Greear <greearb@candelatech.com>
Candela Technologies Inc http://www.candelatech.com
^ permalink raw reply
* Re: [PATCH] macvlan: lockless tx path
From: Eric Dumazet @ 2010-11-10 23:36 UTC (permalink / raw)
To: Ben Greear; +Cc: netdev
In-Reply-To: <4CDB226A.8080903@candelatech.com>
Le mercredi 10 novembre 2010 à 14:53 -0800, Ben Greear a écrit :
> I did similar, and then wrote extra code to detect a 64-bit kernel and if
> so assume that the counters wrap at 64 bits so I didn't have to poll so
> often to make sure I didn't miss a wrap for a 10G NIC. If instead one wraps at 33
> bits and the other at 36, there is no way for me to deal with the wrap
> properly w/out explicitly knowing about that 33 and 36.
>
How do you define 'wrap around' ? Maybe your definition is wrong.
> If the old 32-bit counters in /proc/net/dev instead had a driver that
> managed to wrap them at 28 bits, I can't see how your application could
> have worked properly, so you must have been assuming that the kernel would
> always return a full 32-bit counter.
I suggest you take a look at various SNMP applications that handle this
just fine. RRD for example. You can patch your favorite driver to cap
the stats to 37, 38, ... 31 or 30 bits, it works.
If you sample values faster than half the period, it works.
This has nothing to do with 32 or 64 bits, really.
Read RFC1230 for a good advice
All the statistics are defined using the
syntax Counter as 32 bit wrap around
counters. Thus, if an interface's
hardware chip set maintains these
statistics in 16-bit counters, then the
agent must read the hardware's counters
frequently enough to prevent loss of
significance, in order to maintain
a 32-bit counter in software."
Yes, 16bit counters _are_ fine, even if provided in a 32bit counter, if
you read value fast enough.
^ permalink raw reply
* Re: [PATCH] macvlan: lockless tx path
From: Eric Dumazet @ 2010-11-10 23:24 UTC (permalink / raw)
To: Ben Greear; +Cc: netdev
In-Reply-To: <4CDB226A.8080903@candelatech.com>
Le mercredi 10 novembre 2010 à 14:53 -0800, Ben Greear a écrit :
> Then it's busted. If it claims to return stats64, but instead is returning
> something different, it is wrong. Netlink API still defines a stats32, so
> the kernel should use that if it can't reliably deal with 64-bit counters.
>
It claims nothing like that. You obviously assumed wrong things.
It provides a framework, not a mandatory or exclusive one.
Really this endless discussion is going nowhere.
I suggest you send patches if you want, I am very pleased by current
stats handling.
^ 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