* Re: rwlock recursion on CPU#0, netfilter related?
From: Pekka Pietikainen @ 2005-09-25 20:19 UTC (permalink / raw)
To: Harald Welte; +Cc: netdev
In-Reply-To: <20050925134344.GJ731@sunbeam.de.gnumonks.org>
On Sun, Sep 25, 2005 at 03:43:44PM +0200, Harald Welte wrote:
> 1) how does your kernel .config look like?
http://cvs.fedora.redhat.com/viewcvs/devel/kernel/configs/config-generic?rev=1.60&view=auto
http://cvs.fedora.redhat.com/viewcvs/devel/kernel/configs/config-x86_64-generic?rev=1.16&view=auto
> 2) which modules are loaded
Module Size Used by
w83627hf 46569 0
eeprom 17617 0
i2c_sensor 12225 2 w83627hf,eeprom
i2c_isa 11329 0
rfcomm 61033 0
l2cap 46145 5 rfcomm
bluetooth 73317 4 rfcomm,l2cap
ipv6 325889 16
ppp_synctty 21057 0
ppp_async 22465 1
crc_ccitt 10817 1 ppp_async
ppp_generic 41953 6 ppp_synctty,ppp_async
slhc 16193 1 ppp_generic
ip_conntrack_ftp 82177 0
ipt_ULOG 18913 1
ipt_state 10689 18
ip_conntrack 60053 2 ip_conntrack_ftp,ipt_state
iptable_filter 11969 1
ip_tables 32193 3 ipt_ULOG,ipt_state,iptable_filter
loop 26449 0
video 27977 0
button 16481 0
battery 19657 0
ac 14409 0
ohci1394 46753 0
ieee1394 381273 1 ohci1394
ohci_hcd 33249 0
ehci_hcd 46157 0
parport_pc 40621 0
parport 52557 1 parport_pc
i2c_nforce2 16833 0
i2c_core 34241 5 w83627hf,eeprom,i2c_sensor,i2c_isa,i2c_nforce2
shpchp 108009 0
emu10k1_gp 12865 0
gameport 27089 2 emu10k1_gp
snd_emu10k1 138629 0
snd_rawmidi 39521 1 snd_emu10k1
snd_util_mem 14401 1 snd_emu10k1
snd_hwdep 20321 1 snd_emu10k1
snd_intel8x0 46273 0
snd_ac97_codec 106757 2 snd_emu10k1,snd_intel8x0
snd_seq_dummy 12869 0
snd_seq_oss 47012 0
snd_seq_midi_event 17473 1 snd_seq_oss
snd_seq 74265 5 snd_seq_dummy,snd_seq_oss,snd_seq_midi_event
snd_seq_device 19281 5 snd_emu10k1,snd_rawmidi,snd_seq_dummy,snd_seq_oss,snd_seq
snd_pcm_oss 68465 0
snd_mixer_oss 28225 1 snd_pcm_oss
snd_pcm 115401 4 snd_emu10k1,snd_intel8x0,snd_ac97_codec,snd_pcm_oss
snd_timer 37577 3 snd_emu10k1,snd_seq,snd_pcm
snd 75681 12 snd_emu10k1,snd_rawmidi,snd_hwdep,snd_intel8x0,snd_ac97_codec,snd_seq_oss,snd_seq,snd_seq_device,snd_pcm_oss,snd_mixer_oss,snd_pcm,snd_timer
soundcore 19809 1 snd
snd_page_alloc 21713 3 snd_emu10k1,snd_intel8x0,snd_pcm
r8169 43209 0
forcedeth 30657 0
floppy 77865 0
dm_snapshot 26369 0
dm_zero 10817 0
dm_mirror 32433 0
ext3 154577 3
jbd 76145 1 ext3
dm_mod 73873 7 dm_snapshot,dm_zero,dm_mirror
sata_nv 19141 3
libata 61649 1 sata_nv
sd_mod 29121 4
scsi_mod 167801 2 libata,sd_mod
> 3) how does your ruleset look like?
*filter
:INPUT ACCEPT [0:0]
:FORWARD ACCEPT [0:0]
:OUTPUT ACCEPT [0:0]
:RH-Firewall-1-INPUT - [0:0]
-A INPUT -j RH-Firewall-1-INPUT
-A FORWARD -j RH-Firewall-1-INPUT
-A RH-Firewall-1-INPUT -i lo -j ACCEPT
-A RH-Firewall-1-INPUT -i eth1 -j ACCEPT
-A RH-Firewall-1-INPUT -p icmp --icmp-type echo-request -j ACCEPT
-A RH-Firewall-1-INPUT -p esp -j ACCEPT
-A RH-Firewall-1-INPUT -p ah -j ACCEPT
-A RH-Firewall-1-INPUT -p ipv6 -j ACCEPT
-A RH-Firewall-1-INPUT -m state --state RELATED,ESTABLISHED -j ACCEPT
-A RH-Firewall-1-INPUT -j ULOG
-A RH-Firewall-1-INPUT -p tcp -m state --state NEW -m tcp --dport x -j ACCEPT
-A RH-Firewall-1-INPUT -p udp -m state --state NEW -m udp --dport y -j ACCEPT
(for a bunch of ports, some with -s sourcenet/24 etc.)
-A RH-Firewall-1-INPUT -j DROP
COMMIT
# Completed on Sun Sep 28 10:37:44 2003
So basically a single-host firewall with ULOG and ftp conntracking being the
only fancy things.
> 4) most importantly, have you enabled CONFIG_IP_NF_CONNTRACK_EVENTS ?
> if yes, please disable, it's broken, a fix has been submitted, but I
> don't know if it has propagated to Linus yet (netdev Message-ID:
> <20050922143515.GD8917@rama.de.gnumonks.org>)
Enabled, so this could be it. But 2.6.14-rc2-git4 did crash too (although
it did take a bit longer for that to happen), and the changelog does state:
commit 1dfbab59498d6f227c91988bab6c71af049a5333
tree 6b20409a232ebe8c37f16d06b3fbcde6bec8f328
parent a82b748930fce0dab22c64075c38c830ae116904
author Harald Welte <laforge@netfilter.org> Thu, 22 Sep 2005 23:46:57 -0700
committer David S. Miller <davem@davemloft.net> Thu, 22 Sep 2005 23:46:57
-0700
[NETFILTER] Fix conntrack event cache deadlock/oops
Which is this patch, right? Will verify whether disabling the option makes any
difference tomorrow, as well as your other recommendations.
> Also, I have that Ping time problem on my x86_64 debian unstable (smp).
> But only in 1 out of ten cases on average (when starting ping, ctrl+c,
> pin, ctrl+c, ...). I've always assumed it's some 64bit problem in
> "ping" itself.
Happens for all packets on the "broken" kernels, and works a-ok (few ms
latencies to the same box) on the 2.6.13-era ones that don't crash.
Could be a different bug, sure.
^ permalink raw reply
* Re: [PATCH/RFC] [NETFILTER]: Fix invalid module autoloading by splitting iptable_nat
From: Harald Welte @ 2005-09-25 16:44 UTC (permalink / raw)
To: Patrick McHardy
Cc: Linux Netdev List, Netfilter Development Mailinglist,
David Miller
In-Reply-To: <4336CFB4.6060806@trash.net>
[-- Attachment #1: Type: text/plain, Size: 1376 bytes --]
On Sun, Sep 25, 2005 at 06:26:28PM +0200, Patrick McHardy wrote:
> Harald Welte wrote:
> >Hi Patrick, Dave,
> >I think we really need a solution for the last (known) remaining
> >dependency problem with 2.6.14. Please see the description
> >below. I _think_ the patch is fine, at least I couldn't find any case
> >where we could leak anything by splitting the code in two modules.
> >There's a slight semantic change, though. If the user unloads
> >iptable_nat, all existing connections (including their configured NAT
> >mappings) will continue to work. Only when ip_nat.ko is unloaded, the
> >NAT mappings are evicted from the conntrack table. I like it that way,
> >since it's logical.
>
> I agree, its more logical than having the table and the conntrack
> part in one module.
Ok. Gandalf has reviewed the code, too. This makes three to blame if
something goes wrong ;)
Dave, please apply
(sorry for the missing Signed-off-by: Harald Welte <laforge@netfilter.org>)
--
- Harald Welte <laforge@netfilter.org> http://netfilter.org/
============================================================================
"Fragmentation is like classful addressing -- an interesting early
architectural error that shows how much experimentation was going
on while IP was being designed." -- Paul Vixie
[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply
* Re: [PATCH/RFC] [NETFILTER]: Fix invalid module autoloading by splitting iptable_nat
From: Patrick McHardy @ 2005-09-25 16:26 UTC (permalink / raw)
To: Harald Welte
Cc: Linux Netdev List, Patrick McHardy,
Netfilter Development Mailinglist, David Miller
In-Reply-To: <20050925150755.GK731@sunbeam.de.gnumonks.org>
Harald Welte wrote:
> Hi Patrick, Dave,
>
> I think we really need a solution for the last (known) remaining
> dependency problem with 2.6.14. Please see the description
> below. I _think_ the patch is fine, at least I couldn't find any case
> where we could leak anything by splitting the code in two modules.
>
> There's a slight semantic change, though. If the user unloads
> iptable_nat, all existing connections (including their configured NAT
> mappings) will continue to work. Only when ip_nat.ko is unloaded, the
> NAT mappings are evicted from the conntrack table. I like it that way,
> since it's logical.
I agree, its more logical than having the table and the conntrack
part in one module.
^ permalink raw reply
* [PATCH/RFC] [NETFILTER]: Fix invalid module autoloading by splitting iptable_nat
From: Harald Welte @ 2005-09-25 15:07 UTC (permalink / raw)
To: Patrick McHardy
Cc: Linux Netdev List, Netfilter Development Mailinglist,
David Miller
[-- Attachment #1: Type: text/plain, Size: 11911 bytes --]
Hi Patrick, Dave,
I think we really need a solution for the last (known) remaining
dependency problem with 2.6.14. Please see the description
below. I _think_ the patch is fine, at least I couldn't find any case
where we could leak anything by splitting the code in two modules.
There's a slight semantic change, though. If the user unloads
iptable_nat, all existing connections (including their configured NAT
mappings) will continue to work. Only when ip_nat.ko is unloaded, the
NAT mappings are evicted from the conntrack table. I like it that way,
since it's logical.
If Patrick has no objections, I vote for inclusion in 2.6.14.
Thanks!
[NETFILTER]: Fix invalid module autoloading by splitting iptable_nat
When you've enabled conntrack and NAT as a module (standard case in all
distributions), and you've also enabled the new conntrack netlink
interface, loading ip_conntrack_netlink.ko will auto-load iptable_nat.ko.
This causes a huge performance penalty, since for every packet you iterate
the nat code, even if you don't want it.
This patch splits iptable_nat.ko into the NAT core (ip_nat.ko) and the
iptables frontend (iptable_nat.ko). Threfore, ip_conntrack_netlink.ko will
only pull ip_nat.ko, but not the frontend. ip_nat.ko will "only" allocate
some resources, but not affect runtime performance.
This separation is also a nice step in anticipation of new packet filters
(nf-hipac, ipset, pkttables) being able to use the NAT core.
---
commit 5ad0f056192213b2f6ac9910f754be9e6abea574
tree b06756b046853df61b39139ae58d49d0ca84274c
parent 1699521ab4e23ba4dd9f8bdd894393b8e109fe43
author Harald Welte <laforge@hanuman.de.gnumonks.org> Sun, 25 Sep 2005 17:01:38 +0200
committer Harald Welte <laforge@hanuman.de.gnumonks.org> Sun, 25 Sep 2005 17:01:38 +0200
include/linux/netfilter_ipv4/ip_nat_core.h | 12 ++++------
net/ipv4/netfilter/Makefile | 5 ++--
net/ipv4/netfilter/ip_nat_core.c | 35 +++++++++++++++++++---------
net/ipv4/netfilter/ip_nat_helper.c | 4 +++
net/ipv4/netfilter/ip_nat_standalone.c | 25 +++-----------------
5 files changed, 40 insertions(+), 41 deletions(-)
diff --git a/include/linux/netfilter_ipv4/ip_nat_core.h b/include/linux/netfilter_ipv4/ip_nat_core.h
--- a/include/linux/netfilter_ipv4/ip_nat_core.h
+++ b/include/linux/netfilter_ipv4/ip_nat_core.h
@@ -5,16 +5,14 @@
/* This header used to share core functionality between the standalone
NAT module, and the compatibility layer's use of NAT for masquerading. */
-extern int ip_nat_init(void);
-extern void ip_nat_cleanup(void);
-extern unsigned int nat_packet(struct ip_conntrack *ct,
+extern unsigned int ip_nat_packet(struct ip_conntrack *ct,
enum ip_conntrack_info conntrackinfo,
unsigned int hooknum,
struct sk_buff **pskb);
-extern int icmp_reply_translation(struct sk_buff **pskb,
- struct ip_conntrack *ct,
- enum ip_nat_manip_type manip,
- enum ip_conntrack_dir dir);
+extern int ip_nat_icmp_reply_translation(struct sk_buff **pskb,
+ struct ip_conntrack *ct,
+ enum ip_nat_manip_type manip,
+ enum ip_conntrack_dir dir);
#endif /* _IP_NAT_CORE_H */
diff --git a/net/ipv4/netfilter/Makefile b/net/ipv4/netfilter/Makefile
--- a/net/ipv4/netfilter/Makefile
+++ b/net/ipv4/netfilter/Makefile
@@ -4,7 +4,8 @@
# objects for the standalone - connection tracking / NAT
ip_conntrack-objs := ip_conntrack_standalone.o ip_conntrack_core.o ip_conntrack_proto_generic.o ip_conntrack_proto_tcp.o ip_conntrack_proto_udp.o ip_conntrack_proto_icmp.o
-iptable_nat-objs := ip_nat_standalone.o ip_nat_rule.o ip_nat_core.o ip_nat_helper.o ip_nat_proto_unknown.o ip_nat_proto_tcp.o ip_nat_proto_udp.o ip_nat_proto_icmp.o
+ip_nat-objs := ip_nat_core.o ip_nat_helper.o ip_nat_proto_unknown.o ip_nat_proto_tcp.o ip_nat_proto_udp.o ip_nat_proto_icmp.o
+iptable_nat-objs := ip_nat_rule.o ip_nat_standalone.o
ip_conntrack_pptp-objs := ip_conntrack_helper_pptp.o ip_conntrack_proto_gre.o
ip_nat_pptp-objs := ip_nat_helper_pptp.o ip_nat_proto_gre.o
@@ -40,7 +41,7 @@ obj-$(CONFIG_IP_NF_IPTABLES) += ip_table
# the three instances of ip_tables
obj-$(CONFIG_IP_NF_FILTER) += iptable_filter.o
obj-$(CONFIG_IP_NF_MANGLE) += iptable_mangle.o
-obj-$(CONFIG_IP_NF_NAT) += iptable_nat.o
+obj-$(CONFIG_IP_NF_NAT) += iptable_nat.o ip_nat.o
obj-$(CONFIG_IP_NF_RAW) += iptable_raw.o
# matches
diff --git a/net/ipv4/netfilter/ip_nat_core.c b/net/ipv4/netfilter/ip_nat_core.c
--- a/net/ipv4/netfilter/ip_nat_core.c
+++ b/net/ipv4/netfilter/ip_nat_core.c
@@ -74,12 +74,14 @@ ip_nat_proto_find_get(u_int8_t protonum)
return p;
}
+EXPORT_SYMBOL_GPL(ip_nat_proto_find_get);
void
ip_nat_proto_put(struct ip_nat_protocol *p)
{
module_put(p->me);
}
+EXPORT_SYMBOL_GPL(ip_nat_proto_put);
/* We keep an extra hash for each conntrack, for fast searching. */
static inline unsigned int
@@ -111,6 +113,7 @@ ip_nat_cheat_check(u_int32_t oldvalinv,
return csum_fold(csum_partial((char *)diffs, sizeof(diffs),
oldcheck^0xFFFF));
}
+EXPORT_SYMBOL(ip_nat_cheat_check);
/* Is this tuple already taken? (not by us) */
int
@@ -127,6 +130,7 @@ ip_nat_used_tuple(const struct ip_conntr
invert_tuplepr(&reply, tuple);
return ip_conntrack_tuple_taken(&reply, ignored_conntrack);
}
+EXPORT_SYMBOL(ip_nat_used_tuple);
/* If we source map this tuple so reply looks like reply_tuple, will
* that meet the constraints of range. */
@@ -347,6 +351,7 @@ ip_nat_setup_info(struct ip_conntrack *c
return NF_ACCEPT;
}
+EXPORT_SYMBOL(ip_nat_setup_info);
/* Returns true if succeeded. */
static int
@@ -387,10 +392,10 @@ manip_pkt(u_int16_t proto,
}
/* Do packet manipulations according to ip_nat_setup_info. */
-unsigned int nat_packet(struct ip_conntrack *ct,
- enum ip_conntrack_info ctinfo,
- unsigned int hooknum,
- struct sk_buff **pskb)
+unsigned int ip_nat_packet(struct ip_conntrack *ct,
+ enum ip_conntrack_info ctinfo,
+ unsigned int hooknum,
+ struct sk_buff **pskb)
{
enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
unsigned long statusbit;
@@ -417,12 +422,13 @@ unsigned int nat_packet(struct ip_conntr
}
return NF_ACCEPT;
}
+EXPORT_SYMBOL_GPL(ip_nat_packet);
/* Dir is direction ICMP is coming from (opposite to packet it contains) */
-int icmp_reply_translation(struct sk_buff **pskb,
- struct ip_conntrack *ct,
- enum ip_nat_manip_type manip,
- enum ip_conntrack_dir dir)
+int ip_nat_icmp_reply_translation(struct sk_buff **pskb,
+ struct ip_conntrack *ct,
+ enum ip_nat_manip_type manip,
+ enum ip_conntrack_dir dir)
{
struct {
struct icmphdr icmp;
@@ -509,6 +515,7 @@ int icmp_reply_translation(struct sk_buf
return 1;
}
+EXPORT_SYMBOL_GPL(ip_nat_icmp_reply_translation);
/* Protocol registration. */
int ip_nat_protocol_register(struct ip_nat_protocol *proto)
@@ -525,6 +532,7 @@ int ip_nat_protocol_register(struct ip_n
write_unlock_bh(&ip_nat_lock);
return ret;
}
+EXPORT_SYMBOL(ip_nat_protocol_register);
/* Noone stores the protocol anywhere; simply delete it. */
void ip_nat_protocol_unregister(struct ip_nat_protocol *proto)
@@ -536,6 +544,7 @@ void ip_nat_protocol_unregister(struct i
/* Someone could be still looking at the proto in a bh. */
synchronize_net();
}
+EXPORT_SYMBOL(ip_nat_protocol_unregister);
#if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \
defined(CONFIG_IP_NF_CONNTRACK_NETLINK_MODULE)
@@ -582,7 +591,7 @@ EXPORT_SYMBOL_GPL(ip_nat_port_nfattr_to_
EXPORT_SYMBOL_GPL(ip_nat_port_range_to_nfattr);
#endif
-int __init ip_nat_init(void)
+static int __init ip_nat_init(void)
{
size_t i;
@@ -624,10 +633,14 @@ static int clean_nat(struct ip_conntrack
return 0;
}
-/* Not __exit: called from ip_nat_standalone.c:init_or_cleanup() --RR */
-void ip_nat_cleanup(void)
+static void __exit ip_nat_cleanup(void)
{
ip_ct_iterate_cleanup(&clean_nat, NULL);
ip_conntrack_destroyed = NULL;
vfree(bysource);
}
+
+MODULE_LICENSE("GPL");
+
+module_init(ip_nat_init);
+module_exit(ip_nat_cleanup);
diff --git a/net/ipv4/netfilter/ip_nat_helper.c b/net/ipv4/netfilter/ip_nat_helper.c
--- a/net/ipv4/netfilter/ip_nat_helper.c
+++ b/net/ipv4/netfilter/ip_nat_helper.c
@@ -199,6 +199,7 @@ ip_nat_mangle_tcp_packet(struct sk_buff
}
return 1;
}
+EXPORT_SYMBOL(ip_nat_mangle_tcp_packet);
/* Generic function for mangling variable-length address changes inside
* NATed UDP connections (like the CONNECT DATA XXXXX MESG XXXXX INDEX XXXXX
@@ -256,6 +257,7 @@ ip_nat_mangle_udp_packet(struct sk_buff
return 1;
}
+EXPORT_SYMBOL(ip_nat_mangle_udp_packet);
/* Adjust one found SACK option including checksum correction */
static void
@@ -399,6 +401,7 @@ ip_nat_seq_adjust(struct sk_buff **pskb,
return 1;
}
+EXPORT_SYMBOL(ip_nat_seq_adjust);
/* Setup NAT on this expected conntrack so it follows master. */
/* If we fail to get a free NAT slot, we'll get dropped on confirm */
@@ -425,3 +428,4 @@ void ip_nat_follow_master(struct ip_conn
/* hook doesn't matter, but it has to do destination manip */
ip_nat_setup_info(ct, &range, NF_IP_PRE_ROUTING);
}
+EXPORT_SYMBOL(ip_nat_follow_master);
diff --git a/net/ipv4/netfilter/ip_nat_standalone.c b/net/ipv4/netfilter/ip_nat_standalone.c
--- a/net/ipv4/netfilter/ip_nat_standalone.c
+++ b/net/ipv4/netfilter/ip_nat_standalone.c
@@ -108,8 +108,8 @@ ip_nat_fn(unsigned int hooknum,
case IP_CT_RELATED:
case IP_CT_RELATED+IP_CT_IS_REPLY:
if ((*pskb)->nh.iph->protocol == IPPROTO_ICMP) {
- if (!icmp_reply_translation(pskb, ct, maniptype,
- CTINFO2DIR(ctinfo)))
+ if (!ip_nat_icmp_reply_translation(pskb, ct, maniptype,
+ CTINFO2DIR(ctinfo)))
return NF_DROP;
else
return NF_ACCEPT;
@@ -152,7 +152,7 @@ ip_nat_fn(unsigned int hooknum,
}
IP_NF_ASSERT(info);
- return nat_packet(ct, ctinfo, hooknum, pskb);
+ return ip_nat_packet(ct, ctinfo, hooknum, pskb);
}
static unsigned int
@@ -325,15 +325,10 @@ static int init_or_cleanup(int init)
printk("ip_nat_init: can't setup rules.\n");
goto cleanup_nothing;
}
- ret = ip_nat_init();
- if (ret < 0) {
- printk("ip_nat_init: can't setup rules.\n");
- goto cleanup_rule_init;
- }
ret = nf_register_hook(&ip_nat_in_ops);
if (ret < 0) {
printk("ip_nat_init: can't register in hook.\n");
- goto cleanup_nat;
+ goto cleanup_rule_init;
}
ret = nf_register_hook(&ip_nat_out_ops);
if (ret < 0) {
@@ -374,8 +369,6 @@ static int init_or_cleanup(int init)
nf_unregister_hook(&ip_nat_out_ops);
cleanup_inops:
nf_unregister_hook(&ip_nat_in_ops);
- cleanup_nat:
- ip_nat_cleanup();
cleanup_rule_init:
ip_nat_rule_cleanup();
cleanup_nothing:
@@ -395,14 +388,4 @@ static void __exit fini(void)
module_init(init);
module_exit(fini);
-EXPORT_SYMBOL(ip_nat_setup_info);
-EXPORT_SYMBOL(ip_nat_protocol_register);
-EXPORT_SYMBOL(ip_nat_protocol_unregister);
-EXPORT_SYMBOL_GPL(ip_nat_proto_find_get);
-EXPORT_SYMBOL_GPL(ip_nat_proto_put);
-EXPORT_SYMBOL(ip_nat_cheat_check);
-EXPORT_SYMBOL(ip_nat_mangle_tcp_packet);
-EXPORT_SYMBOL(ip_nat_mangle_udp_packet);
-EXPORT_SYMBOL(ip_nat_used_tuple);
-EXPORT_SYMBOL(ip_nat_follow_master);
MODULE_LICENSE("GPL");
--
- Harald Welte <laforge@netfilter.org> http://netfilter.org/
============================================================================
"Fragmentation is like classful addressing -- an interesting early
architectural error that shows how much experimentation was going
on while IP was being designed." -- Paul Vixie
[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply
* Re: rwlock recursion on CPU#0, netfilter related?
From: Harald Welte @ 2005-09-25 13:43 UTC (permalink / raw)
To: Pekka Pietikainen; +Cc: netdev
In-Reply-To: <20050925105834.GA15243@ee.oulu.fi>
[-- Attachment #1: Type: text/plain, Size: 1790 bytes --]
On Sun, Sep 25, 2005 at 01:58:34PM +0300, Pekka Pietikainen wrote:
> Just to get a wider audience, somewhere between 2.6.13-git4 and current
> (2.6.14-rc2-git4 is the last one I tested, which seems to have some
> fixes in this are wrt. git3, but problem remains) my x86_64
> crashes quite quickly after boot. Using Fedora devel kernels, I can
> probably whip up a vanilla kernel if the maintainers in this area
> prefer that.
>
> https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=167835
Can you please give some more feedback like
1) how does your kernel .config look like?
2) which modules are loaded
3) how does your ruleset look like?
4) most importantly, have you enabled CONFIG_IP_NF_CONNTRACK_EVENTS ?
if yes, please disable, it's broken, a fix has been submitted, but I
don't know if it has propagated to Linus yet (netdev Message-ID:
<20050922143515.GD8917@rama.de.gnumonks.org>)
please also try
a) only loading iptable_filter (and ip_tables), but no other modules
a) only loading ip_conntrack but no other netfilter modules (no nat, no iptables)
b) only loading ip_conntrack and iptable_nat (but no rules)
this kind of debugging helps to locate where it is. netfilter has grown
big ;)
Also, I have that Ping time problem on my x86_64 debian unstable (smp).
But only in 1 out of ten cases on average (when starting ping, ctrl+c,
pin, ctrl+c, ...). I've always assumed it's some 64bit problem in
"ping" itself.
--
- Harald Welte <laforge@gnumonks.org> http://gnumonks.org/
============================================================================
"Privacy in residential applications is a desirable marketing option."
(ETSI EN 300 175-7 Ch. A6)
[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply
* rwlock recursion on CPU#0, netfilter related?
From: Pekka Pietikainen @ 2005-09-25 10:58 UTC (permalink / raw)
To: netdev
Just to get a wider audience, somewhere between 2.6.13-git4 and current
(2.6.14-rc2-git4 is the last one I tested, which seems to have some
fixes in this are wrt. git3, but problem remains) my x86_64
crashes quite quickly after boot. Using Fedora devel kernels, I can
probably whip up a vanilla kernel if the maintainers in this area
prefer that.
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=167835
and
https://bugzilla.redhat.com/bugzilla/attachment.cgi?id=119228
apart from the crashes I get funny ping times on the kernels that
break when they're still up
(64 bytes from 10.10.9.1: icmp_seq=0 ttl=255 time=4294971590968 ms)
^ permalink raw reply
* Re: [PATCH] [NETFILTER] ip_conntrack: Update event cache when status changes
From: David S. Miller @ 2005-09-24 23:56 UTC (permalink / raw)
To: laforge; +Cc: netdev, netfilter-devel
In-Reply-To: <20050924185020.GC731@sunbeam.de.gnumonks.org>
From: Harald Welte <laforge@netfilter.org>
Date: Sat, 24 Sep 2005 20:50:20 +0200
> [NETFILTER] ip_conntrack: Update event cache when status changes
>
> The GRE, SCTP and TCP protocol helpers did not call ip_conntrack_event_cache()
> when updating ct->status. This patch adds the respective calls.
>
> Signed-off-by: Harald Welte <laforge@netfilter.org>
Applied, thanks Harald.
^ permalink raw reply
* Re: [PATCH] fix NFQUEUE Kconfig dependency
From: David S. Miller @ 2005-09-24 23:52 UTC (permalink / raw)
To: laforge; +Cc: kaber, akpm, netdev, linux-kernel
In-Reply-To: <20050923150631.GH731@sunbeam.de.gnumonks.org>
From: Harald Welte <laforge@netfilter.org>
Date: Fri, 23 Sep 2005 17:06:31 +0200
> [NETFILTER]: Fix ip[6]t_NFQUEUE Kconfig dependency
>
> We have to introduce a separate Kconfig menu entry for the NFQUEUE targets.
> They cannot "just" depend on nfnetlink_queue, since nfnetlink_queue could
> be linked into the kernel, whereas iptables can be a module.
>
> Signed-off-by: Harald Welte <laforge@netfilter.org>
Applied, thanks Harald.
^ permalink raw reply
* [PATCH] [NETFILTER] ip_conntrack: Update event cache when status changes
From: Harald Welte @ 2005-09-24 18:50 UTC (permalink / raw)
To: David Miller; +Cc: Linux Netdev List, Netfilter Development Mailinglist
[-- Attachment #1: Type: text/plain, Size: 3094 bytes --]
Hi Dave!
While preparing the first release of libnfnetlink,
libnfnetlink_conntrack and the "conntrack" userspace program, I found
out that some events are never generated within the kernel. Please
apply the patch below, thans.
[NETFILTER] ip_conntrack: Update event cache when status changes
The GRE, SCTP and TCP protocol helpers did not call ip_conntrack_event_cache()
when updating ct->status. This patch adds the respective calls.
Signed-off-by: Harald Welte <laforge@netfilter.org>
---
commit 1699521ab4e23ba4dd9f8bdd894393b8e109fe43
tree 2e0e4becf37fc85de0206eb7d381c3d849075ce8
parent 5a9087bd9bf78e3ca131a84a8eed06d6c31396dc
author Harald Welte <laforge@hanuman.de.gnumonks.org> Sat, 24 Sep 2005 20:47:41 +0200
committer Harald Welte <laforge@hanuman.de.gnumonks.org> Sat, 24 Sep 2005 20:47:41 +0200
net/ipv4/netfilter/ip_conntrack_proto_gre.c | 1 +
net/ipv4/netfilter/ip_conntrack_proto_sctp.c | 1 +
net/ipv4/netfilter/ip_conntrack_proto_tcp.c | 3 ++-
3 files changed, 4 insertions(+), 1 deletions(-)
diff --git a/net/ipv4/netfilter/ip_conntrack_proto_gre.c b/net/ipv4/netfilter/ip_conntrack_proto_gre.c
--- a/net/ipv4/netfilter/ip_conntrack_proto_gre.c
+++ b/net/ipv4/netfilter/ip_conntrack_proto_gre.c
@@ -247,6 +247,7 @@ static int gre_packet(struct ip_conntrac
ct->proto.gre.stream_timeout);
/* Also, more likely to be important, and not a probe. */
set_bit(IPS_ASSURED_BIT, &ct->status);
+ ip_conntrack_event_cache(IPCT_STATUS, skb);
} else
ip_ct_refresh_acct(ct, conntrackinfo, skb,
ct->proto.gre.timeout);
diff --git a/net/ipv4/netfilter/ip_conntrack_proto_sctp.c b/net/ipv4/netfilter/ip_conntrack_proto_sctp.c
--- a/net/ipv4/netfilter/ip_conntrack_proto_sctp.c
+++ b/net/ipv4/netfilter/ip_conntrack_proto_sctp.c
@@ -416,6 +416,7 @@ static int sctp_packet(struct ip_conntra
&& newconntrack == SCTP_CONNTRACK_ESTABLISHED) {
DEBUGP("Setting assured bit\n");
set_bit(IPS_ASSURED_BIT, &conntrack->status);
+ ip_conntrack_event_cache(IPCT_STATUS, skb);
}
return NF_ACCEPT;
diff --git a/net/ipv4/netfilter/ip_conntrack_proto_tcp.c b/net/ipv4/netfilter/ip_conntrack_proto_tcp.c
--- a/net/ipv4/netfilter/ip_conntrack_proto_tcp.c
+++ b/net/ipv4/netfilter/ip_conntrack_proto_tcp.c
@@ -1014,7 +1014,8 @@ static int tcp_packet(struct ip_conntrac
/* Set ASSURED if we see see valid ack in ESTABLISHED
after SYN_RECV or a valid answer for a picked up
connection. */
- set_bit(IPS_ASSURED_BIT, &conntrack->status);
+ set_bit(IPS_ASSURED_BIT, &conntrack->status);
+ ip_conntrack_event_cache(IPCT_STATUS, skb);
}
ip_ct_refresh_acct(conntrack, ctinfo, skb, timeout);
--
- Harald Welte <laforge@netfilter.org> http://netfilter.org/
============================================================================
"Fragmentation is like classful addressing -- an interesting early
architectural error that shows how much experimentation was going
on while IP was being designed." -- Paul Vixie
[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply
* magical e channels teller
From: carol lindzy @ 2005-09-24 11:15 UTC (permalink / raw)
To: Jarred Bienenstock; +Cc: ryan, owner-linux-xfs, netdev, knight
Melany:
http://uk.geocities.com/hjboostpower/?match_your_elegance_and_glamour
nkumpo
I've been cheated by you since I don't know when
finprog esromer dettimmo sz03 extendedblock etargimm
Without some helpThis is the turning point
We got to come together!Give and let live
^ permalink raw reply
* ドキドキすることしよっか♪
From: ふれあい空間 @ 2005-09-23 23:34 UTC (permalink / raw)
To: netdev
最近暗いニュースが多いよね！！
2学期の勉強にも少し飽きてきたし、
週末はパーッと気分転換しない？
関東エリアで遊べる人希望でーーす。
http://specialterm.com
^ permalink raw reply
* [git patches] net driver fixes
From: Jeff Garzik @ 2005-09-23 23:09 UTC (permalink / raw)
To: Andrew Morton, Linus Torvalds; +Cc: netdev, linux-kernel
Please pull from 'upstream-fixes' branch of
master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6.git
to obtain the following fixes:
drivers/net/8390.c | 2 +-
drivers/net/skge.c | 22 ++++++++++++++++++++--
2 files changed, 21 insertions(+), 3 deletions(-)
Paul Gortmaker:
8390 Tx fix for non i386 machines
Stephen Hemminger:
skge: fix Yukon-Lite A0 workaround
diff --git a/drivers/net/8390.c b/drivers/net/8390.c
--- a/drivers/net/8390.c
+++ b/drivers/net/8390.c
@@ -1094,7 +1094,7 @@ static void NS8390_trigger_send(struct n
outb_p(E8390_NODMA+E8390_PAGE0, e8390_base+E8390_CMD);
- if (inb_p(e8390_base) & E8390_TRANS)
+ if (inb_p(e8390_base + E8390_CMD) & E8390_TRANS)
{
printk(KERN_WARNING "%s: trigger_send() called with the transmitter busy.\n",
dev->name);
diff --git a/drivers/net/skge.c b/drivers/net/skge.c
--- a/drivers/net/skge.c
+++ b/drivers/net/skge.c
@@ -1643,6 +1643,22 @@ static void yukon_reset(struct skge_hw *
| GM_RXCR_UCF_ENA | GM_RXCR_MCF_ENA);
}
+/* Apparently, early versions of Yukon-Lite had wrong chip_id? */
+static int is_yukon_lite_a0(struct skge_hw *hw)
+{
+ u32 reg;
+ int ret;
+
+ if (hw->chip_id != CHIP_ID_YUKON)
+ return 0;
+
+ reg = skge_read32(hw, B2_FAR);
+ skge_write8(hw, B2_FAR + 3, 0xff);
+ ret = (skge_read8(hw, B2_FAR + 3) != 0);
+ skge_write32(hw, B2_FAR, reg);
+ return ret;
+}
+
static void yukon_mac_init(struct skge_hw *hw, int port)
{
struct skge_port *skge = netdev_priv(hw->dev[port]);
@@ -1758,9 +1774,11 @@ static void yukon_mac_init(struct skge_h
/* Configure Rx MAC FIFO */
skge_write16(hw, SK_REG(port, RX_GMF_FL_MSK), RX_FF_FL_DEF_MSK);
reg = GMF_OPER_ON | GMF_RX_F_FL_ON;
- if (hw->chip_id == CHIP_ID_YUKON_LITE &&
- hw->chip_rev >= CHIP_REV_YU_LITE_A3)
+
+ /* disable Rx GMAC FIFO Flush for YUKON-Lite Rev. A0 only */
+ if (is_yukon_lite_a0(hw))
reg &= ~GMF_RX_F_FL_ON;
+
skge_write8(hw, SK_REG(port, RX_GMF_CTRL_T), GMF_RST_CLR);
skge_write16(hw, SK_REG(port, RX_GMF_CTRL_T), reg);
/*
^ permalink raw reply
* Re: [PATCH 0/3] netfilter : 3 patches to boost ip_tables performance
From: Dave Hansen @ 2005-09-23 18:04 UTC (permalink / raw)
To: Christoph Lameter
Cc: Eric Dumazet, Harald Welte, Andi Kleen, Christoph Hellwig,
David S. Miller, Linux Kernel Mailing List, netfilter-devel,
netdev
In-Reply-To: <Pine.LNX.4.62.0509231043270.22308@schroedinger.engr.sgi.com>
On Fri, 2005-09-23 at 10:44 -0700, Christoph Lameter wrote:
> for (i = 0; i < area->nr_pages; i++) {
> - area->pages[i] = alloc_page(gfp_mask);
> + if (node < 0)
> + area->pages[i] = alloc_page(gfp_mask);
> + else
> + area->pages[i] = alloc_pages_node(node, gfp_mask, 0);
> if (unlikely(!area->pages[i])) {
> /* Successfully allocated i pages, free them in __vunmap() */
> area->nr_pages = i;
...
> void *vmalloc_exec(unsigned long size)
> {
> - return __vmalloc(size, GFP_KERNEL | __GFP_HIGHMEM, PAGE_KERNEL_EXEC);
> + return __vmalloc(size, GFP_KERNEL | __GFP_HIGHMEM, PAGE_KERNEL_EXEC, -1);
> }
Instead of hard-coding all of those -1's for the node to specify a
default allocation, and changing all of those callers, why not:
void *__vmalloc_node(unsigned long size, unsigned int __nocast gfp_mask,
pgprot_t prot, int node)
{
... existing vmalloc code here
}
void *__vmalloc(unsigned long size, unsigned int __nocast gfp_mask,
pgprot_t prot)
{
__vmalloc_node(size, gfp_mask, prot, -1);
}
A named macro is probably better than -1, but if it is only used in one
place, it is hard to complain.
-- Dave
^ permalink raw reply
* Re: [PATCH 0/3] netfilter : 3 patches to boost ip_tables performance
From: Kyle Moffett @ 2005-09-23 18:00 UTC (permalink / raw)
To: Eric Dumazet
Cc: Harald Welte, Christoph Lameter, Andi Kleen, Christoph Hellwig,
David S. Miller, linux-kernel, netfilter-devel, netdev
In-Reply-To: <43343FC9.5090601@cosmosbay.com>
On Sep 23, 2005, at 13:47:53, Eric Dumazet wrote:
> Harald Welte a écrit :
>> I see a contradiction in your sentence. "a new ip_tables is
>> loaded" every time a user changes a single rule. There are
>> numerous setups that dynamically change the ruleset (e.g. at
>> interface up/down point, or even think of your typical wlan
>> hotspot, where once a user is authorized, he'll get different rules.
>
> But a user changing a single rule usually calls (fork()/exec()) a
> program called iptables. The underlying cost of all this, plus
> copying the rules to user space, so that iptables change them and
> reload them in the kernel is far more important than an
> hypothetical vmalloc_node() performance problem.
Yeah, if you're really worried about the cost of iptables
manipulations, you should probably write your own happy little C
program to atomically load, update, and store the rules. Even then,
the cost of copying the whole ruleset to userspace for modification
is probably greater than that of memory allocation issues, especially
if the ruleset is large enough that memory allocation issues cause
problems :-D
Cheers,
Kyle Moffett
-----BEGIN GEEK CODE BLOCK-----
Version: 3.12
GCM/CS/IT/U d- s++: a18 C++++>$ UB/L/X/*++++(+)>$ P+++(++++)>$ L++++(+
++) E W++(+) N+++(++) o? K? w--- O? M++ V? PS+() PE+(-) Y+ PGP+++ t+(+
++) 5 X R? tv-(--) b++++(++) DI+ D+ G e->++++$ h!*()>++$ r !y?(-)
------END GEEK CODE BLOCK------
^ permalink raw reply
* Re: [PATCH 0/3] netfilter : 3 patches to boost ip_tables performance
From: Eric Dumazet @ 2005-09-23 17:47 UTC (permalink / raw)
To: Harald Welte
Cc: Christoph Lameter, Andi Kleen, Christoph Hellwig, David S. Miller,
linux-kernel, netfilter-devel, netdev
In-Reply-To: <20050923171120.GO731@sunbeam.de.gnumonks.org>
Harald Welte a écrit :
> On Thu, Sep 22, 2005 at 05:50:49PM +0200, Eric Dumazet wrote:
>
>>Christoph Lameter a écrit :
>>
>>>It should really be do_set_mempolicy instead to be cleaner. I got a patch here that fixes the
>>>policy layer.
>>>But still I agree with Christoph that a real vmalloc_node is better. There will be no fuzzing
>>>around with memory policies etc and its certainly better performance wise.
>>
>>vmalloc_node() should be seldom used, at driver init, or when a new
>>ip_tables is loaded. If it happens to be a performance problem, then
>>we can optimize it. Why should we spend days of work for a function
>>that is yet to be used ?
>
>
> I see a contradiction in your sentence. "a new ip_tables is loaded"
> every time a user changes a single rule. There are numerous setups that
> dynamically change the ruleset (e.g. at interface up/down point, or even
> think of your typical wlan hotspot, where once a user is authorized,
> he'll get different rules.
>
But a user changing a single rule usually calls (fork()/exec()) a program
called iptables. The underlying cost of all this, plus copying the rules to
user space, so that iptables change them and reload them in the kernel is far
more important than an hypothetical vmalloc_node() performance problem.
Eric
^ permalink raw reply
* Re: [PATCH 0/3] netfilter : 3 patches to boost ip_tables performance
From: Christoph Lameter @ 2005-09-23 17:44 UTC (permalink / raw)
To: Eric Dumazet
Cc: Harald Welte, Andi Kleen, Christoph Hellwig, David S. Miller,
linux-kernel, netfilter-devel, netdev
In-Reply-To: <20050923171120.GO731@sunbeam.de.gnumonks.org>
Here is an updated version of the vmalloc_node patch:
This patch adds
vmalloc_node(size, node) -> Allocate necessary memory on the specified node
and
get_vm_area_node(size, flags, node)
and the other functions that it depends on.
Index: linux-2.6.14-rc2/include/linux/vmalloc.h
===================================================================
--- linux-2.6.14-rc2.orig/include/linux/vmalloc.h 2005-09-19 20:00:41.000000000 -0700
+++ linux-2.6.14-rc2/include/linux/vmalloc.h 2005-09-23 10:28:37.000000000 -0700
@@ -32,22 +32,35 @@ struct vm_struct {
* Highlevel APIs for driver use
*/
extern void *vmalloc(unsigned long size);
+extern void *vmalloc_node(unsigned long size, int node);
extern void *vmalloc_exec(unsigned long size);
extern void *vmalloc_32(unsigned long size);
-extern void *__vmalloc(unsigned long size, unsigned int __nocast gfp_mask, pgprot_t prot);
-extern void *__vmalloc_area(struct vm_struct *area, unsigned int __nocast gfp_mask, pgprot_t prot);
-extern void vfree(void *addr);
+extern void *__vmalloc(unsigned long size, unsigned int __nocast gfp_mask,
+ pgprot_t prot, int node);
+extern void *__vmalloc_area(struct vm_struct *area, unsigned int __nocast gfp_mask,
+ pgprot_t prot, int node);
+extern void vfree(void *addr);
extern void *vmap(struct page **pages, unsigned int count,
unsigned long flags, pgprot_t prot);
extern void vunmap(void *addr);
-
-/*
- * Lowlevel-APIs (not for driver use!)
+
+/**
+ * get_vm_area - reserve a contingous kernel virtual area
+ *
+ * @size: size of the area
+ * @flags: %VM_IOREMAP for I/O mappings or VM_ALLOC
+ *
+ * Search an area of @size in the kernel virtual mapping area,
+ * and reserved it for out purposes. Returns the area descriptor
+ * on success or %NULL on failure.
*/
-extern struct vm_struct *get_vm_area(unsigned long size, unsigned long flags);
extern struct vm_struct *__get_vm_area(unsigned long size, unsigned long flags,
- unsigned long start, unsigned long end);
+ unsigned long start, unsigned long end, int node);
+#define get_vm_area(__size, __flags) __get_vm_area((__size), (__flags), VMALLOC_START, \
+ VMALLOC_END, -1)
+#define get_vm_area_node(__size, __flags, __node) __get_vm_area((__size), (__flags), \
+ VMALLOC_START, VMALLOC_END, __node)
extern struct vm_struct *remove_vm_area(void *addr);
extern struct vm_struct *__remove_vm_area(void *addr);
extern int map_vm_area(struct vm_struct *area, pgprot_t prot,
Index: linux-2.6.14-rc2/mm/vmalloc.c
===================================================================
--- linux-2.6.14-rc2.orig/mm/vmalloc.c 2005-09-19 20:00:41.000000000 -0700
+++ linux-2.6.14-rc2/mm/vmalloc.c 2005-09-23 10:43:07.000000000 -0700
@@ -5,6 +5,7 @@
* Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999
* SMP-safe vmalloc/vfree/ioremap, Tigran Aivazian <tigran@veritas.com>, May 2000
* Major rework to support vmap/vunmap, Christoph Hellwig, SGI, August 2002
+ * Numa awareness, Christoph Lameter, SGI, June 2005
*/
#include <linux/mm.h>
@@ -159,7 +160,7 @@ int map_vm_area(struct vm_struct *area,
}
struct vm_struct *__get_vm_area(unsigned long size, unsigned long flags,
- unsigned long start, unsigned long end)
+ unsigned long start, unsigned long end, int node)
{
struct vm_struct **p, *tmp, *area;
unsigned long align = 1;
@@ -178,7 +179,7 @@ struct vm_struct *__get_vm_area(unsigned
addr = ALIGN(start, align);
size = PAGE_ALIGN(size);
- area = kmalloc(sizeof(*area), GFP_KERNEL);
+ area = kmalloc_node(sizeof(*area), GFP_KERNEL, node);
if (unlikely(!area))
return NULL;
@@ -231,21 +232,6 @@ out:
return NULL;
}
-/**
- * get_vm_area - reserve a contingous kernel virtual area
- *
- * @size: size of the area
- * @flags: %VM_IOREMAP for I/O mappings or VM_ALLOC
- *
- * Search an area of @size in the kernel virtual mapping area,
- * and reserved it for out purposes. Returns the area descriptor
- * on success or %NULL on failure.
- */
-struct vm_struct *get_vm_area(unsigned long size, unsigned long flags)
-{
- return __get_vm_area(size, flags, VMALLOC_START, VMALLOC_END);
-}
-
/* Caller must hold vmlist_lock */
struct vm_struct *__remove_vm_area(void *addr)
{
@@ -395,7 +381,8 @@ void *vmap(struct page **pages, unsigned
EXPORT_SYMBOL(vmap);
-void *__vmalloc_area(struct vm_struct *area, unsigned int __nocast gfp_mask, pgprot_t prot)
+void *__vmalloc_area(struct vm_struct *area, unsigned int __nocast gfp_mask,
+ pgprot_t prot, int node)
{
struct page **pages;
unsigned int nr_pages, array_size, i;
@@ -406,9 +393,9 @@ void *__vmalloc_area(struct vm_struct *a
area->nr_pages = nr_pages;
/* Please note that the recursion is strictly bounded. */
if (array_size > PAGE_SIZE)
- pages = __vmalloc(array_size, gfp_mask, PAGE_KERNEL);
+ pages = __vmalloc(array_size, gfp_mask, PAGE_KERNEL, node);
else
- pages = kmalloc(array_size, (gfp_mask & ~__GFP_HIGHMEM));
+ pages = kmalloc_node(array_size, (gfp_mask & ~__GFP_HIGHMEM), node);
area->pages = pages;
if (!area->pages) {
remove_vm_area(area->addr);
@@ -418,7 +405,10 @@ void *__vmalloc_area(struct vm_struct *a
memset(area->pages, 0, array_size);
for (i = 0; i < area->nr_pages; i++) {
- area->pages[i] = alloc_page(gfp_mask);
+ if (node < 0)
+ area->pages[i] = alloc_page(gfp_mask);
+ else
+ area->pages[i] = alloc_pages_node(node, gfp_mask, 0);
if (unlikely(!area->pages[i])) {
/* Successfully allocated i pages, free them in __vunmap() */
area->nr_pages = i;
@@ -446,7 +436,7 @@ fail:
* allocator with @gfp_mask flags. Map them into contiguous
* kernel virtual space, using a pagetable protection of @prot.
*/
-void *__vmalloc(unsigned long size, unsigned int __nocast gfp_mask, pgprot_t prot)
+void *__vmalloc(unsigned long size, unsigned int __nocast gfp_mask, pgprot_t prot, int node)
{
struct vm_struct *area;
@@ -454,13 +444,12 @@ void *__vmalloc(unsigned long size, unsi
if (!size || (size >> PAGE_SHIFT) > num_physpages)
return NULL;
- area = get_vm_area(size, VM_ALLOC);
+ area = get_vm_area_node(size, VM_ALLOC, node);
if (!area)
return NULL;
- return __vmalloc_area(area, gfp_mask, prot);
+ return __vmalloc_area(area, gfp_mask, prot, node);
}
-
EXPORT_SYMBOL(__vmalloc);
/**
@@ -476,11 +465,30 @@ EXPORT_SYMBOL(__vmalloc);
*/
void *vmalloc(unsigned long size)
{
- return __vmalloc(size, GFP_KERNEL | __GFP_HIGHMEM, PAGE_KERNEL);
+ return __vmalloc(size, GFP_KERNEL | __GFP_HIGHMEM, PAGE_KERNEL, -1);
}
EXPORT_SYMBOL(vmalloc);
+/**
+ * vmalloc_node - allocate memory on a specific node
+ *
+ * @size: allocation size
+ * @node; numa node
+ *
+ * Allocate enough pages to cover @size from the page level
+ * allocator and map them into contiguous kernel virtual space.
+ *
+ * For tight cotrol over page level allocator and protection flags
+ * use __vmalloc() instead.
+ */
+void *vmalloc_node(unsigned long size, int node)
+{
+ return __vmalloc(size, GFP_KERNEL | __GFP_HIGHMEM, PAGE_KERNEL, node);
+}
+
+EXPORT_SYMBOL(vmalloc_node);
+
#ifndef PAGE_KERNEL_EXEC
# define PAGE_KERNEL_EXEC PAGE_KERNEL
#endif
@@ -500,7 +508,7 @@ EXPORT_SYMBOL(vmalloc);
void *vmalloc_exec(unsigned long size)
{
- return __vmalloc(size, GFP_KERNEL | __GFP_HIGHMEM, PAGE_KERNEL_EXEC);
+ return __vmalloc(size, GFP_KERNEL | __GFP_HIGHMEM, PAGE_KERNEL_EXEC, -1);
}
/**
@@ -513,7 +521,7 @@ void *vmalloc_exec(unsigned long size)
*/
void *vmalloc_32(unsigned long size)
{
- return __vmalloc(size, GFP_KERNEL, PAGE_KERNEL);
+ return __vmalloc(size, GFP_KERNEL, PAGE_KERNEL, -1);
}
EXPORT_SYMBOL(vmalloc_32);
Index: linux-2.6.14-rc2/fs/xfs/linux-2.6/kmem.c
===================================================================
--- linux-2.6.14-rc2.orig/fs/xfs/linux-2.6/kmem.c 2005-09-19 20:00:41.000000000 -0700
+++ linux-2.6.14-rc2/fs/xfs/linux-2.6/kmem.c 2005-09-23 10:17:20.000000000 -0700
@@ -55,7 +55,7 @@ kmem_alloc(size_t size, unsigned int __n
if (size < MAX_SLAB_SIZE || retries > MAX_VMALLOCS)
ptr = kmalloc(size, lflags);
else
- ptr = __vmalloc(size, lflags, PAGE_KERNEL);
+ ptr = __vmalloc(size, lflags, PAGE_KERNEL, -1);
if (ptr || (flags & (KM_MAYFAIL|KM_NOSLEEP)))
return ptr;
if (!(++retries % 100))
Index: linux-2.6.14-rc2/mm/page_alloc.c
===================================================================
--- linux-2.6.14-rc2.orig/mm/page_alloc.c 2005-09-19 20:00:41.000000000 -0700
+++ linux-2.6.14-rc2/mm/page_alloc.c 2005-09-23 10:17:20.000000000 -0700
@@ -2542,7 +2542,7 @@ void *__init alloc_large_system_hash(con
if (flags & HASH_EARLY)
table = alloc_bootmem(size);
else if (hashdist)
- table = __vmalloc(size, GFP_ATOMIC, PAGE_KERNEL);
+ table = __vmalloc(size, GFP_ATOMIC, PAGE_KERNEL, -1);
else {
unsigned long order;
for (order = 0; ((1UL << order) << PAGE_SHIFT) < size; order++)
^ permalink raw reply
* Re: [PATCH 0/3] netfilter : 3 patches to boost ip_tables performance
From: Harald Welte @ 2005-09-23 17:11 UTC (permalink / raw)
To: Eric Dumazet
Cc: Christoph Lameter, Andi Kleen, Christoph Hellwig, David S. Miller,
linux-kernel, netfilter-devel, netdev
In-Reply-To: <4332D2D9.7090802@cosmosbay.com>
[-- Attachment #1: Type: text/plain, Size: 1327 bytes --]
On Thu, Sep 22, 2005 at 05:50:49PM +0200, Eric Dumazet wrote:
> Christoph Lameter a écrit :
> >It should really be do_set_mempolicy instead to be cleaner. I got a patch here that fixes the
> >policy layer.
> >But still I agree with Christoph that a real vmalloc_node is better. There will be no fuzzing
> >around with memory policies etc and its certainly better performance wise.
>
> vmalloc_node() should be seldom used, at driver init, or when a new
> ip_tables is loaded. If it happens to be a performance problem, then
> we can optimize it. Why should we spend days of work for a function
> that is yet to be used ?
I see a contradiction in your sentence. "a new ip_tables is loaded"
every time a user changes a single rule. There are numerous setups that
dynamically change the ruleset (e.g. at interface up/down point, or even
think of your typical wlan hotspot, where once a user is authorized,
he'll get different rules.
--
- Harald Welte <laforge@netfilter.org> http://netfilter.org/
============================================================================
"Fragmentation is like classful addressing -- an interesting early
architectural error that shows how much experimentation was going
on while IP was being designed." -- Paul Vixie
[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply
* Re: [PATCH 0/3] netfilter : 3 patches to boost ip_tables performance
From: Harald Welte @ 2005-09-23 17:09 UTC (permalink / raw)
To: Andi Kleen; +Cc: Eric Dumazet, linux-kernel, netfilter-devel, netdev
In-Reply-To: <200509221503.21650.ak@suse.de>
[-- Attachment #1: Type: text/plain, Size: 1395 bytes --]
On Thu, Sep 22, 2005 at 03:03:21PM +0200, Andi Kleen wrote:
>
> > 1) No more central rwlock protecting each table (filter, nat, mangle, raw),
> > but one lock per CPU. It avoids cache line ping pongs for each packet.
>
> Another useful change would be to not take the lock when there are no
> rules. Currently just loading iptables has a large overhead.
This is partially due to the netfilter hooks that are registered (so we
always take nf_hook_slow() in the NF_HOOK() macro).
The default policies inside an iptables chain are internally implemented
as a rule. Thus, policies as built-in rules have packet/byte counters.
Therefore, without making a semantic change, we cannot do any of the
following optimizations:
1) not take a lock when the chain is empty
2) not register at the netfilter hook when the chain is empty.
This is well-known, but I don't think we can change the semantics for
the user during a stable kernel series. That's one point where not
having 2.7.x really hurts.
--
- Harald Welte <laforge@netfilter.org> http://netfilter.org/
============================================================================
"Fragmentation is like classful addressing -- an interesting early
architectural error that shows how much experimentation was going
on while IP was being designed." -- Paul Vixie
[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply
* [PATCH] fix NFQUEUE Kconfig dependency (was Re: Fw: Kernel 2.6.14-rc2 compile error)
From: Harald Welte @ 2005-09-23 15:06 UTC (permalink / raw)
To: Patrick McHardy
Cc: Andrew Morton, David Miller, Linux Netdev List,
Linux Kernel Mailinglist
In-Reply-To: <433412A6.2090904@trash.net>
[-- Attachment #1: Type: text/plain, Size: 4340 bytes --]
On Fri, Sep 23, 2005 at 04:35:18PM +0200, Patrick McHardy wrote:
> Andrew Morton wrote:
> > I think we fixed this?
>
> Harald made some dependency fixes, but this one looks new to me.
> Harald?
Please apply the following patch, it fixes the problem.
[NETFILTER]: Fix ip[6]t_NFQUEUE Kconfig dependency
We have to introduce a separate Kconfig menu entry for the NFQUEUE targets.
They cannot "just" depend on nfnetlink_queue, since nfnetlink_queue could
be linked into the kernel, whereas iptables can be a module.
Signed-off-by: Harald Welte <laforge@netfilter.org>
---
commit d7252e42c8832cd5695decb1d834e4dcac4bb8d9
tree bb07ec7f79963d6aa1a9c74a99983cd3c27873b3
parent 1cd841ea786e2a74fc0d66299a024ae6b3b7424a
author Harald Welte <laforge@hanuman.de.gnumonks.org> Fri, 23 Sep 2005 17:04:26 +0200
committer Harald Welte <laforge@hanuman.de.gnumonks.org> Fri, 23 Sep 2005 17:04:26 +0200
net/ipv4/netfilter/Kconfig | 11 +++++++++++
net/ipv4/netfilter/Makefile | 2 +-
net/ipv6/netfilter/Kconfig | 11 +++++++++++
net/ipv6/netfilter/Makefile | 2 +-
4 files changed, 24 insertions(+), 2 deletions(-)
diff --git a/net/ipv4/netfilter/Kconfig b/net/ipv4/netfilter/Kconfig
--- a/net/ipv4/netfilter/Kconfig
+++ b/net/ipv4/netfilter/Kconfig
@@ -521,6 +521,17 @@ config IP_NF_TARGET_TCPMSS
To compile it as a module, choose M here. If unsure, say N.
+config IP_NF_TARGET_NFQUEUE
+ tristate "NFQUEUE Target Support"
+ depends on IP_NF_IPTABLES
+ help
+ This Target replaced the old obsolete QUEUE target.
+
+ As opposed to QUEUE, it supports 65535 different queues,
+ not just one.
+
+ To compile it as a module, choose M here. If unsure, say N.
+
# NAT + specific targets
config IP_NF_NAT
tristate "Full NAT"
diff --git a/net/ipv4/netfilter/Makefile b/net/ipv4/netfilter/Makefile
--- a/net/ipv4/netfilter/Makefile
+++ b/net/ipv4/netfilter/Makefile
@@ -87,6 +87,7 @@ obj-$(CONFIG_IP_NF_TARGET_TCPMSS) += ipt
obj-$(CONFIG_IP_NF_TARGET_NOTRACK) += ipt_NOTRACK.o
obj-$(CONFIG_IP_NF_TARGET_CLUSTERIP) += ipt_CLUSTERIP.o
obj-$(CONFIG_IP_NF_TARGET_TTL) += ipt_TTL.o
+obj-$(CONFIG_IP_NF_TARGET_NFQUEUE) += ipt_NFQUEUE.o
# generic ARP tables
obj-$(CONFIG_IP_NF_ARPTABLES) += arp_tables.o
@@ -96,4 +97,3 @@ obj-$(CONFIG_IP_NF_ARP_MANGLE) += arpt_m
obj-$(CONFIG_IP_NF_ARPFILTER) += arptable_filter.o
obj-$(CONFIG_IP_NF_QUEUE) += ip_queue.o
-obj-$(CONFIG_NETFILTER_NETLINK_QUEUE) += ipt_NFQUEUE.o
diff --git a/net/ipv6/netfilter/Kconfig b/net/ipv6/netfilter/Kconfig
--- a/net/ipv6/netfilter/Kconfig
+++ b/net/ipv6/netfilter/Kconfig
@@ -209,6 +209,17 @@ config IP6_NF_TARGET_REJECT
To compile it as a module, choose M here. If unsure, say N.
+config IP6_NF_TARGET_NFQUEUE
+ tristate "NFQUEUE Target Support"
+ depends on IP_NF_IPTABLES
+ help
+ This Target replaced the old obsolete QUEUE target.
+
+ As opposed to QUEUE, it supports 65535 different queues,
+ not just one.
+
+ To compile it as a module, choose M here. If unsure, say N.
+
# if [ "$CONFIG_IP6_NF_FILTER" != "n" ]; then
# dep_tristate ' REJECT target support' CONFIG_IP6_NF_TARGET_REJECT $CONFIG_IP6_NF_FILTER
# if [ "$CONFIG_EXPERIMENTAL" = "y" ]; then
diff --git a/net/ipv6/netfilter/Makefile b/net/ipv6/netfilter/Makefile
--- a/net/ipv6/netfilter/Makefile
+++ b/net/ipv6/netfilter/Makefile
@@ -21,9 +21,9 @@ obj-$(CONFIG_IP6_NF_FILTER) += ip6table_
obj-$(CONFIG_IP6_NF_MANGLE) += ip6table_mangle.o
obj-$(CONFIG_IP6_NF_TARGET_MARK) += ip6t_MARK.o
obj-$(CONFIG_IP6_NF_TARGET_HL) += ip6t_HL.o
+obj-$(CONFIG_IP6_NF_TARGET_NFQUEUE) += ip6t_NFQUEUE.o
obj-$(CONFIG_IP6_NF_QUEUE) += ip6_queue.o
obj-$(CONFIG_IP6_NF_TARGET_LOG) += ip6t_LOG.o
obj-$(CONFIG_IP6_NF_RAW) += ip6table_raw.o
obj-$(CONFIG_IP6_NF_MATCH_HL) += ip6t_hl.o
obj-$(CONFIG_IP6_NF_TARGET_REJECT) += ip6t_REJECT.o
-obj-$(CONFIG_NETFILTER_NETLINK_QUEUE) += ip6t_NFQUEUE.o
--
- Harald Welte <laforge@netfilter.org> http://netfilter.org/
============================================================================
"Fragmentation is like classful addressing -- an interesting early
architectural error that shows how much experimentation was going
on while IP was being designed." -- Paul Vixie
[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply
* Re: [PATCH 2/3] netfilter : 3 patches to boost ip_tables performance
From: Tim Mattox @ 2005-09-23 14:00 UTC (permalink / raw)
To: Willy Tarreau
Cc: Eric Dumazet, Harald Welte, netdev, netfilter-devel, linux-kernel,
Andi Kleen
In-Reply-To: <20050923040234.GC595@alpha.home.local>
AFAIK, the usual reason gcc can generate better code when you write
bool = complex conditional
if (bool) ...
is because of C's short-circuit conditional evaluation rules for || and &&. By
moving the complex expression to an assignment statement, you are telling
gcc that it is okay and safe to evaluate every part of the expression.
If gcc was smart about being able to prove that all parts of the complex
expression where safe (no possibly null pointer references) and had no
side effects, then it could generate the same code with
if (complex conditional) ...
However, there are many situations where even a magical compiler
couldn't prove that there were no possible side effects, etc., and would
have to generate multiple conditional branches to properly meet the
short-circuit conditional evaluation rules for && and ||.
However, in the specific cases in this thread using FWINV without
|| and && operators, an optimizing compiler "should" be smart enough
to generate more linear code for today's heavily pipelined CPUs.
For now, I guess it's still the duty of the programmer to use coding
style to force the compiler to generate more linear machine code.
On 9/23/05, Willy Tarreau <willy@w.ods.org> wrote:
> On Thu, Sep 22, 2005 at 03:05:50PM +0200, Eric Dumazet wrote:
> (...)
> > It was necessary to get the best code with gcc-3.4.4 on i386 and
> > gcc-4.0.1 on x86_64
> >
> > For example :
> >
> > bool1 = FWINV(ret != 0, IPT_INV_VIA_OUT);
> > if (bool1) {
> >
> > gives a better code than :
> >
> > if (FWINV(ret != 0, IPT_INV_VIA_OUT)) {
> >
> > (one less conditional branch)
> >
> > Dont ask me why, it is shocking but true :(
>
> I also noticed many times that gcc's optimization of "if (complex condition)"
> is rather poor and it's often better to put it in a variable before. I even
> remember that if you use an intermediate variable, it can often generate a
> CMOV instruction on processors which support it, while it produces cond tests
> and jumps without the variable. Generally speaking, if you want fast code,
> you have to write it as a long sequence of small instructions, just as if
> you were writing assembly. As you said, shocking but true.
>
> BTW, cheers for your optimizations !
>
> Regards,
> Willy
>
> -
> 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
>
--
Tim Mattox - tmattox@gmail.com
http://homepage.mac.com/tmattox/
I'm a bright... http://www.the-brights.net/
^ permalink raw reply
* Re: [PATCH 2/3] netfilter : 3 patches to boost ip_tables performance
From: Willy Tarreau @ 2005-09-23 11:33 UTC (permalink / raw)
To: Eric Dumazet
Cc: Willy Tarreau, Harald Welte, netdev, netfilter-devel,
linux-kernel, Andi Kleen
In-Reply-To: <43338F30.6070601@cosmosbay.com>
On Fri, Sep 23, 2005 at 07:14:24AM +0200, Eric Dumazet wrote:
> Even without CMOV support, the suggested patch helps :
>
> Here is the code generated with gcc-3.4.4 on a pentium4 (i686) for :
>
> /********************/
> bool1 = ((ip->saddr&ipinfo->smsk.s_addr) != ipinfo->src.s_addr);
> bool1 ^= !!(ipinfo->invflags & IPT_INV_SRCIP);
>
> bool2 = ((ip->daddr&ipinfo->dmsk.s_addr) != ipinfo->dst.s_addr);
> bool2 ^= !!(ipinfo->invflags & IPT_INV_DSTIP);
>
> if ((bool1 | bool2) != 0) {
>
> /********************/
(...)
I totally agree with your demonstration. It would be interesting to compare
she same code on an architecture with more registers (eg: sparc). One of
the reasons of bad optimization of such constructs on x86 seems to be the
lack of registers for the number of variables and intermediate results.
When you write the code like above, you show the workflow to the compiler
(and I often use the same technique).
Regards,
Willy
^ permalink raw reply
* Re: [NETFILTER] fix conntrack event cache deadlock
From: Harald Welte @ 2005-09-23 9:25 UTC (permalink / raw)
To: David S. Miller; +Cc: netdev, netfilter-devel, kaber
In-Reply-To: <20050922.234724.110942393.davem@davemloft.net>
[-- Attachment #1: Type: text/plain, Size: 913 bytes --]
On Thu, Sep 22, 2005 at 11:47:24PM -0700, David S. Miller wrote:
> From: Harald Welte <laforge@netfilter.org>
> Date: Thu, 22 Sep 2005 16:35:15 +0200
>
> > [NETFILTER] Fix conntrack event cache deadlock/oops
>
> Applied. I hope that doing the event cache call outside
> the lock is safe? It was not depending upon that protection
> in any way was it?
no, not at all. It was a mistake to call it within the locked section.
The call was erroneously placed there, because it fitted there without
reorganizing the code flow.
--
- Harald Welte <laforge@netfilter.org> http://netfilter.org/
============================================================================
"Fragmentation is like classful addressing -- an interesting early
architectural error that shows how much experimentation was going
on while IP was being designed." -- Paul Vixie
[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply
* Re: [PATCH 1/8] orinoco: Remove inneeded system includes.
From: Pavel Roskin @ 2005-09-23 8:45 UTC (permalink / raw)
To: Jeff Garzik
Cc: netdev-u79uwXL29TY76Z2rM5mHXA,
orinoco-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f
In-Reply-To: <4333BE87.6060002-e+AXbWqSrlAAvxtiuMwx3w@public.gmane.org>
Quoting Jeff Garzik <jgarzik-e+AXbWqSrlAAvxtiuMwx3w@public.gmane.org>:
> applied 1-8
Thanks!
Sorry for the empty e-mails.
--
Regards,
Pavel Roskin
-------------------------------------------------------
SF.Net email is sponsored by:
Tame your development challenges with Apache's Geronimo App Server.
Download it for free - -and be entered to win a 42" plasma tv or your very
own Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php
^ permalink raw reply
* Re: [PATCH 1/8] orinoco: Remove inneeded system includes.
From: Jeff Garzik @ 2005-09-23 8:36 UTC (permalink / raw)
To: Pavel Roskin
Cc: netdev-u79uwXL29TY76Z2rM5mHXA,
orinoco-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f
In-Reply-To: <E1EIikx-0003sj-5A-mXXj517/zsQ@public.gmane.org>
applied 1-8
-------------------------------------------------------
This SF.Net email is sponsored by:
Power Architecture Resource Center: Free content, downloads, discussions,
and more. http://solutions.newsforge.com/ibmarch.tmpl
^ permalink raw reply
* [PATCH 1/8] orinoco: Remove inneeded system includes.
From: Pavel Roskin @ 2005-09-23 8:18 UTC (permalink / raw)
To: netdev-u79uwXL29TY76Z2rM5mHXA
Cc: orinoco-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f
Signed-off-by: Pavel Roskin <proski-mXXj517/zsQ@public.gmane.org>
Remove inneeded system includes.
Most system includes are not needed. In particular, the hardware
backends don't need anything network related. Some includes have been
moved from local headers to the C files where they are actually used.
Includes that have to be in the local headers are no longer from the C
sources.
diff --git a/drivers/net/wireless/airport.c b/drivers/net/wireless/airport.c
--- a/drivers/net/wireless/airport.c
+++ b/drivers/net/wireless/airport.c
@@ -15,28 +15,11 @@
#define PFX DRIVER_NAME ": "
#include <linux/config.h>
-
#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/init.h>
-#include <linux/ptrace.h>
-#include <linux/slab.h>
-#include <linux/string.h>
-#include <linux/timer.h>
-#include <linux/ioport.h>
-#include <linux/netdevice.h>
-#include <linux/if_arp.h>
-#include <linux/etherdevice.h>
-#include <linux/wireless.h>
-
-#include <asm/io.h>
-#include <asm/system.h>
-#include <asm/current.h>
-#include <asm/prom.h>
-#include <asm/machdep.h>
+#include <linux/delay.h>
#include <asm/pmac_feature.h>
-#include <asm/irq.h>
-#include <asm/uaccess.h>
#include "orinoco.h"
diff --git a/drivers/net/wireless/hermes.c b/drivers/net/wireless/hermes.c
--- a/drivers/net/wireless/hermes.c
+++ b/drivers/net/wireless/hermes.c
@@ -39,17 +39,10 @@
*/
#include <linux/config.h>
-
#include <linux/module.h>
-#include <linux/types.h>
-#include <linux/threads.h>
-#include <linux/smp.h>
-#include <asm/io.h>
-#include <linux/delay.h>
-#include <linux/init.h>
#include <linux/kernel.h>
-#include <linux/net.h>
-#include <asm/errno.h>
+#include <linux/init.h>
+#include <linux/delay.h>
#include "hermes.h"
diff --git a/drivers/net/wireless/hermes.h b/drivers/net/wireless/hermes.h
--- a/drivers/net/wireless/hermes.h
+++ b/drivers/net/wireless/hermes.h
@@ -30,9 +30,8 @@
* access to the hermes_t structure, and to the hardware
*/
-#include <linux/delay.h>
#include <linux/if_ether.h>
-#include <asm/byteorder.h>
+#include <asm/io.h>
/*
* Limits and constants
diff --git a/drivers/net/wireless/orinoco.c b/drivers/net/wireless/orinoco.c
--- a/drivers/net/wireless/orinoco.c
+++ b/drivers/net/wireless/orinoco.c
@@ -77,28 +77,16 @@
#define DRIVER_NAME "orinoco"
#include <linux/config.h>
-
#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/init.h>
-#include <linux/ptrace.h>
-#include <linux/slab.h>
-#include <linux/string.h>
-#include <linux/timer.h>
-#include <linux/ioport.h>
#include <linux/netdevice.h>
-#include <linux/if_arp.h>
#include <linux/etherdevice.h>
#include <linux/ethtool.h>
#include <linux/wireless.h>
#include <net/iw_handler.h>
#include <net/ieee80211.h>
-#include <asm/uaccess.h>
-#include <asm/io.h>
-#include <asm/system.h>
-
-#include "hermes.h"
#include "hermes_rid.h"
#include "orinoco.h"
diff --git a/drivers/net/wireless/orinoco.h b/drivers/net/wireless/orinoco.h
--- a/drivers/net/wireless/orinoco.h
+++ b/drivers/net/wireless/orinoco.h
@@ -9,8 +9,6 @@
#define DRIVER_VERSION "0.15rc2"
-#include <linux/types.h>
-#include <linux/spinlock.h>
#include <linux/netdevice.h>
#include <linux/wireless.h>
#include <net/iw_handler.h>
diff --git a/drivers/net/wireless/orinoco_cs.c b/drivers/net/wireless/orinoco_cs.c
--- a/drivers/net/wireless/orinoco_cs.c
+++ b/drivers/net/wireless/orinoco_cs.c
@@ -14,30 +14,16 @@
#define PFX DRIVER_NAME ": "
#include <linux/config.h>
-
#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/init.h>
-#include <linux/sched.h>
-#include <linux/ptrace.h>
-#include <linux/slab.h>
-#include <linux/string.h>
-#include <linux/ioport.h>
-#include <linux/netdevice.h>
-#include <linux/if_arp.h>
-#include <linux/etherdevice.h>
-#include <linux/wireless.h>
-
+#include <linux/delay.h>
#include <pcmcia/cs_types.h>
#include <pcmcia/cs.h>
#include <pcmcia/cistpl.h>
#include <pcmcia/cisreg.h>
#include <pcmcia/ds.h>
-#include <asm/uaccess.h>
-#include <asm/io.h>
-#include <asm/system.h>
-
#include "orinoco.h"
/********************************************************************/
diff --git a/drivers/net/wireless/orinoco_nortel.c b/drivers/net/wireless/orinoco_nortel.c
--- a/drivers/net/wireless/orinoco_nortel.c
+++ b/drivers/net/wireless/orinoco_nortel.c
@@ -40,29 +40,13 @@
#define PFX DRIVER_NAME ": "
#include <linux/config.h>
-
#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/init.h>
-#include <linux/sched.h>
-#include <linux/ptrace.h>
-#include <linux/slab.h>
-#include <linux/string.h>
-#include <linux/timer.h>
-#include <linux/ioport.h>
-#include <asm/uaccess.h>
-#include <asm/io.h>
-#include <asm/system.h>
-#include <linux/netdevice.h>
-#include <linux/if_arp.h>
-#include <linux/etherdevice.h>
-#include <linux/list.h>
+#include <linux/delay.h>
#include <linux/pci.h>
-#include <linux/fcntl.h>
-
#include <pcmcia/cisreg.h>
-#include "hermes.h"
#include "orinoco.h"
#define COR_OFFSET (0xe0) /* COR attribute offset of Prism2 PC card */
diff --git a/drivers/net/wireless/orinoco_pci.c b/drivers/net/wireless/orinoco_pci.c
--- a/drivers/net/wireless/orinoco_pci.c
+++ b/drivers/net/wireless/orinoco_pci.c
@@ -93,28 +93,12 @@
#define PFX DRIVER_NAME ": "
#include <linux/config.h>
-
#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/init.h>
-#include <linux/sched.h>
-#include <linux/ptrace.h>
-#include <linux/slab.h>
-#include <linux/string.h>
-#include <linux/timer.h>
-#include <linux/ioport.h>
-#include <linux/netdevice.h>
-#include <linux/if_arp.h>
-#include <linux/etherdevice.h>
-#include <linux/list.h>
+#include <linux/delay.h>
#include <linux/pci.h>
-#include <linux/fcntl.h>
-
-#include <asm/uaccess.h>
-#include <asm/io.h>
-#include <asm/system.h>
-#include "hermes.h"
#include "orinoco.h"
/* All the magic there is from wlan-ng */
diff --git a/drivers/net/wireless/orinoco_plx.c b/drivers/net/wireless/orinoco_plx.c
--- a/drivers/net/wireless/orinoco_plx.c
+++ b/drivers/net/wireless/orinoco_plx.c
@@ -117,29 +117,13 @@
#define PFX DRIVER_NAME ": "
#include <linux/config.h>
-
#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/init.h>
-#include <linux/sched.h>
-#include <linux/ptrace.h>
-#include <linux/slab.h>
-#include <linux/string.h>
-#include <linux/timer.h>
-#include <linux/ioport.h>
-#include <asm/uaccess.h>
-#include <asm/io.h>
-#include <asm/system.h>
-#include <linux/netdevice.h>
-#include <linux/if_arp.h>
-#include <linux/etherdevice.h>
-#include <linux/list.h>
+#include <linux/delay.h>
#include <linux/pci.h>
-#include <linux/fcntl.h>
-
#include <pcmcia/cisreg.h>
-#include "hermes.h"
#include "orinoco.h"
#define COR_OFFSET (0x3e0) /* COR attribute offset of Prism2 PC card */
diff --git a/drivers/net/wireless/orinoco_tmd.c b/drivers/net/wireless/orinoco_tmd.c
--- a/drivers/net/wireless/orinoco_tmd.c
+++ b/drivers/net/wireless/orinoco_tmd.c
@@ -53,29 +53,13 @@
#define PFX DRIVER_NAME ": "
#include <linux/config.h>
-
#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/init.h>
-#include <linux/sched.h>
-#include <linux/ptrace.h>
-#include <linux/slab.h>
-#include <linux/string.h>
-#include <linux/timer.h>
-#include <linux/ioport.h>
-#include <asm/uaccess.h>
-#include <asm/io.h>
-#include <asm/system.h>
-#include <linux/netdevice.h>
-#include <linux/if_arp.h>
-#include <linux/etherdevice.h>
-#include <linux/list.h>
+#include <linux/delay.h>
#include <linux/pci.h>
-#include <linux/fcntl.h>
-
#include <pcmcia/cisreg.h>
-#include "hermes.h"
#include "orinoco.h"
#define COR_VALUE (COR_LEVEL_REQ | COR_FUNC_ENA) /* Enable PC card with interrupt in level trigger */
diff --git a/drivers/net/wireless/spectrum_cs.c b/drivers/net/wireless/spectrum_cs.c
--- a/drivers/net/wireless/spectrum_cs.c
+++ b/drivers/net/wireless/spectrum_cs.c
@@ -22,31 +22,17 @@
#define PFX DRIVER_NAME ": "
#include <linux/config.h>
-
#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/init.h>
-#include <linux/sched.h>
-#include <linux/ptrace.h>
-#include <linux/slab.h>
-#include <linux/string.h>
-#include <linux/ioport.h>
-#include <linux/netdevice.h>
-#include <linux/if_arp.h>
-#include <linux/etherdevice.h>
-#include <linux/wireless.h>
+#include <linux/delay.h>
#include <linux/firmware.h>
-
#include <pcmcia/cs_types.h>
#include <pcmcia/cs.h>
#include <pcmcia/cistpl.h>
#include <pcmcia/cisreg.h>
#include <pcmcia/ds.h>
-#include <asm/uaccess.h>
-#include <asm/io.h>
-#include <asm/system.h>
-
#include "orinoco.h"
static unsigned char *primsym;
-------------------------------------------------------
SF.Net email is sponsored by:
Tame your development challenges with Apache's Geronimo App Server.
Download it for free - -and be entered to win a 42" plasma tv or your very
own Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php
^ 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