From: Allan Stephens <allan.stephens@windriver.com> Applied, thanks. --

From: Patrick McHardy <kaber@trash.net> Applied to net-2.6.26, thanks. --

Ah, the perennial add-same-test-to-every-driver conundrum. I think we are far enough along with MSI to _not_ do this anymore in drivers. The platforms with MSI problems should be discovered, made public, and worked around. Otherwise you hide the same problem, just for someone else to discover with another component of the machine. Jeff --

All, While running iperf (transmit) on NIU driver I get several soft lockup errors. I have seen similar soft lockup errors on kernels 2.6.18 and later (with other 10G NIC drivers with LSO disabled) but it is more frequent/pronounced on 2.6.24 kernel. oprofile is not giving much clue as where the cpus are spending time (atleast doesn't indicate any thing abnormal in the driver(s)). The lockup location as indicated by the error message is not consistent and varies from time time and could be...

Jarek, Reproduced the lockup with irqbalance disabled and with single src of interrupt (TX interrupt, UDP transmit). Lockup appears in different location though. Regards matheos irq of interest: 454 (TX interrupt) 454: 19249 93234 907186 2691 0 188 0 160 PCI-MSI-edge eth6 455: 22607 15083 5 13104 25569 161519 62514 25637 PCI-MSI-edge eth6 456: 22390 14921 ...

OOPS! I see - only not cleared. So it's OK! Sorry, Jarek P. --

The commits f3db4851 ([NETNS][IPV6] ip6_fib - fib6_clean_all handle several network namespaces) and 69ddb805 ([NETNS][IPV6] route6 - Make proc entry /proc/net/rt6_stats per namespace) made some proc files per net. Both of them introduced potential OOPS - get_proc_net can return NULL, but this check is lost - and a struct net leak - in case single_open() fails the previously got net is not put. Kill all these bugs with one patch. Signed-off-by: Pavel Emelyanov <xemul@openvz.org> --- ...

From: Daniel Lezcano <dlezcano@fr.ibm.com> Applied, thanks everyone! --

Hello, I have a question regarding disabling sending redirects in IPV6 in the linux kernel: In include/linux/ipv6.h, I see a member called accept_redirects in ipv6_devconf, but there is no send_redirects member. There is /proc/sys/net/ipv6/conf/all/accept_redirects but I could not find /proc/sys/net/ipv6/conf/all/send_redirects. (In IPV4 there is both /proc/sys/net/ipv4/conf/all/send_redirects and /proc/sys/net/ipv4/conf/all/accept_redirects). Is there some reason that there is no way (a...

Make /proc/net/ip6_flowlabel show only flow labels belonging to the current network namespace. Signed-off-by: Benjamin Thery <benjamin.thery@bull.net> Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com> --- net/ipv6/ip6_flowlabel.c | 39 +++++++++++++++++++++++---------------- 1 file changed, 23 insertions(+), 16 deletions(-) Index: net-2.6.26/net/ipv6/ip6_flowlabel.c =================================================================== --- net-2.6.26.orig/net/ipv6/ip6_flowlabel.c ...

This patch makes the anycast and ip6_flowlabel per namespace. -- --

Andrew Paprocki <andrew@ishiboo.com> : Please revert: this register is reserved in the 8168. PHY access and PHY status are ok for the 8168. TBI and PM registers are not. The 8168 should not be included here. -- Ueimor --

Andrew Paprocki <andrew@ishiboo.com> : I am not sure that the 8110/8169 documentation will be of much help but you can still give it a try. Can you elaborate on the kind of hang ? Which kernel do you use ? -- Ueimor --

Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> diff --git a/MAINTAINERS b/MAINTAINERS index 6d628fb..1e14206 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -3512,6 +3512,13 @@ M: pfg@sgi.com L: linux-ia64@vger.kernel.org S: Supported +SFC NETWORK DRIVER +P: Steve Hodgson +P: Ben Hutchings +P: Robert Stonehouse +M: linux-net-drivers@solarflare.com +S: Supported + SGI VISUAL WORKSTATION 320 AND 540 P: Andrey Panin M: pazke@donpac.ru diff --git a/drivers/net/Kconfig b/d...

Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> diff --git a/drivers/net/sfc/ethtool.c b/drivers/net/sfc/ethtool.c new file mode 100644 index 0000000..786c2a2 --- /dev/null +++ b/drivers/net/sfc/ethtool.c @@ -0,0 +1,461 @@ +/**************************************************************************** + * Driver for Solarflare Solarstorm network controllers and boards + * Copyright 2005-2006 Fen Systems Ltd. + * Copyright 2006-2008 Solarflare Communications Inc. + * + * This progra...

Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> diff --git a/drivers/net/sfc/falcon.c b/drivers/net/sfc/falcon.c new file mode 100644 index 0000000..ca2eb86 --- /dev/null +++ b/drivers/net/sfc/falcon.c @@ -0,0 +1,2723 @@ +/**************************************************************************** + * Driver for Solarflare Solarstorm network controllers and boards + * Copyright 2005-2006 Fen Systems Ltd. + * Copyright 2006-2008 Solarflare Communications Inc. + * + * This program ...

Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> diff --git a/drivers/net/sfc/rx.c b/drivers/net/sfc/rx.c new file mode 100644 index 0000000..cf74b5f --- /dev/null +++ b/drivers/net/sfc/rx.c @@ -0,0 +1,875 @@ +/**************************************************************************** + * Driver for Solarflare Solarstorm network controllers and boards + * Copyright 2005-2006 Fen Systems Ltd. + * Copyright 2005-2008 Solarflare Communications Inc. + * + * This program is free softw...

Commit 8b7817f3a959ed99d7443afc12f78a7e1fcc2063 ([IPSEC]: Add ICMP host relookup support) introduced some dst leaks on error paths: the rt pointer can be forgotten to be put. Fix it bu going to a proper label. Found after net namespace's lo refused to unregister :) Many thanks to Den for valuable help during debugging. Herbert pointed out, that xfrm_lookup() will put the rtable in case of error itself, so the first goto fix is redundant. Signed-off-by: Pavel Emelyanov <xemul@openvz.org> ...

I didn't check the kernel build after pulling dev->nd_net access optimization from net-2.6.26 tree :( Sorry. Signed-off-by: Pavel Emelyanov <xemul@openvz.org> --- diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c index 7049b3f..3697e05 100644 --- a/net/ipv4/icmp.c +++ b/net/ipv4/icmp.c @@ -821,7 +821,7 @@ static void icmp_echo(struct sk_buff *skb) { struct net *net; - net = skb->dst->dev->nd_net; + net = dev_net(skb->dst->dev); if (!net->ipv4.sysctl_icm...

Hi, Any comments on this? Would be great to get beet working soon! br, j == original post: This patch fixes the ipsec BEET (Bound End-to-End Tunnel) mode interfamily handling of the net-2.6 (25-rc3) kernel, as specified by the ietf draft found at: http://www.ietf.org/internet-drafts/draft-nikander-esp-beet-mode-08.txt It is sent as a RFC as it doesn't fit that neatly into the newly cleaned-up xfrm model - I suspect that there could be a better way to solve this. Signed-off-by: Joaki...

Why are you disabling this? --

This mostly re-uses the net, used in icmp netnsization patches from Denis. After this ICMP sysctls are completely virtualized. Signed-off-by: Pavel Emelyanov <xemul@openvz.org> --- net/ipv4/icmp.c | 25 ++++++++++++++++--------- 1 files changed, 16 insertions(+), 9 deletions(-) diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c index 958a384..7049b3f 100644 --- a/net/ipv4/icmp.c +++ b/net/ipv4/icmp.c @@ -270,7 +270,8 @@ int xrlim_allow(struct dst_entry *dst, int timeout) return rc;...

Add some flesh to ipv4_sysctl_init_net and ipv4_sysctl_exit_net, i.e. copy the table, alter .data pointers and register it per-net. Other ipv4_table's sysctls are now global, but this is going to change once sysctl permissions patches migrate from -mm tree to mainline in 2.6.26 merge window :) Signed-off-by: Pavel Emelyanov <xemul@openvz.org> --- include/net/netns/ipv4.h | 1 + net/ipv4/sysctl_net_ipv4.c | 138 ++++++++++++++++++++++++++++--------------- 2 files changed, 91 inse...

Initialization is moved to icmp_sk_init, all the places, that refer to them use init_net for now. Signed-off-by: Pavel Emelyanov <xemul@openvz.org> --- include/net/icmp.h | 7 ----- include/net/netns/ipv4.h | 7 +++++ net/ipv4/icmp.c | 60 ++++++++++++++++++++++--------------------- net/ipv4/sysctl_net_ipv4.c | 12 ++++---- 4 files changed, 44 insertions(+), 42 deletions(-) diff --git a/include/net/icmp.h b/include/net/icmp.h index faba64d..dddb839 10064...

This includes adding pernet_operations, empty init and exit hooks and a bit of changes in sysctl_ipv4_init just not to have this part in next patches. Signed-off-by: Pavel Emelyanov <xemul@openvz.org> --- net/ipv4/sysctl_net_ipv4.c | 24 +++++++++++++++++++++++- 1 files changed, 23 insertions(+), 1 deletions(-) diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c index 88286f3..c2fca30 100644 --- a/net/ipv4/sysctl_net_ipv4.c +++ b/net/ipv4/sysctl_net_ipv4.c @@ -811,...