1) TIPC and ipv6 endianness cures from Al Viro. 2) Fix copy_from_user() results being used as error returns, from Sam Ravnborg, based upon a patch from Pavel E. 3) When asking for null authentication algorithm in ESP, we get a null hmac, instead of just a plain null digest. Fix from Herbert Xu. 4) Missing SunRPC kernel-doc annotation from Randy Dunlap. 5) IPSEC auditing prints out the flowlabel incorrectly. Fix from Yoshifuji HIDEAKI. 6) Compat 32-bit support for multicast soc...

From: Russell King <rmk@arm.linux.org.uk> The ReasmTimeout and ReasmFails look interesting. Maybe it was the namespace bits? Pavel, could you take a quick look? Thanks. --

who cares about ATM? Well anyone using ADSL/2/2+ high speed internet access is stuck with it, and that's not a small number of people. Diagnosing problems is often complicated by the fact that large backhaul networks appear as only 1 IP hop in traceroute. The goal is to support sending and receiving F5 OAM cells on a Linux host, from in-use PVCs bound to a br2684 bridge, your typical USB ADSL modem. The ATM OAM segment "ping" can be useful in troubleshooting, yet few modems support it, and *non...

Backgroud: OAM cells are like the ICMP of ATM networking, they are separate from payload cells, not part of the headers. They do share the VPI/VCI of the data circuit, but have a different cell Packet Type Indication (PTI) With br2684, it exposes an ethernet-like device. Tcpdump can work on that no problem, and perhaps it could expose the headers of the first/last cell of each "frame", but I don't see how the OAM cells would fit into capturing packets that way. Or can an AF_PACKET socket b...

Update the collision counter as well. Signed-off-by: Francois Romieu <romieu@fr.zoreil.com> --- drivers/net/sis190.c | 38 +++++++++++++++++++++++++++++++++++--- 1 files changed, 35 insertions(+), 3 deletions(-) diff --git a/drivers/net/sis190.c b/drivers/net/sis190.c index 20f4829..abc63b0 100644 --- a/drivers/net/sis190.c +++ b/drivers/net/sis190.c @@ -212,6 +212,12 @@ enum _DescStatusBit { THOL2 = 0x20000000, THOL1 = 0x10000000, THOL0 = 0x00000000, + + WND = 0x0008000...

- remove the function pointer to help gcc optimizing the inline pci_dma functions - pci_dma_sync_single_for_cpu is not needed for a single large packet - convert rtl8169_try_rx_copy to bool b449655ff52ff8a29c66c5fc3fc03617e61182ee did the same for the r8169 driver. Signed-off-by: Francois Romieu <romieu@fr.zoreil.com> --- drivers/net/sis190.c | 54 +++++++++++++++++++++++++------------------------ 1 files changed, 28 insertions(+), 26 deletions(-) diff --git a/drivers/net/sis19...

There is no DMA involved here. Align the IP header without condition. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Acked-by: Francois Romieu <romieu@fr.zoreil.com> --- drivers/net/sis190.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/sis190.c b/drivers/net/sis190.c index 0d6aa1f..248c385 100644 --- a/drivers/net/sis190.c +++ b/drivers/net/sis190.c @@ -526,9 +526,9 @@ static inline int sis190_try_rx_copy(struct sk_buff **sk_buff...

The following series fixes some bugs as well as enabling peer-2-peer applications including OpenMPI and HPMPI. I hope this can make 2.6.26. NOTE: The changes in patch 3 require a new firmware version. I added the version change to drivers/net/cxgb3/version.h in this patch so that the changes that require the new firmware as well as the version bump are all in one git commit. This keeps things like 'git bisect' from leaving the driver broken. -- Steve. --

What are the interoperability implications of this? Looking closer I see that iw_nes has the send_first module parameter. How does this interact with that? I guess it's fine to apply this, but do we have a plan for how we want to handle this issue in the long-term? - R. --

OpenMPI and other stress testing exposed a few bad bugs in handling aborts in the middle of a normal close. - serialize abort reply and peer abort processing with disconnect processing - warn (and ignore) if ep timer is stopped when it wasn't running - cleaned up disconnect path to correctly deal with aborting and dead endpoints - in iwch_modify_qp(), add a ref to the ep before releasing the qp lock if iwch_ep_disconnect() will be called. The dref after calling disconnect. Signed-off-b...

[NET_SCHED] sch_sfq: fix queue limiting while enqueuing Current way to use q->limit both as per flow and total queue limit doesn't make much sense. There would be no reason for sfq's limit parameter if the same could be done with ifconfig's txqueuelen, and it's too small for this anyway: with a number of flow queues around maximum (SFQ_DEPTH) each queue would be ~1 packet long, which means no queuing... There is also changed a place for checking the total limit: let's do it at the beginning...

OK, you are right: it would break some scripts... But IMHO this way of treating tx_queue_len isn't right: it can be changed after initialization too, and looks like perfect way to control the queue size globally. Anyway, current use of "limit" parameter in sfq warps its idea. Another possibility would be like this: sch->q.qlen >= max_t(__u32, sch->dev->tx_queue_len, q->limit) or removing this global sch->q.qlen check at all. Or maybe we need to add one more tc parameter for sf...

This is broken, we hold dev->queue_lock in ->destroy. --

My point was that it might sleep and can thus not be called in atomic context. This was a misunderstanding on my side though and your patch is fine. --

What about PCI posting? --

Hi Dave: This patch is a minor optimisation to avoid using hmac on null authentication when explicitly specified by the user. [IPSEC]: Use digest_null directly for auth Previously digest_null had no setkey function which meant that we used hmac(digest_null) for IPsec since IPsec always calls setkey. Now that digest_null has a setkey we no longer need to do that. In fact when only confidentiality is specified for ESP we already use digest_null directly. However, when the null algorithm is ...

Found by sparse dubious !x & y warning...hidden in the GetBit macro why !Word doesn't make any sense. Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> --- drivers/net/eepro.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/drivers/net/eepro.c b/drivers/net/eepro.c index 83bda6c..56f5049 100644 --- a/drivers/net/eepro.c +++ b/drivers/net/eepro.c @@ -633,7 +633,7 @@ static void __init printEEPROMInfo(struct net_device *dev) printk(KERN_DEBUG " PC...

From: Al Viro <viro@ZenIV.linux.org.uk> Applied, thanks Al. --

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> --- net/ipv6/ip6mr.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/net/ipv6/ip6mr.c b/net/ipv6/ip6mr.c index c8c6e33..2de3c46 100644 --- a/net/ipv6/ip6mr.c +++ b/net/ipv6/ip6mr.c @@ -358,7 +358,7 @@ static int pim6_rcv(struct sk_buff *skb) if (pim->type != ((PIM_VERSION << 4) | PIM_REGISTER) || (pim->flags & PIM_NULL_REGISTER) || (ip_compute_csum((void *)pim, sizeof(*pim)) != 0 &...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> --- net/mac80211/mesh.h | 1 + net/mac80211/mesh_hwmp.c | 1 - 2 files changed, 1 insertions(+), 1 deletions(-) diff --git a/net/mac80211/mesh.h b/net/mac80211/mesh.h index 742003d..9ee3aff 100644 --- a/net/mac80211/mesh.h +++ b/net/mac80211/mesh.h @@ -13,6 +13,7 @@ #include <linux/types.h> #include <linux/jhash.h> +#include <asm/unaligned.h> #include "ieee80211_i.h" diff --git a/net/mac80211/mes...

It has been discussed or it has been agreed? IMO it's a bullshit - especially for low-level stuff like that. As for the way some do includes... It's not exactly a family-friendly list, but still I'd rather not comment in details on the existing practices... --

From: Al Viro <viro@ZenIV.linux.org.uk> Patch applied, thanks a lot Al. --

From: YOSHIFUJI Hideaki / 吉藤英明 <yoshfuji@linux-ipv6.org> You'll have to be more specific than just saying "fixed" when you submit a patch like this. What about the format is it fixing? Also, can this possibly break existing audit log scanning tools and parsers? --

From: YOSHIFUJI Hideaki / 吉藤英明 <yoshfuji@linux-ipv6.org> Patch applied, James if you have some objections let me know, otherwise I'll toss this to -stable too. Thanks! --

From: Jason Riedy <jason@acm.org> No, the "part 1" is exactly the patch posted by Tomas Winkler and referenced in the 49186b4a083655 commit, which is: commit d7d313000ba2fc94a5383511a17ff38a39bab928 Author: Tomas Winkler <tomas.winkler@intel.com> Date: Wed Apr 23 03:48:57 2008 -0700 iwlwifi: Fix built-in compilation of iwlcore This patch fixes problem in Makefile that prevented built-in compilation of iwlcore Commit that caused this problem: eadd3c4...