netdev - RE: [PATCH] tcp: tcp_release_cb() should release socket ownership

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Fri, 07 Mar 2014 08:01:05 -0800
From:	Eric Dumazet <eric.dumazet@...il.com>
To:	Lars Persson <lars.persson@...s.com>,
	David Miller <davem@...emloft.net>
Cc:	"netdev@...r.kernel.org" <netdev@...r.kernel.org>
Subject: RE: [PATCH] tcp: tcp_release_cb() should release socket ownership

On Fri, 2014-03-07 at 16:45 +0100, Lars Persson wrote:
> So far no deadlock on the socket lock is seen with this patch. I have seen two out-of-memory crashes which leaves the system deadlocked trying to allocate memory.
> 
> Stack traces and kernel log from one of the oom crashes is attached. I was thinking this may be caused by having the dirty_ratio set too high given our product's heavy use of memory. Currently re-testing with dirty_ratio reduced to 5%.
> 
> 
> [11515.530000] Recording maint invoked oom-killer: gfp_mask=0x201da, order=0, oom_score_adj=0
> [11515.540000] CPU: 0 PID: 1304 Comm: Recording maint Tainted: G           O 3.10.32 #10
> [11515.550000] Stack : 80666e52 00000049 80660000 00000000 805563dc 8b1876f8 8054af80 805d2227
> 	  00000518 00000000 80656d40 8b1876f8 000007ce 00000002 00000000 8047d858
> 	  00000000 8003bb64 00000006 00000000 8054d024
>  [11515.580000]  8b189a84 8b189a84 8054af80
> 	  00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
> 	  00000000 00000000 00000000 00000000 00000000 00000000 00000000 8b189a18
> 	  ...


Well, this is 3.10.32, with local changes to dirty_ratio....

so I will simply ignore mm issues for the moment.

Thanks a lot Lars

Tested-by: Lars Persson <lars.persson@...s.com>



--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html