linux-kernel - Re: [2.6.36-rc regression] occasional complete system hangs on sparc64 SMP

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]

Message-ID: <19637.64300.446447.318000@pilspetsen.it.uu.se>
Date:	Wed, 13 Oct 2010 20:32:12 +0200
From:	Mikael Pettersson <mikpe@...uu.se>
To:	David Miller <davem@...emloft.net>
Cc:	mikpe@...uu.se, sparclinux@...r.kernel.org,
	linux-kernel@...r.kernel.org
Subject: Re: [2.6.36-rc regression] occasional complete system hangs on
 sparc64 SMP

David Miller writes:
 > From: Mikael Pettersson <mikpe@...uu.se>
 > Date: Sun, 3 Oct 2010 23:36:16 +0200
 > 
 > > I've been testing older kernels and can now say that 2.6.35-git5
 > > and newer kernels are definitely affected, but 2.6.35-git4 seems
 > > solid.  I'll dig deeper into the .35-git4->git5 changes towards
 > > the end of next week.
 > > 
 > > I never got any data out of sysrq-y or any other sysrq when a hang
 > > occurs; usually they'd just print the name of the command but no data,
 > > sometimes they'd oops and crash the kernel even harder.
 > 
 > So I haven't been ignoring this report, but I cannot reproduce
 > it on the SB2500 I have access to right now.
 > 
 > I've been doing gcc builds and testsuite runs in a loop for
 > several days.
 > 
 > Occaisionally I get a dying 'as' which I have seen before and
 > plan to investigate, but no hard hangs and no crashes.
 > 
 > I'll keep prodding, but if you can narrow it down further
 > please let us know.

I've looked at the .35-git4 to -git5 changes, and unfortunately
they cannot be relevant for the sparc64 hangs: the bulk of the
changes are x86-only, and the remaining few are totally harmless.

So either .35-git4 is also broken, even though I haven't been
able to break it after a week of gcc bootstraps and test runs,
or some of the hangs I've seen in later kernels have been caused
by some other bug.  Either way, I don't think I can make progress
bisecting & testing older kernels.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/