linux-kernel - [2.6.27-rc6 regression] Intermittent boot problem with 2.6.27-rc6

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-next>] [day] [month] [year] [list]

Message-ID: <20080916200414.GB16993@localdomain>
Date:	Tue, 16 Sep 2008 13:04:14 -0700
From:	Ravikiran G Thirumalai <kiran@...lex86.org>
To:	linux-kernel@...r.kernel.org, "Rafael J. Wysocki" <rjw@...k.pl>
Cc:	torvalds@...ux-foundation.org, Ingo Molnar <mingo@...e.hu>
Subject: [2.6.27-rc6 regression] Intermittent boot problem with 2.6.27-rc6

Hi,
Booting 2.6.27-rc6, or the current Linus' git on large x86_64 NUMA wedges
intermittently.  This happens on machines with more than 4 NUMA nodes.  The
same kernel boots on similar hardware with fewer NUMA nodes (or boards).
2.6.27-rc3 seemed to boot well on similar configurations, so looks like a
regression crept in between 2.6.27-rc3 and rc6.

I am trying a few things -- commandline options to eliminate popular
suspects (such as unsynced tsc clocksources)  as well as git bisect, but
wanted to report the regression in the meanwhile.
Here's the dmesg of a 24 core 6 node wedged system.

Thanks,
Kiran

View attachment "paw40.dmesg" of type "text/plain" (57207 bytes)