Subject: Re: fftw_test failing on Core 2 Duo/GENERIC.MP
To: Andrew Doran <email@example.com>
From: Brad du Plessis <firstname.lastname@example.org>
Date: 12/03/2007 11:07:57
Andrew Doran wrote:
> On Fri, Nov 30, 2007 at 09:22:39AM +0200, Brad du Plessis wrote:
>> On this system a few of days ago (Intel DG33FB, Intel 2180, 4.0_RC3
>> GENERIC.MP) I encountered random segmentation faults when trying to
>> build pkgsrc packages.
>> You'll note that what is a bit concerning is that no system running with
>> both cores enabled passed this test. Has anyone else by any chance tried
>> this on a Core 2 Duo system and seen similar results?
>> Is this something to worry about, or is this a bogus test?
> As Matthias mentioned it is likely to be a TLB management bug in the CPU,
> possibly not one that can be patched by the BIOS.
> Do your tests fork a lot? Can you try adding a "tlbflushg();" to the end
> of pmap_destroy() in src/sys/arch/x86/x86/pmap.c and see if that helps?
The test does no forking that I'm aware of. I did however run 2
instances of the application to occupy both CPU cores. I tried this
pmap.c change and the test ran for about 30 minutes before failing. Is
there anything else I can check/try?