linux-kernel - Re: [RFC EDAC/GHES] edac: lock module owner to avoid error report conflicts

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:	Thu, 1 Nov 2012 15:47:30 -0200
From:	Mauro Carvalho Chehab <mchehab@...hat.com>
To:	Tony Luck <tony.luck@...el.com>
Cc:	Borislav Petkov <bp@...en8.de>,
	Linux Edac Mailing List <linux-edac@...r.kernel.org>,
	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [RFC EDAC/GHES] edac: lock module owner to avoid error report
 conflicts

Em Thu, 1 Nov 2012 10:25:23 -0700
Tony Luck <tony.luck@...el.com> escreveu:

> On Thu, Nov 1, 2012 at 4:47 AM, Mauro Carvalho Chehab
> <mchehab@...hat.com> wrote:
> > Take a look at arch/x86/kernel/cpu/mcheck/mce-apei.c:
> >
> >         void apei_mce_report_mem_error(int corrected, struct cper_sec_mem_err *mem_err)
> >         {
> >                 struct mce m;
> >
> >                 /* Only corrected MC is reported */
> >                 if (!corrected || !(mem_err->validation_bits &
> >                                         CPER_MEM_VALID_PHYSICAL_ADDRESS))
> >                         return;
> >
> >                 mce_setup(&m);
> >                 m.bank = 1;
> >                 /* Fake a memory read corrected error with unknown channel */
> >                 m.status = MCI_STATUS_VAL | MCI_STATUS_EN | MCI_STATUS_ADDRV | 0x9f;
> >                 m.addr = mem_err->physical_addr;
> >                 mce_log(&m);
> >                 mce_notify_irq();
> >         }
> >
> > Bank information there is fake; status is fake. Only addr is really filled
> > there; it works only for corrected errors.
> 
> This went in like this to help out the Westmere-EX processors that
> didn't fill out MCi_ADDR for corrected errors. APEI could get the
> address from some platform CSRs ... reporting via /dev/mcelog
> so that predictive analysis in mcelog(8) would work on these machines.

Ok, but it is broken on other platforms like Sandy Bridge.

> I don't think we can rip it out yet ... not until those machines are
> shuffled off to recycle heaven.

Perhaps then we could add a logic at apei-mce to only forward errors to
MCE on the platforms where the MCE log is known to be right.

> But perhaps we should get smarter about which machines we enable
> APEI on?

That makes sense. IMO, APEI should be on by default only if no other driver
exists, like in the case of Nehalem-EX. For platforms supported by i7core_edac,
sb_edac and amd64_edac, we could add a parameter to explicitly force it to
be on, otherwise, APEI will be disabled.

> If we get everything we need from the machine check banks,
> then the detour via the BIOS to report the same thing again isn't helpful.

Agreed.

Regards,
Mauro
--
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/