Re: WHEA Support in Linux

[Chen Gong <gong.chen@linux.intel.com>]

[-- Attachment #1: Type: text/plain, Size: 2751 bytes --]

On Wed, Feb 27, 2013 at 07:00:53PM +0000, Manish wrote:
> Date:	Wed, 27 Feb 2013 19:00:53 +0000 (UTC)
> From: Manish <manishmj1111@gmail.com>
> To: linux-acpi@vger.kernel.org
> Subject: Re: WHEA Support in Linux
> User-Agent: Loom/3.14 (http://gmane.org/)
> 
> 
> 
> 
> Thanks a lot for your quick response.
> 
> How can we verify that a particular version of Linux supports WHEA or not?
> In Windows, OSC method returns supported for the Whea Guid. Is it applicable
> also for Linux?

Check drivers/acpi/apei/ghes.c, in function ghes_init:
...
        rc = apei_osc_setup();
        if (rc == 0 && osc_sb_apei_support_acked)
                pr_info(GHES_PFX "APEI firmware first mode is enabled by APEI bit and WHEA _OSC.\n");
        else if (rc == 0 && !osc_sb_apei_support_acked)
                pr_info(GHES_PFX "APEI firmware first mode is enabled by WHEA _OSC.\n");
        else if (rc && osc_sb_apei_support_acked)
                pr_info(GHES_PFX "APEI firmware first mode is enabled by APEI bit.\n");
        else
                pr_info(GHES_PFX "Failed to enable APEI firmware first mode.\n");
...
search above output from dmesg.

> Also, In Windows, Error Injection is verifed using WHEA HCT tool and the Errors
> are logged in Windows Event Viewer.
> Is there any such facility/tool available to verify Error Injection (EINJ) stack
> in Linux?
modprob einj and use it. Or you can use mce-test to do this:
git://git.kernel.org/pub/scm/linux/kernel/git/gong.chen/mce-test.git

read it and use it by youself.

> Also, Is there any thing similar to Event Viewer of Windows in Linux where the
> WHEA errors are logged?
> 
mcelog

> (I'm getting the following message in dmesg log from RHEL:
> "ERST: Could not register with persistent Store"
> 
> You don't compile pstore support/not loaded it as a module/no NVRAM, UEFI
> runtime space etc. to support pstore.)
> 
> As far as the above ERST dmesg is concerned, We are allocating a Flash Block for
> ERST persistent store.
> The persistent store will be used for Writing/Reading an Error Record by the OS
> in case of NMI when fatal error is generated.
Yes.

> In case of fatal error from Windows, it is getting written as well as Read
> correctly from the persistent store and error is avialable in Event Viewer.
Same as in Linux.

> Can you please help us to narrow down this issue or in what could be reason for
> this error in dmesg log.
In simple, pstore doesn't work, so the root cause in pstore. You need to
debug it. Use kernel command parameter pstore.backend=erst to try again.
The prerequisite is your platform supports erst.
(Check dmesg to find "Error Record Serialization Table (ERST) support is
initialized.")

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 836 bytes --]