mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: William Roche <william.roche@oracle.com>
To: Borislav Petkov <bp@alien8.de>
Cc: yazen.ghannam@amd.com, tony.luck@intel.com, tglx@kernel.org,
	mingo@redhat.com, dave.hansen@linux.intel.com, x86@kernel.org,
	hpa@zytor.com, linux-edac@vger.kernel.org,
	linux-kernel@vger.kernel.org, John.Allen@amd.com,
	jane.chu@oracle.com
Subject: Re: [PATCH v2 1/1] x86/mce/amd: Fix VM crash during deferred error handling
Date: Thu, 12 Mar 2026 16:11:10 +0100	[thread overview]
Message-ID: <8e35298b-7511-4f7f-8f13-9b03738b286c@oracle.com> (raw)
In-Reply-To: <20260312144203.GCabLQuwFySHkkCyBO@fat_crate.local>

Thank you for your reply,


On 3/12/26 15:42, Borislav Petkov wrote:
> On Wed, Feb 18, 2026 at 04:30:25PM +0000, “William Roche wrote:
>> From: William Roche <william.roche@oracle.com>
>>
>> A non Scalable MCA system may prevent access to SMCA specific registers
> 
> "may prevent"?
> 
> Please explain in the commit message the whole scenario how you're triggering
> this in detail.
> 

 From the kernel point of view (regardless if it is running on bare 
metal or in a VM), access to these registers registers is provided by 
the platform: either the Hardware or the emulation framework.

Yazen indicated on Feb 12 that "AMD systems generally have a 
Read-as-Zero/Writes-Ignored behavior when accessing unimplemented MCA 
registers", but you rightly indicated on Feb 9 that "KVM works as 
advertized" and so prevents access to unimplemented SMCA specific 
registers. That's the reason why I had to say "may".

This access crashes on AMD VMs and "may" work on AMD hardware according 
to Yazen.


>> like MCA_DESTAT. This is the case of QEMU/KVM VMs, where the kernel
>> has to check for the SMCA feature before accessing MCA_DESTAT.
>>
>> Fixes: 7cb735d7c0cb ("x86/mce: Unify AMD DFR handler with MCA Polling")
>> Signed-off-by: William Roche <william.roche@oracle.com>
>> Reviewed-by: Yazen Ghannam <yazen.ghannam@amd.com>
>> Cc: stable@vger.kernel.org
> 
> AFAIR, you're injecting errors. This is not really a critical fix that
> warrants this going to stable.

Errors are injected into VMs by the hypervisor when real memory hardware 
errors occur on the system that impact the VM address space.
This is not only a test, this is real life mechanism. With the fix 
7cb735d7c0cb that has been integrated, VMs kernel running on AMD now 
crashes on Deferred errors, where it used to be able to deal with them 
before this commit.
That's the reason why we need this additional fix.

> 
>> ---
>>   arch/x86/kernel/cpu/mce/amd.c | 17 +++++++++++------
>>   1 file changed, 11 insertions(+), 6 deletions(-)
>>
>> diff --git a/arch/x86/kernel/cpu/mce/amd.c b/arch/x86/kernel/cpu/mce/amd.c
>> index 3f1dda355307..7b9932f13bca 100644
>> --- a/arch/x86/kernel/cpu/mce/amd.c
>> +++ b/arch/x86/kernel/cpu/mce/amd.c
>> @@ -875,13 +875,18 @@ void amd_clear_bank(struct mce *m)
>>   {
>>   	amd_reset_thr_limit(m->bank);
>>   
>> -	/* Clear MCA_DESTAT for all deferred errors even those logged in MCA_STATUS. */
>> -	if (m->status & MCI_STATUS_DEFERRED)
>> -		mce_wrmsrq(MSR_AMD64_SMCA_MCx_DESTAT(m->bank), 0);
>> +	if (mce_flags.smca) {
> 
> All this code should not run in a VM. So why does it?

Why do you say that this code should not run in a VM ?
Error injection mechanism has been running for several years with QEMU/KVM.
I must be missing something here. Please let me know.

> 
> What is the use case we're supposed to support here?
> 

Dealing with real life deferred memory errors impacting VMs address space.

I hope this clarifies the need for this new kernel fix.

Thanks again,
William.


  reply	other threads:[~2026-03-12 15:11 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-02-18 16:30 [PATCH v2 0/1] AMD VM crashing on deferred memory error injection “William Roche
2026-02-18 16:30 ` [PATCH v2 1/1] x86/mce/amd: Fix VM crash during deferred error handling “William Roche
2026-03-12 14:42   ` Borislav Petkov
2026-03-12 15:11     ` William Roche [this message]
2026-03-12 16:04       ` Borislav Petkov
2026-03-12 22:44         ` William Roche
2026-03-13 20:10           ` Borislav Petkov
2026-03-16 15:27             ` William Roche
2026-03-13 20:26           ` Yazen Ghannam
2026-03-16 15:26             ` William Roche
2026-03-19 14:25               ` Yazen Ghannam
2026-03-12 14:23 ` [PATCH v2 0/1] AMD VM crashing on deferred memory error injection William Roche

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=8e35298b-7511-4f7f-8f13-9b03738b286c@oracle.com \
    --to=william.roche@oracle.com \
    --cc=John.Allen@amd.com \
    --cc=bp@alien8.de \
    --cc=dave.hansen@linux.intel.com \
    --cc=hpa@zytor.com \
    --cc=jane.chu@oracle.com \
    --cc=linux-edac@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=tglx@kernel.org \
    --cc=tony.luck@intel.com \
    --cc=x86@kernel.org \
    --cc=yazen.ghannam@amd.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®