From: Lancelot SIX <Lancelot.Six@amd.com>
To: Jay Cornwall <jay.cornwall@amd.com>,
"amd-gfx@lists.freedesktop.org" <amd-gfx@lists.freedesktop.org>
Subject: Re: [PATCH 2/2] drm/amdkfd: Clear MODE.VSKIP in gfx9 trap handler
Date: Fri, 10 Jan 2025 19:45:11 +0000 [thread overview]
Message-ID: <fed57f7d-a0a5-4946-9f87-1dffde14eb21@amd.com> (raw)
In-Reply-To: <d72b85bc-9fd3-472e-b5cf-e5f83d4aa6c3@amd.com>
On 10/01/2025 19:37, Jay Cornwall wrote:
> On 1/10/2025 12:14, Six, Lancelot wrote:
>
>>> If user shader issues S_SETVSKIP then this state will persist when
>>> executing the trap handler, causing vector instructions to be
>>> skipped.
>>>
>>> Restore VSKIP state before resuming the user shader.
>>
>> Hi,
>>
>> I agree with the patch, I just find the text above confusing.
>>
>> Reading this, I'd expect to see vskip to be resumed before s_rfe in
>> the "non-save"
>> path. However, that patch touches the save path (disable vskip as
>> soon as MODE
>> has been saved) so the rest of the save procedure can work as
>> expected. VSKIP
>> will eventually be restored in the restore path when restoring MODE
>> (no change
>> needed), which is done after any instruction that could have been
>> impacted by
>> VSKIP has already been executed.
>
> How about:
>
> VSKIP state is already saved/restored through the MODE register.
Perfect, that works for me.
Thanks,
Lancelot.
next prev parent reply other threads:[~2025-01-10 19:45 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-01-09 17:56 [PATCH 1/2] drm/amdkfd: Sync trap handler binary with source Jay Cornwall
2025-01-09 17:56 ` [PATCH 2/2] drm/amdkfd: Clear MODE.VSKIP in gfx9 trap handler Jay Cornwall
2025-01-10 18:14 ` Six, Lancelot
2025-01-10 19:37 ` Jay Cornwall
2025-01-10 19:45 ` Lancelot SIX [this message]
2025-01-10 15:03 ` [PATCH 1/2] drm/amdkfd: Sync trap handler binary with source Lancelot SIX
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=fed57f7d-a0a5-4946-9f87-1dffde14eb21@amd.com \
--to=lancelot.six@amd.com \
--cc=amd-gfx@lists.freedesktop.org \
--cc=jay.cornwall@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