Bug 486398 - SSE4a: unhandled instruction bytes: 0xF2 0xF 0x78 0xC0 0x8 0x18 0xC5 0xF9 0x7E 0x84 (insertq $24, $8, %xmm0, %xmm0)
Summary: SSE4a: unhandled instruction bytes: 0xF2 0xF 0x78 0xC0 0x8 0x18 0xC5 0xF9 0x7...
Status: ASSIGNED
Alias: None
Product: valgrind
Classification: Developer tools
Component: vex (show other bugs)
Version: 3.23.0
Platform: Arch Linux Linux
: NOR normal
Target Milestone: ---
Assignee: Paul Floyd
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2024-05-01 11:04 UTC by Pol Marcet Sarda
Modified: 2025-03-23 14:51 UTC (History)
1 user (show)

See Also:
Latest Commit:
Version Fixed In:
Sentry Crash Report:


Attachments
Patch to add support for the hole SSE4a instruction set (compressed due to diff being quite large at ~4MB) (221.10 KB, application/gzip)
2025-01-31 23:02 UTC, Pol Marcet Sarda
Details
attachment-1788766-0.html (1.15 KB, text/html)
2025-02-23 22:47 UTC, Pol Marcet Sarda
Details
attachment-985123-0.html (1.75 KB, text/html)
2025-03-22 16:47 UTC, Pol Marcet Sarda
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Pol Marcet Sarda 2024-05-01 11:04:52 UTC
Currently experiencing this on a release build of a program

The instruction seems to be an SSE4a instruction, a subset only available on AMD CPUs

If this instruction were supported, it would be greatly appreciated!

This instruction is issued by clang in 19.0.0
Comment 1 Pol Marcet Sarda 2024-05-01 14:33:33 UTC
link to godot example:
https://godbolt.org/z/WdorK6bWx
Comment 2 Pol Marcet Sarda 2025-01-25 01:43:21 UTC
I'd be interested in fixing this bug, as I am facing it again. I will take it and I'll take a look at contributing documentation
Comment 3 Pol Marcet Sarda 2025-01-31 23:02:36 UTC
Created attachment 177871 [details]
Patch to add support for the hole SSE4a instruction set (compressed due to diff being quite large at ~4MB)
Comment 4 Pol Marcet Sarda 2025-01-31 23:02:56 UTC
Please see attached patch with the SSE4a support
Comment 5 Paul Floyd 2025-02-23 17:33:04 UTC
Thanks for the patch. I'm busy at the moment setting up a new PC. It might take some time. I'll take a look when that's all done.
Comment 6 Pol Marcet Sarda 2025-02-23 22:47:54 UTC
Created attachment 178782 [details]
attachment-1788766-0.html

Thanks for the response!

I totally understand, best of luck with the new PC

Regards,

Pol Marcet Sardà

El dom, 23 feb 2025, 18:33, Paul Floyd <bugzilla_noreply@kde.org> escribió:

> https://bugs.kde.org/show_bug.cgi?id=486398
>
> --- Comment #5 from Paul Floyd <pjfloyd@wanadoo.fr> ---
> Thanks for the patch. I'm busy at the moment setting up a new PC. It might
> take
> some time. I'll take a look when that's all done.
>
> --
> You are receiving this mail because:
> You reported the bug.
Comment 7 Pol Marcet Sarda 2025-03-22 16:47:25 UTC
Created attachment 179646 [details]
attachment-985123-0.html

Hello,

is there any update on this, perchance? No pressure, I understand that life
gets on its way, but having this would make my life a lot easier.

Thanks,
Pol Marcet Sardà

Missatge de Pol Marcet Sardà <polmarcetsarda@gmail.com> del dia dg., 23 de
febr. 2025 a les 23:47:

> Thanks for the response!
>
> I totally understand, best of luck with the new PC
>
> Regards,
>
> Pol Marcet Sardà
>
> El dom, 23 feb 2025, 18:33, Paul Floyd <bugzilla_noreply@kde.org>
> escribió:
>
>> https://bugs.kde.org/show_bug.cgi?id=486398
>>
>> --- Comment #5 from Paul Floyd <pjfloyd@wanadoo.fr> ---
>> Thanks for the patch. I'm busy at the moment setting up a new PC. It
>> might take
>> some time. I'll take a look when that's all done.
>>
>> --
>> You are receiving this mail because:
>> You reported the bug.
>
>
Comment 8 Paul Floyd 2025-03-23 14:51:55 UTC
The patch builds OK with no warnings on FreeBSD with clang 18.1.6. No problems with the new tests either.

OK with sourceware build macines.