[valgrind] [Bug 295974] vex-x86 omits some SSE4.1 opcodes such as PEXTRD

2026-05-20 Thread Mark Wielaard
https://bugs.kde.org/show_bug.cgi?id=295974

Mark Wielaard  changed:

   What|Removed |Added

 Resolution|--- |FIXED
 Status|ASSIGNED|RESOLVED

--- Comment #15 from Mark Wielaard  ---
commit 3556cd926eca4f79b6e822abea57a9bdf088bfea
Author: Alexandra Hájková 
Date:   Thu Apr 30 10:44:27 2026 -0400

Add SSE4.1 PEXTRD instruction for x86 32 bit

Support pextrd instruction in guest_x86_toIR.c and
host_x86_isel.c. Add test function to sse4-common.h
and update none/tests/x86/sse4-x86.c to test the
instruction.

BZ: https://bugs.kde.org/show_bug.cgi?id=295974

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 295974] vex-x86 omits some SSE4.1 opcodes such as PEXTRD

2026-05-13 Thread Mark Wielaard
https://bugs.kde.org/show_bug.cgi?id=295974

--- Comment #14 from Mark Wielaard  ---
(In reply to Alexandra Hajkova from comment #13)
> Created attachment 192084 [details]
> patch

This one looks good to me. Please apply.

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 295974] vex-x86 omits some SSE4.1 opcodes such as PEXTRD

2026-05-06 Thread Alexandra Hajkova
https://bugs.kde.org/show_bug.cgi?id=295974

Alexandra Hajkova  changed:

   What|Removed |Added

 Attachment #192050|0   |1
is obsolete||

--- Comment #13 from Alexandra Hajkova  ---
Created attachment 192084
  --> https://bugs.kde.org/attachment.cgi?id=192084&action=edit
patch

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 295974] vex-x86 omits some SSE4.1 opcodes such as PEXTRD

2026-05-06 Thread Mark Wielaard
https://bugs.kde.org/show_bug.cgi?id=295974

Mark Wielaard  changed:

   What|Removed |Added

 Ever confirmed|0   |1
 Status|REPORTED|ASSIGNED

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 295974] vex-x86 omits some SSE4.1 opcodes such as PEXTRD

2026-05-06 Thread Mark Wielaard
https://bugs.kde.org/show_bug.cgi?id=295974

--- Comment #12 from Mark Wielaard  ---
(In reply to mcermak from comment #10)
> Hello Sasshka,
> 
> the attached patch makes sense to me.  Comparing patched and unpatched
> sources, I see in both cases:  
> 
> $ tests/vg_regtest none/tests/amd64/sse4-64.vgtest
> none/tests/x86/sse4-x86.vgtest
> sse4-64: valgrind   -q ./sse4-64 
> sse4-x86:valgrind   -q ./sse4-x86 
> 
> == 2 tests, 0 stderr failures, 0 stdout failures, 0 stderrB failures, 0
> stdoutB failures, 0 post failures ==
> 
> Which is expected.  Looking closer, the patched binary shows the expected
> logs (while the unpatched doesn't):
> 
> $ ../../../vg-in-place ./sse4-x86 |& fgrep pextrd
> r pextrd $0 60b160857d45c48447b8d8c0eeef1e50 
> eeef1e50
> m pextrd $0 60b160857d45c48447b8d8c0eeef1e50 
> eeef1e50
> r pextrd $1 60b160857d45c48447b8d8c0eeef1e50 
> 47b8d8c0
> m pextrd $1 60b160857d45c48447b8d8c0eeef1e50 
> 47b8d8c0
> r pextrd $2 60b160857d45c48447b8d8c0eeef1e50 
> 7d45c484
> m pextrd $2 60b160857d45c48447b8d8c0eeef1e50 
> 7d45c484
> r pextrd $3 60b160857d45c48447b8d8c0eeef1e50 
> 60b16085
> m pextrd $3 60b160857d45c48447b8d8c0eeef1e50 
> 60b16085
> $ 
> 
> Nice.  No regression spotted.

So there is a lot of change in the .exp file because the new test is added in
the middle:

// -- SSE 4.1 --
test_PINSRD();
+   test_PEXTRD();
test_PMAXSB();
test_PMAXSD();
test_PMAXUD();

Please add it to the end, so you only have to add a few new lines to the .exp
file.

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 295974] vex-x86 omits some SSE4.1 opcodes such as PEXTRD

2026-05-06 Thread Paul Floyd
https://bugs.kde.org/show_bug.cgi?id=295974

Paul Floyd  changed:

   What|Removed |Added

 CC||[email protected]

--- Comment #11 from Paul Floyd  ---

>  519574  valgrind 3.27 "--fair-sched=yes" does not work
>  519613  Valgrind incorrectly unpacks the result of sys_port (port_getn)
>  on error, leading to a ~60s wallclock time delay on every call
> -295974  Add SSE4.1 PEXTRD instructions for x86 32 bit
> +295974  Add SSE4.1 PEXTRD instruction for x86 32 bit

And in numerical order.

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 295974] vex-x86 omits some SSE4.1 opcodes such as PEXTRD

2026-05-06 Thread bugzilla_noreply
https://bugs.kde.org/show_bug.cgi?id=295974

[email protected] changed:

   What|Removed |Added

 CC||[email protected]

--- Comment #10 from [email protected] ---
Hello Sasshka,

the attached patch makes sense to me.  Comparing patched and unpatched sources,
I see in both cases:  

$ tests/vg_regtest none/tests/amd64/sse4-64.vgtest
none/tests/x86/sse4-x86.vgtest
sse4-64: valgrind   -q ./sse4-64 
sse4-x86:valgrind   -q ./sse4-x86 

== 2 tests, 0 stderr failures, 0 stdout failures, 0 stderrB failures, 0 stdoutB
failures, 0 post failures ==

Which is expected.  Looking closer, the patched binary shows the expected logs
(while the unpatched doesn't):

$ ../../../vg-in-place ./sse4-x86 |& fgrep pextrd
r pextrd $0 60b160857d45c48447b8d8c0eeef1e50 
eeef1e50
m pextrd $0 60b160857d45c48447b8d8c0eeef1e50 
eeef1e50
r pextrd $1 60b160857d45c48447b8d8c0eeef1e50 
47b8d8c0
m pextrd $1 60b160857d45c48447b8d8c0eeef1e50 
47b8d8c0
r pextrd $2 60b160857d45c48447b8d8c0eeef1e50 
7d45c484
m pextrd $2 60b160857d45c48447b8d8c0eeef1e50 
7d45c484
r pextrd $3 60b160857d45c48447b8d8c0eeef1e50 
60b16085
m pextrd $3 60b160857d45c48447b8d8c0eeef1e50 
60b16085
$ 

Nice.  No regression spotted.  The only nitpick I can find is a typo in your
NEWS entry:

$ git diff
diff --git a/NEWS b/NEWS
index 0f28fcb87..2e517b62c 100644
--- a/NEWS
+++ b/NEWS
@@ -28,7 +28,7 @@ are not entered into bugzilla tend to get forgotten about or
ignored.
 519574  valgrind 3.27 "--fair-sched=yes" does not work
 519613  Valgrind incorrectly unpacks the result of sys_port (port_getn)
 on error, leading to a ~60s wallclock time delay on every call
-295974  Add SSE4.1 PEXTRD instructions for x86 32 bit
+295974  Add SSE4.1 PEXTRD instruction for x86 32 bit

 To see details of a given bug, visit
   https://bugs.kde.org/show_bug.cgi?id=XX
$

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 295974] vex-x86 omits some SSE4.1 opcodes such as PEXTRD

2026-05-05 Thread Alexandra Hajkova
https://bugs.kde.org/show_bug.cgi?id=295974

--- Comment #9 from Alexandra Hajkova  ---
Created attachment 192050
  --> https://bugs.kde.org/attachment.cgi?id=192050&action=edit
patch

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 295974] vex-x86 omits some SSE4.1 opcodes such as PEXTRD

2026-04-01 Thread Mark Wielaard
https://bugs.kde.org/show_bug.cgi?id=295974

Mark Wielaard  changed:

   What|Removed |Added

 CC|[email protected]|[email protected]

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 295974] vex-x86 omits some SSE4.1 opcodes such as PEXTRD

2026-03-27 Thread bugzilla_noreply
https://bugs.kde.org/show_bug.cgi?id=295974

--- Comment #8 from [email protected] ---
unsubscribe

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 295974] vex-x86 omits some SSE4.1 opcodes such as PEXTRD

2026-03-27 Thread Mark Wielaard
https://bugs.kde.org/show_bug.cgi?id=295974

Mark Wielaard  changed:

   What|Removed |Added

 Blocks||518222


Referenced Bugs:

https://bugs.kde.org/show_bug.cgi?id=518222
[Bug 518222] SSE4.1 support for x86 (32bit)
-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 295974] vex-x86 omits some SSE4.1 opcodes such as PEXTRD

2025-11-12 Thread Alexandra Hajkova
https://bugs.kde.org/show_bug.cgi?id=295974

Alexandra Hajkova  changed:

   What|Removed |Added

   Assignee|[email protected] |[email protected]
 CC||[email protected]

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 295974] vex-x86 omits some SSE4.1 opcodes such as PEXTRD

2017-05-07 Thread Jeffrey Walton
https://bugs.kde.org/show_bug.cgi?id=295974

--- Comment #7 from Jeffrey Walton  ---
(In reply to Ivo Raisr from comment #6)
> Please do not revive old bugs. Leave them R.I.P. :-)
> 
> If this is a new problem for you, please report a new bug as per
> http://valgrind.org/support/bug_reports.html
> including all the required details.

Thanks Iva.

The bug is still open. How do we know when to add to an existing bug report,
and open a new one?

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 295974] vex-x86 omits some SSE4.1 opcodes such as PEXTRD

2017-05-07 Thread Ivo Raisr
https://bugs.kde.org/show_bug.cgi?id=295974

Ivo Raisr  changed:

   What|Removed |Added

 CC||[email protected]

--- Comment #6 from Ivo Raisr  ---
Please do not revive old bugs. Leave them R.I.P. :-)

If this is a new problem for you, please report a new bug as per
http://valgrind.org/support/bug_reports.html
including all the required details.

-- 
You are receiving this mail because:
You are watching all bug changes.

[valgrind] [Bug 295974] vex-x86 omits some SSE4.1 opcodes such as PEXTRD

2017-05-07 Thread Jeffrey Walton
https://bugs.kde.org/show_bug.cgi?id=295974

Jeffrey Walton  changed:

   What|Removed |Added

 CC||[email protected]

--- Comment #5 from Jeffrey Walton  ---
We've got a user reporting a similar issue for unhandled instructions "0x66 0xF
0x3A 0x22". Only the last byte differs, and I'm guessing its the same or nearly
the same issue. I don't know the compiler. Also see
https://groups.google.com/d/msg/cryptopp-users/rwRSMrxJVcw/68BESEbCAQAJ.

Others have reported the same unhandled instructions "0x66 0xF 0x3A 0x22". It
looks like the compiler is Clang. Also see
https://github.com/miloyip/rapidjson/issues/266.

-- 
You are receiving this mail because:
You are watching all bug changes.