[mono][2/2] Add SIMD Support for s390x by saitama951 · Pull Request #116779 · dotnet/runtime

saitama951 · 2025-06-18T07:00:01Z

This is a follow-up patch to #116669 to add vector support to s390x

saitama951 · 2025-06-18T07:01:29Z

@uweigand @nealef @iii-i , Can you please review the architecture specific code?

saitama951 · 2025-06-18T07:02:41Z

cc: @steveisok @giritrivedi @akoeplinger

This is a followup patch to dotnet#116669 to add vector support to s390x

dotnet-policy-service · 2025-06-18T12:27:07Z

Tagging subscribers to this area: @steveisok, @vitek-karas
See info in area-owners.md if you want to be subscribed.

dotnet-policy-service · 2025-06-18T12:27:07Z

Tagging subscribers to this area: @steveisok, @vitek-karas
See info in area-owners.md if you want to be subscribed.

* remove SIY_1 (duplicacy) * remove locgrnle and locghinle use locghiho instead * reformat whole patch * remove vflc use vfpso instead * move common op's to a common ifdef in mini-ops * remove NEW_INS , it's used no-where * rewrite the whole logic for Vector conditional ops for floats * update ANDN with vnc instruction * add couple of comments * remove some pseudo op in simd-intrinsics * add aligned loads and stores

saitama951 · 2025-07-03T14:28:56Z

@uweigand Thank you for the in-depth review for the patch, much appreciated. I have addressed most of your review comments

risc-vv · 2025-07-04T08:11:11Z

@dotnet/samsung Could you please take a look? These changes may be related to riscv64.

uweigand

Thanks for the updates! A few more comments inline ...

risc-vv · 2025-07-09T11:37:39Z

RISC-V Release-CLR-QEMU: 9082 / 9112 (99.67%)

=======================
passed: 9082
failed: 2
skipped: 597
killed: 28
------------------------
TOTAL tests: 9709
VIRTUAL time: 37h 28min 3s 692ms
REAL time: 38min 18s 629ms
=======================

report.xml, report.md, failures.xml, testclr_details.tar.zst

Build information and commands

GIT: bc1d70105ad01d003319b00eb44214b5bcd8c481
CI: d6c9c1ab3a7411819463edc05ded301e89ba586a
REPO: dotnet/runtime
BRANCH: main
CONFIG: Release
LIB_CONFIG: Release

* reformat patch - rename opcode from OP_ to OP_S390 * rewrite OP_XCOMPARE and OP_XEXTRACT into one opcode OP_S390_OPXCOMPARE_XEXTRACT * introduce compare neumonics without setting the condition code * rewrite OP_VEC_ABS to handle float and integer separately

uweigand

Thanks! Hopefully final set of comments inline.

* remove hard-coded s390_vr16 to allocate a temp_reg instead * omit move incase ins->sreg and ins->dreg

uweigand

I don't think I can formally approve patches in this repo, but from a s390x architecture perspective this now LGTM. Thanks!

akoeplinger

This only touches s390x code.

akoeplinger · 2025-07-21T15:59:07Z

/ba-g failures on other architectures are not related

jkotas · 2025-07-22T00:14:27Z

There are build breaks in CrossAOT_Mono crossaot CI legs on other PRs that seems to have been introduced by this change.

From
https://dev.azure.com/dnceng-public/cbb18261-c48f-4abb-8651-8cdcb5474649/_apis/build/builds/1101302/logs/472 :

2025-07-21T19:54:52.0156626Z [ 98%] Building CXX object mono/mini/CMakeFiles/monosgen-objects.dir/mini-llvm-cpp.cpp.o
2025-07-21T19:54:52.3475830Z /__w/1/s/src/mono/mono/mini/mini-llvm.c:10245:8: error: use of undeclared identifier 'OP_VECTOR_ANDN'
2025-07-21T19:54:52.3478786Z 10245 | case OP_VECTOR_ANDN: {
2025-07-21T19:54:52.3481551Z | ^
2025-07-21T19:54:52.3538995Z /__w/1/s/src/mono/mono/mini/mini-llvm.c:10257:8: error: use of undeclared identifier 'OP_VECTOR_IABS'
2025-07-21T19:54:52.3539699Z 10257 | case OP_VECTOR_IABS: {
2025-07-21T19:54:52.3545686Z |

saitama951 · 2025-07-22T00:41:49Z

@jkotas looks like TARGET_X86 is missing from my change in here

https://github.com/saitama951/runtime/blob/main/src/mono/mono/mini/mini-ops.h#L2008-L2015

I'll raise a fix shortly

saitama951 requested review from steveisok and vitek-karas as code owners June 18, 2025 07:00

dotnet-policy-serviceBot added the community-contribution Indicates that the PR has been added by a community member label Jun 18, 2025

github-actionsBot added the needs-area-label An area label is needed to ensure this gets routed to the appropriate area owners label Jun 18, 2025

saitama951 commented Jun 18, 2025
View reviewed changes

Comment threadsrc/mono/mono/mini/simd-intrinsics.c

[mono][2/2] Add SIMD Support for s390x
8dcc633
This is a followup patch to dotnet#116669 to add vector support to s390x

saitama951 force-pushed the simd-patch-2 branch from 42440c6 to 8dcc633Compare June 18, 2025 07:14

This was referenced Jun 18, 2025
browser-wasm Windows build error #116746
Closed
The Operation will be canceled. The next steps may not contain expected logs. dotnet/dnceng#3008
Open
iOS.Device/Simulator.LibraryMode.Test: failed to determine exit code - RETURN_CODE_NOT_SET #116558
Open

jkotas added area-Codegen-JIT-mono labels Jun 18, 2025

jkotas added arch-s390x Related to s390x architecture (unsupported) and removed needs-area-label An area label is needed to ensure this gets routed to the appropriate area owners labels Jun 18, 2025

nealef reviewed Jun 19, 2025
View reviewed changes

Comment threadsrc/mono/mono/mini/mini-s390x.c Outdated

build-analysisBot mentioned this pull request Jun 23, 2025
[mono] AOT compiling UnsafeAccessorsTests.dll exited with code 134. #116937
Closed

uweigand reviewed Jun 25, 2025
View reviewed changes

This was referenced Jul 3, 2025
Occasional failure in "browser-wasm windows Release LibraryTests: Build Product" #116671
Closed
wasm build failure in CI #117017
Closed

uweigand reviewed Jul 8, 2025
View reviewed changes

saitama951 force-pushed the simd-patch-2 branch from 289e0c8 to 781a25cCompare July 18, 2025 13:56

Adress Review Comments - 2
ca450d2
* reformat patch - rename opcode from OP_ to OP_S390 * rewrite OP_XCOMPARE and OP_XEXTRACT into one opcode OP_S390_OPXCOMPARE_XEXTRACT * introduce compare neumonics without setting the condition code * rewrite OP_VEC_ABS to handle float and integer separately

saitama951 force-pushed the simd-patch-2 branch from 781a25c to ca450d2Compare July 18, 2025 13:59

saitama951 requested a review from uweigand July 18, 2025 14:05

saitama951 force-pushed the simd-patch-2 branch 2 times, most recently from aa9194f to ca08838Compare July 18, 2025 14:46

uweigand reviewed Jul 18, 2025
View reviewed changes

build-analysisBot mentioned this pull request Jul 18, 2025
[WASM] Hundreds of tests failing without log #117772
Closed

Address Review Comments - 3
b802d1b
* remove hard-coded s390_vr16 to allocate a temp_reg instead * omit move incase ins->sreg and ins->dreg

saitama951 force-pushed the simd-patch-2 branch from ca08838 to b802d1bCompare July 19, 2025 07:32

Merge branch 'main' into simd-patch-2
3affb7a

saitama951 requested a review from uweigand July 19, 2025 08:05

build-analysisBot mentioned this pull request Jul 19, 2025
TimeZoneInfoTests.NoBackwardTimeZones tests are failing on Android #117731
Open

uweigand reviewed Jul 21, 2025
View reviewed changes

akoeplinger approved these changes Jul 21, 2025
View reviewed changes

akoeplinger merged commit c6b844e into dotnet:mainJul 21, 2025
72 of 85 checks passed

saitama951 mentioned this pull request Jul 22, 2025
fix undefined ops for AOT legs #117915
Merged

saitama951 mentioned this pull request Aug 5, 2025
[S390X] Add simd fallback support for unsupported s390x architectures #118376
Merged

github-actionsBot locked and limited conversation to collaborators Aug 21, 2025

Uh oh!

Conversation

saitama951 commented Jun 18, 2025

Uh oh!

saitama951 commented Jun 18, 2025

Uh oh!

saitama951 commented Jun 18, 2025

Uh oh!

Uh oh!

dotnet-policy-serviceBot commented Jun 18, 2025

Uh oh!

dotnet-policy-serviceBot commented Jun 18, 2025

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

saitama951 commented Jul 3, 2025

Uh oh!

risc-vv commented Jul 4, 2025

Uh oh!

uweigand left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

risc-vv commented Jul 9, 2025• edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

uweigand left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

uweigand left a comment

Choose a reason for hiding this comment

Uh oh!

akoeplinger left a comment

Choose a reason for hiding this comment

Uh oh!

akoeplinger commented Jul 21, 2025

Uh oh!

Uh oh!

jkotas commented Jul 22, 2025

Uh oh!

saitama951 commented Jul 22, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

risc-vv commented Jul 9, 2025•
edited
Loading