Skip to content

[AArch64] Update feature dep. for Armv9.6 extensions #113466

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Merged
merged 4 commits into from
Oct 24, 2024
Merged

[AArch64] Update feature dep. for Armv9.6 extensions #113466

merged 4 commits into from
Oct 24, 2024

Conversation

Lukacma
Copy link
Contributor

@Lukacma Lukacma commented Oct 23, 2024

No description provided.

@llvmbot
Copy link
Member

llvmbot commented Oct 23, 2024

@llvm/pr-subscribers-backend-aarch64

Author: None (Lukacma)

Changes

Full diff: https://github.com/llvm/llvm-project/pull/113466.diff

2 Files Affected:

  • (modified) llvm/lib/Target/AArch64/AArch64Features.td (+7-6)
  • (modified) llvm/unittests/TargetParser/TargetParserTest.cpp (+39-3)
diff --git a/llvm/lib/Target/AArch64/AArch64Features.td b/llvm/lib/Target/AArch64/AArch64Features.td
index 476d5a4e093efa..afa96821984b33 100644
--- a/llvm/lib/Target/AArch64/AArch64Features.td
+++ b/llvm/lib/Target/AArch64/AArch64Features.td
@@ -526,23 +526,23 @@ def FeatureCMPBR : ExtensionWithMArch<"cmpbr", "CMPBR", "FEAT_CMPBR",
   "Enable Armv9.6-A base compare and branch instructions">;
 
 def FeatureF8F32MM: ExtensionWithMArch<"f8f32mm", "F8F32MM", "FEAT_F8F32MM",
-  "Enable Armv9.6-A FP8 to Single-Precision Matrix Multiplication">;
+  "Enable Armv9.6-A FP8 to Single-Precision Matrix Multiplication", [FeatureNEON, FeatureFP8]>;
 
 def FeatureF8F16MM: ExtensionWithMArch<"f8f16mm", "F8F16MM", "FEAT_F8F16MM",
-  "Enable Armv9.6-A FP8 to Half-Precision Matrix Multiplication">;
+  "Enable Armv9.6-A FP8 to Half-Precision Matrix Multiplication", [FeatureNEON, FeatureFP8]>;
 
 def FeatureFPRCVT: ExtensionWithMArch<"fprcvt", "FPRCVT", "FEAT_FPRCVT",
   "Enable Armv9.6-A base convert instructions for SIMD&FP scalar register operands of"
-  " different input and output sizes">;
+  " different input and output sizes", [FeatureFPARMv8]>;
 
 def FeatureLSFE : ExtensionWithMArch<"lsfe", "LSFE", "FEAT_LSFE",
-  "Enable Armv9.6-A base Atomic floating-point in-memory instructions">;
+  "Enable Armv9.6-A base Atomic floating-point in-memory instructions", [FeatureFPARMv8]>;
 
 def FeatureSME2p2: ExtensionWithMArch<"sme2p2", "SME2p2", "FEAT_SME2p2",
   "Enable Armv9.6-A Scalable Matrix Extension 2.2 instructions", [FeatureSME2p1]>;
 
 def FeatureSSVE_AES : ExtensionWithMArch<"ssve-aes", "SSVE_AES", "FEAT_SSVE_AES",
-  "Enable Armv9.6-A SVE2 AES support in streaming SVE mode">;
+  "Enable Armv9.6-A SVE2 AES support in streaming SVE mode", [FeatureSME2, FeatureSVE2AES]>;
 
 def FeatureSVE2p2 : ExtensionWithMArch<"sve2p2", "SVE2p2", "FEAT_SVE2p2",
   "Enable Armv9.6-A Scalable Vector Extension 2.2 instructions", [FeatureSVE2p1]>;
@@ -554,7 +554,8 @@ def FeatureSVEBFSCALE: ExtensionWithMArch<"sve-bfscale", "SVE_BFSCALE", "FEAT_SV
   "Enable Armv9.6-A SVE BFloat16 scaling instructions">;
 
 def FeatureSVE_F16F32MM: ExtensionWithMArch<"sve-f16f32mm", "SVE_F16F32MM", "FEAT_SVE_F16F32MM",
-  "Enable Armv9.6-A FP16 to FP32 Matrix Multiply instructions">;
+  "Enable Armv9.6-A FP16 to FP32 Matrix Multiply instructions", [FeatureSVE]>;
+
 def FeatureLSUI: ExtensionWithMArch<"lsui", "LSUI", "FEAT_LSUI",
   "Enable Armv9.6-A unprivileged load/store instructions">;
 
diff --git a/llvm/unittests/TargetParser/TargetParserTest.cpp b/llvm/unittests/TargetParser/TargetParserTest.cpp
index 572006166d0d8b..07e3742a7e9f99 100644
--- a/llvm/unittests/TargetParser/TargetParserTest.cpp
+++ b/llvm/unittests/TargetParser/TargetParserTest.cpp
@@ -1782,7 +1782,15 @@ AArch64ExtensionDependenciesBaseArchTestParams
         {AArch64::ARMV8A, {"nofp", "jscvt"}, {"fp-armv8", "jsconv"}, {}},
         {AArch64::ARMV8A, {"jscvt", "nofp"}, {}, {"fp-armv8", "jsconv"}},
 
-        // simd -> {aes, sha2, sha3, sm4}
+        // fp -> lsfe
+        {AArch64::ARMV9_6A, {"nofp", "lsfe"}, {"fp-armv8", "lsfe"}, {}},
+        {AArch64::ARMV9_6A, {"lsfe", "nofp"}, {}, {"fp-armv8", "lsfe"}},
+
+        // fp -> fprcvt
+        {AArch64::ARMV9_6A, {"nofp", "fprcvt"}, {"fp-armv8", "fprcvt"}, {}},
+        {AArch64::ARMV9_6A, {"fprcvt", "nofp"}, {}, {"fp-armv8", "fprcvt"}},
+
+        // simd -> {aes, sha2, sha3, sm4, f8f16mm, f8f32mm}
         {AArch64::ARMV8A, {"nosimd", "aes"}, {"neon", "aes"}, {}},
         {AArch64::ARMV8A, {"aes", "nosimd"}, {}, {"neon", "aes"}},
         {AArch64::ARMV8A, {"nosimd", "sha2"}, {"neon", "sha2"}, {}},
@@ -1791,6 +1799,10 @@ AArch64ExtensionDependenciesBaseArchTestParams
         {AArch64::ARMV8A, {"sha3", "nosimd"}, {}, {"neon", "sha3"}},
         {AArch64::ARMV8A, {"nosimd", "sm4"}, {"neon", "sm4"}, {}},
         {AArch64::ARMV8A, {"sm4", "nosimd"}, {}, {"neon", "sm4"}},
+        {AArch64::ARMV9_6A, {"nosimd", "f8f16mm"}, {"neon", "f8f16mm"}, {}},
+        {AArch64::ARMV9_6A, {"f8f16mm", "nosimd"}, {}, {"neon", "f8f16mm"}},
+        {AArch64::ARMV9_6A, {"nosimd", "f8f32mm"}, {"neon", "f8f32mm"}, {}},
+        {AArch64::ARMV9_6A, {"f8f32mm", "nosimd"}, {}, {"neon", "f8f32mm"}},
 
         // simd -> {rdm, dotprod, fcma}
         {AArch64::ARMV8A, {"nosimd", "rdm"}, {"neon", "rdm"}, {}},
@@ -1810,13 +1822,21 @@ AArch64ExtensionDependenciesBaseArchTestParams
         {AArch64::ARMV8A, {"nobf16", "sme"}, {"bf16", "sme"}, {}},
         {AArch64::ARMV8A, {"sme", "nobf16"}, {}, {"bf16", "sme"}},
 
-        // sve -> {sve2, f32mm, f64mm}
+        // sve -> {sve2, f32mm, f64mm, sve-f16f32mm}
         {AArch64::ARMV8A, {"nosve", "sve2"}, {"sve", "sve2"}, {}},
         {AArch64::ARMV8A, {"sve2", "nosve"}, {}, {"sve", "sve2"}},
         {AArch64::ARMV8A, {"nosve", "f32mm"}, {"sve", "f32mm"}, {}},
         {AArch64::ARMV8A, {"f32mm", "nosve"}, {}, {"sve", "f32mm"}},
         {AArch64::ARMV8A, {"nosve", "f64mm"}, {"sve", "f64mm"}, {}},
         {AArch64::ARMV8A, {"f64mm", "nosve"}, {}, {"sve", "f64mm"}},
+        {AArch64::ARMV9_6A,
+         {"nosve", "sve-f16f32mm"},
+         {"sve", "sve-f16f32mm"},
+         {}},
+        {AArch64::ARMV9_6A,
+         {"sve-f16f32mm", "nosve"},
+         {},
+         {"sve", "sve-f16f32mm"}},
 
         // sve2 -> {sve2p1, sve2-bitperm, sve2-sha3, sve2-sm4}
         {AArch64::ARMV8A, {"nosve2", "sve2p1"}, {"sve2", "sve2p1"}, {}},
@@ -1861,7 +1881,7 @@ AArch64ExtensionDependenciesBaseArchTestParams
         {AArch64::ARMV8A, {"sme-fa64", "nosme"}, {}, {"sme", "sme-fa64"}},
 
         // sme2 -> {sme2p1, ssve-fp8fma, ssve-fp8dot2, ssve-fp8dot4, sme-f8f16,
-        // sme-f8f32, sme-b16b16}
+        // sme-f8f32, sme-b16b16i, ssve-aes}
         {AArch64::ARMV8A, {"nosme2", "sme2p1"}, {"sme2", "sme2p1"}, {}},
         {AArch64::ARMV8A, {"sme2p1", "nosme2"}, {}, {"sme2", "sme2p1"}},
         {AArch64::ARMV8A,
@@ -1894,6 +1914,8 @@ AArch64ExtensionDependenciesBaseArchTestParams
         {AArch64::ARMV8A, {"sme-f8f32", "nosme2"}, {}, {"sme2", "sme-f8f32"}},
         {AArch64::ARMV8A, {"nosme2", "sme-b16b16"}, {"sme2", "sme-b16b16"}, {}},
         {AArch64::ARMV8A, {"sme-b16b16", "nosme2"}, {}, {"sme2", "sme-b16b16"}},
+        {AArch64::ARMV9_6A, {"nosme2", "ssve-aes"}, {"sme2", "ssve-aes"}, {}},
+        {AArch64::ARMV9_6A, {"ssve-aes", "nosme2"}, {}, {"ssve-aes", "sme2"}},
 
         // sme2p1 -> {sme2p2}
         {AArch64::ARMV9_6A, {"nosme2p1", "sme2p2"}, {"sme2p2", "sme2p1"}, {}},
@@ -1904,6 +1926,10 @@ AArch64ExtensionDependenciesBaseArchTestParams
         {AArch64::ARMV8A, {"sme-f8f16", "nofp8"}, {}, {"fp8", "sme-f8f16"}},
         {AArch64::ARMV8A, {"nofp8", "sme-f8f32"}, {"fp8", "sme-f8f32"}, {}},
         {AArch64::ARMV8A, {"sme-f8f32", "nofp8"}, {}, {"fp8", "sme-f8f32"}},
+        {AArch64::ARMV9_6A, {"nofp8", "f8f16mm"}, {"fp8", "f8f16mm"}, {}},
+        {AArch64::ARMV9_6A, {"f8f16mm", "nofp8"}, {}, {"fp8", "f8f16mm"}},
+        {AArch64::ARMV9_6A, {"nofp8", "f8f32mm"}, {"fp8", "f8f32mm"}, {}},
+        {AArch64::ARMV9_6A, {"f8f32mm", "nofp8"}, {}, {"fp8", "f8f32mm"}},
 
         // lse -> lse128
         {AArch64::ARMV8A, {"nolse", "lse128"}, {"lse", "lse128"}, {}},
@@ -1926,6 +1952,16 @@ AArch64ExtensionDependenciesBaseArchTestParams
         // rcpc -> rcpc3
         {AArch64::ARMV8A, {"norcpc", "rcpc3"}, {"rcpc", "rcpc3"}, {}},
         {AArch64::ARMV8A, {"rcpc3", "norcpc"}, {}, {"rcpc", "rcpc3"}},
+
+        // sve-aes -> ssve-aes
+        {AArch64::ARMV9_6A,
+         {"nosve2-aes", "ssve-aes"},
+         {"sve2-aes", "ssve-aes"},
+         {}},
+        {AArch64::ARMV9_6A,
+         {"ssve-aes", "nosve2-aes"},
+         {},
+         {"ssve-aes", "sve2-aes"}},
 };
 
 INSTANTIATE_TEST_SUITE_P(

SpencerAbson
SpencerAbson approved these changes Oct 24, 2024
View reviewed changes
Copy link
Contributor

@SpencerAbson SpencerAbson left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM, thanks.

jthackray
jthackray approved these changes Oct 24, 2024
View reviewed changes
Copy link
Contributor

@jthackray jthackray left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks, LGTM

@Lukacma Lukacma merged commit 9575ab2 into llvm:main Oct 24, 2024
8 checks passed
@llvm-ci
Copy link
Collaborator

llvm-ci commented Oct 24, 2024

LLVM Buildbot has detected a new failure on builder openmp-offload-libc-amdgpu-runtime running on omp-vega20-1 while building llvm at step 11 "Add check check-libc-amdgcn-amd-amdhsa".

Full details are available at: https://lab.llvm.org/buildbot/#/builders/73/builds/7501

Here is the relevant piece of the build log for the reference 
Step 11 (Add check check-libc-amdgcn-amd-amdhsa) failure: 1200 seconds without output running [b'ninja', b'-j 32', b'check-libc-amdgcn-amd-amdhsa'], attempting to kill
...
[2440/2681] Linking CXX executable libc/test/src/string/libc.test.src.string.memcmp_test.__hermetic__.__build__
[2441/2681] Linking CXX executable libc/test/src/locale/libc.test.src.locale.localeconv_test.__hermetic__.__build__
[2442/2681] Linking CXX executable libc/test/src/inttypes/libc.test.src.inttypes.strtoimax_test.__hermetic__.__build__
[2443/2681] Linking CXX executable libc/test/src/string/libc.test.src.string.strxfrm_test.__hermetic__.__build__
[2444/2681] Linking CXX executable libc/test/src/time/libc.test.src.time.clock_test.__hermetic__.__build__
[2445/2681] Linking CXX executable libc/test/src/inttypes/libc.test.src.inttypes.strtoumax_test.__hermetic__.__build__
[2446/2681] Linking CXX executable libc/test/src/string/libc.test.src.string.memmove_test.__hermetic__.__build__
[2447/2681] Linking CXX executable libc/test/src/time/libc.test.src.time.clock_gettime_test.__hermetic__.__build__
[2448/2681] Linking CXX executable libc/test/src/stdio/libc.test.src.stdio.asprintf_test.__hermetic__.__build__
[2449/2681] Linking CXX executable libc/test/src/stdio/libc.test.src.stdio.vasprintf_test.__hermetic__.__build__
command timed out: 1200 seconds without output running [b'ninja', b'-j 32', b'check-libc-amdgcn-amd-amdhsa'], attempting to kill
process killed by signal 9
program finished with exit code -1
elapsedTime=1469.563407

@frobtech frobtech mentioned this pull request Oct 25, 2024
Closed
NoumanAmir657 pushed a commit to NoumanAmir657/llvm-project that referenced this pull request Nov 4, 2024
@Lukacma @jthackray
@SpencerAbson @NoumanAmir657
[AArch64] Update feature dep. for Armv9.6 extensions (llvm#113466)
424dbcd 
Co-authored-by: Jonathan Thackray <[email protected]>
Co-authored-by: SpencerAbson <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@jthackray jthackray jthackray approved these changes

@SpencerAbson SpencerAbson SpencerAbson approved these changes

@CarolineConcatto CarolineConcatto Awaiting requested review from CarolineConcatto

Assignees
No one assigned
Labels
backend:AArch64
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
5 participants
@Lukacma @llvmbot @llvm-ci @jthackray @SpencerAbson