[clang] [llvm] [TargetParser][AArch64] Add alias for FEAT_RDM. (PR #80540)
https://github.com/labrinea closed https://github.com/llvm/llvm-project/pull/80540 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [TargetParser][AArch64] Add alias for FEAT_RDM. (PR #80540)
https://github.com/labrinea updated https://github.com/llvm/llvm-project/pull/80540 >From 5354d6c6736f84881466b6b5b99479137eddbf29 Mon Sep 17 00:00:00 2001 From: Alexandros Lamprineas Date: Tue, 30 Jan 2024 11:17:55 + Subject: [PATCH] [TargetParser][AArch64] Add alias for FEAT_RDM. This patch allows using the name "rdma" as an alias for "rdm". The name makes its way to target attributes as well as the command line via the -march and -mcpu options. The motivation was originally to support this in Function Multi Versioning but it also makes sense to align with GCC on the command line. --- clang/docs/ReleaseNotes.rst | 5 + clang/test/CodeGen/attr-target-version.c | 12 ++-- clang/test/Driver/aarch64-rdm.c | 3 +++ clang/test/Sema/attr-target-clones-aarch64.c | 2 +- clang/test/SemaCXX/attr-target-version.cpp| 1 + .../llvm/TargetParser/AArch64TargetParser.h | 13 - llvm/lib/TargetParser/AArch64TargetParser.cpp | 15 +-- 7 files changed, 37 insertions(+), 14 deletions(-) diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst index cfe0ac6a5dca61..6f6ce7c68a7a71 100644 --- a/clang/docs/ReleaseNotes.rst +++ b/clang/docs/ReleaseNotes.rst @@ -346,6 +346,11 @@ Arm and AArch64 Support improvements for most targets. We have not changed the default behavior for ARMv6, but may revisit that decision in the future. Users can restore the old behavior with -m[no-]unaligned-access. +- An alias identifier (rdma) has been added for targeting the AArch64 + Architecture Extension which uses Rounding Doubling Multiply Accumulate + instructions (rdm). The identifier is available on the command line as + a feature modifier for -march and -mcpu as well as via target attributes + like ``target_version`` or ``target_clones``. Android Support ^^^ diff --git a/clang/test/CodeGen/attr-target-version.c b/clang/test/CodeGen/attr-target-version.c index ae97977a9144f6..56a42499d0a7ca 100644 --- a/clang/test/CodeGen/attr-target-version.c +++ b/clang/test/CodeGen/attr-target-version.c @@ -25,7 +25,7 @@ int foo() { } inline int __attribute__((target_version("sha1+pmull+f64mm"))) fmv_inline(void) { return 1; } -inline int __attribute__((target_version("fp16+fcma+sme+ fp16 "))) fmv_inline(void) { return 2; } +inline int __attribute__((target_version("fp16+fcma+rdma+sme+ fp16 "))) fmv_inline(void) { return 2; } inline int __attribute__((target_version("sha3+i8mm+f32mm"))) fmv_inline(void) { return 12; } inline int __attribute__((target_version("dit+sve-ebf16"))) fmv_inline(void) { return 8; } inline int __attribute__((target_version("dpb+rcpc2 "))) fmv_inline(void) { return 6; } @@ -261,12 +261,12 @@ int hoo(void) { // CHECK-NEXT: resolver_entry: // CHECK-NEXT:call void @__init_cpu_features_resolver() // CHECK-NEXT:[[TMP0:%.*]] = load i64, ptr @__aarch64_cpu_features, align 8 -// CHECK-NEXT:[[TMP1:%.*]] = and i64 [[TMP0]], 4398048608256 -// CHECK-NEXT:[[TMP2:%.*]] = icmp eq i64 [[TMP1]], 4398048608256 +// CHECK-NEXT:[[TMP1:%.*]] = and i64 [[TMP0]], 4398048608320 +// CHECK-NEXT:[[TMP2:%.*]] = icmp eq i64 [[TMP1]], 4398048608320 // CHECK-NEXT:[[TMP3:%.*]] = and i1 true, [[TMP2]] // CHECK-NEXT:br i1 [[TMP3]], label [[RESOLVER_RETURN:%.*]], label [[RESOLVER_ELSE:%.*]] // CHECK: resolver_return: -// CHECK-NEXT:ret ptr @fmv_inline._MfcmaMfp16Mfp16Msme +// CHECK-NEXT:ret ptr @fmv_inline._MfcmaMfp16Mfp16MrdmMsme // CHECK: resolver_else: // CHECK-NEXT:[[TMP4:%.*]] = load i64, ptr @__aarch64_cpu_features, align 8 // CHECK-NEXT:[[TMP5:%.*]] = and i64 [[TMP4]], 864726312827224064 @@ -575,7 +575,7 @@ int hoo(void) { // // // CHECK: Function Attrs: noinline nounwind optnone -// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MfcmaMfp16Mfp16Msme +// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MfcmaMfp16Mfp16MrdmMsme // CHECK-SAME: () #[[ATTR13:[0-9]+]] { // CHECK-NEXT: entry: // CHECK-NEXT:ret i32 2 @@ -829,7 +829,7 @@ int hoo(void) { // CHECK: attributes #[[ATTR10]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+bf16,+fullfp16,+ls64,+sme,+sme2" } // CHECK: attributes #[[ATTR11]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+ccpp,+fullfp16,+ls64" } // CHECK: attributes #[[ATTR12]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+aes,+f64mm,+fp-armv8,+fullfp16,+ls64,+neon,+sve" } -// CHECK: attributes #[[ATTR13]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+bf16,+complxnum,+fp-armv8,+fullfp16,+ls64,+neon,+sme" } +// CHECK: attributes #[[ATTR13]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-feature
[clang] [llvm] [TargetParser][AArch64] Add alias for FEAT_RDM. (PR #80540)
labrinea wrote: Thanks! https://github.com/llvm/llvm-project/pull/80540 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [TargetParser][AArch64] Add alias for FEAT_RDM. (PR #80540)
https://github.com/jroelofs approved this pull request. https://github.com/llvm/llvm-project/pull/80540 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [TargetParser][AArch64] Add alias for FEAT_RDM. (PR #80540)
labrinea wrote: ping https://github.com/llvm/llvm-project/pull/80540 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [TargetParser][AArch64] Add alias for FEAT_RDM. (PR #80540)
https://github.com/labrinea updated https://github.com/llvm/llvm-project/pull/80540 >From d2c973c8ebd7605b47a8c5fc928d2d85426c8a6d Mon Sep 17 00:00:00 2001 From: Alexandros Lamprineas Date: Tue, 30 Jan 2024 11:17:55 + Subject: [PATCH] [TargetParser][AArch64] Add alias for FEAT_RDM. This patch allows using the name "rdma" as an alias for "rdm". The name makes its way to target attributes as well as the command line via the -march and -mcpu options. The motivation was originally to support this in Function Multi Versioning but it also makes sense to align with GCC on the command line. --- clang/docs/ReleaseNotes.rst | 5 + clang/test/CodeGen/attr-target-version.c | 12 ++-- clang/test/Driver/aarch64-rdm.c | 3 +++ clang/test/Sema/attr-target-clones-aarch64.c | 2 +- clang/test/SemaCXX/attr-target-version.cpp| 1 + .../llvm/TargetParser/AArch64TargetParser.h | 13 - llvm/lib/TargetParser/AArch64TargetParser.cpp | 15 +-- 7 files changed, 37 insertions(+), 14 deletions(-) diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst index f44fef28b9f17f..530158dda66689 100644 --- a/clang/docs/ReleaseNotes.rst +++ b/clang/docs/ReleaseNotes.rst @@ -333,6 +333,11 @@ Arm and AArch64 Support improvements for most targets. We have not changed the default behavior for ARMv6, but may revisit that decision in the future. Users can restore the old behavior with -m[no-]unaligned-access. +- An alias identifier (rdma) has been added for targeting the AArch64 + Architecture Extension which uses Rounding Doubling Multiply Accumulate + instructions (rdm). The identifier is available on the command line as + a feature modifier for -march and -mcpu as well as via target attributes + like ``target_version`` or ``target_clones``. Android Support ^^^ diff --git a/clang/test/CodeGen/attr-target-version.c b/clang/test/CodeGen/attr-target-version.c index ae97977a9144f6..56a42499d0a7ca 100644 --- a/clang/test/CodeGen/attr-target-version.c +++ b/clang/test/CodeGen/attr-target-version.c @@ -25,7 +25,7 @@ int foo() { } inline int __attribute__((target_version("sha1+pmull+f64mm"))) fmv_inline(void) { return 1; } -inline int __attribute__((target_version("fp16+fcma+sme+ fp16 "))) fmv_inline(void) { return 2; } +inline int __attribute__((target_version("fp16+fcma+rdma+sme+ fp16 "))) fmv_inline(void) { return 2; } inline int __attribute__((target_version("sha3+i8mm+f32mm"))) fmv_inline(void) { return 12; } inline int __attribute__((target_version("dit+sve-ebf16"))) fmv_inline(void) { return 8; } inline int __attribute__((target_version("dpb+rcpc2 "))) fmv_inline(void) { return 6; } @@ -261,12 +261,12 @@ int hoo(void) { // CHECK-NEXT: resolver_entry: // CHECK-NEXT:call void @__init_cpu_features_resolver() // CHECK-NEXT:[[TMP0:%.*]] = load i64, ptr @__aarch64_cpu_features, align 8 -// CHECK-NEXT:[[TMP1:%.*]] = and i64 [[TMP0]], 4398048608256 -// CHECK-NEXT:[[TMP2:%.*]] = icmp eq i64 [[TMP1]], 4398048608256 +// CHECK-NEXT:[[TMP1:%.*]] = and i64 [[TMP0]], 4398048608320 +// CHECK-NEXT:[[TMP2:%.*]] = icmp eq i64 [[TMP1]], 4398048608320 // CHECK-NEXT:[[TMP3:%.*]] = and i1 true, [[TMP2]] // CHECK-NEXT:br i1 [[TMP3]], label [[RESOLVER_RETURN:%.*]], label [[RESOLVER_ELSE:%.*]] // CHECK: resolver_return: -// CHECK-NEXT:ret ptr @fmv_inline._MfcmaMfp16Mfp16Msme +// CHECK-NEXT:ret ptr @fmv_inline._MfcmaMfp16Mfp16MrdmMsme // CHECK: resolver_else: // CHECK-NEXT:[[TMP4:%.*]] = load i64, ptr @__aarch64_cpu_features, align 8 // CHECK-NEXT:[[TMP5:%.*]] = and i64 [[TMP4]], 864726312827224064 @@ -575,7 +575,7 @@ int hoo(void) { // // // CHECK: Function Attrs: noinline nounwind optnone -// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MfcmaMfp16Mfp16Msme +// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MfcmaMfp16Mfp16MrdmMsme // CHECK-SAME: () #[[ATTR13:[0-9]+]] { // CHECK-NEXT: entry: // CHECK-NEXT:ret i32 2 @@ -829,7 +829,7 @@ int hoo(void) { // CHECK: attributes #[[ATTR10]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+bf16,+fullfp16,+ls64,+sme,+sme2" } // CHECK: attributes #[[ATTR11]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+ccpp,+fullfp16,+ls64" } // CHECK: attributes #[[ATTR12]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+aes,+f64mm,+fp-armv8,+fullfp16,+ls64,+neon,+sve" } -// CHECK: attributes #[[ATTR13]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+bf16,+complxnum,+fp-armv8,+fullfp16,+ls64,+neon,+sme" } +// CHECK: attributes #[[ATTR13]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-feature
[clang] [llvm] [TargetParser][AArch64] Add alias for FEAT_RDM. (PR #80540)
https://github.com/labrinea updated https://github.com/llvm/llvm-project/pull/80540 >From 6f1f4e18de7ebad5e090ea268f3f053562db444c Mon Sep 17 00:00:00 2001 From: Alexandros Lamprineas Date: Tue, 30 Jan 2024 11:17:55 + Subject: [PATCH] [TargetParser][AArch64] Add alias for FEAT_RDM. This patch allows using the name "rdma" as an alias for "rdm". The name makes its way to target attributes as well as the command line via the -march and -mcpu options. The motivation was originally to support this in Function Multi Versioning but it also makes sense to align with GCC on the command line. --- clang/docs/ReleaseNotes.rst | 5 + clang/test/CodeGen/attr-target-version.c | 12 ++-- clang/test/Driver/aarch64-rdm.c | 3 +++ clang/test/Sema/attr-target-clones-aarch64.c | 2 +- clang/test/SemaCXX/attr-target-version.cpp| 1 + .../llvm/TargetParser/AArch64TargetParser.h | 13 - llvm/lib/TargetParser/AArch64TargetParser.cpp | 15 +-- 7 files changed, 37 insertions(+), 14 deletions(-) diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst index f44fef28b9f17f..530158dda66689 100644 --- a/clang/docs/ReleaseNotes.rst +++ b/clang/docs/ReleaseNotes.rst @@ -333,6 +333,11 @@ Arm and AArch64 Support improvements for most targets. We have not changed the default behavior for ARMv6, but may revisit that decision in the future. Users can restore the old behavior with -m[no-]unaligned-access. +- An alias identifier (rdma) has been added for targeting the AArch64 + Architecture Extension which uses Rounding Doubling Multiply Accumulate + instructions (rdm). The identifier is available on the command line as + a feature modifier for -march and -mcpu as well as via target attributes + like ``target_version`` or ``target_clones``. Android Support ^^^ diff --git a/clang/test/CodeGen/attr-target-version.c b/clang/test/CodeGen/attr-target-version.c index ae97977a9144f6..56a42499d0a7ca 100644 --- a/clang/test/CodeGen/attr-target-version.c +++ b/clang/test/CodeGen/attr-target-version.c @@ -25,7 +25,7 @@ int foo() { } inline int __attribute__((target_version("sha1+pmull+f64mm"))) fmv_inline(void) { return 1; } -inline int __attribute__((target_version("fp16+fcma+sme+ fp16 "))) fmv_inline(void) { return 2; } +inline int __attribute__((target_version("fp16+fcma+rdma+sme+ fp16 "))) fmv_inline(void) { return 2; } inline int __attribute__((target_version("sha3+i8mm+f32mm"))) fmv_inline(void) { return 12; } inline int __attribute__((target_version("dit+sve-ebf16"))) fmv_inline(void) { return 8; } inline int __attribute__((target_version("dpb+rcpc2 "))) fmv_inline(void) { return 6; } @@ -261,12 +261,12 @@ int hoo(void) { // CHECK-NEXT: resolver_entry: // CHECK-NEXT:call void @__init_cpu_features_resolver() // CHECK-NEXT:[[TMP0:%.*]] = load i64, ptr @__aarch64_cpu_features, align 8 -// CHECK-NEXT:[[TMP1:%.*]] = and i64 [[TMP0]], 4398048608256 -// CHECK-NEXT:[[TMP2:%.*]] = icmp eq i64 [[TMP1]], 4398048608256 +// CHECK-NEXT:[[TMP1:%.*]] = and i64 [[TMP0]], 4398048608320 +// CHECK-NEXT:[[TMP2:%.*]] = icmp eq i64 [[TMP1]], 4398048608320 // CHECK-NEXT:[[TMP3:%.*]] = and i1 true, [[TMP2]] // CHECK-NEXT:br i1 [[TMP3]], label [[RESOLVER_RETURN:%.*]], label [[RESOLVER_ELSE:%.*]] // CHECK: resolver_return: -// CHECK-NEXT:ret ptr @fmv_inline._MfcmaMfp16Mfp16Msme +// CHECK-NEXT:ret ptr @fmv_inline._MfcmaMfp16Mfp16MrdmMsme // CHECK: resolver_else: // CHECK-NEXT:[[TMP4:%.*]] = load i64, ptr @__aarch64_cpu_features, align 8 // CHECK-NEXT:[[TMP5:%.*]] = and i64 [[TMP4]], 864726312827224064 @@ -575,7 +575,7 @@ int hoo(void) { // // // CHECK: Function Attrs: noinline nounwind optnone -// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MfcmaMfp16Mfp16Msme +// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MfcmaMfp16Mfp16MrdmMsme // CHECK-SAME: () #[[ATTR13:[0-9]+]] { // CHECK-NEXT: entry: // CHECK-NEXT:ret i32 2 @@ -829,7 +829,7 @@ int hoo(void) { // CHECK: attributes #[[ATTR10]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+bf16,+fullfp16,+ls64,+sme,+sme2" } // CHECK: attributes #[[ATTR11]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+ccpp,+fullfp16,+ls64" } // CHECK: attributes #[[ATTR12]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+aes,+f64mm,+fp-armv8,+fullfp16,+ls64,+neon,+sve" } -// CHECK: attributes #[[ATTR13]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+bf16,+complxnum,+fp-armv8,+fullfp16,+ls64,+neon,+sme" } +// CHECK: attributes #[[ATTR13]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-feature
[clang] [llvm] [TargetParser][AArch64] Add alias for FEAT_RDM. (PR #80540)
mstorsjo wrote: FYI, see 4b8d9abca7d0280878fb12de331e688ee85d7cd8 for another existing case where we already support both `rdm` and `rdma`. But I don't think that case can share any of the aliasing logic from here. https://github.com/llvm/llvm-project/pull/80540 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [TargetParser][AArch64] Add alias for FEAT_RDM. (PR #80540)
https://github.com/labrinea updated https://github.com/llvm/llvm-project/pull/80540 >From fb56f1c31f7f6fb847b95e0ffe97a39ac76759a9 Mon Sep 17 00:00:00 2001 From: Alexandros Lamprineas Date: Tue, 30 Jan 2024 11:17:55 + Subject: [PATCH] [TargetParser][AArch64] Add alias for FEAT_RDM. This patch allows using the name "rdma" as an alias for "rdm". The name makes its way to target attributes as well as the command line via the -march and -mcpu options. The motivation was originally to support this in Function Multi Versioning but it also makes sense to align with GCC on the command line. --- clang/docs/ReleaseNotes.rst | 5 + clang/test/CodeGen/attr-target-version.c | 12 ++-- clang/test/Driver/aarch64-rdm.c | 3 +++ clang/test/Sema/attr-target-clones-aarch64.c | 2 +- clang/test/SemaCXX/attr-target-version.cpp| 1 + .../llvm/TargetParser/AArch64TargetParser.h | 13 - llvm/lib/TargetParser/AArch64TargetParser.cpp | 15 +-- 7 files changed, 37 insertions(+), 14 deletions(-) diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst index 7e16b9f0c67dbd1..6ab31498ff33b9d 100644 --- a/clang/docs/ReleaseNotes.rst +++ b/clang/docs/ReleaseNotes.rst @@ -325,6 +325,11 @@ Arm and AArch64 Support improvements for most targets. We have not changed the default behavior for ARMv6, but may revisit that decision in the future. Users can restore the old behavior with -m[no-]unaligned-access. +- An alias identifier (rdma) has been added for targeting the AArch64 + Architecture Extension which uses Rounding Doubling Multiply Accumulate + instructions (rdm). The identifier is available on the command line as + a feature modifier for -march and -mcpu as well as via target attributes + like ``target_version`` or ``target_clones``. Android Support ^^^ diff --git a/clang/test/CodeGen/attr-target-version.c b/clang/test/CodeGen/attr-target-version.c index c27d48f3ecf681d..9eb9f3c73178479 100644 --- a/clang/test/CodeGen/attr-target-version.c +++ b/clang/test/CodeGen/attr-target-version.c @@ -25,7 +25,7 @@ int foo() { } inline int __attribute__((target_version("sha1+pmull+f64mm"))) fmv_inline(void) { return 1; } -inline int __attribute__((target_version("fp16+fcma+sme+ fp16 "))) fmv_inline(void) { return 2; } +inline int __attribute__((target_version("fp16+fcma+rdma+sme+ fp16 "))) fmv_inline(void) { return 2; } inline int __attribute__((target_version("sha3+i8mm+f32mm"))) fmv_inline(void) { return 12; } inline int __attribute__((target_version("dit+sve-ebf16"))) fmv_inline(void) { return 8; } inline int __attribute__((target_version("dpb+rcpc2 "))) fmv_inline(void) { return 6; } @@ -261,12 +261,12 @@ int hoo(void) { // CHECK-NEXT: resolver_entry: // CHECK-NEXT:call void @__init_cpu_features_resolver() // CHECK-NEXT:[[TMP0:%.*]] = load i64, ptr @__aarch64_cpu_features, align 8 -// CHECK-NEXT:[[TMP1:%.*]] = and i64 [[TMP0]], 4398048608256 -// CHECK-NEXT:[[TMP2:%.*]] = icmp eq i64 [[TMP1]], 4398048608256 +// CHECK-NEXT:[[TMP1:%.*]] = and i64 [[TMP0]], 4398048608320 +// CHECK-NEXT:[[TMP2:%.*]] = icmp eq i64 [[TMP1]], 4398048608320 // CHECK-NEXT:[[TMP3:%.*]] = and i1 true, [[TMP2]] // CHECK-NEXT:br i1 [[TMP3]], label [[RESOLVER_RETURN:%.*]], label [[RESOLVER_ELSE:%.*]] // CHECK: resolver_return: -// CHECK-NEXT:ret ptr @fmv_inline._Mfp16Mfp16MfcmaMsme +// CHECK-NEXT:ret ptr @fmv_inline._MrdmMfp16Mfp16MfcmaMsme // CHECK: resolver_else: // CHECK-NEXT:[[TMP4:%.*]] = load i64, ptr @__aarch64_cpu_features, align 8 // CHECK-NEXT:[[TMP5:%.*]] = and i64 [[TMP4]], 864726312827224064 @@ -575,7 +575,7 @@ int hoo(void) { // // // CHECK: Function Attrs: noinline nounwind optnone -// CHECK-LABEL: define {{[^@]+}}@fmv_inline._Mfp16Mfp16MfcmaMsme +// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MrdmMfp16Mfp16MfcmaMsme // CHECK-SAME: () #[[ATTR13:[0-9]+]] { // CHECK-NEXT: entry: // CHECK-NEXT:ret i32 2 @@ -829,7 +829,7 @@ int hoo(void) { // CHECK: attributes #[[ATTR10]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+bf16,+fullfp16,+ls64,+sme,+sme2" } // CHECK: attributes #[[ATTR11]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+ccpp,+fullfp16,+ls64" } // CHECK: attributes #[[ATTR12]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+aes,+f64mm,+fp-armv8,+fullfp16,+ls64,+neon,+sve" } -// CHECK: attributes #[[ATTR13]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+bf16,+complxnum,+fp-armv8,+fullfp16,+ls64,+neon,+sme" } +// CHECK: attributes #[[ATTR13]] = { noinline nounwind optnone "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-fea
[clang] [llvm] [TargetParser][AArch64] Add alias for FEAT_RDM. (PR #80540)
@@ -673,7 +673,7 @@ int hoo(void) { // // // CHECK: Function Attrs: noinline nounwind optnone -// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MlseMrdm +// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MlseMrdma jroelofs wrote: FWIW, I think it's okay to change this one: I doubt it will actually break anyone but we should get more conservative with this as you get closer to finalizing this part of the ACLE. https://github.com/llvm/llvm-project/pull/80540 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [TargetParser][AArch64] Add alias for FEAT_RDM. (PR #80540)
@@ -673,7 +673,7 @@ int hoo(void) { // // // CHECK: Function Attrs: noinline nounwind optnone -// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MlseMrdm +// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MlseMrdma labrinea wrote: Right, I didn't know that the resolver contains references to every target-version declaration that is visible in its translation unit even if the definition actually resides in another translation unit. That would be a problem if the definition has a mangled name that doesn't match because of the alias. Thanks for clarifying! https://github.com/llvm/llvm-project/pull/80540 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [TargetParser][AArch64] Add alias for FEAT_RDM. (PR #80540)
@@ -673,7 +673,7 @@ int hoo(void) { // // // CHECK: Function Attrs: noinline nounwind optnone -// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MlseMrdm +// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MlseMrdma jroelofs wrote: I think they are part of ABI, and giving them internal linkage is undesirable, IMO, as that would break an important use case: splitting the implementation across multiple TUs. https://godbolt.org/z/3zaM9MzPb https://github.com/llvm/llvm-project/pull/80540 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [TargetParser][AArch64] Add alias for FEAT_RDM. (PR #80540)
@@ -673,7 +673,7 @@ int hoo(void) { // // // CHECK: Function Attrs: noinline nounwind optnone -// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MlseMrdm +// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MlseMrdma labrinea wrote: I raised this with the GCC team, I don't have an answer yet but I believe GCC only accepts "rdma" on the attribute (so I suppose that makes its way to the mangled name?). I remember you raised this matter again in a previous review, when we changed priorities and the mangling was affected. That is a good point but I think first we need to agree whether those functions are part of the ABI or not. Right now I don't see why they need to be. We could even make then have internal linkage, no? If I am mistaken a counter example would help, thanks! https://github.com/llvm/llvm-project/pull/80540 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [TargetParser][AArch64] Add alias for FEAT_RDM. (PR #80540)
@@ -39,7 +39,7 @@ inline int __attribute__((target_version("memtag3+rcpc3+mops"))) fmv_inline(void inline int __attribute__((target_version("aes+dotprod"))) fmv_inline(void) { return 13; } inline int __attribute__((target_version("simd+fp16fml"))) fmv_inline(void) { return 14; } inline int __attribute__((target_version("fp+sm4"))) fmv_inline(void) { return 15; } -inline int __attribute__((target_version("lse+rdm"))) fmv_inline(void) { return 16; } +inline int __attribute__((target_version("lse+rdma"))) fmv_inline(void) { return 16; } jroelofs wrote: If there isn't one, mind adding another test to make sure that the old name still works? https://github.com/llvm/llvm-project/pull/80540 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [TargetParser][AArch64] Add alias for FEAT_RDM. (PR #80540)
@@ -673,7 +673,7 @@ int hoo(void) { // // // CHECK: Function Attrs: noinline nounwind optnone -// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MlseMrdm +// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MlseMrdma jroelofs wrote: What does GCC do? https://github.com/llvm/llvm-project/pull/80540 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [TargetParser][AArch64] Add alias for FEAT_RDM. (PR #80540)
@@ -673,7 +673,7 @@ int hoo(void) { // // // CHECK: Function Attrs: noinline nounwind optnone -// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MlseMrdm +// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MlseMrdma jroelofs wrote: It seems potentially problematic that this leaks into the mangling. Maybe these ought to be canonicalized in `AppendTargetVersionMangling` and `AppendTargetClonesMangling` with a `resolveExtAlias` call. https://github.com/llvm/llvm-project/pull/80540 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [llvm] [TargetParser][AArch64] Add alias for FEAT_RDM. (PR #80540)
llvmbot wrote: @llvm/pr-subscribers-clang @llvm/pr-subscribers-clang-driver Author: Alexandros Lamprineas (labrinea) Changes This patch allows using the name "rdma" as an alias for "rdm". The name makes its way to target attributes as well as the command line via the -march and -mcpu options. The motivation was originally to support this in Function Multi Versioning but it also makes sense to align with GCC on the command line. --- Full diff: https://github.com/llvm/llvm-project/pull/80540.diff 7 Files Affected: - (modified) clang/docs/ReleaseNotes.rst (+6) - (modified) clang/test/CodeGen/attr-target-version.c (+3-3) - (modified) clang/test/Driver/aarch64-rdm.c (+3) - (modified) clang/test/Sema/attr-target-clones-aarch64.c (+1-1) - (modified) clang/test/SemaCXX/attr-target-version.cpp (+1) - (modified) llvm/include/llvm/TargetParser/AArch64TargetParser.h (+8-5) - (modified) llvm/lib/TargetParser/AArch64TargetParser.cpp (+13-2) ``diff diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst index e634db3c718c9..5d525b74c056d 100644 --- a/clang/docs/ReleaseNotes.rst +++ b/clang/docs/ReleaseNotes.rst @@ -222,6 +222,12 @@ X86 Support Arm and AArch64 Support ^^^ +- An alias identifier (rdma) has been added for targeting the AArch64 + Architecture Extension which uses Rounding Doubling Multiply Accumulate + instructions (rdm). The identifier is available on the command line as + a feature modifier for -march and -mcpu as well as via target attributes + like ``target_version`` or ``target_clones``. + Android Support ^^^ diff --git a/clang/test/CodeGen/attr-target-version.c b/clang/test/CodeGen/attr-target-version.c index 2a96697e4291b..2ad6f3a5b0c44 100644 --- a/clang/test/CodeGen/attr-target-version.c +++ b/clang/test/CodeGen/attr-target-version.c @@ -39,7 +39,7 @@ inline int __attribute__((target_version("memtag3+rcpc3+mops"))) fmv_inline(void inline int __attribute__((target_version("aes+dotprod"))) fmv_inline(void) { return 13; } inline int __attribute__((target_version("simd+fp16fml"))) fmv_inline(void) { return 14; } inline int __attribute__((target_version("fp+sm4"))) fmv_inline(void) { return 15; } -inline int __attribute__((target_version("lse+rdm"))) fmv_inline(void) { return 16; } +inline int __attribute__((target_version("lse+rdma"))) fmv_inline(void) { return 16; } inline int __attribute__((target_version("default"))) fmv_inline(void) { return 3; } __attribute__((target_version("ls64"))) int fmv_e(void); @@ -385,7 +385,7 @@ int hoo(void) { // CHECK-NEXT:[[TMP59:%.*]] = and i1 true, [[TMP58]] // CHECK-NEXT:br i1 [[TMP59]], label [[RESOLVER_RETURN27:%.*]], label [[RESOLVER_ELSE28:%.*]] // CHECK: resolver_return27: -// CHECK-NEXT:ret ptr @fmv_inline._MlseMrdm +// CHECK-NEXT:ret ptr @fmv_inline._MlseMrdma // CHECK: resolver_else28: // CHECK-NEXT:[[TMP60:%.*]] = load i64, ptr @__aarch64_cpu_features, align 8 // CHECK-NEXT:[[TMP61:%.*]] = and i64 [[TMP60]], 32 @@ -673,7 +673,7 @@ int hoo(void) { // // // CHECK: Function Attrs: noinline nounwind optnone -// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MlseMrdm +// CHECK-LABEL: define {{[^@]+}}@fmv_inline._MlseMrdma // CHECK-SAME: () #[[ATTR25:[0-9]+]] { // CHECK-NEXT: entry: // CHECK-NEXT:ret i32 16 diff --git a/clang/test/Driver/aarch64-rdm.c b/clang/test/Driver/aarch64-rdm.c index f2542b381e7c2..62e1a4def4ce1 100644 --- a/clang/test/Driver/aarch64-rdm.c +++ b/clang/test/Driver/aarch64-rdm.c @@ -1,13 +1,16 @@ // RUN: %clang --target=aarch64-none-elf -march=armv8a+rdm -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-RDM %s +// RUN: %clang --target=aarch64-none-elf -march=armv8a+rdma -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-RDM %s // RUN: %clang --target=aarch64-none-elf -mcpu=generic+rdm -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-RDM %s // RUN: %clang --target=aarch64-none-elf -mcpu=falkor -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-RDM %s // RUN: %clang --target=aarch64-none-elf -mcpu=thunderx2t99 -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-RDM %s // CHECK-RDM: "-target-feature" "+rdm" // RUN: %clang --target=aarch64-none-elf -march=armv8a+nordm -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-NORDM %s +// RUN: %clang --target=aarch64-none-elf -march=armv8a+nordma -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-NORDM %s // RUN: %clang --target=aarch64-none-elf -mcpu=generic+nordm -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-NORDM %s // CHECK-NORDM-NOT: "-target-feature" "+rdm" // // RUN: %clang --target=aarch64-none-elf -march=armv8.1a -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-RDM %s // RUN: %clang --target=aarch64-none-elf -march=armv8.1a+nordm -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-NORDM-DEFAULT %s +// RUN: %clang --target=aarch64-none-elf -march=armv8.1a+nordma -### -c %s 2>&1 | FileCheck --check-prefix=CHECK-NORDM-DEFAULT %s //