[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
https://github.com/shiltian closed https://github.com/llvm/llvm-project/pull/130963 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
https://github.com/shiltian updated
https://github.com/llvm/llvm-project/pull/130963
>From 61eac4e7d7f8604021f67c48384f8c09bedd647f Mon Sep 17 00:00:00 2001
From: Shilei Tian
Date: Fri, 21 Mar 2025 12:16:30 -0400
Subject: [PATCH] Reapply "[AMDGPU] Use COV6 by default (#118515)"
This reverts commit 68bcba6d7a1cc18996c0bcb7c62267c62d2040d0.
---
clang/docs/ReleaseNotes.rst| 2 ++
clang/include/clang/Driver/Options.td | 4 ++--
clang/lib/Driver/ToolChains/CommonArgs.cpp | 2 +-
clang/test/CodeGen/amdgpu-address-spaces.cpp | 2 +-
clang/test/CodeGenCUDA/amdgpu-code-object-version.cu | 2 +-
clang/test/CodeGenCXX/dynamic-cast-address-space.cpp | 6 +++---
clang/test/CodeGenHIP/default-attributes.hip | 4 ++--
clang/test/CodeGenOpenCL/amdgpu-enqueue-kernel.cl | 6 +++---
.../amdgcn/bitcode/oclc_abi_version_600.bc | 0
.../lib/amdgcn/bitcode/oclc_abi_version_600.bc | 0
.../lib64/amdgcn/bitcode/oclc_abi_version_600.bc | 0
clang/test/Driver/hip-device-libs.hip | 2 +-
clang/test/OpenMP/amdgcn_target_global_constructor.cpp | 4 ++--
libc/cmake/modules/prepare_libc_gpu_build.cmake| 2 +-
llvm/docs/ReleaseNotes.md | 2 ++
llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp| 2 +-
.../CodeGen/AMDGPU/default_amdhsa_code_object_version.ll | 7 +++
17 files changed, 29 insertions(+), 18 deletions(-)
create mode 100644
clang/test/Driver/Inputs/rocm-spack/llvm-amdgpu-4.0.0-ieagcs7inf7runpyfvepqkurasoglq4z/amdgcn/bitcode/oclc_abi_version_600.bc
create mode 100644
clang/test/Driver/Inputs/rocm_resource_dir/lib/amdgcn/bitcode/oclc_abi_version_600.bc
create mode 100644
clang/test/Driver/Inputs/rocm_resource_dir/lib64/amdgcn/bitcode/oclc_abi_version_600.bc
create mode 100644
llvm/test/CodeGen/AMDGPU/default_amdhsa_code_object_version.ll
diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index 159991e8db981..5b5c4b23e7068 100644
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -381,6 +381,8 @@ Target Specific Changes
AMDGPU Support
^^
+- Bump the default code object version to 6. ROCm 6.3 is required to run any
program compiled with COV6.
+
NVPTX Support
^^
diff --git a/clang/include/clang/Driver/Options.td
b/clang/include/clang/Driver/Options.td
index 762e57be34eb0..fbd5cf632c350 100644
--- a/clang/include/clang/Driver/Options.td
+++ b/clang/include/clang/Driver/Options.td
@@ -5161,12 +5161,12 @@ defm amdgpu_ieee : BoolMOption<"amdgpu-ieee",
NegFlag>;
def mcode_object_version_EQ : Joined<["-"], "mcode-object-version=">,
Group,
- HelpText<"Specify code object ABI version. Defaults to 5. (AMDGPU only)">,
+ HelpText<"Specify code object ABI version. Defaults to 6. (AMDGPU only)">,
Visibility<[ClangOption, FlangOption, CC1Option, FC1Option]>,
Values<"none,4,5,6">,
NormalizedValuesScope<"llvm::CodeObjectVersionKind">,
NormalizedValues<["COV_None", "COV_4", "COV_5", "COV_6"]>,
- MarshallingInfoEnum, "COV_5">;
+ MarshallingInfoEnum, "COV_6">;
defm cumode : SimpleMFlag<"cumode",
"Specify CU wavefront", "Specify WGP wavefront",
diff --git a/clang/lib/Driver/ToolChains/CommonArgs.cpp
b/clang/lib/Driver/ToolChains/CommonArgs.cpp
index d8c72b98ada25..157b9ff971add 100644
--- a/clang/lib/Driver/ToolChains/CommonArgs.cpp
+++ b/clang/lib/Driver/ToolChains/CommonArgs.cpp
@@ -2760,7 +2760,7 @@ void tools::checkAMDGPUCodeObjectVersion(const Driver &D,
unsigned tools::getAMDGPUCodeObjectVersion(const Driver &D,
const llvm::opt::ArgList &Args) {
- unsigned CodeObjVer = 5; // default
+ unsigned CodeObjVer = 6; // default
if (auto *CodeObjArg = getAMDGPUCodeObjectArgument(D, Args))
StringRef(CodeObjArg->getValue()).getAsInteger(0, CodeObjVer);
return CodeObjVer;
diff --git a/clang/test/CodeGen/amdgpu-address-spaces.cpp
b/clang/test/CodeGen/amdgpu-address-spaces.cpp
index ae2c61439f4ca..b121b559f58dc 100644
--- a/clang/test/CodeGen/amdgpu-address-spaces.cpp
+++ b/clang/test/CodeGen/amdgpu-address-spaces.cpp
@@ -29,7 +29,7 @@ int [[clang::address_space(999)]] bbb = 1234;
// CHECK: @u = addrspace(5) global i32 undef, align 4
// CHECK: @aaa = addrspace(6) global i32 1000, align 4
// CHECK: @bbb = addrspace(999) global i32 1234, align 4
-// CHECK: @__oclc_ABI_version = weak_odr hidden local_unnamed_addr
addrspace(4) constant i32 500
+// CHECK: @__oclc_ABI_version = weak_odr hidden local_unnamed_addr
addrspace(4) constant i32 600
//.
// CHECK-LABEL: define dso_local amdgpu_kernel void @foo(
// CHECK-SAME: ) #[[ATTR0:[0-9]+]] {
diff --git a/clang/test/CodeGenCUDA/amdgpu-code-object-version.cu
b/clang/test/CodeGenCUDA/amdgpu-code-object-version.cu
index ffe12544917f7..aa0e3edec3f6a 100644
--- a/clang/test/CodeGenCUDA/amdgpu-code-objec
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
https://github.com/shiltian updated
https://github.com/llvm/llvm-project/pull/130963
>From 576596fb09e3f497858da0f922d746914a0c5c3d Mon Sep 17 00:00:00 2001
From: Shilei Tian
Date: Wed, 12 Mar 2025 09:39:45 -0400
Subject: [PATCH] Reapply "[AMDGPU] Use COV6 by default (#118515)"
This reverts commit 68bcba6d7a1cc18996c0bcb7c62267c62d2040d0.
---
clang/docs/ReleaseNotes.rst| 2 ++
clang/include/clang/Driver/Options.td | 4 ++--
clang/lib/Driver/ToolChains/CommonArgs.cpp | 2 +-
clang/test/CodeGen/amdgpu-address-spaces.cpp | 2 +-
clang/test/CodeGenCUDA/amdgpu-code-object-version.cu | 2 +-
clang/test/CodeGenCXX/dynamic-cast-address-space.cpp | 6 +++---
clang/test/CodeGenHIP/default-attributes.hip | 4 ++--
clang/test/CodeGenOpenCL/amdgpu-enqueue-kernel.cl | 6 +++---
.../amdgcn/bitcode/oclc_abi_version_600.bc | 0
.../lib/amdgcn/bitcode/oclc_abi_version_600.bc | 0
.../lib64/amdgcn/bitcode/oclc_abi_version_600.bc | 0
clang/test/Driver/hip-device-libs.hip | 2 +-
clang/test/OpenMP/amdgcn_target_global_constructor.cpp | 4 ++--
libc/cmake/modules/prepare_libc_gpu_build.cmake| 2 +-
llvm/docs/ReleaseNotes.md | 2 ++
llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp| 2 +-
.../CodeGen/AMDGPU/default_amdhsa_code_object_version.ll | 7 +++
17 files changed, 29 insertions(+), 18 deletions(-)
create mode 100644
clang/test/Driver/Inputs/rocm-spack/llvm-amdgpu-4.0.0-ieagcs7inf7runpyfvepqkurasoglq4z/amdgcn/bitcode/oclc_abi_version_600.bc
create mode 100644
clang/test/Driver/Inputs/rocm_resource_dir/lib/amdgcn/bitcode/oclc_abi_version_600.bc
create mode 100644
clang/test/Driver/Inputs/rocm_resource_dir/lib64/amdgcn/bitcode/oclc_abi_version_600.bc
create mode 100644
llvm/test/CodeGen/AMDGPU/default_amdhsa_code_object_version.ll
diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index 666bbf22acc93..afb729c514108 100644
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -372,6 +372,8 @@ Target Specific Changes
AMDGPU Support
^^
+- Bump the default code object version to 6. ROCm 6.3 is required to run any
program compiled with COV6.
+
NVPTX Support
^^
diff --git a/clang/include/clang/Driver/Options.td
b/clang/include/clang/Driver/Options.td
index ee783f8b09c24..591cc0a3e13ee 100644
--- a/clang/include/clang/Driver/Options.td
+++ b/clang/include/clang/Driver/Options.td
@@ -5162,12 +5162,12 @@ defm amdgpu_ieee : BoolMOption<"amdgpu-ieee",
NegFlag>;
def mcode_object_version_EQ : Joined<["-"], "mcode-object-version=">,
Group,
- HelpText<"Specify code object ABI version. Defaults to 5. (AMDGPU only)">,
+ HelpText<"Specify code object ABI version. Defaults to 6. (AMDGPU only)">,
Visibility<[ClangOption, FlangOption, CC1Option, FC1Option]>,
Values<"none,4,5,6">,
NormalizedValuesScope<"llvm::CodeObjectVersionKind">,
NormalizedValues<["COV_None", "COV_4", "COV_5", "COV_6"]>,
- MarshallingInfoEnum, "COV_5">;
+ MarshallingInfoEnum, "COV_6">;
defm cumode : SimpleMFlag<"cumode",
"Specify CU wavefront", "Specify WGP wavefront",
diff --git a/clang/lib/Driver/ToolChains/CommonArgs.cpp
b/clang/lib/Driver/ToolChains/CommonArgs.cpp
index bb553806d2eef..3075907365a60 100644
--- a/clang/lib/Driver/ToolChains/CommonArgs.cpp
+++ b/clang/lib/Driver/ToolChains/CommonArgs.cpp
@@ -2757,7 +2757,7 @@ void tools::checkAMDGPUCodeObjectVersion(const Driver &D,
unsigned tools::getAMDGPUCodeObjectVersion(const Driver &D,
const llvm::opt::ArgList &Args) {
- unsigned CodeObjVer = 5; // default
+ unsigned CodeObjVer = 6; // default
if (auto *CodeObjArg = getAMDGPUCodeObjectArgument(D, Args))
StringRef(CodeObjArg->getValue()).getAsInteger(0, CodeObjVer);
return CodeObjVer;
diff --git a/clang/test/CodeGen/amdgpu-address-spaces.cpp
b/clang/test/CodeGen/amdgpu-address-spaces.cpp
index ae2c61439f4ca..b121b559f58dc 100644
--- a/clang/test/CodeGen/amdgpu-address-spaces.cpp
+++ b/clang/test/CodeGen/amdgpu-address-spaces.cpp
@@ -29,7 +29,7 @@ int [[clang::address_space(999)]] bbb = 1234;
// CHECK: @u = addrspace(5) global i32 undef, align 4
// CHECK: @aaa = addrspace(6) global i32 1000, align 4
// CHECK: @bbb = addrspace(999) global i32 1234, align 4
-// CHECK: @__oclc_ABI_version = weak_odr hidden local_unnamed_addr
addrspace(4) constant i32 500
+// CHECK: @__oclc_ABI_version = weak_odr hidden local_unnamed_addr
addrspace(4) constant i32 600
//.
// CHECK-LABEL: define dso_local amdgpu_kernel void @foo(
// CHECK-SAME: ) #[[ATTR0:[0-9]+]] {
diff --git a/clang/test/CodeGenCUDA/amdgpu-code-object-version.cu
b/clang/test/CodeGenCUDA/amdgpu-code-object-version.cu
index ffe12544917f7..aa0e3edec3f6a 100644
--- a/clang/test/CodeGenCUDA/amdgpu-code-objec
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
@@ -332,6 +332,8 @@ Target Specific Changes AMDGPU Support ^^ +- Bump the default code object version to 6. ROCm 6.3 is required to run any program compiled with COV6. jhuber6 wrote: Yes, I have no clue why we don't provide that in the HSA header. Probably some ill-conceived notion that HSA is independent to ROCm. https://github.com/llvm/llvm-project/pull/130963 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
@@ -332,6 +332,8 @@ Target Specific Changes AMDGPU Support ^^ +- Bump the default code object version to 6. ROCm 6.3 is required to run any program compiled with COV6. arsenm wrote: We probably should have a defined policy for version support. It's also a long standing frustration of mine that we do not version the amdhsa OS to correspond to the rocm runtime version https://github.com/llvm/llvm-project/pull/130963 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
@@ -332,6 +332,8 @@ Target Specific Changes AMDGPU Support ^^ +- Bump the default code object version to 6. ROCm 6.3 is required to run any program compiled with COV6. jhuber6 wrote: Is it really? What runtime changes actually were required. Figured it'd still work so long as you're not using the generic targets. https://github.com/llvm/llvm-project/pull/130963 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
https://github.com/jhuber6 approved this pull request. https://github.com/llvm/llvm-project/pull/130963 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
shiltian wrote: > > The OpenMP runtime doesn't know how to handle `generic` ISAs right? > > I don't understand why that is tied to the version It is probably no longer tied to the version since we no longer build one device runtime per target. https://github.com/llvm/llvm-project/pull/130963 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
arsenm wrote: > The OpenMP runtime doesn't know how to handle `generic` ISAs right? I don't understand why that is tied to the version https://github.com/llvm/llvm-project/pull/130963 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
shiltian wrote: > The OpenMP runtime doesn't know how to handle `generic` ISAs right? I think people are working on it? https://github.com/llvm/llvm-project/pull/130963 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
jhuber6 wrote: The OpenMP runtime doesn't know how to handle `generic` ISAs right? https://github.com/llvm/llvm-project/pull/130963 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
jhuber6 wrote: > CC @jdoerfert @ye-luo Once this is merged, ROCm 6.3 will be needed to run any > program compiled for AMDGPU. Unless you pass `-mcode-object-version=5` right? https://github.com/llvm/llvm-project/pull/130963 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
shiltian wrote: > > CC @jdoerfert @ye-luo Once this is merged, ROCm 6.3 will be needed to run > > any program compiled for AMDGPU. > > Unless you pass `-mcode-object-version=5` right? Yes. https://github.com/llvm/llvm-project/pull/130963 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
llvmbot wrote:
@llvm/pr-subscribers-libc
Author: Shilei Tian (shiltian)
Changes
This reverts commit 68bcba6d7a1cc18996c0bcb7c62267c62d2040d0.
---
Full diff: https://github.com/llvm/llvm-project/pull/130963.diff
17 Files Affected:
- (modified) clang/docs/ReleaseNotes.rst (+3-1)
- (modified) clang/include/clang/Driver/Options.td (+2-2)
- (modified) clang/lib/Driver/ToolChains/CommonArgs.cpp (+1-1)
- (modified) clang/test/CodeGen/amdgpu-address-spaces.cpp (+1-1)
- (modified) clang/test/CodeGenCUDA/amdgpu-code-object-version.cu (+1-1)
- (modified) clang/test/CodeGenCXX/dynamic-cast-address-space.cpp (+3-3)
- (modified) clang/test/CodeGenHIP/default-attributes.hip (+2-2)
- (modified) clang/test/CodeGenOpenCL/amdgpu-enqueue-kernel.cl (+3-3)
- (added)
clang/test/Driver/Inputs/rocm-spack/llvm-amdgpu-4.0.0-ieagcs7inf7runpyfvepqkurasoglq4z/amdgcn/bitcode/oclc_abi_version_600.bc
()
- (added)
clang/test/Driver/Inputs/rocm_resource_dir/lib/amdgcn/bitcode/oclc_abi_version_600.bc
()
- (added)
clang/test/Driver/Inputs/rocm_resource_dir/lib64/amdgcn/bitcode/oclc_abi_version_600.bc
()
- (modified) clang/test/Driver/hip-device-libs.hip (+1-1)
- (modified) clang/test/OpenMP/amdgcn_target_global_constructor.cpp (+2-2)
- (modified) libc/cmake/modules/prepare_libc_gpu_build.cmake (+1-1)
- (modified) llvm/docs/ReleaseNotes.md (+2)
- (modified) llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp (+1-1)
- (added) llvm/test/CodeGen/AMDGPU/default_amdhsa_code_object_version.ll (+7)
``diff
diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index f6e4d6a34e5dc..2ed05ce28a1bc 100644
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -244,7 +244,7 @@ Improvements to Clang's diagnostics
as function arguments or return value respectively. Note that
:doc:`ThreadSafetyAnalysis` still does not perform alias analysis. The
feature will be default-enabled with ``-Wthread-safety`` in a future release.
-- The ``-Wsign-compare`` warning now treats expressions with bitwise not(~)
and minus(-) as signed integers
+- The ``-Wsign-compare`` warning now treats expressions with bitwise not(~)
and minus(-) as signed integers
except for the case where the operand is an unsigned integer
and throws warning if they are compared with unsigned integers (##18878).
@@ -332,6 +332,8 @@ Target Specific Changes
AMDGPU Support
^^
+- Bump the default code object version to 6. ROCm 6.3 is required to run any
program compiled with COV6.
+
NVPTX Support
^^
diff --git a/clang/include/clang/Driver/Options.td
b/clang/include/clang/Driver/Options.td
index e69cd6b833c3a..6fd09ec371e58 100644
--- a/clang/include/clang/Driver/Options.td
+++ b/clang/include/clang/Driver/Options.td
@@ -5153,12 +5153,12 @@ defm amdgpu_ieee : BoolMOption<"amdgpu-ieee",
NegFlag>;
def mcode_object_version_EQ : Joined<["-"], "mcode-object-version=">,
Group,
- HelpText<"Specify code object ABI version. Defaults to 5. (AMDGPU only)">,
+ HelpText<"Specify code object ABI version. Defaults to 6. (AMDGPU only)">,
Visibility<[ClangOption, FlangOption, CC1Option, FC1Option]>,
Values<"none,4,5,6">,
NormalizedValuesScope<"llvm::CodeObjectVersionKind">,
NormalizedValues<["COV_None", "COV_4", "COV_5", "COV_6"]>,
- MarshallingInfoEnum, "COV_5">;
+ MarshallingInfoEnum, "COV_6">;
defm cumode : SimpleMFlag<"cumode",
"Specify CU wavefront", "Specify WGP wavefront",
diff --git a/clang/lib/Driver/ToolChains/CommonArgs.cpp
b/clang/lib/Driver/ToolChains/CommonArgs.cpp
index b43472a52038b..0f6fd11e2837b 100644
--- a/clang/lib/Driver/ToolChains/CommonArgs.cpp
+++ b/clang/lib/Driver/ToolChains/CommonArgs.cpp
@@ -2756,7 +2756,7 @@ void tools::checkAMDGPUCodeObjectVersion(const Driver &D,
unsigned tools::getAMDGPUCodeObjectVersion(const Driver &D,
const llvm::opt::ArgList &Args) {
- unsigned CodeObjVer = 5; // default
+ unsigned CodeObjVer = 6; // default
if (auto *CodeObjArg = getAMDGPUCodeObjectArgument(D, Args))
StringRef(CodeObjArg->getValue()).getAsInteger(0, CodeObjVer);
return CodeObjVer;
diff --git a/clang/test/CodeGen/amdgpu-address-spaces.cpp
b/clang/test/CodeGen/amdgpu-address-spaces.cpp
index ae2c61439f4ca..b121b559f58dc 100644
--- a/clang/test/CodeGen/amdgpu-address-spaces.cpp
+++ b/clang/test/CodeGen/amdgpu-address-spaces.cpp
@@ -29,7 +29,7 @@ int [[clang::address_space(999)]] bbb = 1234;
// CHECK: @u = addrspace(5) global i32 undef, align 4
// CHECK: @aaa = addrspace(6) global i32 1000, align 4
// CHECK: @bbb = addrspace(999) global i32 1234, align 4
-// CHECK: @__oclc_ABI_version = weak_odr hidden local_unnamed_addr
addrspace(4) constant i32 500
+// CHECK: @__oclc_ABI_version = weak_odr hidden local_unnamed_addr
addrspace(4) constant i32 600
//.
// CHECK-LABEL: define dso_local amdgpu_kernel void @foo(
// CHECK-SAME: ) #[[ATTR0:[0-9]+]] {
diff --git a/clang/test/CodeGenCUDA/amdgpu
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
shiltian wrote: CC @jdoerfert @ye-luo Once this is merged, ROCm 6.3 will be needed to run any program compiled for AMDGPU. https://github.com/llvm/llvm-project/pull/130963 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
shiltian wrote: We will need to wait for the AMD bots to be ready. https://github.com/llvm/llvm-project/pull/130963 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
llvmbot wrote:
@llvm/pr-subscribers-backend-amdgpu
Author: Shilei Tian (shiltian)
Changes
This reverts commit 68bcba6d7a1cc18996c0bcb7c62267c62d2040d0.
---
Full diff: https://github.com/llvm/llvm-project/pull/130963.diff
17 Files Affected:
- (modified) clang/docs/ReleaseNotes.rst (+3-1)
- (modified) clang/include/clang/Driver/Options.td (+2-2)
- (modified) clang/lib/Driver/ToolChains/CommonArgs.cpp (+1-1)
- (modified) clang/test/CodeGen/amdgpu-address-spaces.cpp (+1-1)
- (modified) clang/test/CodeGenCUDA/amdgpu-code-object-version.cu (+1-1)
- (modified) clang/test/CodeGenCXX/dynamic-cast-address-space.cpp (+3-3)
- (modified) clang/test/CodeGenHIP/default-attributes.hip (+2-2)
- (modified) clang/test/CodeGenOpenCL/amdgpu-enqueue-kernel.cl (+3-3)
- (added)
clang/test/Driver/Inputs/rocm-spack/llvm-amdgpu-4.0.0-ieagcs7inf7runpyfvepqkurasoglq4z/amdgcn/bitcode/oclc_abi_version_600.bc
()
- (added)
clang/test/Driver/Inputs/rocm_resource_dir/lib/amdgcn/bitcode/oclc_abi_version_600.bc
()
- (added)
clang/test/Driver/Inputs/rocm_resource_dir/lib64/amdgcn/bitcode/oclc_abi_version_600.bc
()
- (modified) clang/test/Driver/hip-device-libs.hip (+1-1)
- (modified) clang/test/OpenMP/amdgcn_target_global_constructor.cpp (+2-2)
- (modified) libc/cmake/modules/prepare_libc_gpu_build.cmake (+1-1)
- (modified) llvm/docs/ReleaseNotes.md (+2)
- (modified) llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp (+1-1)
- (added) llvm/test/CodeGen/AMDGPU/default_amdhsa_code_object_version.ll (+7)
``diff
diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index f6e4d6a34e5dc..2ed05ce28a1bc 100644
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -244,7 +244,7 @@ Improvements to Clang's diagnostics
as function arguments or return value respectively. Note that
:doc:`ThreadSafetyAnalysis` still does not perform alias analysis. The
feature will be default-enabled with ``-Wthread-safety`` in a future release.
-- The ``-Wsign-compare`` warning now treats expressions with bitwise not(~)
and minus(-) as signed integers
+- The ``-Wsign-compare`` warning now treats expressions with bitwise not(~)
and minus(-) as signed integers
except for the case where the operand is an unsigned integer
and throws warning if they are compared with unsigned integers (##18878).
@@ -332,6 +332,8 @@ Target Specific Changes
AMDGPU Support
^^
+- Bump the default code object version to 6. ROCm 6.3 is required to run any
program compiled with COV6.
+
NVPTX Support
^^
diff --git a/clang/include/clang/Driver/Options.td
b/clang/include/clang/Driver/Options.td
index e69cd6b833c3a..6fd09ec371e58 100644
--- a/clang/include/clang/Driver/Options.td
+++ b/clang/include/clang/Driver/Options.td
@@ -5153,12 +5153,12 @@ defm amdgpu_ieee : BoolMOption<"amdgpu-ieee",
NegFlag>;
def mcode_object_version_EQ : Joined<["-"], "mcode-object-version=">,
Group,
- HelpText<"Specify code object ABI version. Defaults to 5. (AMDGPU only)">,
+ HelpText<"Specify code object ABI version. Defaults to 6. (AMDGPU only)">,
Visibility<[ClangOption, FlangOption, CC1Option, FC1Option]>,
Values<"none,4,5,6">,
NormalizedValuesScope<"llvm::CodeObjectVersionKind">,
NormalizedValues<["COV_None", "COV_4", "COV_5", "COV_6"]>,
- MarshallingInfoEnum, "COV_5">;
+ MarshallingInfoEnum, "COV_6">;
defm cumode : SimpleMFlag<"cumode",
"Specify CU wavefront", "Specify WGP wavefront",
diff --git a/clang/lib/Driver/ToolChains/CommonArgs.cpp
b/clang/lib/Driver/ToolChains/CommonArgs.cpp
index b43472a52038b..0f6fd11e2837b 100644
--- a/clang/lib/Driver/ToolChains/CommonArgs.cpp
+++ b/clang/lib/Driver/ToolChains/CommonArgs.cpp
@@ -2756,7 +2756,7 @@ void tools::checkAMDGPUCodeObjectVersion(const Driver &D,
unsigned tools::getAMDGPUCodeObjectVersion(const Driver &D,
const llvm::opt::ArgList &Args) {
- unsigned CodeObjVer = 5; // default
+ unsigned CodeObjVer = 6; // default
if (auto *CodeObjArg = getAMDGPUCodeObjectArgument(D, Args))
StringRef(CodeObjArg->getValue()).getAsInteger(0, CodeObjVer);
return CodeObjVer;
diff --git a/clang/test/CodeGen/amdgpu-address-spaces.cpp
b/clang/test/CodeGen/amdgpu-address-spaces.cpp
index ae2c61439f4ca..b121b559f58dc 100644
--- a/clang/test/CodeGen/amdgpu-address-spaces.cpp
+++ b/clang/test/CodeGen/amdgpu-address-spaces.cpp
@@ -29,7 +29,7 @@ int [[clang::address_space(999)]] bbb = 1234;
// CHECK: @u = addrspace(5) global i32 undef, align 4
// CHECK: @aaa = addrspace(6) global i32 1000, align 4
// CHECK: @bbb = addrspace(999) global i32 1234, align 4
-// CHECK: @__oclc_ABI_version = weak_odr hidden local_unnamed_addr
addrspace(4) constant i32 500
+// CHECK: @__oclc_ABI_version = weak_odr hidden local_unnamed_addr
addrspace(4) constant i32 600
//.
// CHECK-LABEL: define dso_local amdgpu_kernel void @foo(
// CHECK-SAME: ) #[[ATTR0:[0-9]+]] {
diff --git a/clang/test/CodeGenC
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
shiltian wrote: * **#130963** https://app.graphite.dev/github/pr/llvm/llvm-project/130963?utm_source=stack-comment-icon"; target="_blank">https://static.graphite.dev/graphite-32x32-black.png"; alt="Graphite" width="10px" height="10px"/> 👈 https://app.graphite.dev/github/pr/llvm/llvm-project/130963?utm_source=stack-comment-view-in-graphite"; target="_blank">(View in Graphite) * `main` This stack of pull requests is managed by https://graphite.dev?utm-source=stack-comment";>Graphite. Learn more about https://stacking.dev/?utm_source=stack-comment";>stacking. https://github.com/llvm/llvm-project/pull/130963 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
llvmbot wrote:
@llvm/pr-subscribers-clang-driver
Author: Shilei Tian (shiltian)
Changes
This reverts commit 68bcba6d7a1cc18996c0bcb7c62267c62d2040d0.
---
Full diff: https://github.com/llvm/llvm-project/pull/130963.diff
17 Files Affected:
- (modified) clang/docs/ReleaseNotes.rst (+3-1)
- (modified) clang/include/clang/Driver/Options.td (+2-2)
- (modified) clang/lib/Driver/ToolChains/CommonArgs.cpp (+1-1)
- (modified) clang/test/CodeGen/amdgpu-address-spaces.cpp (+1-1)
- (modified) clang/test/CodeGenCUDA/amdgpu-code-object-version.cu (+1-1)
- (modified) clang/test/CodeGenCXX/dynamic-cast-address-space.cpp (+3-3)
- (modified) clang/test/CodeGenHIP/default-attributes.hip (+2-2)
- (modified) clang/test/CodeGenOpenCL/amdgpu-enqueue-kernel.cl (+3-3)
- (added)
clang/test/Driver/Inputs/rocm-spack/llvm-amdgpu-4.0.0-ieagcs7inf7runpyfvepqkurasoglq4z/amdgcn/bitcode/oclc_abi_version_600.bc
()
- (added)
clang/test/Driver/Inputs/rocm_resource_dir/lib/amdgcn/bitcode/oclc_abi_version_600.bc
()
- (added)
clang/test/Driver/Inputs/rocm_resource_dir/lib64/amdgcn/bitcode/oclc_abi_version_600.bc
()
- (modified) clang/test/Driver/hip-device-libs.hip (+1-1)
- (modified) clang/test/OpenMP/amdgcn_target_global_constructor.cpp (+2-2)
- (modified) libc/cmake/modules/prepare_libc_gpu_build.cmake (+1-1)
- (modified) llvm/docs/ReleaseNotes.md (+2)
- (modified) llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp (+1-1)
- (added) llvm/test/CodeGen/AMDGPU/default_amdhsa_code_object_version.ll (+7)
``diff
diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index f6e4d6a34e5dc..2ed05ce28a1bc 100644
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -244,7 +244,7 @@ Improvements to Clang's diagnostics
as function arguments or return value respectively. Note that
:doc:`ThreadSafetyAnalysis` still does not perform alias analysis. The
feature will be default-enabled with ``-Wthread-safety`` in a future release.
-- The ``-Wsign-compare`` warning now treats expressions with bitwise not(~)
and minus(-) as signed integers
+- The ``-Wsign-compare`` warning now treats expressions with bitwise not(~)
and minus(-) as signed integers
except for the case where the operand is an unsigned integer
and throws warning if they are compared with unsigned integers (##18878).
@@ -332,6 +332,8 @@ Target Specific Changes
AMDGPU Support
^^
+- Bump the default code object version to 6. ROCm 6.3 is required to run any
program compiled with COV6.
+
NVPTX Support
^^
diff --git a/clang/include/clang/Driver/Options.td
b/clang/include/clang/Driver/Options.td
index e69cd6b833c3a..6fd09ec371e58 100644
--- a/clang/include/clang/Driver/Options.td
+++ b/clang/include/clang/Driver/Options.td
@@ -5153,12 +5153,12 @@ defm amdgpu_ieee : BoolMOption<"amdgpu-ieee",
NegFlag>;
def mcode_object_version_EQ : Joined<["-"], "mcode-object-version=">,
Group,
- HelpText<"Specify code object ABI version. Defaults to 5. (AMDGPU only)">,
+ HelpText<"Specify code object ABI version. Defaults to 6. (AMDGPU only)">,
Visibility<[ClangOption, FlangOption, CC1Option, FC1Option]>,
Values<"none,4,5,6">,
NormalizedValuesScope<"llvm::CodeObjectVersionKind">,
NormalizedValues<["COV_None", "COV_4", "COV_5", "COV_6"]>,
- MarshallingInfoEnum, "COV_5">;
+ MarshallingInfoEnum, "COV_6">;
defm cumode : SimpleMFlag<"cumode",
"Specify CU wavefront", "Specify WGP wavefront",
diff --git a/clang/lib/Driver/ToolChains/CommonArgs.cpp
b/clang/lib/Driver/ToolChains/CommonArgs.cpp
index b43472a52038b..0f6fd11e2837b 100644
--- a/clang/lib/Driver/ToolChains/CommonArgs.cpp
+++ b/clang/lib/Driver/ToolChains/CommonArgs.cpp
@@ -2756,7 +2756,7 @@ void tools::checkAMDGPUCodeObjectVersion(const Driver &D,
unsigned tools::getAMDGPUCodeObjectVersion(const Driver &D,
const llvm::opt::ArgList &Args) {
- unsigned CodeObjVer = 5; // default
+ unsigned CodeObjVer = 6; // default
if (auto *CodeObjArg = getAMDGPUCodeObjectArgument(D, Args))
StringRef(CodeObjArg->getValue()).getAsInteger(0, CodeObjVer);
return CodeObjVer;
diff --git a/clang/test/CodeGen/amdgpu-address-spaces.cpp
b/clang/test/CodeGen/amdgpu-address-spaces.cpp
index ae2c61439f4ca..b121b559f58dc 100644
--- a/clang/test/CodeGen/amdgpu-address-spaces.cpp
+++ b/clang/test/CodeGen/amdgpu-address-spaces.cpp
@@ -29,7 +29,7 @@ int [[clang::address_space(999)]] bbb = 1234;
// CHECK: @u = addrspace(5) global i32 undef, align 4
// CHECK: @aaa = addrspace(6) global i32 1000, align 4
// CHECK: @bbb = addrspace(999) global i32 1234, align 4
-// CHECK: @__oclc_ABI_version = weak_odr hidden local_unnamed_addr
addrspace(4) constant i32 500
+// CHECK: @__oclc_ABI_version = weak_odr hidden local_unnamed_addr
addrspace(4) constant i32 600
//.
// CHECK-LABEL: define dso_local amdgpu_kernel void @foo(
// CHECK-SAME: ) #[[ATTR0:[0-9]+]] {
diff --git a/clang/test/CodeGenCUD
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
llvmbot wrote:
@llvm/pr-subscribers-clang
Author: Shilei Tian (shiltian)
Changes
This reverts commit 68bcba6d7a1cc18996c0bcb7c62267c62d2040d0.
---
Full diff: https://github.com/llvm/llvm-project/pull/130963.diff
17 Files Affected:
- (modified) clang/docs/ReleaseNotes.rst (+3-1)
- (modified) clang/include/clang/Driver/Options.td (+2-2)
- (modified) clang/lib/Driver/ToolChains/CommonArgs.cpp (+1-1)
- (modified) clang/test/CodeGen/amdgpu-address-spaces.cpp (+1-1)
- (modified) clang/test/CodeGenCUDA/amdgpu-code-object-version.cu (+1-1)
- (modified) clang/test/CodeGenCXX/dynamic-cast-address-space.cpp (+3-3)
- (modified) clang/test/CodeGenHIP/default-attributes.hip (+2-2)
- (modified) clang/test/CodeGenOpenCL/amdgpu-enqueue-kernel.cl (+3-3)
- (added)
clang/test/Driver/Inputs/rocm-spack/llvm-amdgpu-4.0.0-ieagcs7inf7runpyfvepqkurasoglq4z/amdgcn/bitcode/oclc_abi_version_600.bc
()
- (added)
clang/test/Driver/Inputs/rocm_resource_dir/lib/amdgcn/bitcode/oclc_abi_version_600.bc
()
- (added)
clang/test/Driver/Inputs/rocm_resource_dir/lib64/amdgcn/bitcode/oclc_abi_version_600.bc
()
- (modified) clang/test/Driver/hip-device-libs.hip (+1-1)
- (modified) clang/test/OpenMP/amdgcn_target_global_constructor.cpp (+2-2)
- (modified) libc/cmake/modules/prepare_libc_gpu_build.cmake (+1-1)
- (modified) llvm/docs/ReleaseNotes.md (+2)
- (modified) llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp (+1-1)
- (added) llvm/test/CodeGen/AMDGPU/default_amdhsa_code_object_version.ll (+7)
``diff
diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index f6e4d6a34e5dc..2ed05ce28a1bc 100644
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -244,7 +244,7 @@ Improvements to Clang's diagnostics
as function arguments or return value respectively. Note that
:doc:`ThreadSafetyAnalysis` still does not perform alias analysis. The
feature will be default-enabled with ``-Wthread-safety`` in a future release.
-- The ``-Wsign-compare`` warning now treats expressions with bitwise not(~)
and minus(-) as signed integers
+- The ``-Wsign-compare`` warning now treats expressions with bitwise not(~)
and minus(-) as signed integers
except for the case where the operand is an unsigned integer
and throws warning if they are compared with unsigned integers (##18878).
@@ -332,6 +332,8 @@ Target Specific Changes
AMDGPU Support
^^
+- Bump the default code object version to 6. ROCm 6.3 is required to run any
program compiled with COV6.
+
NVPTX Support
^^
diff --git a/clang/include/clang/Driver/Options.td
b/clang/include/clang/Driver/Options.td
index e69cd6b833c3a..6fd09ec371e58 100644
--- a/clang/include/clang/Driver/Options.td
+++ b/clang/include/clang/Driver/Options.td
@@ -5153,12 +5153,12 @@ defm amdgpu_ieee : BoolMOption<"amdgpu-ieee",
NegFlag>;
def mcode_object_version_EQ : Joined<["-"], "mcode-object-version=">,
Group,
- HelpText<"Specify code object ABI version. Defaults to 5. (AMDGPU only)">,
+ HelpText<"Specify code object ABI version. Defaults to 6. (AMDGPU only)">,
Visibility<[ClangOption, FlangOption, CC1Option, FC1Option]>,
Values<"none,4,5,6">,
NormalizedValuesScope<"llvm::CodeObjectVersionKind">,
NormalizedValues<["COV_None", "COV_4", "COV_5", "COV_6"]>,
- MarshallingInfoEnum, "COV_5">;
+ MarshallingInfoEnum, "COV_6">;
defm cumode : SimpleMFlag<"cumode",
"Specify CU wavefront", "Specify WGP wavefront",
diff --git a/clang/lib/Driver/ToolChains/CommonArgs.cpp
b/clang/lib/Driver/ToolChains/CommonArgs.cpp
index b43472a52038b..0f6fd11e2837b 100644
--- a/clang/lib/Driver/ToolChains/CommonArgs.cpp
+++ b/clang/lib/Driver/ToolChains/CommonArgs.cpp
@@ -2756,7 +2756,7 @@ void tools::checkAMDGPUCodeObjectVersion(const Driver &D,
unsigned tools::getAMDGPUCodeObjectVersion(const Driver &D,
const llvm::opt::ArgList &Args) {
- unsigned CodeObjVer = 5; // default
+ unsigned CodeObjVer = 6; // default
if (auto *CodeObjArg = getAMDGPUCodeObjectArgument(D, Args))
StringRef(CodeObjArg->getValue()).getAsInteger(0, CodeObjVer);
return CodeObjVer;
diff --git a/clang/test/CodeGen/amdgpu-address-spaces.cpp
b/clang/test/CodeGen/amdgpu-address-spaces.cpp
index ae2c61439f4ca..b121b559f58dc 100644
--- a/clang/test/CodeGen/amdgpu-address-spaces.cpp
+++ b/clang/test/CodeGen/amdgpu-address-spaces.cpp
@@ -29,7 +29,7 @@ int [[clang::address_space(999)]] bbb = 1234;
// CHECK: @u = addrspace(5) global i32 undef, align 4
// CHECK: @aaa = addrspace(6) global i32 1000, align 4
// CHECK: @bbb = addrspace(999) global i32 1234, align 4
-// CHECK: @__oclc_ABI_version = weak_odr hidden local_unnamed_addr
addrspace(4) constant i32 500
+// CHECK: @__oclc_ABI_version = weak_odr hidden local_unnamed_addr
addrspace(4) constant i32 600
//.
// CHECK-LABEL: define dso_local amdgpu_kernel void @foo(
// CHECK-SAME: ) #[[ATTR0:[0-9]+]] {
diff --git a/clang/test/CodeGenCUDA/amdgp
[clang] [libc] [llvm] Reapply "[AMDGPU] Use COV6 by default (#118515)" (PR #130963)
https://github.com/shiltian created
https://github.com/llvm/llvm-project/pull/130963
This reverts commit 68bcba6d7a1cc18996c0bcb7c62267c62d2040d0.
>From 0f831a4a78fefcdf0ac973173397325a1f53d393 Mon Sep 17 00:00:00 2001
From: Shilei Tian
Date: Wed, 12 Mar 2025 09:39:45 -0400
Subject: [PATCH] Reapply "[AMDGPU] Use COV6 by default (#118515)"
This reverts commit 68bcba6d7a1cc18996c0bcb7c62267c62d2040d0.
---
clang/docs/ReleaseNotes.rst| 4 +++-
clang/include/clang/Driver/Options.td | 4 ++--
clang/lib/Driver/ToolChains/CommonArgs.cpp | 2 +-
clang/test/CodeGen/amdgpu-address-spaces.cpp | 2 +-
clang/test/CodeGenCUDA/amdgpu-code-object-version.cu | 2 +-
clang/test/CodeGenCXX/dynamic-cast-address-space.cpp | 6 +++---
clang/test/CodeGenHIP/default-attributes.hip | 4 ++--
clang/test/CodeGenOpenCL/amdgpu-enqueue-kernel.cl | 6 +++---
.../amdgcn/bitcode/oclc_abi_version_600.bc | 0
.../lib/amdgcn/bitcode/oclc_abi_version_600.bc | 0
.../lib64/amdgcn/bitcode/oclc_abi_version_600.bc | 0
clang/test/Driver/hip-device-libs.hip | 2 +-
clang/test/OpenMP/amdgcn_target_global_constructor.cpp | 4 ++--
libc/cmake/modules/prepare_libc_gpu_build.cmake| 2 +-
llvm/docs/ReleaseNotes.md | 2 ++
llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp| 2 +-
.../CodeGen/AMDGPU/default_amdhsa_code_object_version.ll | 7 +++
17 files changed, 30 insertions(+), 19 deletions(-)
create mode 100644
clang/test/Driver/Inputs/rocm-spack/llvm-amdgpu-4.0.0-ieagcs7inf7runpyfvepqkurasoglq4z/amdgcn/bitcode/oclc_abi_version_600.bc
create mode 100644
clang/test/Driver/Inputs/rocm_resource_dir/lib/amdgcn/bitcode/oclc_abi_version_600.bc
create mode 100644
clang/test/Driver/Inputs/rocm_resource_dir/lib64/amdgcn/bitcode/oclc_abi_version_600.bc
create mode 100644
llvm/test/CodeGen/AMDGPU/default_amdhsa_code_object_version.ll
diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index f6e4d6a34e5dc..2ed05ce28a1bc 100644
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -244,7 +244,7 @@ Improvements to Clang's diagnostics
as function arguments or return value respectively. Note that
:doc:`ThreadSafetyAnalysis` still does not perform alias analysis. The
feature will be default-enabled with ``-Wthread-safety`` in a future release.
-- The ``-Wsign-compare`` warning now treats expressions with bitwise not(~)
and minus(-) as signed integers
+- The ``-Wsign-compare`` warning now treats expressions with bitwise not(~)
and minus(-) as signed integers
except for the case where the operand is an unsigned integer
and throws warning if they are compared with unsigned integers (##18878).
@@ -332,6 +332,8 @@ Target Specific Changes
AMDGPU Support
^^
+- Bump the default code object version to 6. ROCm 6.3 is required to run any
program compiled with COV6.
+
NVPTX Support
^^
diff --git a/clang/include/clang/Driver/Options.td
b/clang/include/clang/Driver/Options.td
index e69cd6b833c3a..6fd09ec371e58 100644
--- a/clang/include/clang/Driver/Options.td
+++ b/clang/include/clang/Driver/Options.td
@@ -5153,12 +5153,12 @@ defm amdgpu_ieee : BoolMOption<"amdgpu-ieee",
NegFlag>;
def mcode_object_version_EQ : Joined<["-"], "mcode-object-version=">,
Group,
- HelpText<"Specify code object ABI version. Defaults to 5. (AMDGPU only)">,
+ HelpText<"Specify code object ABI version. Defaults to 6. (AMDGPU only)">,
Visibility<[ClangOption, FlangOption, CC1Option, FC1Option]>,
Values<"none,4,5,6">,
NormalizedValuesScope<"llvm::CodeObjectVersionKind">,
NormalizedValues<["COV_None", "COV_4", "COV_5", "COV_6"]>,
- MarshallingInfoEnum, "COV_5">;
+ MarshallingInfoEnum, "COV_6">;
defm cumode : SimpleMFlag<"cumode",
"Specify CU wavefront", "Specify WGP wavefront",
diff --git a/clang/lib/Driver/ToolChains/CommonArgs.cpp
b/clang/lib/Driver/ToolChains/CommonArgs.cpp
index b43472a52038b..0f6fd11e2837b 100644
--- a/clang/lib/Driver/ToolChains/CommonArgs.cpp
+++ b/clang/lib/Driver/ToolChains/CommonArgs.cpp
@@ -2756,7 +2756,7 @@ void tools::checkAMDGPUCodeObjectVersion(const Driver &D,
unsigned tools::getAMDGPUCodeObjectVersion(const Driver &D,
const llvm::opt::ArgList &Args) {
- unsigned CodeObjVer = 5; // default
+ unsigned CodeObjVer = 6; // default
if (auto *CodeObjArg = getAMDGPUCodeObjectArgument(D, Args))
StringRef(CodeObjArg->getValue()).getAsInteger(0, CodeObjVer);
return CodeObjVer;
diff --git a/clang/test/CodeGen/amdgpu-address-spaces.cpp
b/clang/test/CodeGen/amdgpu-address-spaces.cpp
index ae2c61439f4ca..b121b559f58dc 100644
--- a/clang/test/CodeGen/amdgpu-address-spaces.cpp
+++ b/clang/test/CodeGen/amdgpu-address-spaces.cpp
@@ -29,7 +29,7 @@ int [[clang::address
