https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/155818
>From efdd979d6189b986ea09dec7c0c0da26725ab1a1 Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Wed, 27 Aug 2025 10:24:51 -0500
Subject: [PATCH 1/5] [Flang][OpenMP] Enable no-loop kernels
Enable the
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/155818
>From efdd979d6189b986ea09dec7c0c0da26725ab1a1 Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Wed, 27 Aug 2025 10:24:51 -0500
Subject: [PATCH 1/5] [Flang][OpenMP] Enable no-loop kernels
Enable the
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/155818
>From efdd979d6189b986ea09dec7c0c0da26725ab1a1 Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Wed, 27 Aug 2025 10:24:51 -0500
Subject: [PATCH 1/4] [Flang][OpenMP] Enable no-loop kernels
Enable the
@@ -2106,6 +2106,29 @@ Operation *TargetOp::getInnermostCapturedOmpOp() {
});
}
+/// Check if we can promote SPMD kernel to No-Loop kernel
+static bool canPromoteToNoLoop(Operation *capturedOp, TeamsOp teamsOp,
+ WsloopOp *wsLoopOp) {
+ //
@@ -0,0 +1,43 @@
+! Check if the first OpenMP GPU kernel is promoted to no-loop mode.
DominikAdamski wrote:
Done. Please see updated test.
https://github.com/llvm/llvm-project/pull/155818
___
cfe-commits mailing list
c
@@ -2590,13 +2590,27 @@ convertOmpWsloop(Operation &opInst, llvm::IRBuilderBase
&builder,
}
builder.SetInsertPoint(*regionBlock, (*regionBlock)->begin());
+
+ bool noLoopMode = false;
+ omp::TargetOp targetOp = wsloopOp->getParentOfType();
+ if (targetOp) {
+Operat
https://github.com/DominikAdamski edited
https://github.com/llvm/llvm-project/pull/155818
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/155818
>From efdd979d6189b986ea09dec7c0c0da26725ab1a1 Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Wed, 27 Aug 2025 10:24:51 -0500
Subject: [PATCH 1/3] [Flang][OpenMP] Enable no-loop kernels
Enable the
@@ -2106,6 +2106,29 @@ Operation *TargetOp::getInnermostCapturedOmpOp() {
});
}
+/// Check if we can promote SPMD kernel to No-Loop kernel
+static bool canPromoteToNoLoop(Operation *capturedOp, TeamsOp teamsOp,
+ WsloopOp *wsLoopOp) {
+ //
@@ -2106,6 +2106,29 @@ Operation *TargetOp::getInnermostCapturedOmpOp() {
});
}
+/// Check if we can promote SPMD kernel to No-Loop kernel
DominikAdamski wrote:
Done
https://github.com/llvm/llvm-project/pull/155818
_
@@ -2106,6 +2106,29 @@ Operation *TargetOp::getInnermostCapturedOmpOp() {
});
}
+/// Check if we can promote SPMD kernel to No-Loop kernel
+static bool canPromoteToNoLoop(Operation *capturedOp, TeamsOp teamsOp,
+ WsloopOp *wsLoopOp) {
+ //
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/155818
>From efdd979d6189b986ea09dec7c0c0da26725ab1a1 Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Wed, 27 Aug 2025 10:24:51 -0500
Subject: [PATCH 1/3] [Flang][OpenMP] Enable no-loop kernels
Enable the
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/155818
>From efdd979d6189b986ea09dec7c0c0da26725ab1a1 Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Wed, 27 Aug 2025 10:24:51 -0500
Subject: [PATCH 1/2] [Flang][OpenMP] Enable no-loop kernels
Enable the
https://github.com/DominikAdamski closed
https://github.com/llvm/llvm-project/pull/96742
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/DominikAdamski edited
https://github.com/llvm/llvm-project/pull/96742
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/96742
>From 5b487aac3c8414b6f37f6888f361ca7488094048 Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Fri, 21 Jun 2024 18:03:53 +0200
Subject: [PATCH 1/5] [Flang-new][OpenMP] Add offload related flags for AM
DominikAdamski wrote:
@jhuber6 You are right. Flang-new for AMD GPU requires `-mlink-builtin-bitcode`
for math functions.
https://github.com/llvm/llvm-project/pull/96742
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cg
DominikAdamski wrote:
> > > Who could be the right person to ask?
> >
> > I don't know. Open-source LLVM Flang meetings can be good place to ask this
> > question.
>
> Did you ask? What feedback did you get?
@banach-space I asked question on flang-slack, I mentioned the issue on the
latest Fl
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/96742
>From 5b487aac3c8414b6f37f6888f361ca7488094048 Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Fri, 21 Jun 2024 18:03:53 +0200
Subject: [PATCH 1/5] [Flang-new][OpenMP] Add offload related flags for AM
@@ -8024,7 +8024,7 @@ def source_date_epoch : Separate<["-"],
"source-date-epoch">,
// CUDA Options
//===--===//
-let Visibility = [CC1Option] in {
+let Visibility = [CC1Option, FC1Option] in {
-
DominikAdamski wrote:
> > Would it be possible for you to investigate that? It really shouldn't be
> > required if we can't help it.
>
> +1
Fixed in PR https://github.com/llvm/llvm-project/pull/99002
https://github.com/llvm/llvm-project/pull/96742
_
https://github.com/DominikAdamski edited
https://github.com/llvm/llvm-project/pull/96742
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -333,6 +333,9 @@ void Flang::AddAMDGPUTargetArgs(const ArgList &Args,
StringRef Val = A->getValue();
CmdArgs.push_back(Args.MakeArgString("-mcode-object-version=" + Val));
}
+
+ const ToolChain &TC = getToolChain();
+ TC.addClangTargetOptions(Args, CmdArgs, Actio
https://github.com/DominikAdamski closed
https://github.com/llvm/llvm-project/pull/99002
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/DominikAdamski created
https://github.com/llvm/llvm-project/pull/99002
-fcuda-is-device flag is not used for OpenMP offloading for AMD GPUs and it
does not need to be added as clang cc1 option for OpenMP code.
This PR has the same functionality as
https://github.com/llvm/ll
DominikAdamski wrote:
@jhuber6 I'm working on that.
https://github.com/llvm/llvm-project/pull/96742
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
DominikAdamski wrote:
@tblah Thanks for your review.
Unfortunately, I had to restore adding fcuda-is-device option (
https://github.com/llvm/llvm-project/pull/97531 ) because of regression related
to handling by clang virtual functions in OpenMP target region.
https://github.com/llvm/llvm-proj
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/96742
>From 5b487aac3c8414b6f37f6888f361ca7488094048 Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Fri, 21 Jun 2024 18:03:53 +0200
Subject: [PATCH 1/4] [Flang-new][OpenMP] Add offload related flags for AM
https://github.com/DominikAdamski closed
https://github.com/llvm/llvm-project/pull/97531
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/DominikAdamski created
https://github.com/llvm/llvm-project/pull/97531
Reverts llvm/llvm-project#96909
It breaks CI:
https://gitlab.e4s.io/uo-public/llvm-openmp-offloading/-/jobs/283716
>From f165b2279dc663cee7dc31d8213afb8c8fb48ab7 Mon Sep 17 00:00:00 2001
From: Dominik A
@@ -333,6 +333,9 @@ void Flang::AddAMDGPUTargetArgs(const ArgList &Args,
StringRef Val = A->getValue();
CmdArgs.push_back(Args.MakeArgString("-mcode-object-version=" + Val));
}
+
+ const ToolChain &TC = getToolChain();
+ TC.addClangTargetOptions(Args, CmdArgs, Actio
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/96742
>From 5b487aac3c8414b6f37f6888f361ca7488094048 Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Fri, 21 Jun 2024 18:03:53 +0200
Subject: [PATCH 1/3] [Flang-new][OpenMP] Add offload related flags for AM
https://github.com/DominikAdamski edited
https://github.com/llvm/llvm-project/pull/96742
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/DominikAdamski edited
https://github.com/llvm/llvm-project/pull/96742
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
DominikAdamski wrote:
Updated PR after: https://github.com/llvm/llvm-project/pull/96909/ .
Scope of changes:
`-fcuda-is-device` is not attached by OpenMP AMD GPU toolchain any more, so we
do not need to accept this flag by Flang-new. This flag remains HIP/CUDA
specific.
- OpenMP AMD GPU toolcha
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/96742
>From 5b487aac3c8414b6f37f6888f361ca7488094048 Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Fri, 21 Jun 2024 18:03:53 +0200
Subject: [PATCH 1/2] [Flang-new][OpenMP] Add offload related flags for AM
https://github.com/DominikAdamski closed
https://github.com/llvm/llvm-project/pull/96909
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/DominikAdamski created
https://github.com/llvm/llvm-project/pull/96909
`-fcuda-is-device` flag is not used for OpenMP offloading for AMD GPUs and it
does not need to be added as clang cc1 option for OpenMP code.
>From 6bd9a2e440dd61184bf7f4061ea78fa8ed31e17c Mon Sep 17 00:00
@@ -333,6 +333,9 @@ void Flang::AddAMDGPUTargetArgs(const ArgList &Args,
StringRef Val = A->getValue();
CmdArgs.push_back(Args.MakeArgString("-mcode-object-version=" + Val));
}
+
+ const ToolChain &TC = getToolChain();
+ TC.addClangTargetOptions(Args, CmdArgs, Actio
DominikAdamski wrote:
> > fcuda-is-device flag is not used by Flang currently. In the future it will
> > be needed for Flang equivalent functions:
> > AMDGPUTargetCodeGenInfo::getGlobalVarAddressSpace
> > AMDGPUTargetInfo::getTargetDefines .
>
> I don't follow - why would anything related to
@@ -333,6 +333,9 @@ void Flang::AddAMDGPUTargetArgs(const ArgList &Args,
StringRef Val = A->getValue();
CmdArgs.push_back(Args.MakeArgString("-mcode-object-version=" + Val));
}
+
+ const ToolChain &TC = getToolChain();
+ TC.addClangTargetOptions(Args, CmdArgs, Actio
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/96742
>From 5b487aac3c8414b6f37f6888f361ca7488094048 Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Fri, 21 Jun 2024 18:03:53 +0200
Subject: [PATCH] [Flang-new][OpenMP] Add offload related flags for AMDGPU
https://github.com/DominikAdamski edited
https://github.com/llvm/llvm-project/pull/96742
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/DominikAdamski created
https://github.com/llvm/llvm-project/pull/96742
Flang-new needs to add `mlink-builtin-bitcode` objects to properly support
offload code generation for AMD GPU.
fcuda-is-device flag is not used by Flang currently. In the future it will be
needed for Fl
https://github.com/DominikAdamski closed
https://github.com/llvm/llvm-project/pull/88190
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -345,7 +345,13 @@ void Flang::AddAMDGPUTargetArgs(const ArgList &Args,
// Check ROCm path if specified
const ToolChain &TC = getToolChain();
- TC.getDeviceLibs(Args);
+ std::string HIPVersion;
+ llvm::raw_string_ostream HIPInfo(HIPVersion);
+ TC.printVerboseInfo(HIP
@@ -201,3 +201,16 @@
! RUN: -nogpulibc %s 2>&1 \
! RUN: | FileCheck --check-prefix=NO-LIBC-GPU-AMDGPU %s
! NO-LIBC-GPU-AMDGPU-NOT: "-lcgpu-amdgpu"
+
+! RUN: rm -rf %t/Inputs
+! RUN: mkdir -p %t/Inputs
+! RUN: cp -r %S/../../../clang/test/Driver/Inputs/rocm %t/Input
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/88190
>From 44def17f36e8e27eb4232681e5ae7eff5de6d90f Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Tue, 9 Apr 2024 14:35:26 -0500
Subject: [PATCH 1/5] [Flang][AMDGPU] Add rocm-path flag
ROCm installation
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/88190
>From 44def17f36e8e27eb4232681e5ae7eff5de6d90f Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Tue, 9 Apr 2024 14:35:26 -0500
Subject: [PATCH 1/4] [Flang][AMDGPU] Add rocm-path flag
ROCm installation
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/88190
>From 44def17f36e8e27eb4232681e5ae7eff5de6d90f Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Tue, 9 Apr 2024 14:35:26 -0500
Subject: [PATCH 1/4] [Flang][AMDGPU] Add rocm-path flag
ROCm installation
@@ -345,7 +345,13 @@ void Flang::AddAMDGPUTargetArgs(const ArgList &Args,
// Check ROCm path if specified
const ToolChain &TC = getToolChain();
- TC.getDeviceLibs(Args);
+ std::string HIPVersion;
+ llvm::raw_string_ostream HIPInfo(HIPVersion);
+ TC.printVerboseInfo(HIP
https://github.com/DominikAdamski closed
https://github.com/llvm/llvm-project/pull/88280
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -345,7 +345,13 @@ void Flang::AddAMDGPUTargetArgs(const ArgList &Args,
// Check ROCm path if specified
const ToolChain &TC = getToolChain();
- TC.getDeviceLibs(Args);
+ std::string HIPVersion;
+ llvm::raw_string_ostream HIPInfo(HIPVersion);
+ TC.printVerboseInfo(HIP
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/88190
>From 44def17f36e8e27eb4232681e5ae7eff5de6d90f Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Tue, 9 Apr 2024 14:35:26 -0500
Subject: [PATCH 1/3] [Flang][AMDGPU] Add rocm-path flag
ROCm installation
DominikAdamski wrote:
> clang already tests this pretty well, but I'd still like to have at least one
> check to make sure that flang is actually putting -rpath in the linker.
Done. I added test.
https://github.com/llvm/llvm-project/pull/88280
___
cf
@@ -0,0 +1,32 @@
+// REQUIRES: x86-registered-target
DominikAdamski wrote:
Done
https://github.com/llvm/llvm-project/pull/88280
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mai
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/88280
>From 1330d076904d98a0a594700cca1c3e4a6b15dd58 Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Wed, 10 Apr 2024 09:42:09 -0500
Subject: [PATCH 1/3] [Flang] Add option frtlib-add-rpath
This option is
https://github.com/DominikAdamski edited
https://github.com/llvm/llvm-project/pull/88280
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/DominikAdamski edited
https://github.com/llvm/llvm-project/pull/88280
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/88280
>From 1330d076904d98a0a594700cca1c3e4a6b15dd58 Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Wed, 10 Apr 2024 09:42:09 -0500
Subject: [PATCH 1/2] [Flang] Add option frtlib-add-rpath
This option is
https://github.com/DominikAdamski created
https://github.com/llvm/llvm-project/pull/88280
This option is used by clang and should also be visible in flang. It is already
handled by the toolchains used by both clang and flang.
Reported issue: https://github.com/llvm/llvm-project/issues/82553
>
@@ -345,7 +345,13 @@ void Flang::AddAMDGPUTargetArgs(const ArgList &Args,
// Check ROCm path if specified
const ToolChain &TC = getToolChain();
- TC.getDeviceLibs(Args);
+ std::string HIPVersion;
+ llvm::raw_string_ostream HIPInfo(HIPVersion);
+ TC.printVerboseInfo(HIP
@@ -342,6 +342,10 @@ void Flang::AddAMDGPUTargetArgs(const ArgList &Args,
StringRef Val = A->getValue();
CmdArgs.push_back(Args.MakeArgString("-mcode-object-version=" + Val));
}
+
+ // Check ROCm path if specified
+ const ToolChain &TC = getToolChain();
+ TC.getDev
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/88190
>From 44def17f36e8e27eb4232681e5ae7eff5de6d90f Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Tue, 9 Apr 2024 14:35:26 -0500
Subject: [PATCH 1/3] [Flang][AMDGPU] Add rocm-path flag
ROCm installation
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/88190
>From 44def17f36e8e27eb4232681e5ae7eff5de6d90f Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Tue, 9 Apr 2024 14:35:26 -0500
Subject: [PATCH 1/2] [Flang][AMDGPU] Add rocm-path flag
ROCm installation
@@ -201,3 +201,24 @@
! RUN: -nogpulibc %s 2>&1 \
! RUN: | FileCheck --check-prefix=NO-LIBC-GPU-AMDGPU %s
! NO-LIBC-GPU-AMDGPU-NOT: "-lcgpu-amdgpu"
+
+! RUN: rm -rf %t/Inputs
+
+! RUN: not %flang -### -v --target=x86_64-unknown-linux-gnu -fopenmp \
+! RUN: --off
@@ -1340,7 +1340,8 @@ def hip_link : Flag<["--"], "hip-link">,
Group,
HelpText<"Link clang-offload-bundler bundles for HIP">;
def no_hip_rt: Flag<["-"], "no-hip-rt">, Group,
HelpText<"Do not link against HIP runtime libraries">;
-def rocm_path_EQ : Joined<["--"], "rocm-pat
@@ -342,6 +342,10 @@ void Flang::AddAMDGPUTargetArgs(const ArgList &Args,
StringRef Val = A->getValue();
CmdArgs.push_back(Args.MakeArgString("-mcode-object-version=" + Val));
}
+
+ // Check ROCm path if specified
+ const ToolChain &TC = getToolChain();
+ TC.getDev
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/88190
>From 8782af25c5946dc33342798c36c7d64569d16ab5 Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Tue, 9 Apr 2024 14:35:26 -0500
Subject: [PATCH 1/2] [Flang][AMDGPU] Add rocm-path flag
ROCm installation
https://github.com/DominikAdamski created
https://github.com/llvm/llvm-project/pull/88190
ROCm installation path is used for finding and automatically linking required
bitcode libraries for OpenMP AMDGPU offload.
>From 8782af25c5946dc33342798c36c7d64569d16ab5 Mon Sep 17 00:00:00 2001
From: Dom
https://github.com/DominikAdamski closed
https://github.com/llvm/llvm-project/pull/71045
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/DominikAdamski closed
https://github.com/llvm/llvm-project/pull/73944
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -0,0 +1,31 @@
+//=== AMDGPUAddrSpace.h -*- C++
-*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM
Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apa
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/73944
>From 60ceda3d1025891f5037f020a2efe35108f62ca3 Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Thu, 30 Nov 2023 08:06:12 -0600
Subject: [PATCH 1/4] [NFC][AMDGPU] Move address space enum to LLVM direct
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/73944
>From 60ceda3d1025891f5037f020a2efe35108f62ca3 Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Thu, 30 Nov 2023 08:06:12 -0600
Subject: [PATCH 1/4] [NFC][AMDGPU] Move address space enum to LLVM direct
@@ -0,0 +1,31 @@
+//=== AMDGPUAddrSpace.h -*- C++
-*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM
Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apa
DominikAdamski wrote:
@lenary Thank you for your input.
@kparzysz @lenary Shall I add all address spaces which are mentioned in [other
LLVM
header](https://github.com/llvm/llvm-project/blob/main/llvm/lib/Target/AMDGPU/AMDGPU.h#L395-L456)?
Currently I added address spaces which were mentioned i
DominikAdamski wrote:
The address spaces for AMDGPU defined
[here](https://github.com/llvm/llvm-project/blob/main/llvm/lib/Target/AMDGPU/AMDGPU.h#L395-L456)
contain more types of address spaces in comparison to the enum defined in
clang. Is it ok to extend number of address space types for cla
@@ -31,6 +31,15 @@ class Triple;
// back-end to TableGen to create these clean tables.
namespace AMDGPU {
+/// Address space values for AMD GPUs
+enum AddrSpace {
+ Generic = 0,
DominikAdamski wrote:
Done. Moved to enum class.
https://github.com/llvm/llvm-p
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/73944
>From 60ceda3d1025891f5037f020a2efe35108f62ca3 Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Thu, 30 Nov 2023 08:06:12 -0600
Subject: [PATCH 1/2] [NFC][AMDGPU] Move address space enum to LLVM direct
Author: Dominik Adamski
Date: 2023-11-29T03:01:01-06:00
New Revision: 95943d2fab7e6f8dcea216df2d56a0512201b467
URL:
https://github.com/llvm/llvm-project/commit/95943d2fab7e6f8dcea216df2d56a0512201b467
DIFF:
https://github.com/llvm/llvm-project/commit/95943d2fab7e6f8dcea216df2d56a0512201b467.dif
Author: Dominik Adamski
Date: 2023-11-28T13:18:46-06:00
New Revision: f00ffcdb58d6db902a8f86b0ce83a03874d113ad
URL:
https://github.com/llvm/llvm-project/commit/f00ffcdb58d6db902a8f86b0ce83a03874d113ad
DIFF:
https://github.com/llvm/llvm-project/commit/f00ffcdb58d6db902a8f86b0ce83a03874d113ad.dif
https://github.com/DominikAdamski closed
https://github.com/llvm/llvm-project/pull/72638
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/72638
>From 1cc1d9d3480f750980ea0a395cca3b202a606f57 Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Fri, 17 Nov 2023 03:02:49 -0600
Subject: [PATCH 1/2] [Flang] Add code-object-version option
Information
@@ -85,6 +85,19 @@ class CodeGenOptions : public CodeGenOptionsBase {
RK_WithPattern, // Remark pattern specified via '-Rgroup=regexp'.
};
+ /// \brief Enumeration value for AMDGPU code object version, which is the
+ /// code object version times 100.
+ enum class Cod
@@ -264,6 +263,37 @@ static void addDepdendentLibs(mlir::ModuleOp &mlirModule,
}
}
+// Add to MLIR code target specific items which are dependent on target
+// configuration specified by the user
+static void addTargetSpecificMLIRItems(mlir::ModuleOp &mlirModule,
+
@@ -85,6 +85,19 @@ class CodeGenOptions : public CodeGenOptionsBase {
RK_WithPattern, // Remark pattern specified via '-Rgroup=regexp'.
};
+ /// \brief Enumeration value for AMDGPU code object version, which is the
+ /// code object version times 100.
+ enum class Cod
@@ -264,6 +263,37 @@ static void addDepdendentLibs(mlir::ModuleOp &mlirModule,
}
}
+// Add to MLIR code target specific items which are dependent on target
+// configuration specified by the user
+static void addTargetSpecificMLIRItems(mlir::ModuleOp &mlirModule,
+
DominikAdamski wrote:
Done
https://github.com/llvm/llvm-project/pull/72638
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -264,6 +263,37 @@ static void addDepdendentLibs(mlir::ModuleOp &mlirModule,
}
}
+// Add to MLIR code target specific items which are dependent on target
+// configuration specified by the user
+static void addTargetSpecificMLIRItems(mlir::ModuleOp &mlirModule,
+
@@ -264,6 +263,37 @@ static void addDepdendentLibs(mlir::ModuleOp &mlirModule,
}
}
+// Add to MLIR code target specific items which are dependent on target
+// configuration specified by the user
+static void addTargetSpecificMLIRItems(mlir::ModuleOp &mlirModule,
+
@@ -264,6 +263,37 @@ static void addDepdendentLibs(mlir::ModuleOp &mlirModule,
}
}
+// Add to MLIR code target specific items which are dependent on target
+// configuration specified by the user
+static void addTargetSpecificMLIRItems(mlir::ModuleOp &mlirModule,
+
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/72638
>From e5d339c24193e4e37013b3b25460009418d6ce6d Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Fri, 17 Nov 2023 03:02:49 -0600
Subject: [PATCH 1/2] [Flang] Add code-object-version option
Information
https://github.com/DominikAdamski updated
https://github.com/llvm/llvm-project/pull/72638
>From e5d339c24193e4e37013b3b25460009418d6ce6d Mon Sep 17 00:00:00 2001
From: Dominik Adamski
Date: Fri, 17 Nov 2023 03:02:49 -0600
Subject: [PATCH] [Flang] Add code-object-version option
Information abou
https://github.com/DominikAdamski created
https://github.com/llvm/llvm-project/pull/72638
Information about code object version can be configured by the user for AMD GPU
target and it needs to be placed in LLVM IR generated by Flang.
Information about code object version in MLIR generated by t
Author: Dominik Adamski
Date: 2023-11-03T05:13:06-05:00
New Revision: 2b1948c2be0e935bdd2e764ae1a2b84c405fdc04
URL:
https://github.com/llvm/llvm-project/commit/2b1948c2be0e935bdd2e764ae1a2b84c405fdc04
DIFF:
https://github.com/llvm/llvm-project/commit/2b1948c2be0e935bdd2e764ae1a2b84c405fdc04.dif
Author: Dominik Adamski
Date: 2023-11-03T05:13:06-05:00
New Revision: 2b1948c2be0e935bdd2e764ae1a2b84c405fdc04
URL:
https://github.com/llvm/llvm-project/commit/2b1948c2be0e935bdd2e764ae1a2b84c405fdc04
DIFF:
https://github.com/llvm/llvm-project/commit/2b1948c2be0e935bdd2e764ae1a2b84c405fdc04.dif
https://github.com/DominikAdamski created
https://github.com/llvm/llvm-project/pull/71045
If -nogpulib option is passed by the user, then the OpenMP device runtime is
not used and we should not emit globals to configure debugging at compile-time
for the device runtime.
Link to -nogpulib flag
@@ -1126,6 +1133,185 @@ void OpenMPIRBuilder::emitCancelationCheckImpl(Value
*CancelFlag,
Builder.SetInsertPoint(NonCancellationBlock, NonCancellationBlock->begin());
}
+// Callback used to create OpenMP runtime calls to support
+// omp parallel clause for the device.
+// W
@@ -1126,6 +1133,185 @@ void OpenMPIRBuilder::emitCancelationCheckImpl(Value
*CancelFlag,
Builder.SetInsertPoint(NonCancellationBlock, NonCancellationBlock->begin());
}
+// Callback used to create OpenMP runtime calls to support
+// omp parallel clause for the device.
+// W
1 - 100 of 119 matches
Mail list logo
