https://github.com/frasercrmck created
https://github.com/llvm/llvm-project/pull/85634
Having a mix of tabs and spaces makes the diff of any changes to the build
system noisier than necessary. This commit unifies them to two spaces.
This includes some minor cosmetic changes such as with joinin
frasercrmck wrote:
@tstellar are you the code owner for `libclc`?
https://github.com/llvm/llvm-project/pull/85634
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
frasercrmck wrote:
FWIW I don't think the build bots are actually building `libclc`.
Locally, at least, it's not sufficient to specify `LLVM_ENABLE_PROJECTS=libclc`
- you _also_ need `LLVM_EXTERNAL_PROJECTS=libclc`. Even then, it doesn't build
for me in-tree, because it assumes `CMAKE_SOURCE_D
https://github.com/frasercrmck closed
https://github.com/llvm/llvm-project/pull/85634
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
frasercrmck wrote:
> > FWIW I don't think the build bots are actually building `libclc`.
> > Locally, at least, it's not sufficient to specify
> > `LLVM_ENABLE_PROJECTS=libclc` - you _also_ need
> > `LLVM_EXTERNAL_PROJECTS=libclc`. Even then, it doesn't build for me
> > in-tree, because it ass
Author: Fraser Cormack
Date: 2024-03-20T11:43:38Z
New Revision: 4b369ff076d9205ddbcfcdef89276a737a83ce0e
URL:
https://github.com/llvm/llvm-project/commit/4b369ff076d9205ddbcfcdef89276a737a83ce0e
DIFF:
https://github.com/llvm/llvm-project/commit/4b369ff076d9205ddbcfcdef89276a737a83ce0e.diff
LOG
https://github.com/frasercrmck created
https://github.com/llvm/llvm-project/pull/86945
The libclc project is currently only properly supported as an external project.
However, when trying to get it to also build in-tree, the CMake configuration
messages it outputs stand out amongst the rest of
https://github.com/frasercrmck closed
https://github.com/llvm/llvm-project/pull/86945
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/frasercrmck created
https://github.com/llvm/llvm-project/pull/86965
This commit fixes the problem of missing build dependencies between libclc
source files and their various includes (namely headers and .inc files).
We would like to do this with compiler-generated dependency
frasercrmck wrote:
> The build here seems to be trying to define clc as a language, which then
> results in needing to rely on language support magic like this. I think it
> would be better if this did what rocm-device-libs does and treat these as
> custom targets. I don't think it's particula
https://github.com/frasercrmck closed
https://github.com/llvm/llvm-project/pull/86965
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Author: Erik Tomusk
Date: 2020-11-27T13:16:39Z
New Revision: 7793db35ca2c1fe63687c0a7140cbec540a9aded
URL:
https://github.com/llvm/llvm-project/commit/7793db35ca2c1fe63687c0a7140cbec540a9aded
DIFF:
https://github.com/llvm/llvm-project/commit/7793db35ca2c1fe63687c0a7140cbec540a9aded.diff
LOG: [
Author: Fraser Cormack
Date: 2023-08-31T14:28:16+01:00
New Revision: 37a3de1e2eedf848b8442217ef3790436f69a7db
URL:
https://github.com/llvm/llvm-project/commit/37a3de1e2eedf848b8442217ef3790436f69a7db
DIFF:
https://github.com/llvm/llvm-project/commit/37a3de1e2eedf848b8442217ef3790436f69a7db.diff
Author: Fraser Cormack
Date: 2024-04-04T17:56:40+01:00
New Revision: 338ecfbac351e4b211836ad73d75bbdf729e9134
URL:
https://github.com/llvm/llvm-project/commit/338ecfbac351e4b211836ad73d75bbdf729e9134
DIFF:
https://github.com/llvm/llvm-project/commit/338ecfbac351e4b211836ad73d75bbdf729e9134.diff
Author: Fraser Cormack
Date: 2024-04-08T11:05:38+01:00
New Revision: 8461d901a770516cf2069fe3bce979a6f8fc8d76
URL:
https://github.com/llvm/llvm-project/commit/8461d901a770516cf2069fe3bce979a6f8fc8d76
DIFF:
https://github.com/llvm/llvm-project/commit/8461d901a770516cf2069fe3bce979a6f8fc8d76.diff
frasercrmck wrote:
> This change broke standalone build against LLVM dylib — it now insists on
> linking to non-existing static libraries:
>
> ```
> FAILED: prepare_builtins
> : && /usr/lib/ccache/bin/x86_64-pc-linux-gnu-g++ -march=znver2
> --param=l1-cache-size=32 --param=l1-cache-line-size=
@@ -27,31 +36,51 @@ set( LIBCLC_TARGETS_TO_BUILD "all"
option( ENABLE_RUNTIME_SUBNORMAL "Enable runtime linking of subnormal
support." OFF )
-find_package(LLVM REQUIRED HINTS "${LLVM_CMAKE_DIR}")
-include(AddLLVM)
+if(CMAKE_SOURCE_DIR STREQUAL CMAKE_CURRENT_SOURCE_DIR)
-
@@ -358,3 +399,7 @@ foreach( t ${LIBCLC_TARGETS_TO_BUILD} )
endif()
endforeach( d )
endforeach( t )
+
+if( NOT LIBCLC_STANDALONE_BUILD )
+ add_subdirectory( test )
frasercrmck wrote:
Ach yeah sorry, I have another path locally which adds some LIT tests
https://github.com/frasercrmck updated
https://github.com/llvm/llvm-project/pull/87622
>From b87c2862491aed03f883d6b124a845b0d7c47699 Mon Sep 17 00:00:00 2001
From: Fraser Cormack
Date: Wed, 3 Apr 2024 16:09:30 +0100
Subject: [PATCH] [libclc] Refactor build system to allow in-tree builds
The p
https://github.com/frasercrmck edited
https://github.com/llvm/llvm-project/pull/87622
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/frasercrmck created
https://github.com/llvm/llvm-project/pull/87989
None
>From b87c2862491aed03f883d6b124a845b0d7c47699 Mon Sep 17 00:00:00 2001
From: Fraser Cormack
Date: Wed, 3 Apr 2024 16:09:30 +0100
Subject: [PATCH 1/2] [libclc] Refactor build system to allow in-tree bui
frasercrmck wrote:
> ```
> set( spv_suffix ${arch_suffix}.spv )
> add_custom_command( OUTPUT "${spv_suffix}"
> COMMAND ${LLVM_SPIRV} ${spvflags} -o "${spv_suffix}"
> ${builtins_opt_lib_tgt}
> DEPENDS ${builtins_link_lib_tgt} )
> ```
>
> I think that ought to be `DEPE
frasercrmck wrote:
> > > ```
> > > set( spv_suffix ${arch_suffix}.spv )
> > > add_custom_command( OUTPUT "${spv_suffix}"
> > > COMMAND ${LLVM_SPIRV} ${spvflags} -o "${spv_suffix}"
> > > ${builtins_opt_lib_tgt}
> > > DEPENDS ${builtins_link_lib_tgt} )
> > > ```
> > >
Author: Fraser Cormack
Date: 2024-04-08T15:48:15+01:00
New Revision: f46f6465062bd6ddc96e3838c50e1a0f85f92dd4
URL:
https://github.com/llvm/llvm-project/commit/f46f6465062bd6ddc96e3838c50e1a0f85f92dd4
DIFF:
https://github.com/llvm/llvm-project/commit/f46f6465062bd6ddc96e3838c50e1a0f85f92dd4.diff
frasercrmck wrote:
> > > > ```
> > > > set( spv_suffix ${arch_suffix}.spv )
> > > > add_custom_command( OUTPUT "${spv_suffix}"
> > > > COMMAND ${LLVM_SPIRV} ${spvflags} -o "${spv_suffix}"
> > > > ${builtins_opt_lib_tgt}
> > > > DEPENDS ${builtins_link_lib_tgt} )
> > >
https://github.com/frasercrmck updated
https://github.com/llvm/llvm-project/pull/87989
>From b87c2862491aed03f883d6b124a845b0d7c47699 Mon Sep 17 00:00:00 2001
From: Fraser Cormack
Date: Wed, 3 Apr 2024 16:09:30 +0100
Subject: [PATCH 1/4] [libclc] Refactor build system to allow in-tree builds
T
Author: Fraser Cormack
Date: 2024-04-08T16:51:30+01:00
New Revision: b439140e2982dd77ef28a9069e16ae77bbe2bc5a
URL:
https://github.com/llvm/llvm-project/commit/b439140e2982dd77ef28a9069e16ae77bbe2bc5a
DIFF:
https://github.com/llvm/llvm-project/commit/b439140e2982dd77ef28a9069e16ae77bbe2bc5a.diff
frasercrmck wrote:
> Thanks. Unfortunately, I'm still getting a build failure:
>
> ```
> [1792/1922] cd /tmp/portage/dev-libs/libclc-19.0.0./work/libclc_build &&
> /usr/lib/llvm/17/bin/llvm-spirv --spirv-max-version=1.1 -o spirv-mesa3d-.spv
> builtins.link.spirv-mesa3d-
> FAILED: spirv-mes
https://github.com/frasercrmck updated
https://github.com/llvm/llvm-project/pull/87622
>From 47cd6ddbf292fd8a50534e091ac64ebee3c40ee5 Mon Sep 17 00:00:00 2001
From: Fraser Cormack
Date: Wed, 3 Apr 2024 16:09:30 +0100
Subject: [PATCH] [libclc] Refactor build system to allow in-tree builds
The p
frasercrmck wrote:
CC @mgorny - I was unsure whether you are actively using and interested
`libclc` or you were just looking after build bots or some other reason. I've
successfully built the SPIR-V libraries out-of-tree but another pair of eyes
can't hurt.
I was wondering - if we `libclc` b
https://github.com/frasercrmck updated
https://github.com/llvm/llvm-project/pull/87622
>From 47cd6ddbf292fd8a50534e091ac64ebee3c40ee5 Mon Sep 17 00:00:00 2001
From: Fraser Cormack
Date: Wed, 3 Apr 2024 16:09:30 +0100
Subject: [PATCH 1/2] [libclc] Refactor build system to allow in-tree builds
T
https://github.com/frasercrmck updated
https://github.com/llvm/llvm-project/pull/87622
>From 47cd6ddbf292fd8a50534e091ac64ebee3c40ee5 Mon Sep 17 00:00:00 2001
From: Fraser Cormack
Date: Wed, 3 Apr 2024 16:09:30 +0100
Subject: [PATCH 1/3] [libclc] Refactor build system to allow in-tree builds
T
https://github.com/frasercrmck updated
https://github.com/llvm/llvm-project/pull/87622
>From 47cd6ddbf292fd8a50534e091ac64ebee3c40ee5 Mon Sep 17 00:00:00 2001
From: Fraser Cormack
Date: Wed, 3 Apr 2024 16:09:30 +0100
Subject: [PATCH 1/4] [libclc] Refactor build system to allow in-tree builds
T
https://github.com/frasercrmck closed
https://github.com/llvm/llvm-project/pull/87622
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
frasercrmck wrote:
Hmm, isn't CMake fun? Thanks for looking into this.
Unfortunately the file-level dependencies are there for a reason. With this
patch, if I `rm builtins.link.nvptx64--.bc` and build `builtins.opt.nvptx64--`
(with `ninja`), it builds the `link` target but doesn't actually reb
https://github.com/frasercrmck approved this pull request.
LGTM, thank you!
https://github.com/llvm/llvm-project/pull/95018
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
frasercrmck wrote:
> Yes, the article you link to is highly relevant. Without this PR we're
> hitting the problem described in example #4.
>
> We can promote the issue to a build failure by choosing a custom command
> which can't run concurrently with itself:
>
> ```cmake
> cmake_minimum_requ
https://github.com/frasercrmck updated
https://github.com/llvm/llvm-project/pull/87989
>From c53dc3564822f305d09d19dc3b7327b33028fe11 Mon Sep 17 00:00:00 2001
From: Fraser Cormack
Date: Thu, 4 Apr 2024 17:49:13 +0100
Subject: [PATCH] [libclc] Add initial LIT tests
---
libclc/CMakeLists.txt
Author: Fraser Cormack
Date: 2024-04-16T15:20:26+01:00
New Revision: a0f8191af3945482f0f7a7c8f030e8c519a795b7
URL:
https://github.com/llvm/llvm-project/commit/a0f8191af3945482f0f7a7c8f030e8c519a795b7
DIFF:
https://github.com/llvm/llvm-project/commit/a0f8191af3945482f0f7a7c8f030e8c519a795b7.diff
Author: Fraser Cormack
Date: 2024-04-16T15:20:26+01:00
New Revision: 3d118f92081ea0c7048749dc5d08c8e8217be4eb
URL:
https://github.com/llvm/llvm-project/commit/3d118f92081ea0c7048749dc5d08c8e8217be4eb
DIFF:
https://github.com/llvm/llvm-project/commit/3d118f92081ea0c7048749dc5d08c8e8217be4eb.diff
Author: Fraser Cormack
Date: 2024-04-16T16:50:14+01:00
New Revision: 9d111286322ec99b32467eef3aeec6b588c49f18
URL:
https://github.com/llvm/llvm-project/commit/9d111286322ec99b32467eef3aeec6b588c49f18
DIFF:
https://github.com/llvm/llvm-project/commit/9d111286322ec99b32467eef3aeec6b588c49f18.diff
https://github.com/frasercrmck created
https://github.com/llvm/llvm-project/pull/88922
Building the libclc project in-tree with debug tools can be very slow.
This commit adds an option for a user to specify a dierctory from which
to import (e.g., release-built) tools. All tools required by the p
https://github.com/frasercrmck updated
https://github.com/llvm/llvm-project/pull/87989
>From 9048dbc6547fa14c78649d02f27d68fc27ca7e70 Mon Sep 17 00:00:00 2001
From: Fraser Cormack
Date: Thu, 4 Apr 2024 17:49:13 +0100
Subject: [PATCH] [libclc] Add initial LIT tests
These tests aren't very meani
https://github.com/frasercrmck edited
https://github.com/llvm/llvm-project/pull/87989
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/frasercrmck edited
https://github.com/llvm/llvm-project/pull/87989
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/frasercrmck updated
https://github.com/llvm/llvm-project/pull/88922
>From bff3b97b599e1c9529a95ead9d61f9ecc9cad129 Mon Sep 17 00:00:00 2001
From: Fraser Cormack
Date: Tue, 16 Apr 2024 15:30:20 +0100
Subject: [PATCH 1/5] [libclc] Convert llvm-spirv to imported executable
This
@@ -1,3 +1,6 @@
-__kernel void foo(int *i) {
+// RUN: %clang -emit-llvm -S -o - %s | FileCheck %s
frasercrmck wrote:
Yes, I was hoping to explore in that direction. I've kind of just copied the
skeleton of the LIT tests we have downstream, but as you say, this i
@@ -1,3 +1,6 @@
-__kernel void foo(int *i) {
+// RUN: %clang -emit-llvm -S -o - %s | FileCheck %s
frasercrmck wrote:
Ach I misunderstood how `update_cc_test_checks` works. It doesn't actually go
through the regular LIT infrastructure, so we can't do anything inv
@@ -1,3 +1,6 @@
-__kernel void foo(int *i) {
+// RUN: %clang -emit-llvm -S -o - %s | FileCheck %s
frasercrmck wrote:
I wasn't so much thinking that it would pretend to be generic, but rather that
it'd use a custom test format (something like
[analyzer_test.py](
@@ -1,3 +1,6 @@
-__kernel void foo(int *i) {
+// RUN: %clang -emit-llvm -S -o - %s | FileCheck %s
frasercrmck wrote:
Fair enough, I hadn't really appreciated that `update_cc_test_checks` was
orthogonal to all of these other LIT concepts like `REQUIRES` and featu
https://github.com/frasercrmck created
https://github.com/llvm/llvm-project/pull/89147
Commit #87622 broke the build. Ninja was happy with creating the output
directories as necessary, but Unix Makefiles isn't. Ensure they are always
created.
Fixes #88626.
>From 7bd2001064a45c87ab9094a041a95
Author: Fraser Cormack
Date: 2024-04-18T07:01:13+01:00
New Revision: 06f54e7c2b4f26fbcad906a6b4aa2e46bf60b8af
URL:
https://github.com/llvm/llvm-project/commit/06f54e7c2b4f26fbcad906a6b4aa2e46bf60b8af
DIFF:
https://github.com/llvm/llvm-project/commit/06f54e7c2b4f26fbcad906a6b4aa2e46bf60b8af.diff
Author: Fraser Cormack
Date: 2024-04-18T07:01:13+01:00
New Revision: 93d51194b1df8229268953ec94063fc4194a320b
URL:
https://github.com/llvm/llvm-project/commit/93d51194b1df8229268953ec94063fc4194a320b
DIFF:
https://github.com/llvm/llvm-project/commit/93d51194b1df8229268953ec94063fc4194a320b.diff
Author: Fraser Cormack
Date: 2024-04-18T07:01:13+01:00
New Revision: 0aeeff3059e79b86f55ed92a4488bdee8fa66e12
URL:
https://github.com/llvm/llvm-project/commit/0aeeff3059e79b86f55ed92a4488bdee8fa66e12
DIFF:
https://github.com/llvm/llvm-project/commit/0aeeff3059e79b86f55ed92a4488bdee8fa66e12.diff
Author: Fraser Cormack
Date: 2024-04-18T07:01:13+01:00
New Revision: dc74c69f016fa65949bf0cf691061e12bef93d29
URL:
https://github.com/llvm/llvm-project/commit/dc74c69f016fa65949bf0cf691061e12bef93d29
DIFF:
https://github.com/llvm/llvm-project/commit/dc74c69f016fa65949bf0cf691061e12bef93d29.diff
https://github.com/frasercrmck updated
https://github.com/llvm/llvm-project/pull/89147
>From 0f102a202e4a80fbc106303df5005a2b7922b0cf Mon Sep 17 00:00:00 2001
From: Fraser Cormack
Date: Wed, 17 Apr 2024 22:28:03 +0100
Subject: [PATCH] [libclc] Fix build with Unix Makefiles
Commit #87622 broke
https://github.com/frasercrmck closed
https://github.com/llvm/llvm-project/pull/89147
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -39,6 +39,10 @@ function(compile_to_bc)
set( TARGET_ARG "-target" ${ARG_TRIPLE} )
endif()
+ # Ensure the directory we are told to output to exists
+ get_filename_component( ARG_OUTPUT_DIR ${ARG_OUTPUT} DIRECTORY )
frasercrmck wrote:
Yeah, there ver
https://github.com/frasercrmck created
https://github.com/llvm/llvm-project/pull/89756
We've recently seen the libclc llvm-link invocations become so long that they
exceed the character limits on certain platforms.
Using a 'response file' should solve this by offloading the list of inputs into
@@ -88,10 +88,25 @@ function(link_bc)
${ARGN}
)
+ set( LINK_INPUT_ARG ${ARG_INPUTS} )
+ if( WIN32 OR CYGWIN )
+# Create a response file in case the number of inputs exceeds command-line
+# character limits on certain platforms.
+file( TO_CMAKE_PATH ${LIBCLC
https://github.com/frasercrmck approved this pull request.
LGTM, thanks for the screenshot.
https://github.com/llvm/llvm-project/pull/89746
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-comm
https://github.com/frasercrmck approved this pull request.
LGTM, thanks
https://github.com/llvm/llvm-project/pull/98149
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/frasercrmck edited
https://github.com/llvm/llvm-project/pull/98149
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -181,3 +181,15 @@ _CLC_OVERLOAD _CLC_DEF double atan(double x)
_CLC_UNARY_VECTORIZE(_CLC_OVERLOAD _CLC_DEF, double, atan, double);
#endif // cl_khr_fp64
+
+#ifdef cl_khr_fp16
frasercrmck wrote:
Can this be `_CLC_DEFINE_UNARY_BUILTIN_FP16(atan)`?
https://g
frasercrmck wrote:
> Also, just leave that thought here: A next step might be to use `native`
> functions instead (when available). They should provide enough precision for
> half, but be more performance. Also as the accuracy is undefined for those
> functions, maybe it should be an opt-in op
frasercrmck wrote:
@rjodinchr I was about to merge and noticed your email is hidden. I just wanted
to check with you whether you want to make it public, as per the discussion
https://discourse.llvm.org/t/hidden-emails-on-github-should-we-do-something-about-it/74223/67.
AFAICT there's still no
https://github.com/frasercrmck closed
https://github.com/llvm/llvm-project/pull/98149
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/frasercrmck approved this pull request.
https://github.com/llvm/llvm-project/pull/99481
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/frasercrmck approved this pull request.
LGTM
https://github.com/llvm/llvm-project/pull/99841
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/frasercrmck closed
https://github.com/llvm/llvm-project/pull/99841
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/frasercrmck closed
https://github.com/llvm/llvm-project/pull/89756
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -1,3 +1,6 @@
-__kernel void foo(int *i) {
+// RUN: %clang -emit-llvm -S -o - %s | FileCheck %s
frasercrmck wrote:
That's true, but I'm still sceptical that we can make the tests simple enough
for `update_cc_test_checks` to be viable. We'd have to embed libclc
frasercrmck wrote:
> @frasercrmck Could you review it please?
Sure, will do, thanks for this. We've done something similar downstream so I
can hopefully verify against what we've done.
https://github.com/llvm/llvm-project/pull/98149
___
cfe-commits m
https://github.com/frasercrmck closed
https://github.com/llvm/llvm-project/pull/95018
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -394,6 +401,7 @@ foreach( t ${LIBCLC_TARGETS_TO_BUILD} )
DEPENDS ${builtins_link_lib}
)
add_custom_target( "prepare-${spv_suffix}" ALL DEPENDS "${spv_suffix}" )
+ set_target_properties( "prepare-${spv_suffix}" PROPERTIES FOLDER
"libclc/Sourcegenning
frasercrmck wrote:
> > I'm not sure why we would ever want the current default if this is an
> > option. I'm trying to see it, but I can't work out a case where a 64bit
> > pointer would make sense, since the even tens-of-thousands of money
> > supercomputer cards have less than 256KiB of addr
https://github.com/frasercrmck edited
https://github.com/llvm/llvm-project/pull/111682
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/frasercrmck closed
https://github.com/llvm/llvm-project/pull/111682
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/frasercrmck approved this pull request.
LGTM, thanks
https://github.com/llvm/llvm-project/pull/111369
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -137,10 +137,6 @@ if( llvm-spirv_exe )
list( APPEND LIBCLC_TARGETS_ALL spirv-mesa3d- spirv64-mesa3d- )
endif()
-if( LIBCLC_TARGETS_TO_BUILD STREQUAL "all" )
- set( LIBCLC_TARGETS_TO_BUILD ${LIBCLC_TARGETS_ALL} )
-endif()
-
list( SORT LIBCLC_TARGETS_TO_BUILD )
-
https://github.com/frasercrmck approved this pull request.
https://github.com/llvm/llvm-project/pull/111549
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/frasercrmck created
https://github.com/llvm/llvm-project/pull/111682
When padded -nocudalib/-nogpulib, Cuda's argument handling would bail out
before handling -fcuda-short-ptr, meaning the frontend and backend data layouts
would mismatch.
>From 67b4b2acf2a54a3a6e2d765714f32
frasercrmck wrote:
> Seems reasonable, which architectures require this? I know that NVIDIA
> deprecated the 32-bit `nvptx` target in CUDA 12 or something.
I'm not an expert on CUDA but, AFAICT, even in 64-bit CUDA, certain pointers
such as those pointing to shared memory are 32 bit, because t
frasercrmck wrote:
> > > Seems reasonable, which architectures require this? I know that NVIDIA
> > > deprecated the 32-bit `nvptx` target in CUDA 12 or something.
> >
> >
> > I'm not an expert on CUDA but, AFAICT, even in 64-bit CUDA, certain
> > pointers such as those pointing to shared mem
frasercrmck wrote:
> > > I'm not sure why we would ever want the current default if this is an
> > > option. I'm trying to see it, but I can't work out a case where a 64bit
> > > pointer would make sense, since the even tens-of-thousands of money
> > > supercomputer cards have less than 256KiB
frasercrmck wrote:
> > CC @rjodinchr. I realise now that this idea may prove problematic for
> > `clspv`/`clspv64` targets. If this idea were to taken further, things like
> > OpenCL `minmag` would call `__clc_minmag` which would call `__clc_fabs`
> > (e.g.), but I notice `fabs` is not impleme
https://github.com/frasercrmck closed
https://github.com/llvm/llvm-project/pull/109926
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/frasercrmck commented:
Thanks for doing this! I've had something like this at the back of my mind for
a while.
https://github.com/llvm/llvm-project/pull/111369
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.ll
@@ -23,29 +23,41 @@ generic implementations of most library requirements,
allowing the target
to override the generic implementation at the granularity of individual
functions.
-libclc currently only supports the PTX target, but support for more
+libclc currently supports NVP
https://github.com/frasercrmck edited
https://github.com/llvm/llvm-project/pull/111369
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -23,29 +23,41 @@ generic implementations of most library requirements,
allowing the target
to override the generic implementation at the granularity of individual
functions.
-libclc currently only supports the PTX target, but support for more
+libclc currently supports NVP
@@ -23,29 +23,41 @@ generic implementations of most library requirements,
allowing the target
to override the generic implementation at the granularity of individual
functions.
-libclc currently only supports the PTX target, but support for more
+libclc currently supports NVP
@@ -23,29 +23,41 @@ generic implementations of most library requirements,
allowing the target
to override the generic implementation at the granularity of individual
functions.
-libclc currently only supports the PTX target, but support for more
+libclc currently supports NVP
https://github.com/frasercrmck approved this pull request.
https://github.com/llvm/llvm-project/pull/111371
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/frasercrmck updated
https://github.com/llvm/llvm-project/pull/109985
>From f54b7322913a24bca6e611ec1b12b0be606ee1a0 Mon Sep 17 00:00:00 2001
From: Fraser Cormack
Date: Mon, 23 Sep 2024 12:54:02 +0100
Subject: [PATCH 1/2] [libclc] Split off library build system into helpers
T
https://github.com/frasercrmck updated
https://github.com/llvm/llvm-project/pull/109985
>From 454110573c6fcb472f5a87d3de31d2a0addd546d Mon Sep 17 00:00:00 2001
From: Fraser Cormack
Date: Mon, 23 Sep 2024 12:54:02 +0100
Subject: [PATCH 1/2] [libclc] Split off library build system into helpers
T
https://github.com/frasercrmck updated
https://github.com/llvm/llvm-project/pull/109985
>From 454110573c6fcb472f5a87d3de31d2a0addd546d Mon Sep 17 00:00:00 2001
From: Fraser Cormack
Date: Mon, 23 Sep 2024 12:54:02 +0100
Subject: [PATCH 1/2] [libclc] Split off library build system into helpers
T
frasercrmck wrote:
> Also do you know if libclc has any tests worth running in CI? It doesn't have
> a check-libclc so I'm not sure if that's because there's not much to run, or
> running the tests requires a GPU or something else.
>
> Context is I've been looking at the pipeline we use for PR
https://github.com/frasercrmck approved this pull request.
LGTM other than the question about sorting the list
https://github.com/llvm/llvm-project/pull/111528
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailm
@@ -151,6 +147,17 @@ if( "spirv-mesa3d-" IN_LIST LIBCLC_TARGETS_TO_BUILD OR
"spirv64-mesa3d-" IN_LIST
endif()
endif()
+if( LIBCLC_TARGETS_TO_BUILD STREQUAL "all" )
frasercrmck wrote:
Right yeah, because `spirv-mesa3d-` is only added to `ALL` if the tool is
1 - 100 of 415 matches
Mail list logo
