This revision was automatically updated to reflect the committed changes.
Closed by commit rL274220: AMDGPU: Set amdgpu_kernel calling convention for
OpenCL kernels. (authored by nhaustov).
Changed prior to commit:
http://reviews.llvm.org/D21367?vs=62198=62343#toc
Repository:
rL LLVM
nhaustov updated this revision to Diff 62198.
nhaustov added a comment.
Update after discussion with Sam and review from Anastasia.
http://reviews.llvm.org/D21367
Files:
include/clang/Basic/Specifiers.h
lib/AST/ItaniumMangle.cpp
lib/AST/Type.cpp
lib/AST/TypePrinter.cpp
yaxunl added a comment.
> Maybe we can use the fact that calling conventions are separate in AST and in
> LLVM IR. So in AST maybe we can have opencl_kernel and in LLVM it can map to
> spir_kernel for SPIR and amdgpu_kernel for AMDGPU.
This approach looks good to me. Thanks.
Anastasia added inline comments.
Comment at: lib/Basic/Targets.cpp:2122
@@ +2121,3 @@
+
+
+ CallingConvCheckResult checkCallingConvention(CallingConv CC) const override
{
One extra line?
http://reviews.llvm.org/D21367
nhaustov added a comment.
In http://reviews.llvm.org/D21367#467894, @yaxunl wrote:
> I am wondering what's the difference between this calling convention and
> spir_kernel.
spir_kernel has only effect and was created only for SPIR target. In theory it
could perhaps be reused for AMDGPU
yaxunl added a comment.
In http://reviews.llvm.org/D21367#467982, @nhaustov wrote:
> In http://reviews.llvm.org/D21367#467894, @yaxunl wrote:
>
> > I am wondering what's the difference between this calling convention and
> > spir_kernel.
>
>
> spir_kernel has only effect and was created only
nhaustov added a comment.
In http://reviews.llvm.org/D21367#467894, @yaxunl wrote:
> I am wondering what's the difference between this calling convention and
> spir_kernel.
spir_kernel has only effect and was created only for SPIR target. In theory it
could perhaps be reused for AMDGPU
yaxunl added a comment.
I am wondering what's the difference between this calling convention and
spir_kernel.
http://reviews.llvm.org/D21367
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
tstellarAMD added a comment.
This seems OK to me.
http://reviews.llvm.org/D21367
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
nhaustov updated this revision to Diff 61247.
nhaustov added a comment.
Add test for calling OpenCL kernel from kernel.
http://reviews.llvm.org/D21367
Files:
include/clang/AST/Type.h
include/clang/Basic/Specifiers.h
lib/AST/ItaniumMangle.cpp
lib/AST/Type.cpp
lib/AST/TypePrinter.cpp
nhaustov added a comment.
In http://reviews.llvm.org/D21367#461480, @tstellarAMD wrote:
> Does this new patch fix the OpenCL regression?
Yes, it fixes the problem with calling kernel from kernel. I'll add a test too.
http://reviews.llvm.org/D21367
tstellarAMD added a comment.
Does this new patch fix the OpenCL regression?
http://reviews.llvm.org/D21367
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
nhaustov created this revision.
nhaustov added reviewers: tstellarAMD, rsmith.
nhaustov added a subscriber: cfe-commits.
Herald added a subscriber: kzhuravl.
Extend CC field of FunctionType to 5 bits and ExtInfo to 10 bits
to make space for new calling convention.
Add new Clang calling convention
13 matches
Mail list logo
