Add a new flag "--no-entry" #2585
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…ion. This is a workaround for some OpenCL drivers that are incompatible with entry points, leading them to have twice the number of kernel arguments. See KhronosGroup#1486 for more details.
|
|
MrSidims
approved these changes
May 29, 2024
svenvh
reviewed
May 29, 2024
| @@ -278,6 +278,10 @@ static cl::opt<SPIRV::BuiltinFormat> SPIRVBuiltinFormat( | |||
| clEnumValN(SPIRV::BuiltinFormat::Global, "global", | |||
| "Use globals to represent SPIR-V builtin variables"))); | |||
|
|
|||
| static cl::opt<bool> | |||
| NoEntry("no-entry", cl::init(false), | |||
There was a problem hiding this comment.
For consistency with other llvm-spirv options:
Suggested change
| NoEntry("no-entry", cl::init(false), | |
| NoEntry("spirv-no-entry", cl::init(false), |
LU-JOHN
reviewed
May 29, 2024
| BM->addEntryPoint(ExecutionModelKernel, BF->getId(), BF->getName(), | ||
| auto Name = | ||
| BM->getGenerateKernelEntryPoints() ? BF->getName() : I->getName().str(); | ||
| BM->addEntryPoint(ExecutionModelKernel, BF->getId(), Name, |
There was a problem hiding this comment.
Is this change necessary? If !getGenerateKernelEntryPoints, then BF->getName()==I->getName().str() since BF->getName() will be set by line 885.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Add a new flag "--no-entry" which disables kernel entry point generation. This is a workaround for some OpenCL drivers that are incompatible with entry points, leading them to have twice the number of kernel arguments. See #1486 for more details.
Adding this option allows a single toolchain to be used when compiling kernels for all devices. Some real world instances of this problem:
https://community.intel.com/t5/GPU-Compute-Software/OpenCL-CL-INVALID-KERNEL-ARGS-with-driver-gt-30-0-101-1340/td-p/1420021
https://community.intel.com/t5/GPU-Compute-Software/OpenCL-Kernel-Name-appends-quot-1-quot/m-p/1419720#M644
The workaround is to use ocloc but that adds another toolchain to dev environments and it doesn't have CLC++ support in master just yet.
This might be useful to backport all the way to version 14. I'll open those PRs if this gets approved instead of spamming everything upfront.