Revert "Update sycl-docs.yml" #23
Open
Conversation
This file contains hidden or 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
…l#15142) Requesting ownership for libdevice nativecpu files on behalf of @intel/dpcpp-nativecpu-pi-reviewers. Thank you
This commit reenables a selection of tests for ze_debug as they no longer seem to be failing for this option. Signed-off-by: Larsen, Steffen <[email protected]>
CONFLICT (content): Merge conflict in clang/test/Driver/linker-wrapper.c
CONFLICT (content): Merge conflict in .github/workflows/pr-code-format.yml
CONFLICT (content): Merge conflict in llvm/include/llvm/InitializePasses.h CONFLICT (content): Merge conflict in llvm/include/llvm/LinkAllPasses.h
CONFLICT (content): Merge conflict in clang/include/clang/Sema/Sema.h CONFLICT (content): Merge conflict in clang/include/clang/Sema/SemaSYCL.h CONFLICT (content): Merge conflict in clang/lib/Sema/SemaSYCL.cpp
By storing the env variables in the config helper we eliminate the need for extra function parameters while also making it accessible throughout the codebase.
…r uplift OCLCPURT to 2024.18.6.0.02 (intel#15021)
… operations (intel#15140) Multiple declarations were missing for shuffle and broadcast operations, in particular work group broadcast ones. This adds them.
…ion (intel#14992)" This reverts commit 0a9db37.
intel#15099) Currently if there is only single device in the context then kernel compiler passes ip version of that device via -device option to ocloc when compiling OpenCL program to spirv to let ocloc enable all extensions supported by that device. Problem is that ocloc -spv_only doesn't produce spirv file when multiple devices are provided via -device option. That's why in this case enable common extensions supported by all devices manually. To do that use ocloc query to get common supported features for the list of devices and then process the return and enable features via ocloc -internal_options -cl-ext=+feature1,...
Signed-off-by: Raiyan Latif <[email protected]>
As we are working towards upstreaming the SYCL support in the compiler, a few areas need to be cleaned up for readability and improved logic to reduce some repetitious output. Does not impact behaviors in general.
Restores missed context analysis for host pointer update. Regression is caused by intel@3dc75a7 --------- Signed-off-by: Tikhomirova, Kseniya <[email protected]>
I don't know if this change the intention of this test, but there is a problem of buffer out-of-bound access in the test. `Buf` is originally a size `1` buffer, but in line 186, the range is set to 10, which would lead to buffer access out-of-bound in `FunctorMulti F`.
…4839) We added some codes to prevent "LoopIdiomRecognize" from converting loop to memset, so that it won't conflict with host Asan. Now we needn't this fix anymore.
This commit reenables a selection of tests that have previously been disabled, most for different reasons. Additionally, the subbuffer test still seems to fail on Gen12 OpenCL, so this commit adds a comment with a link to the GH issue. --------- Signed-off-by: Larsen, Steffen <[email protected]>
Wrap the tests in `REQUIRE: asserts` in order to make sure that the compiler can generate the messages. Also make sure we don't run the e2e tests on windows, as JIT is not supported there.
Running the check-sycl-dumps target sets the SYCL_LIB_DUMPS_ONLY param for lit which should only enable the ABI dump tests. However, since 1fde656 we've added a non `lit.formats.ShTest()` which doesn't respect config.suffixes, so we need to account for this.
…ntel#15160) 1) It keeps all `convert`-related code in one place 2) It allows not to include that extra functionality if this particular method isn't used 3) Potentially, it moves all instances where `half`/`bfloat16` needs to be "complete" in `vector.hpp`. Can't really verify that because `generic_type_traits.hpp` includes definitions for both at the moment
Previously, OffloadWrapper::ConstructJob would call llc to convert LLVM IR into machine code. With this change, it calls clang instead. This should be mostly NFC for x86, there may be some small changes in exactly which passes run, but they are intended to be roughly the same, and enables future changes passing along more clang options. For other targets, this is a bugfix. Specifically, for RISC-V, llc and clang disagree on which subtarget features to enable by default, resulting in linker errors due to subtarget feature mismatches between the llc-compiled wrapper and other clang-compiled object files.
…tel#15307) Do not internalize kernels when supporting dynamic linking. Kernels must be visible so that host code can find them. --------- Signed-off-by: Lu, John <[email protected]>
Follow up from intel#15346. Signed-off-by: Sarnie, Nick <[email protected]>
…ntel#15408) Improving maintainability: - Remove dependency of Matrix tests on LIT's params: `gpu-intel-pvc=True`, `matrix=1`, `matrix-tf32=1`, `matrix-fp16=1`. - Remove `matrix`, `matrix-xmx8`, `matrix-tf32`, `matrix-fp16` as not used - Utilize auto-detection of architecture, aspects and runtime query instead. - add handling of `igc-dev` parameter, in case it is passed in the command line. - minor clean up from previous "typos" in `REQUIRES` and `XFAIL` directives
…option to clang-linker-wrapper (intel#15374) Previously clang-linker-wrapper used -sycl-module-split-mode option to decide whether to use sycl-post-link tool or not to use. This patch adds options for a explicit specifiyng usage of tool and library.
…V path (intel#15384) From SYCL 2020 specification: > The sycl::atomic_ref class also has a template parameter AddressSpace, > which allows the application to make an assertion about the address > space of the object of type T that it references. The default value > for this parameter is access::address_space::generic_space, which > indicates that the object could be in either the global or local > address spaces. If the application knows the address space, it can set > this template parameter to either access::address_space::global_space > or access::address_space::local_space as an assertion to the > implementation. Specifying the address space via this template > parameter may allow the implementation to perform certain > optimizations. Specifying an address space that does not match the > object’s actual address space results in undefined behavior We use `ext::oneapi::experimental::static_address_cast` to do that. It's not implemented for CUDA/HIP yet, that path continues using `sycl::address_space_cast` that performs runtime checks: > An implementation must return nullptr if the run-time value of pointer > is not compatible with Space, and must issue a compiletime diagnostic > if the deduced address space for pointer is not compatible with Space.
Update sycl-docs.yml
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.
Reverts #22