[SYCL] Extend broadcast to TriviallyCopyable types #2160
Merged
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
Uses the TriviallyCopyable shuffle approach for broadcasts. Note that this works for both work-groups and sub-groups, because OpGroupBroadcast is defined as supporting both groups in SPIR-V. Signed-off-by: John Pennycook <john.pennycook@intel.com>
Signed-off-by: John Pennycook <john.pennycook@intel.com>
Pennycook
added
enhancement
vladimirlaz
reviewed
Jul 23, 2020
Signed-off-by: John Pennycook <john.pennycook@intel.com>
Signed-off-by: John Pennycook <john.pennycook@intel.com>
|
Resolved merge conflicts. |
|
@vladimirlaz, @AlexeySachkov: I think this one is ready for review now. |
AlexeySachkov
previously approved these changes
Jul 30, 2020
Functionality is not specific to shuffles. Signed-off-by: John Pennycook <john.pennycook@intel.com>
Resolve build failures from level0 changes. Signed-off-by: John Pennycook <john.pennycook@intel.com>
AlexeySachkov
approved these changes
Jul 31, 2020
jsji
pushed a commit
that referenced
this pull request
Oct 5, 2023
…2160) In some cases, we will see IR with the following @__spirv_BuiltInGlobalInvocationId = external dso_local local_unnamed_addr addrspace(1) constant <3 x i64>, align 32 ... %0 = load <6 x i32>, ptr addrspace(1) @__spirv_BuiltInGlobalInvocationId, align 32 %1 = extractelement <6 x i32> %0, i64 0 Note the global type and load type are different. Change the handling of vector loads from vector globals to reconstruct the global vector type and then bitcast to the load type. Thanks to @jcranmer-intel for helping me find the simplest solution. Original commit: KhronosGroup/SPIRV-LLVM-Translator@c7ec63a
Chenyang-L
pushed a commit
that referenced
this pull request
Feb 18, 2025
Query out and use local size set in program IL in CL adapter.
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.
Uses the TriviallyCopyable shuffle approach for broadcasts.
Note that this works for both work-groups and sub-groups, because
OpGroupBroadcast is defined as supporting both groups in SPIR-V.
Signed-off-by: John Pennycook john.pennycook@intel.com
Closes #1885. It's become clear while implementing this functionality that some refactoring of
spirv.hppandgroup_algorithms.hppwould be nice, since there is a lot of code duplication to handle cases that are essentially identical except for different types of cast. I think such refactoring is beyond the scope of this PR, as the requirements of exactly which types should be supported and what casts are necessary is still evolving.