This repository has been archived by the owner on Jun 10, 2024. It is now read-only.
Llvm3.6.21 #73
Open
Llvm3.6.21 #73
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
…t not yet fully functional. Simple tests work. Dominance frontiers need to be shaken out. This commit is not about running code though, just a clean port.
…ter results. This gets the same results as LLVM3.4 for IR, and apparently the same results for metadata, minus all the "metadata" text.
since we are not generating GLSL for LLVM 3.6.2 yet
Tests: Update runtests and baseResults for LLVM3.6.2, to be Top IR only.
Previously Stores were done at vector granularity but these were problematic for parallel threads storing into same vector.
TopBuilder: generate component granularity Stores
fix for multi-dimension runtime-sized arrays at bottom of buffer blocks
the metadata corresponds to the argument that is operated on atomically.
GLSL Backend: fix method to extract typeproxy's name from IO metadata
add uniform metadata for atomic memory intrinsic calls
test: enable Bottom IR and GLSL generation and comparison
GLSL back end: switch to single walk for io metadata
Remove assert that the rows of the left operand must equal the columns of the right operand.
TopBuilder: allow non-square matrix multiplies
To more accurately implement IEEE semantics for float !=
now it reflects change to single walk io metadata
fix base results for matrix3 test
Use LLVM fcmp une instead of fcmp one
test: update base results for changes in precision generated by glslang
use native select for integer result mix()
Previously the interpolant was loaded and passed as an rvalue. This created the potential that the interpolant argument could map to more than one input variable in the presence of control flow.
Pass interpolant arg to InterpolateAt* as lvalue in TopIR
Enable AMD_EXTENSIONS (to be in sync with glslang)
Previously an empty string was used. Now the Backend will recognize the names.
Update baseresults to pick up glslang precision changes
Pass anon@ names through metadata for anonymous object.
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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.
No description provided.