Enable MLIR by default for more cases #2274
Merged
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
Codecov Report
@@ Coverage Diff @@
## develop #2274 +/- ##
========================================
Coverage 91.29% 91.29%
========================================
Files 436 436
Lines 16335 16335
========================================
Hits 14913 14913
Misses 1422 1422 |
This build is not recommended to merge 🔴 |
🔴torchvision-inceptionv3_1: FAILED: MIGraphX is not within tolerance - check verbose output🔴slim-inceptionv4_1: FAILED: MIGraphX is not within tolerance - check verbose output🔴bert_base_cased_fp16: FAILED: MIGraphX is not within tolerance - check verbose output🔴bert_large_uncased_fp16: FAILED: MIGraphX is not within tolerance - check verbose output🔴distilgpt2_fp16: FAILED: MIGraphX is not within tolerance - check verbose output |
umangyadav
reviewed
Oct 3, 2023
|
Hello, would it be possible to have a |
|
@pfultz2 Wanted to poke the status of improving this PR to match the heuristic ticket, since we're getting close to the branch date |
Which ticket is that? |
pfultz2
changed the title
|
@krzysz00 I updated the PR. |
Using |
krzysz00
reviewed
Oct 10, 2023
src/targets/gpu/fuse_mlir.cpp
Outdated
| { | ||
| match::find_matches(mpm, find_mlir_standalone_dot_op{}); | ||
| } | ||
| mlir_mode mode = enabled(MIGRAPHX_ENABLE_MLIR{}) ? mlir_mode::fast : mlir_mode::none; |
There was a problem hiding this comment.
Hold on, since we now have MIGRAPHX_DISABLE_MLIR, shouldn't this be keying off of _DISABLE_MLIR and not _ENABLE_MLIR?
There was a problem hiding this comment.
No, DISABLE_MLIR will disable MLIR completely(and its already handled in target.cpp), whereas ENABLE_MLIR will enable it for gemm fusions(because it is not enabled by default because its not always faster).
There was a problem hiding this comment.
Ah. So maybe we should rename it to something like MIGRAPHX_ENABLE_MLIR_GEMM_FUSION because it's a very confusing variable name
| } | ||
|
|
||
| struct find_mlir_fused_ops | ||
| { | ||
| mlir_mode conv_mode = mlir_mode::none; | ||
| mlir_mode dot_mode = mlir_mode::none; |
There was a problem hiding this comment.
Nit: shouldn't we have an explicit constructor? Or /*conv_mode=*/ comments when we create this?
There was a problem hiding this comment.
Let me see if designated initializers work here(an explicit constructor wont allow that though).
krzysz00
approved these changes
Oct 12, 2023
src/targets/gpu/fuse_mlir.cpp
Outdated
| @@ -37,23 +37,13 @@ struct module; | |||
| namespace gpu { | |||
|
|
|||
| MIGRAPHX_DECLARE_ENV_VAR(MIGRAPHX_ENABLE_MLIR); | |||
There was a problem hiding this comment.
Wanted to note the agreement from yesterday's meeting that this becomes MIGRAPHX_ENABLE_EXTRA_MLIR
umangyadav
approved these changes
Oct 12, 2023
|
@causten Fixed the mlir tests. |
|
So the CI failure looks like a bug in MLIR, here is the backtrace using a debug build of mlir: From the mlir trace it comes from programs like this: Also running with helgrind, I do see some data races in MLIR: So I am not sure if this is a multihreaded issue, since it still produces the error when |
…MIGraphX into mlir-fast-check
|
@pfultz2 Because rocMLIR uses LLVM internally, ubsan builds of rocMLIR needs to use LLVM's ubsan flags, which include That is, an invalid vptr error is expected and should be disabled. It is not possible, to my knowledge, to fix this. I set up the MLIR debug CI to pass those options in MLIR builds. |
|
Also, I'm not sure the race is related |
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.
This will enable MLIR by default for these cases:
MIGRAPHX_ENABLE_MLIR) to enable MLIR for floating-point gemm fusionsExcept:
Also there is
MIGRAPHX_DISABLE_MLIRto disable MLIR completely.