This is an archive of the discontinued LLVM Phabricator instance.

[mlir][sparse] Implementing sparse=>dense conversion.
ClosedPublic

Authored by wrengr on Sep 29 2021, 4:47 PM.

Event Timeline

wrengr created this revision.Sep 29 2021, 4:47 PM
wrengr requested review of this revision.Sep 29 2021, 4:47 PM
wrengr updated this revision to Diff 376339.Sep 30 2021, 1:29 PM

Implementing some of my todo notes

wrengr updated this revision to Diff 376355.Sep 30 2021, 1:55 PM

Splitting off D110882

wrengr updated this revision to Diff 376365.Sep 30 2021, 2:14 PM

Splitting out D110883 and D110884

wrengr updated this revision to Diff 376407.Sep 30 2021, 6:03 PM

Moving the new functions to the right section.

wrengr updated this revision to Diff 376408.Sep 30 2021, 6:18 PM

Updating commentary about the code generated for sparse=>dense conversion.

aartbik added inline comments.Oct 1 2021, 10:19 AM
mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorConversion.cpp
438–471

Technically, we also fall into this branch for the strange dense to dense conversion. We will have typically folded those away, but I would not completely rely on this having taken place always and defend against that in the code.

444–446

Note that in the context of another project, we may migrate library code to actual codegen (which has the advantage of a smaller memory footprint potentially and allows for "unforeseen" type combinations). Such a migration may take care of all such performance concerns.

452

I think we need a subtle rewriting of the new call utility since we need the sparse encoding for some info, but the id permutation for the "new" tensor (ie. the dense result)

mlir/lib/ExecutionEngine/SparseUtils.cpp
145–146

this sentence does not flow quite well, or am I reading it wrong?

603–604

We really should not need permutation here. If you call toCOO with ID permutation, it restores the indices to original order.

Note that toCOO takes the permutation of the *target*, and internally restores permutation, if there was one from *source* (the internally stored inverse permutation). So if you call toCOO with ID perm, you don't need the perm here anymore, since indices are in the natural "dense" order MLIR expects.

Also, please add tests for this

(1) in mlir/test/Dialect/SparseTensor conversion, a CHECK test on the expected loop structure (see examples there)
(2) in mlir/test/Integration/Dialect/SparseTensor/CPU, add a new "end-to-end" test that does a dense->sparse->dense roundtrip

bixia added inline comments.Oct 1 2021, 3:12 PM
mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorConversion.cpp
452

this should be encSrc, not encDst, right?

mlir/lib/ExecutionEngine/SparseUtils.cpp
111

nested

wrengr marked 2 inline comments as done.Oct 1 2021, 3:30 PM
wrengr added inline comments.
mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorConversion.cpp
438–471

Just to be clear, that should only ever happen when the source is a "sparse" tensor that happens to use dense storage for all dimensions, right? Which is to say, the destination is always a bog-standard dense tensor, right? If so then the code should still work; if not then we'll have to figure out how to tighten up the guards for detecting the different cases.

I'll update the commentary (on the assumption that the answer to the first question is yes; of course, if that's true, then shouldn't the "sparse=>sparse" case have the same caveat?)

444–446

Yep, that's what I was thinking re "good enough for now" :) I mainly added the comment since when discussing the design with Tatiana, she raised some concerns about the performance implications of the method/function calls. For our current goals, I can't imagine this branch would be taken often enough to constitute a performance bottleneck. (And whenever it does, it's easy enough to fix at that point.)

452

Yeah, I've been mulling over a few ways to clean genNewCall up. Did you want me to do that in/before this differential, or is it okay to do afterwards?

mlir/lib/ExecutionEngine/SparseUtils.cpp
145–146

I'll try rewording

603–604

Oh good. I thought that's how things worked, but I wasn't quite certain.

wrengr updated this revision to Diff 376644.Oct 1 2021, 3:32 PM
wrengr marked an inline comment as done.

Addressing comments

aartbik added inline comments.Oct 4 2021, 9:14 AM
mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorConversion.cpp
438–471

No, I really means no encoding at all. There is a subtle difference between an un-annotated tensor and a tensor with all-dense annotations. All conversions work for the all-dense annotated case (it is treated as a sort of sparse tensors). But the logic on falling into a branch based on encDst, !encDest, srcDest, !srcDest (so four truth values) fell into this branch for two cases, but you only implemented the sparse->dense.

So you will have to add one if-test and return failure.

wrengr updated this revision to Diff 377031.Oct 4 2021, 2:06 PM
wrengr marked 2 inline comments as done.

Re-updating commentary about covering dense=>sparse not dense=>dense. Also some preliminary WIP towards adding tests.

mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorConversion.cpp
438–471

Got it

wrengr marked 2 inline comments as not done.Oct 4 2021, 2:09 PM
wrengr updated this revision to Diff 377045.Oct 4 2021, 3:51 PM
wrengr marked an inline comment as done.

Debugging some errors

wrengr added inline comments.Oct 4 2021, 3:52 PM
mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorConversion.cpp
452

Good catch :) That lead to an utterly inscrutable crash stacktrace

wrengr updated this revision to Diff 377317.Oct 5 2021, 11:33 AM

(more debugging attempt; not much progress)

aartbik added inline comments.Oct 5 2021, 1:40 PM
mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorConversion.cpp
441

using the "zero" data value itself has a bit of a risk that a sparse data structure with a few explicitly stored zeros will bail this loop too soon; having a 0/1 result and passing the value as another ref parameter seems safer

467

you cannot just pass "indices" to next call and use it here; you will need the IR to load the contents returned through the memref by the getNext() call, using explicit memref::LoadOp ops on the elements in the memref and passing this to the StoreOp

470

You are replacing the op (which returns a tensor) with the result of an alloc (which is a memref).
That type mismatch will fail. You need a buffer cast in between.

470

This replacement will also need some legality check changes. Up to now, we were replacing sparse tensors with opague pointers, and the checks/rewriting did all the work

But now we have

dense_tensor = convert ....
return dense_tenor

and the mechanism will need some "love" to make it accept the rewriting even though the types were already legal to start with

wrengr updated this revision to Diff 379851.Oct 14 2021, 2:45 PM
wrengr marked 2 inline comments as done.

rebased, and factored out D111763 and D111766

wrengr added inline comments.Oct 14 2021, 2:57 PM
mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorConversion.cpp
441

Gross; but yes, you're right.

I think I'll leave the commentary here as-is, however, since —like the field accesses on line 452, and like the commentary for other branches— it's reflecting more what the C++ code in the ExecutionEngine does, rather than reflecting what the generated MLIR code does; and there's no ambiguity about what the getNext() method returns, even though IMPL_COO_GETNEXT introduces the problem you mention.

wrengr marked an inline comment as done.Oct 14 2021, 2:57 PM
wrengr updated this revision to Diff 382805.Oct 27 2021, 2:54 PM

Everything should now be ready for review. I've added integration and codegen tests, fixed various infelicities, and rebased.

N.B., I'm planning to add support for dynamic sizes, but I want to do that in a separate differential

aartbik added inline comments.Oct 27 2021, 3:40 PM
mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorConversion.cpp
128

returns (to be consistent with Generates)

also "as" instead of "at"? (2x)

130

is the "inline" really important (here and below)

internal linkage gives the compiler sufficient room to make that decision using built-in heuristics?

262

we typically do not show that much detail of the C++ lib in this file

269

This is not a style we use in MLIR, and especially for an internally linked method, and compared to the surrounding methods, it feels a bit heavy

270

let's not document SparseUtils inside this file

300

not in this revision yet? seems a relatively minor addition (and then we are done for all cases!)

314

same

444–446

remove this; it more or less applies to all method calls here, and it will fall under the general umbrella of perhaps moving to 100% codegen over support lib usage....

450

we don't emit errors when rewriting rules fail to apply

459

all this block scoping makes the code more lengthy than it could be

I would either break this up in methods where it makes sense, or otherwise just take the scoping hit for readability

480

shouldn't we bail out if there are dynamic sizes?
or, better yet, just add those in this revision

mlir/lib/ExecutionEngine/SparseUtils.cpp
150–151

this class adds an enormous amount of code for a very thin iterator

how about just having the very simple start()/next() iteration inside the COO class itself?
That way, we can also assert an error if you try to insert while iterating

557

assert(action == kEmpty); disappeared in your rewrite

557

this is of course inside a macro now, but LLVM wants braces {} on all branches if one of them is branched

644

if you use a shorter name, perhaps it fits on one line

757–766

I prefer not to have this function at all; but if we keep it, it should go to CRunnerUtils.cpp since we have some other printing utilities there

but this is not related to sparse

mlir/test/Dialect/SparseTensor/conversion_sparse2dense.mlir
16 ↗(On Diff #382805)

Ah, you made this much more accurate checking than conversion.mlir does (good).
But, let's also add a check in that file in the more informal way there

mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_conversion_sparse2dense.mlir
45 ↗(On Diff #382805)

from dense to sparse

64–69 ↗(On Diff #382805)

I think you are overthinking this

I agree that the silent exit(1) is not good (so perhaps we should just go back to fully FileChecked cases)
But how about just printing -12345 on error and exiting and then also add a CHECK-NOT -12345 or something like that

but again, I actually prefer the fully FileChecked cases and just rely on printing and verifying the output

132–133 ↗(On Diff #382805)

yeah, we need to do that or asan will fail

Proper way, I think

(1) memref.buffer_cast from tensor to memref
(2) dealloc memref

N.B., I'm planning to add support for dynamic sizes, but I want to do that in a separate differential

Ah, I saw this after writing my comments. So you can ignore the part where I asked about that ;-)

wrengr updated this revision to Diff 382836.EditedOct 27 2021, 4:07 PM

Added child differential D112674 for handling dynamic sizes

wrengr updated this revision to Diff 382862.Oct 27 2021, 5:24 PM
wrengr marked 19 inline comments as done.

Addressing comments

mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorConversion.cpp
300

It's done in D112674.

452

Ha, actually I was right(-ish) the first time. We want the destination encoding here, so we don't apply the dimOrdering permutation twice. The problem was that I was passing a nullptr rather than explicitly constructing the SparseTensorEncodingAttr

480

Done in D112674

mlir/lib/ExecutionEngine/SparseUtils.cpp
757–766

Yeah, I'd rather not have this function too. Unfortunately, when I tried using the CRunnerUtils.cpp functions that vector.print does, I could never get it to link right: it either complained about re-defining a symbol, or about using an un-defined symbol. As for implementing it in the integration test itself, I can't seem to find a way to define string values (not attributes) for passing to fputs()

Moved to CRunnerUtils.cpp for now. Will try to rip it out in a future differential.

mlir/test/Dialect/SparseTensor/conversion_sparse2dense.mlir
16 ↗(On Diff #382805)

I'm not sure I follow. I was just trying to do the same as conversion.mlir (but in the opposite direction, naturally); and elsewhere you suggested breaking things out into a separate test rather than reusing the ones already there.

aartbik added inline comments.Oct 27 2021, 6:42 PM
mlir/lib/ExecutionEngine/SparseUtils.cpp
757–766

How about just leaving it at the exit(1) for not and think about this in a future differential.
Rather than introducing something we want to rip again anyway?

mlir/test/Dialect/SparseTensor/conversion_sparse2dense.mlir
16 ↗(On Diff #382805)

Ok, fair enough. I was suggesting to have one sparse->dense check in the conversion.mlir in the more concise style of that test, but having this more rigid tests is indeed sufficient. You can ignore the suggestion ;-)

aartbik added inline comments.Oct 27 2021, 8:42 PM
mlir/lib/ExecutionEngine/CRunnerUtils.cpp
47–58 ↗(On Diff #382862)

Let's not do this at all, but let's go with just using CHECKs
I moved the sparse_conversion test back to this format as well:

https://reviews.llvm.org/D112688
wrengr updated this revision to Diff 383129.Oct 28 2021, 12:45 PM
wrengr marked 7 inline comments as done.

Addressing comments

yeah, looking good. few last nits before we submit...

mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorConversion.cpp
311

Inserts

(I usually use the s-form of a verb in the top level comment, but the imperative form in the inlined code)

318–319

is this comment still relevant? I think we can safely remove it?

478

you still have some block scoping left?
none of these really release stuff early to keep memory lower
so I would opt for readability over block scoping

mlir/lib/ExecutionEngine/SparseUtils.cpp
96

yeah, this is awesome!

120

iteratorLocked = false?

752

feel free to rename this into more intuitive names in a follow up revision too, btw

wrengr updated this revision to Diff 383144.Oct 28 2021, 1:25 PM
wrengr marked 6 inline comments as done.

Addressing nits

mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorConversion.cpp
478

Imo the blocks improve readability rather than detract.

mlir/lib/ExecutionEngine/SparseUtils.cpp
752

Will do

aartbik added inline comments.Oct 28 2021, 1:34 PM
mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorConversion.cpp
452–463

how about removing this comment block and simply using

// Setup a synthetic all-dense, no-permutation encoding for the dense destination.
encDst = SparseTensorEncodingAttr::get(

      op->getContext(),
      SmallVector<SparseTensorEncodingAttr::DimLevelType>(
          rank, SparseTensorEncodingAttr::DimLevelType::Dense),
AffineMap(), 0, 0);

we don't need anything copied from src here.

wrengr marked an inline comment as done.Oct 28 2021, 1:59 PM
wrengr added inline comments.
mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorConversion.cpp
452–463

We need the (ptrTp,indTp,valTp) values for _mlir_ciface_newSparseTensor to enter the right CASE. But I can abbreviate the commentary

wrengr updated this revision to Diff 383158.Oct 28 2021, 1:59 PM
wrengr marked an inline comment as done.

Addressing nits

aartbik added inline comments.Oct 28 2021, 2:09 PM
mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorConversion.cpp
452–463

this last one may have crossed. I just feel L601-612 takes too much real estate explaining what could be a onliner

478

Ok, acceptable ;-)

mlir/lib/ExecutionEngine/SparseUtils.cpp
604

tensor is a bit more consistent name with IMPL3

613

oh, we will leak memory right now!

When iteration is done, we should release the COO tensor/
The easiest way would be to do this

if (elem == nullptr) {

delete iter;  
return false;

}

and document that tensor can no longer be used once getNext returns false

mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_conversion_sparse2dense.mlir
79–81 ↗(On Diff #383144)

this is taken from the original sparse_conversion.mlir
but with your addition it feels like we should

(1) remove this comment, or
(2) add a comment for tensor4,5,6 as well

aartbik added inline comments.Oct 28 2021, 2:12 PM
mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorConversion.cpp
452–463

Oh yeah, right you are. Okay, then indeed just a bit less comment, but same code ;-)

mlir/lib/ExecutionEngine/SparseUtils.cpp
613

this is the most important issue, since without this fix, asan will break the test

wrengr marked 7 inline comments as done.Oct 28 2021, 2:46 PM
wrengr updated this revision to Diff 383170.Oct 28 2021, 2:47 PM

Fixing memory leak

aartbik accepted this revision.Oct 28 2021, 3:02 PM

Ship it, Wren!

This revision is now accepted and ready to land.Oct 28 2021, 3:02 PM
This revision was automatically updated to reflect the committed changes.