[InstCombine] Fold (X / C) < X and (X >> C) < X into X > 0 #85555

Poseydon42 · 2024-03-16T22:16:56Z

Proofs:
https://alive2.llvm.org/ce/z/52droC
This resolves #85313

github-actions · 2024-03-16T22:17:13Z

Thank you for submitting a Pull Request (PR) to the LLVM Project!

This PR will be automatically labeled and the relevant teams will be
notified.

If you wish to, you can add reviewers by using the "Reviewers" section on this page.

If this is not working for you, it is probably because you do not have write
permissions for the repository. In which case you can instead tag reviewers by
name in a comment by using @ followed by their GitHub username.

If you have received no comments on your PR for a week, you can request a review
by "ping"ing the PR by adding a comment “Ping”. The common courtesy "ping" rate
is once a week. Please remember that you are asking for valuable time from other developers.

If you have further questions, they may be answered by the LLVM GitHub User Guide.

You can also ask questions in a comment on this PR, on the LLVM Discord or on the forums.

llvmbot · 2024-03-16T22:17:47Z

@llvm/pr-subscribers-llvm-transforms

Author: None (Poseydon42)

Changes

Proofs:
https://alive2.llvm.org/ce/z/yG_wxp
https://alive2.llvm.org/ce/z/iNBZVy
https://alive2.llvm.org/ce/z/XA_HG5
This resolves #85313

Full diff: https://github.com/llvm/llvm-project/pull/85555.diff

2 Files Affected:

(modified) llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp (+25)
(modified) llvm/test/Transforms/InstCombine/icmp-div-constant.ll (+76)

diff --git a/llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp b/llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp
index 0dce0077bf1588..415ea61362776d 100644
--- a/llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp
+++ b/llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp
@@ -7406,6 +7406,31 @@ Instruction *InstCombinerImpl::visitICmpInst(ICmpInst &I) {
   if (Instruction *Res = foldReductionIdiom(I, Builder, DL))
     return Res;
 
+  // Folding (X / Y) < X => X > 0 for some constant Y other than 0 or 1
+  {
+    Constant *Divisor;
+    Value *Dividend;
+    if (match(Op0,
+              m_CombineOr(m_SDiv(m_Value(Dividend), m_ImmConstant(Divisor)),
+                          m_UDiv(m_Value(Dividend), m_ImmConstant(Divisor)))) &&
+        match(Op1, m_Deferred(Dividend)) &&
+        !(Divisor->isZeroValue() || Divisor->isOneValue())) {
+      return new ICmpInst(ICmpInst::ICMP_SGT, Dividend,
+                          Constant::getNullValue(Dividend->getType()));
+    }
+  }
+
+  // Another case of this fold is (X >> Y) < X => X > 0 if Y != 0
+  {
+    Constant *Shift;
+    Value *V;
+    if (match(Op0, m_LShr(m_Value(V), m_ImmConstant(Shift))) &&
+        match(Op1, m_Deferred(V)) && !Shift->isZeroValue()) {
+      return new ICmpInst(ICmpInst::ICMP_SGT, V,
+                          Constant::getNullValue(V->getType()));
+    }
+  }
+
   return Changed ? &I : nullptr;
 }
 
diff --git a/llvm/test/Transforms/InstCombine/icmp-div-constant.ll b/llvm/test/Transforms/InstCombine/icmp-div-constant.ll
index 8dcb96284685ff..0bdfa5daa547a5 100644
--- a/llvm/test/Transforms/InstCombine/icmp-div-constant.ll
+++ b/llvm/test/Transforms/InstCombine/icmp-div-constant.ll
@@ -375,3 +375,79 @@ define i1 @sdiv_eq_smin_use(i32 %x, i32 %y) {
   %r = icmp eq i32 %d, -2147483648
   ret i1 %r
 }
+
+; Folding (X / C) < X => X > 0 for C != 1
+
+define i1 @sdiv_x_by_const_cmp_x(i32 %x) {
+; CHECK-LABEL: @sdiv_x_by_const_cmp_x(
+; CHECK-NEXT:    [[TMP1:%.*]] = icmp sgt i32 [[X:%.*]], 0
+; CHECK-NEXT:    ret i1 [[TMP1]]
+;
+  %1 = sdiv i32 %x, 123
+  %2 = icmp slt i32 %1, %x
+  ret i1 %2
+}
+
+define i1 @udiv_x_by_const_cmp_x(i32 %x) {
+; CHECK-LABEL: @udiv_x_by_const_cmp_x(
+; CHECK-NEXT:    [[TMP1:%.*]] = icmp sgt i32 [[X:%.*]], 0
+; CHECK-NEXT:    ret i1 [[TMP1]]
+;
+  %1 = udiv i32 %x, 123
+  %2 = icmp slt i32 %1, %x
+  ret i1 %2
+}
+
+; Same as above but with right shift instead of division (C != 0)
+
+define i1 @lshr_x_by_const_cmp_x(i32 %x) {
+; CHECK-LABEL: @lshr_x_by_const_cmp_x(
+; CHECK-NEXT:    [[TMP1:%.*]] = icmp sgt i32 [[X:%.*]], 0
+; CHECK-NEXT:    ret i1 [[TMP1]]
+;
+  %1 = lshr i32 %x, 1
+  %2 = icmp slt i32 %1, %x
+  ret i1 %2
+}
+
+; Negative tests for the folds above
+
+define i1 @sdiv_x_by_1_cmp_x_negative(i32 %x) {
+; CHECK-LABEL: @sdiv_x_by_1_cmp_x_negative(
+; CHECK-NEXT:    ret i1 false
+;
+  %1 = sdiv i32 %x, 1
+  %2 = icmp slt i32 %1, %x
+  ret i1 %2
+}
+
+define i1 @sdiv_x_by_1_cmp_y(i32 %x, i32 %y) {
+; CHECK-LABEL: @sdiv_x_by_1_cmp_y(
+; CHECK-NEXT:    [[TMP1:%.*]] = sdiv i32 [[X:%.*]], 123
+; CHECK-NEXT:    [[TMP2:%.*]] = icmp slt i32 [[TMP1]], [[Y:%.*]]
+; CHECK-NEXT:    ret i1 [[TMP2]]
+;
+  %1 = sdiv i32 %x, 123
+  %2 = icmp slt i32 %1, %y
+  ret i1 %2
+}
+
+define i1 @lshr_x_by_0_cmp_x(i32 %x) {
+; CHECK-LABEL: @lshr_x_by_0_cmp_x(
+; CHECK-NEXT:    ret i1 false
+;
+  %1 = lshr i32 %x, 0
+  %2 = icmp slt i32 %1, %x
+  ret i1 %2
+}
+
+define i1 @lshr_x_by_const_cmp_y(i32 %x, i32 %y) {
+; CHECK-LABEL: @lshr_x_by_const_cmp_y(
+; CHECK-NEXT:    [[X:%.*]] = lshr i32 [[X1:%.*]], 1
+; CHECK-NEXT:    [[TMP1:%.*]] = icmp slt i32 [[X]], [[Y:%.*]]
+; CHECK-NEXT:    ret i1 [[TMP1]]
+;
+  %1 = lshr i32 %x, 1
+  %2 = icmp slt i32 %1, %y
+  ret i1 %2
+}

goldsteinn · 2024-03-16T22:43:45Z

Just a note on the proofs.
You really only proved for the exact constant chosen i.e lshr x, 1 in the case of the lshr proofs. That doesn't necessarily transfer to all other constants (in fact 0 is incorrect).
See some general proofs: https://alive2.llvm.org/ce/z/4Q448E
I filled in the sdiv cases, can you please complete the udiv/lshr/ashr ones.

Edit: added proofs for eq/ne case.

llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp

goldsteinn · 2024-03-16T22:58:30Z

Also welcome and thank you for the commit. There a rough draft of a contribution guide for InstCombine here: #79007

You might find it useful.

Poseydon42 · 2024-03-17T19:03:37Z

I have modified the code to work with all predicates for udiv/lshr and unsigned predicates for sdiv, however I believe that ashr does not work with this type of fold. As you can see here, the only two predicates that it does work with are slt and sge. Should I include these two folds for ashr or should I ignore it altogether and only focus on the other 3 instructions?

goldsteinn · 2024-03-17T19:23:21Z

I have modified the code to work with all predicates for udiv/lshr and unsigned predicates for sdiv, however I believe that ashr does not work with this type of fold. As you can see here, the only two predicates that it does work with are slt and sge. Should I include these two folds for ashr or should I ignore it altogether and only focus on the other 3 instructions?

Think made a mistake when merging your updates.
You are right about ashr. Up to you if you want to handle it in this patch.

Also please update your proofs with all cases you plan to handle.

Poseydon42 · 2024-03-17T19:29:42Z

Yeah, that merge was an accident, I'll remove it before pushing the most recent changes. It seems to me like handling ashr would just unnecessarily increase the complexity of the code without providing significant benefit, so I'll omit it. The only question I have now is whether I should add tests for all three instructions I'm working with and all 10 (or 6) comparison predicates, making it 26 tests in total, which, together with negative tests, will bring it closer to 30, or should I simply test 2-3 predicates for each of the three instructions?

goldsteinn · 2024-03-17T20:05:14Z

Yeah, that merge was an accident, I'll remove it before pushing the most recent changes. It seems to me like handling ashr would just unnecessarily increase the complexity of the code without providing significant benefit, so I'll omit it. The only question I have now is whether I should add tests for all three instructions I'm working with and all 10 (or 6) comparison predicates, making it 26 tests in total, which, together with negative tests, will bring it closer to 30, or should I simply test 2-3 predicates for each of the three instructions?

So generally you should be testing different control flows in your code moreso than IR variants.

In this case you have all predicates have the same control flow for lshr/udiv, so maybe between the two test one equality pred, one signed, and one unsigned (i.e eq, sle, ugt). Then a negative test for each maybe with an out of bounds constant.

For sdiv maybe ne / sgt with two negative tests (out of bounds/unsigned pred).

Then you can pretty randomly make 1-2 of the tests use vec.

Please be sure to have a negative test with a vec where one vec element is invalid and the other is valid.

github-actions · 2024-03-17T22:43:56Z

✅ With the latest revision this PR passed the C/C++ code formatter.

Poseydon42 · 2024-03-17T22:52:10Z

Updated proofs (26 in total) can be found here. I've fixed most of the issues mentioned here, the only two things I'm worried about are:
1.The current code does not work with vector constants that have different values in each component (e.g. take a look at test case lshr_by_const_cmp_sle_value). I believe this is caused by m_APInt not matching when a vector constant has different components, but I am unsure about that. Can something be done about this or should I leave this particular scenario unaffected by the optimization?
2. It appears that lshr already has a similar fold implemented, since test case lshr_by_const_cmp_ule_value was not affected by the changes that I've made (it folded to ret i1 true both before and after the changes to the code). Can someone confirm that this is the case, and if so, should I remove the lshr code path from my PR?

llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp

goldsteinn · 2024-03-17T23:23:49Z

Updated proofs (26 in total) can be found here. I've fixed most of the issues mentioned here, the only two things I'm worried about are: 1.The current code does not work with vector constants that have different values in each component (e.g. take a look at test case lshr_by_const_cmp_sle_value). I believe this is caused by m_APInt not matching when a vector constant has different components, but I am unsure about that. Can something be done about this or should I leave this particular scenario unaffected by the optimization?

Yes that is one of the drawbacks of APInt, it only match if the vector is a splat (all elements that same).
You could use: m_SpecificInt_ICMP (see PatternMatch.h).
So for the div cases: m_SpecificInt_ICMP(ICmpInst::ICMP_UGT, APInt::getOneBitSet(Op0->getType()->getScalarSizeInBits(), 1)) (not the cleanest API to use).

It appears that lshr already has a similar fold implemented, since test case lshr_by_const_cmp_ule_value was not affected by the changes that I've made (it folded to ret i1 true both before and after the changes to the code). Can someone confirm that this is the case, and if so, should I remove the lshr code path from my PR?

See: https://github.com/llvm/llvm-project/blob/main/llvm/lib/Analysis/InstructionSimplify.cpp#L3172
No need to go out of your way to avoid that case. Its not extra code on your end to handle (and its correct). I wouldn't use it as a test, however, as its not going through your code path.

Poseydon42 · 2024-03-18T20:58:59Z

Right, after a painful hour of rewriting git history multiple times, it seems to me that I've covered everything that was mentioned here before. Please let me know if there's anything else for me to fix or improve.

Poseydon42 · 2024-03-18T22:10:55Z

Oops, looks like I forgot to run tests on my machine before pushing the changes. I've localised the source of one of many failing tests to this function in getelementptr.ll:

define i1 @test10(ptr %x, ptr %y) {
; CHECK-LABEL: @test10(
; CHECK-NEXT:    [[T4:%.*]] = icmp eq ptr [[X:%.*]], [[Y:%.*]]
; CHECK-NEXT:    ret i1 [[T4]]
;
  %t1 = getelementptr { i32, i32 }, ptr %x, i32 0, i32 1
  %t3 = getelementptr { i32, i32 }, ptr %y, i32 0, i32 1
  %t4 = icmp eq ptr %t1, %t3
  ret i1 %t4
}

The crash report indicates assertion failure in APInt.h within setBit(). It appears to be caused by the fact that for ptr types, bit size is reported as 0, so later in the matching code, a call to APInt::getOneBitSet essentially asks for a number that has 0 bits, with bit #0 set. I'm unsure if I should simply ignore cases like this (e.g. by adding if (Op0->getType()->getScalarSizeInBits() == 0) return nullptr), or if there's a nicer/more usual way of handling this?

goldsteinn · 2024-03-19T15:17:11Z

Oops, looks like I forgot to run tests on my machine before pushing the changes. I've localised the source of one of many failing tests to this function in getelementptr.ll:
define i1 @test10(ptr %x, ptr %y) {
; CHECK-LABEL: @test10(
; CHECK-NEXT:    [[T4:%.*]] = icmp eq ptr [[X:%.*]], [[Y:%.*]]
; CHECK-NEXT:    ret i1 [[T4]]
;
  %t1 = getelementptr { i32, i32 }, ptr %x, i32 0, i32 1
  %t3 = getelementptr { i32, i32 }, ptr %y, i32 0, i32 1
  %t4 = icmp eq ptr %t1, %t3
  ret i1 %t4
}
The crash report indicates assertion failure in APInt.h within setBit(). It appears to be caused by the fact that for ptr types, bit size is reported as 0, so later in the matching code, a call to APInt::getOneBitSet essentially asks for a number that has 0 bits, with bit #0 set. I'm unsure if I should simply ignore cases like this (e.g. by adding if (Op0->getType()->getScalarSizeInBits() == 0) return nullptr), or if there's a nicer/more usual way of handling this?

I'd say just handle the splat vector case for now. If #85676 gets in before this, you could use that API.

Poseydon42 · 2024-03-19T19:30:55Z

I'd say just handle the splat vector case for now. If #85676 gets in before this, you could use that API.

Done, should be passing tests now.

llvm/test/Transforms/InstCombine/icmp-div-constant.ll

llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp

dtcxzyw

https://github.com/llvm/llvm-project/pull/79007/files may help :)

llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp

Poseydon42 · 2024-03-19T21:30:26Z

Updated proofs with ashr: https://alive2.llvm.org/ce/z/52droC

goldsteinn · 2024-03-19T23:13:12Z

llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp

+                          Constant::getNullValue(Op1->getType()));
+    }
+
+    if (match(Op0, m_AShr(m_Specific(Op1), m_APInt(Shift))) && Shift->ugt(0) &&


nit: !Shift->isZero()

It is not fixed.

I believe it is fixed now.

goldsteinn · 2024-03-19T23:15:26Z

llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp

+      return new ICmpInst(ICmpInst::getSwappedPredicate(Pred), Op1,
+                          Constant::getNullValue(Op1->getType()));
+    }
+  }


nit: There is a bit of redundancy in that each of the return cases is identical, but at least imo its fine.

goldsteinn · 2024-03-19T23:16:04Z

LGTM, wait on one more approval.

Poseydon42 · 2024-03-28T08:15:06Z

Ping @nikic @dtcxzyw

llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp

PR Link: llvm/llvm-project#85555

nikic

LGTM

llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp

llvm/test/Transforms/InstCombine/icmp-div-constant.ll

Poseydon42 · 2024-04-10T22:30:19Z

Thanks nikic, looks like it is now ready for another review.
Ping @dtcxzyw

github-actions · 2024-04-11T09:41:09Z

@Poseydon42 Congratulations on having your first Pull Request (PR) merged into the LLVM Project!

Your changes will be combined with recent changes from other authors, then tested
by our build bots. If there is a problem with a build, you may receive a report in an email or a comment on this PR.

Please check whether problems have been caused by your change specifically, as
the builds can include changes from many authors. It is not uncommon for your
change to be included in a build that fails due to someone else's changes, or
infrastructure issues.

How to do this, and the rest of the post-merge process, is covered in detail here.

If your change does cause a problem, it may be reverted, or you can revert it yourself.
This is a normal part of LLVM development. You can fix your changes and open a new PR to merge them again.

If you don't get any reports, no action is required from you. Your changes are working as expected, well done!

Poseydon42 requested a review from nikic as a code owner March 16, 2024 22:16

llvmbot added the llvm:transforms label Mar 16, 2024

goldsteinn reviewed Mar 16, 2024

View reviewed changes

llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp Outdated Show resolved Hide resolved

goldsteinn reviewed Mar 16, 2024

View reviewed changes

llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp Outdated Show resolved Hide resolved

goldsteinn reviewed Mar 16, 2024

View reviewed changes

llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp Outdated Show resolved Hide resolved

goldsteinn reviewed Mar 16, 2024

View reviewed changes

llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp Outdated Show resolved Hide resolved

goldsteinn reviewed Mar 16, 2024

View reviewed changes

llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp Outdated Show resolved Hide resolved

goldsteinn requested a review from dtcxzyw March 16, 2024 22:53

Poseydon42 force-pushed the icmp-div-by-constant-compare-with-dividend branch from 3bf2533 to 3f08ea9 Compare March 17, 2024 22:41

Poseydon42 force-pushed the icmp-div-by-constant-compare-with-dividend branch from 3f08ea9 to 65de414 Compare March 17, 2024 22:48

goldsteinn reviewed Mar 17, 2024

View reviewed changes

llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp Outdated Show resolved Hide resolved

Poseydon42 force-pushed the icmp-div-by-constant-compare-with-dividend branch from 65de414 to 47cc91b Compare March 17, 2024 23:15

Poseydon42 force-pushed the icmp-div-by-constant-compare-with-dividend branch from 47cc91b to 634bbc3 Compare March 18, 2024 20:58

Poseydon42 force-pushed the icmp-div-by-constant-compare-with-dividend branch from 634bbc3 to 92dd571 Compare March 19, 2024 19:30

Poseydon42 force-pushed the icmp-div-by-constant-compare-with-dividend branch from 92dd571 to 0d36e34 Compare March 19, 2024 19:35

goldsteinn reviewed Mar 19, 2024

View reviewed changes

llvm/test/Transforms/InstCombine/icmp-div-constant.ll Outdated Show resolved Hide resolved

goldsteinn reviewed Mar 19, 2024

View reviewed changes

llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp Outdated Show resolved Hide resolved

dtcxzyw requested changes Mar 19, 2024

View reviewed changes

Poseydon42 force-pushed the icmp-div-by-constant-compare-with-dividend branch from 0d36e34 to 52da7a7 Compare March 19, 2024 22:24

goldsteinn reviewed Mar 19, 2024

View reviewed changes

Poseydon42 requested a review from dtcxzyw March 23, 2024 19:03

dtcxzyw reviewed Mar 28, 2024

View reviewed changes

llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp Outdated Show resolved Hide resolved

dtcxzyw added a commit to dtcxzyw/llvm-opt-benchmark that referenced this pull request Mar 28, 2024

pre-commit: test PR85555

142b55b

PR Link: llvm/llvm-project#85555

dtcxzyw mentioned this pull request Mar 28, 2024

pre-commit: test PR85555 dtcxzyw/llvm-opt-benchmark#459

Closed

Poseydon42 force-pushed the icmp-div-by-constant-compare-with-dividend branch from 52da7a7 to 51d44fd Compare March 28, 2024 17:37

Poseydon42 requested a review from dtcxzyw March 28, 2024 21:42

nikic approved these changes Apr 9, 2024

View reviewed changes

llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp Outdated Show resolved Hide resolved

llvm/lib/Transforms/InstCombine/InstCombineCompares.cpp Show resolved Hide resolved

llvm/test/Transforms/InstCombine/icmp-div-constant.ll Outdated Show resolved Hide resolved

Poseydon42 added 2 commits April 10, 2024 23:22

[InstCombine] Add test cases for folds (X / C) cmp X and (X >> C) cmp X

31a1058

[InstCombine] Fold (X / C pred X) and (X >> C pred X) into X ~pred 0

c203d1f

Poseydon42 force-pushed the icmp-div-by-constant-compare-with-dividend branch from 51d44fd to c203d1f Compare April 10, 2024 22:28

nikic merged commit 462e102 into llvm:main Apr 11, 2024
4 checks passed

dtcxzyw mentioned this pull request May 9, 2024

Opt should know x/2 <= x - x/2 is true for udiv #91527

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[InstCombine] Fold (X / C) < X and (X >> C) < X into X > 0 #85555

[InstCombine] Fold (X / C) < X and (X >> C) < X into X > 0 #85555

Poseydon42 commented Mar 16, 2024 •

edited

github-actions bot commented Mar 16, 2024

llvmbot commented Mar 16, 2024

goldsteinn commented Mar 16, 2024 •

edited

goldsteinn commented Mar 16, 2024

Poseydon42 commented Mar 17, 2024

goldsteinn commented Mar 17, 2024 •

edited

Poseydon42 commented Mar 17, 2024

goldsteinn commented Mar 17, 2024 •

edited

github-actions bot commented Mar 17, 2024 •

edited

Poseydon42 commented Mar 17, 2024

goldsteinn commented Mar 17, 2024

Poseydon42 commented Mar 18, 2024

Poseydon42 commented Mar 18, 2024

goldsteinn commented Mar 19, 2024

Poseydon42 commented Mar 19, 2024 •

edited

dtcxzyw left a comment

Poseydon42 commented Mar 19, 2024

goldsteinn Mar 19, 2024

dtcxzyw Mar 28, 2024

Poseydon42 Apr 10, 2024

goldsteinn Mar 19, 2024

goldsteinn commented Mar 19, 2024

Poseydon42 commented Mar 28, 2024

nikic left a comment

Poseydon42 commented Apr 10, 2024

github-actions bot commented Apr 11, 2024

[InstCombine] Fold (X / C) < X and (X >> C) < X into X > 0 #85555

[InstCombine] Fold (X / C) < X and (X >> C) < X into X > 0 #85555

Conversation

Poseydon42 commented Mar 16, 2024 • edited

github-actions bot commented Mar 16, 2024

llvmbot commented Mar 16, 2024

goldsteinn commented Mar 16, 2024 • edited

goldsteinn commented Mar 16, 2024

Poseydon42 commented Mar 17, 2024

goldsteinn commented Mar 17, 2024 • edited

Poseydon42 commented Mar 17, 2024

goldsteinn commented Mar 17, 2024 • edited

github-actions bot commented Mar 17, 2024 • edited

Poseydon42 commented Mar 17, 2024

goldsteinn commented Mar 17, 2024

Poseydon42 commented Mar 18, 2024

Poseydon42 commented Mar 18, 2024

goldsteinn commented Mar 19, 2024

Poseydon42 commented Mar 19, 2024 • edited

dtcxzyw left a comment

Choose a reason for hiding this comment

Poseydon42 commented Mar 19, 2024

goldsteinn Mar 19, 2024

Choose a reason for hiding this comment

dtcxzyw Mar 28, 2024

Choose a reason for hiding this comment

Poseydon42 Apr 10, 2024

Choose a reason for hiding this comment

goldsteinn Mar 19, 2024

Choose a reason for hiding this comment

goldsteinn commented Mar 19, 2024

Poseydon42 commented Mar 28, 2024

nikic left a comment

Choose a reason for hiding this comment

Poseydon42 commented Apr 10, 2024

github-actions bot commented Apr 11, 2024

Poseydon42 commented Mar 16, 2024 •

edited

goldsteinn commented Mar 16, 2024 •

edited

goldsteinn commented Mar 17, 2024 •

edited

goldsteinn commented Mar 17, 2024 •

edited

github-actions bot commented Mar 17, 2024 •

edited

Poseydon42 commented Mar 19, 2024 •

edited