Page MenuHomePhabricator

skatkov (Serguei Katkov)
User

Projects

User does not belong to any projects.

User Details

User Since
Jan 25 2017, 1:38 AM (284 w, 1 d)

Recent Activity

Fri, Jul 1

skatkov updated subscribers of D128155: [GuardWidening] Use freeze to widen using possible poison value..
Fri, Jul 1, 2:13 AM · Restricted Project, Restricted Project
skatkov updated the diff for D128155: [GuardWidening] Use freeze to widen using possible poison value..

I come up with this version of the solution.

Fri, Jul 1, 1:51 AM · Restricted Project, Restricted Project
skatkov added a comment to D128322: [GuardWidening] Use logical and in widenCondCommon as it stated in doc.

Abandon due to it is wrong way to fix this.

Fri, Jul 1, 1:44 AM · Restricted Project, Restricted Project

Wed, Jun 29

skatkov added a comment to D128322: [GuardWidening] Use logical and in widenCondCommon as it stated in doc.

Dear reviewers, after deeper investigation the issue I guess I come up with the tests https://reviews.llvm.org/D128779 showing the logical and does not resolve the issue and correctly used freeze the only choice.
I would very appreciate if you could review the tests and confirm that my understanding is valid.

Yes, I agree with your analysis. The @simple_case test is clearest one I think, because it hoists a condition across other control flow.

Basically, if you have two adjacent llvm.experimental.guard (or separated by must-exec instructions), then they can be combined with logical and. But if they're separated by any (normal or abnormal) control flow, then the second condition must be frozen.

So yeah, we should go back to your original solution using freeze. Sorry for the red herring about logical and.

No need for sorry. It helped to go deeper into optimization. Simple freeze for combine checks does not work as well. Need to figure out how to deal with that.

Wed, Jun 29, 1:54 AM · Restricted Project, Restricted Project
skatkov closed D128779: [GuardWidening] Add tests showing the incorrect behavior of GW wrt poison.

Landed.

Wed, Jun 29, 12:09 AM · Restricted Project, Restricted Project

Tue, Jun 28

skatkov updated the diff for D128779: [GuardWidening] Add tests showing the incorrect behavior of GW wrt poison.

handle nits.

Tue, Jun 28, 11:41 PM · Restricted Project, Restricted Project
skatkov updated the diff for D128779: [GuardWidening] Add tests showing the incorrect behavior of GW wrt poison.

Add test with side effect for combine range checks.

Tue, Jun 28, 11:39 PM · Restricted Project, Restricted Project
skatkov added inline comments to D128779: [GuardWidening] Add tests showing the incorrect behavior of GW wrt poison.
Tue, Jun 28, 10:41 PM · Restricted Project, Restricted Project
skatkov added a comment to D128322: [GuardWidening] Use logical and in widenCondCommon as it stated in doc.

Dear reviewers, after deeper investigation the issue I guess I come up with the tests https://reviews.llvm.org/D128779 showing the logical and does not resolve the issue and correctly used freeze the only choice.
I would very appreciate if you could review the tests and confirm that my understanding is valid.

Tue, Jun 28, 9:21 PM · Restricted Project, Restricted Project
skatkov requested review of D128779: [GuardWidening] Add tests showing the incorrect behavior of GW wrt poison.
Tue, Jun 28, 9:19 PM · Restricted Project, Restricted Project

Fri, Jun 24

skatkov added inline comments to D128322: [GuardWidening] Use logical and in widenCondCommon as it stated in doc.
Fri, Jun 24, 10:44 AM · Restricted Project, Restricted Project
skatkov updated the diff for D128322: [GuardWidening] Use logical and in widenCondCommon as it stated in doc.

After long offline conversation with Max, we came to conclusion that combineCheck case does not require logical and.
It based on the facts that all conditions in this case are nuw/nsw if first condition is true and the first condition nuw/nsw flags cannot be based on other conditions (due to its constant is minimal one) and can be safely hoisted before the guard as its poison flags (if they exist) are based on something else then on other conditions.

Fri, Jun 24, 4:28 AM · Restricted Project, Restricted Project

Thu, Jun 23

skatkov committed rG9ffe1b0a792b: [GuardWidening] Update all tests with update_test_checks.py (authored by skatkov).
[GuardWidening] Update all tests with update_test_checks.py
Thu, Jun 23, 10:05 PM · Restricted Project, Restricted Project
skatkov updated the diff for D128322: [GuardWidening] Use logical and in widenCondCommon as it stated in doc.

Step forward trying to take into account Philip's comments.

  1. Add LagicAnd parsing in range check parser.
  2. Utilize the isGuaranteedNotToBeUndefOrPoison function.
Thu, Jun 23, 2:55 AM · Restricted Project, Restricted Project

Wed, Jun 22

skatkov committed rG5e1ccdf960d5: [RS4GC] Handle freeze case for vector (authored by skatkov).
[RS4GC] Handle freeze case for vector
Wed, Jun 22, 9:59 PM · Restricted Project, Restricted Project
skatkov closed D128254: [RS4GC] Handle freeze case for vector.
Wed, Jun 22, 9:59 PM · Restricted Project, Restricted Project
skatkov updated the diff for D128254: [RS4GC] Handle freeze case for vector.

Update pat h before landing to handle reviewer's comment.

Wed, Jun 22, 9:58 PM · Restricted Project, Restricted Project
skatkov added a comment to D128322: [GuardWidening] Use logical and in widenCondCommon as it stated in doc.

I don't follow the implementation details here, but based on test changes this looks like the right approach.

What worries me is that usage of select may break parseRangeChecks logic
where it expects arithmetic and to split already widened condition...

Possibly you want to replace the matcher there with m_LogicalAnd. This is safe if the range checks in question are of the form X pred C1 && X pred C2, i.e. both work on the same variable, and the other side is a constant. If not, then more care might be needed.

Wed, Jun 22, 4:03 AM · Restricted Project, Restricted Project

Tue, Jun 21

skatkov accepted D127992: InlineSpiller: Don't fold spills into undef reads.

lgtm.

Tue, Jun 21, 10:59 PM · Restricted Project, Restricted Project
skatkov added a comment to D128155: [GuardWidening] Use freeze to widen using possible poison value..

Please take a look the alternative solution with logical and: https://reviews.llvm.org/D128322

Tue, Jun 21, 10:55 PM · Restricted Project, Restricted Project
skatkov requested review of D128322: [GuardWidening] Use logical and in widenCondCommon as it stated in doc.
Tue, Jun 21, 10:54 PM · Restricted Project, Restricted Project
skatkov committed rG8f891b7c391e: [LoopVectorize] Uninitialized phi node leads to a crash in SSAUpdater. (authored by skatkov).
[LoopVectorize] Uninitialized phi node leads to a crash in SSAUpdater.
Tue, Jun 21, 9:20 PM · Restricted Project, Restricted Project
skatkov closed D128033: [LoopVectorize] Uninitialized phi node leads to a crash in SSAUpdater..
Tue, Jun 21, 9:19 PM · Restricted Project, Restricted Project
skatkov updated the summary of D128033: [LoopVectorize] Uninitialized phi node leads to a crash in SSAUpdater..
Tue, Jun 21, 8:47 PM · Restricted Project, Restricted Project
skatkov updated the summary of D128033: [LoopVectorize] Uninitialized phi node leads to a crash in SSAUpdater..
Tue, Jun 21, 8:47 PM · Restricted Project, Restricted Project
skatkov updated the summary of D128033: [LoopVectorize] Uninitialized phi node leads to a crash in SSAUpdater..
Tue, Jun 21, 8:46 PM · Restricted Project, Restricted Project
skatkov updated the diff for D128033: [LoopVectorize] Uninitialized phi node leads to a crash in SSAUpdater..

Update test before landing.

Tue, Jun 21, 8:46 PM · Restricted Project, Restricted Project
skatkov added inline comments to D128033: [LoopVectorize] Uninitialized phi node leads to a crash in SSAUpdater..
Tue, Jun 21, 8:31 PM · Restricted Project, Restricted Project
skatkov added a comment to D128033: [LoopVectorize] Uninitialized phi node leads to a crash in SSAUpdater..

The test was obtained by bugpoint. I'll do my best to reduce it more.

Tue, Jun 21, 8:43 AM · Restricted Project, Restricted Project
skatkov planned changes to D128155: [GuardWidening] Use freeze to widen using possible poison value..
Tue, Jun 21, 7:41 AM · Restricted Project, Restricted Project
skatkov added a comment to D128155: [GuardWidening] Use freeze to widen using possible poison value..

I tend to re-write the patch and introduce the option to use freeze or logical and then choose the best one.
To have improved version of this patch for freeze and implement select solution I'll need to do some preparation work to know what condition comes from the widening guard (it should be first in logcal and and there is no need to freeze it) and what comes from other guard.
So, please be tuned.

Tue, Jun 21, 7:40 AM · Restricted Project, Restricted Project
skatkov requested review of D128254: [RS4GC] Handle freeze case for vector.
Tue, Jun 21, 3:05 AM · Restricted Project, Restricted Project

Mon, Jun 20

skatkov added a reviewer for D128033: [LoopVectorize] Uninitialized phi node leads to a crash in SSAUpdater.: Ayal.
Mon, Jun 20, 8:44 PM · Restricted Project, Restricted Project
skatkov added a reviewer for D128033: [LoopVectorize] Uninitialized phi node leads to a crash in SSAUpdater.: dmgreen.

ping

Mon, Jun 20, 8:43 PM · Restricted Project, Restricted Project
skatkov committed rG163c77b2e083: [AARCH64 folding] Do not fold any copy with NZCV (authored by skatkov).
[AARCH64 folding] Do not fold any copy with NZCV
Mon, Jun 20, 8:39 PM · Restricted Project, Restricted Project
skatkov closed D127294: [AARCH64 folding] Do not fold any copy with NZCV.
Mon, Jun 20, 8:39 PM · Restricted Project, Restricted Project
skatkov added a comment to D128155: [GuardWidening] Use freeze to widen using possible poison value..

Does this actually need freeze, rather than just logical and instead of bitwise and?

Mon, Jun 20, 8:04 PM · Restricted Project, Restricted Project
skatkov added inline comments to D128155: [GuardWidening] Use freeze to widen using possible poison value..
Mon, Jun 20, 9:35 AM · Restricted Project, Restricted Project
skatkov updated the diff for D128155: [GuardWidening] Use freeze to widen using possible poison value..
Mon, Jun 20, 9:34 AM · Restricted Project, Restricted Project

Sun, Jun 19

skatkov requested review of D128155: [GuardWidening] Use freeze to widen using possible poison value..
Sun, Jun 19, 11:20 PM · Restricted Project, Restricted Project

Fri, Jun 17

skatkov requested review of D128033: [LoopVectorize] Uninitialized phi node leads to a crash in SSAUpdater..
Fri, Jun 17, 12:08 AM · Restricted Project, Restricted Project

Tue, Jun 14

skatkov added a comment to D127294: [AARCH64 folding] Do not fold any copy with NZCV.

ping

Tue, Jun 14, 9:31 PM · Restricted Project, Restricted Project
skatkov accepted D127744: freeze does not change the constant property.

lgtm

Tue, Jun 14, 7:52 AM · Restricted Project, Restricted Project
skatkov added a reverting change for rG6bf27918144c: [MachineSSAUpdater] compile time improvement in GetValueInMiddleOfBlock: rGd713f0eab8c4: Revert "[MachineSSAUpdater] compile time improvement in GetValueInMiddleOfBlock".
Tue, Jun 14, 6:27 AM · Restricted Project, Restricted Project
skatkov committed rGd713f0eab8c4: Revert "[MachineSSAUpdater] compile time improvement in GetValueInMiddleOfBlock" (authored by skatkov).
Revert "[MachineSSAUpdater] compile time improvement in GetValueInMiddleOfBlock"
Tue, Jun 14, 6:27 AM · Restricted Project, Restricted Project
skatkov added a reverting change for D126523: [MachineSSAUpdater] compile time improvement in GetValueInMiddleOfBlock: rGd713f0eab8c4: Revert "[MachineSSAUpdater] compile time improvement in GetValueInMiddleOfBlock".
Tue, Jun 14, 6:27 AM · Restricted Project, Restricted Project
skatkov committed rG6bf27918144c: [MachineSSAUpdater] compile time improvement in GetValueInMiddleOfBlock (authored by skatkov).
[MachineSSAUpdater] compile time improvement in GetValueInMiddleOfBlock
Tue, Jun 14, 4:30 AM · Restricted Project, Restricted Project
skatkov closed D126523: [MachineSSAUpdater] compile time improvement in GetValueInMiddleOfBlock.
Tue, Jun 14, 4:30 AM · Restricted Project, Restricted Project
skatkov updated the diff for D126523: [MachineSSAUpdater] compile time improvement in GetValueInMiddleOfBlock.

Comments updated.

Tue, Jun 14, 1:44 AM · Restricted Project, Restricted Project
skatkov added a comment to D126523: [MachineSSAUpdater] compile time improvement in GetValueInMiddleOfBlock.

Looks straightforward enough to me. If the value is a PHI, it must be at the beginning of the current block. If it is outside this block, then whoever registered it is confident that it dominates this block. Is it correct that the SSAUpdater will not end up in a state where it tries to use a phi to produce another phi in the same block? Because that might race against the assumption that the used phi is available earlier in the block.

Tue, Jun 14, 1:37 AM · Restricted Project, Restricted Project

Mon, Jun 13

skatkov added a comment to D126523: [MachineSSAUpdater] compile time improvement in GetValueInMiddleOfBlock.

ping.

Mon, Jun 13, 10:08 PM · Restricted Project, Restricted Project
skatkov committed rG095bf6be28c5: [Greedy RegAlloc] Fix the handling of split register in last chance re-coloring. (authored by skatkov).
[Greedy RegAlloc] Fix the handling of split register in last chance re-coloring.
Mon, Jun 13, 10:05 PM · Restricted Project, Restricted Project
skatkov closed D127281: [Greedy RegAlloc] Fix the handling of split register in last chance re-coloring..
Mon, Jun 13, 10:05 PM · Restricted Project, Restricted Project
skatkov updated the diff for D127281: [Greedy RegAlloc] Fix the handling of split register in last chance re-coloring..

Update test before landing.

Mon, Jun 13, 9:49 PM · Restricted Project, Restricted Project

Thu, Jun 9

skatkov added a reviewer for D127294: [AARCH64 folding] Do not fold any copy with NZCV: danilaml.
Thu, Jun 9, 9:04 PM · Restricted Project, Restricted Project
skatkov updated the diff for D127281: [Greedy RegAlloc] Fix the handling of split register in last chance re-coloring..
Thu, Jun 9, 8:53 PM · Restricted Project, Restricted Project
skatkov accepted D127308: [Statepoints] FixupStatepoint: Clear isKill flag if COPY is not deleted..

Thanks!

Thu, Jun 9, 8:43 PM · Restricted Project, Restricted Project
skatkov added inline comments to D127281: [Greedy RegAlloc] Fix the handling of split register in last chance re-coloring..
Thu, Jun 9, 8:26 PM · Restricted Project, Restricted Project

Wed, Jun 8

skatkov updated the diff for D127281: [Greedy RegAlloc] Fix the handling of split register in last chance re-coloring..

Handled comments about tests.

Wed, Jun 8, 11:10 PM · Restricted Project, Restricted Project
skatkov added a comment to D127308: [Statepoints] FixupStatepoint: Clear isKill flag if COPY is not deleted..

Can you please land the test first and in this patch upload the diff what changed.

Wed, Jun 8, 10:32 PM · Restricted Project, Restricted Project
skatkov added inline comments to D127281: [Greedy RegAlloc] Fix the handling of split register in last chance re-coloring..
Wed, Jun 8, 9:09 PM · Restricted Project, Restricted Project
skatkov requested review of D127294: [AARCH64 folding] Do not fold any copy with NZCV.
Wed, Jun 8, 5:31 AM · Restricted Project, Restricted Project
skatkov requested review of D127281: [Greedy RegAlloc] Fix the handling of split register in last chance re-coloring..
Wed, Jun 8, 2:58 AM · Restricted Project, Restricted Project
skatkov added a comment to D126523: [MachineSSAUpdater] compile time improvement in GetValueInMiddleOfBlock.

ping

Wed, Jun 8, 2:43 AM · Restricted Project, Restricted Project

Jun 2 2022

skatkov updated the diff for D126523: [MachineSSAUpdater] compile time improvement in GetValueInMiddleOfBlock.
Jun 2 2022, 10:58 PM · Restricted Project, Restricted Project
skatkov committed rG24e16e4af208: [SSAUpdaterImpl] Do not generate phi node with all the same incoming values (authored by skatkov).
[SSAUpdaterImpl] Do not generate phi node with all the same incoming values
Jun 2 2022, 10:25 PM · Restricted Project, Restricted Project
skatkov closed D126525: [SSAUpdaterImpl] Do not generate phi node with all the same incoming values.
Jun 2 2022, 10:24 PM · Restricted Project, Restricted Project
skatkov added a comment to D126525: [SSAUpdaterImpl] Do not generate phi node with all the same incoming values.

ok, I take a liberty to land this and ready to revert or follow-up basing on @arsenm feedback.

Jun 2 2022, 10:13 PM · Restricted Project, Restricted Project
skatkov committed rGc4d955dd7f68: [MachineSSAUpdate] Add a test for redundant phi generation. (authored by skatkov).
[MachineSSAUpdate] Add a test for redundant phi generation.
Jun 2 2022, 9:28 PM · Restricted Project, Restricted Project

May 30 2022

skatkov updated the diff for D126525: [SSAUpdaterImpl] Do not generate phi node with all the same incoming values.

Handled comments

May 30 2022, 11:57 PM · Restricted Project, Restricted Project
skatkov added a comment to D126525: [SSAUpdaterImpl] Do not generate phi node with all the same incoming values.

Thanks! I'll wait for 1-2 days for @arsenm feedback before landing the patch.

May 30 2022, 9:46 PM · Restricted Project, Restricted Project
skatkov added a reviewer for D126523: [MachineSSAUpdater] compile time improvement in GetValueInMiddleOfBlock: sameerds.
May 30 2022, 9:46 PM · Restricted Project, Restricted Project
skatkov updated the diff for D126525: [SSAUpdaterImpl] Do not generate phi node with all the same incoming values.

remove redundant asserts, add usage of defBB.

May 30 2022, 2:40 AM · Restricted Project, Restricted Project

May 29 2022

skatkov added a comment to D126525: [SSAUpdaterImpl] Do not generate phi node with all the same incoming values.

This is interesting! Can you please make a simple/dedicated LLVM IR case for this only, so we can go over the example to get the whole picture?

llvm/test/CodeGen/AArch64/tail-dup-redundant-phi.mir is dedicated test for this patch and https://reviews.llvm.org/D126523. This test is not landed yet. I just uploaded it here as diff to show the difference.

Even then, a target independent test in LLVM IR will certainly be useful.

May 29 2022, 10:07 PM · Restricted Project, Restricted Project
skatkov added inline comments to D126525: [SSAUpdaterImpl] Do not generate phi node with all the same incoming values.
May 29 2022, 10:03 PM · Restricted Project, Restricted Project
skatkov added a comment to D126525: [SSAUpdaterImpl] Do not generate phi node with all the same incoming values.

This is interesting! Can you please make a simple/dedicated LLVM IR case for this only, so we can go over the example to get the whole picture?

May 29 2022, 7:53 PM · Restricted Project, Restricted Project

May 27 2022

skatkov requested review of D126525: [SSAUpdaterImpl] Do not generate phi node with all the same incoming values.
May 27 2022, 1:52 AM · Restricted Project, Restricted Project
skatkov requested review of D126523: [MachineSSAUpdater] compile time improvement in GetValueInMiddleOfBlock.
May 27 2022, 12:09 AM · Restricted Project, Restricted Project

May 25 2022

skatkov committed rGc2eccc67ce07: [GuardWidening] Remove nuw/nsw flags for hoisted instructions (authored by skatkov).
[GuardWidening] Remove nuw/nsw flags for hoisted instructions
May 25 2022, 11:32 PM · Restricted Project, Restricted Project
skatkov closed D126354: [GuardWidening] Remove nuw/nsw flags for hoisted instructions.
May 25 2022, 11:32 PM · Restricted Project, Restricted Project
skatkov added a comment to D126354: [GuardWidening] Remove nuw/nsw flags for hoisted instructions.

Looks good, but please regenerate checks here. It's not clear at all what this test is trying to ensure (e.g. what is wide.chk).

May 25 2022, 11:20 PM · Restricted Project, Restricted Project
skatkov updated the diff for D126354: [GuardWidening] Remove nuw/nsw flags for hoisted instructions.
May 25 2022, 11:20 PM · Restricted Project, Restricted Project
skatkov committed rG0a838ad51707: [GuardWidening] Add test showing incorrect behavior with nuw/nsw flags (authored by skatkov).
[GuardWidening] Add test showing incorrect behavior with nuw/nsw flags
May 25 2022, 11:05 PM · Restricted Project, Restricted Project
skatkov committed rG6f3f8b669b61: [GuardWidening] Update test with update_test_checks.py (authored by skatkov).
[GuardWidening] Update test with update_test_checks.py
May 25 2022, 10:30 PM · Restricted Project, Restricted Project
skatkov added a comment to D126354: [GuardWidening] Remove nuw/nsw flags for hoisted instructions.

Not familiar with this pass, but I suspect the right fix for this might be to use CreateLogicalAnd instead of CreateAnd.

May 25 2022, 9:39 PM · Restricted Project, Restricted Project
skatkov requested review of D126354: [GuardWidening] Remove nuw/nsw flags for hoisted instructions.
May 25 2022, 12:49 AM · Restricted Project, Restricted Project

May 14 2022

skatkov accepted D125538: [StatepointLowering] Properly handle local and non-local relocates of the same value..

Good catch.

May 14 2022, 10:50 AM · Restricted Project, Restricted Project

Apr 26 2022

skatkov added a comment to D124247: [Trivial Dead] Consider any non volatile load as trivially dead independent on ordering.

And thanks for the discussion anyway - the topic is pretty complex and discuss it with someone is really helpful.

Apr 26 2022, 8:22 PM · Restricted Project, Restricted Project
skatkov added a comment to D124247: [Trivial Dead] Consider any non volatile load as trivially dead independent on ordering.

Sequential consistency adds total order of all seq_cst operations to release-acquire semantic. However If load is not used its total order does not matter, so only acquire semantic plays role here.

Apr 26 2022, 8:21 PM · Restricted Project, Restricted Project
skatkov added a comment to D123848: [RS4GC] Don't clone BDV if its inputs are not derived.

Please go through all comments in findBasePointer and verify that they are up to date with you changes...

Apr 26 2022, 3:58 AM · Restricted Project, Restricted Project

Apr 25 2022

skatkov added a comment to D123530: Optimize cross block gc.relocate lowering. NFC..

Well.. Do I understand correctly that the original problem comes from the fact that during lowering of statepoint if value is relocated through register then this register is exported unconditionally?

Apr 25 2022, 11:33 PM · Restricted Project, Restricted Project
skatkov added a comment to D124247: [Trivial Dead] Consider any non volatile load as trivially dead independent on ordering.

Thanks, for you comment. Let me add details about my reasoning.
Let's we have load atomic %a with acquire or seq_cst semantic and this load has no uses.
According to specification this load has a ordering semantic only if it observes the value stored with release semantic to the same memory location.
If there is no uses then explicit check for is impossible. But it can be a implicit check like in example you mentioned.
So to do this implicit check we should load a value before our load and check its value. If our load must have the same value as previous load the previously loaded value may be used for such check.
Side note, let's consider there is only one store release to avoid mentioning "observe release store or later store" all the time.

Apr 25 2022, 9:23 PM · Restricted Project, Restricted Project
skatkov added a comment to D124247: [Trivial Dead] Consider any non volatile load as trivially dead independent on ordering.

Hi Eli, nice example. Do I understand correctly that here is the problem with exactly using of monotonic? If we used unordered (aka relaxed) then load2 can be safely removed?

Apr 25 2022, 4:35 AM · Restricted Project, Restricted Project

Apr 22 2022

skatkov added a comment to D124247: [Trivial Dead] Consider any non volatile load as trivially dead independent on ordering.

This isn't trivially correct.

For example, say you have a sequence "%load1 = load seq_cst %a; %load2 = load relaxed %a". I think the later load can sort of "inherit" the sequential consistency of the earlier load; the values it can contain are restricted. (I mean, maybe I'm missing something, but it's not as simple as "no one will check what actually value has been read".)

Apr 22 2022, 11:51 AM · Restricted Project, Restricted Project
skatkov added a comment to D123530: Optimize cross block gc.relocate lowering. NFC..

Generally looks good. Could please land test first in the current implementation to see the difference caused by patch.

Apr 22 2022, 4:33 AM · Restricted Project, Restricted Project
skatkov requested review of D124247: [Trivial Dead] Consider any non volatile load as trivially dead independent on ordering.
Apr 22 2022, 4:23 AM · Restricted Project, Restricted Project

Apr 18 2022

skatkov added a comment to D123848: [RS4GC] Don't clone BDV if its inputs are not derived.

I'm also looking into this. The idea is clear but I worry this adds a complexity with decoding the state.
I'm looking into the way to make it more straightforward. Probably it will require some re-factoring.

Apr 18 2022, 1:03 AM · Restricted Project, Restricted Project

Mar 29 2022

skatkov committed rG6444a65514b5: [LSR] Fixup canonicalization formula and its checker. (authored by skatkov).
[LSR] Fixup canonicalization formula and its checker.
Mar 29 2022, 12:06 AM · Restricted Project, Restricted Project
skatkov closed D122457: [LSR] Fix canonicalization formula and its checker..
Mar 29 2022, 12:05 AM · Restricted Project, Restricted Project

Mar 28 2022

skatkov updated the diff for D122457: [LSR] Fix canonicalization formula and its checker..
Mar 28 2022, 12:19 AM · Restricted Project, Restricted Project