8280600: C2: assert(!had_error) failed: bad dominance #7307

rwestrel · 2022-02-01T15:03:33Z

For the inner loop of the test, c2, first, creates a counted loop and
assigns the type -400..1 to the trip count Phi. The loop body has a
range check that guards a CastII on the trip count Phi. That CastII
has type: 0..1 as a result.

Pre/main/post loops are then created. In the process,
PhaseIdealLoop::cast_incr_before_loop() inserts a CastII right before
the main loop. That CastII's input is the add node of the pre
loop. The loop is also unrolled once. The CastII on the loop entry is
pushed through the add node and replaced by the range check CastII
that has the same input and dominates. As a result, the main loop
tripcount Phi has type -400..1 but init value, 1..2. This causes the
Phi to constant fold to 1. The amount of unrolling and the main loop
bounds are inconsistent (the loop shouldn't be unrolled if it executes
for one iteration) but when the unrolling decision is made the type of
the lower bound of the main loop is not yet accurately known.

The actual crash is the result of a chain of nodes sunk out of the
inner loop and pinned out of the loop with a CastII node. The type of
the CastII and the type of its input after unrolling conflict.

The fix I propose is to change the type of the CastII that's added
before the main loop so it can't be replaced by a dominating
CastII. Given that CastII's role is to carry dependencies for nodes in
the loop body on the loop entry test, I think it's the right thing to
do. I've also been working on 8275202 (C2: optimize out more redundant
conditions) as a way to avoid type inconsistencies like this one.

Progress

Change must not contain extraneous whitespace
Commit message must refer to an issue
Change must be properly reviewed

Issue

JDK-8280600: C2: assert(!had_error) failed: bad dominance

Reviewers

Vladimir Kozlov (@vnkozlov - Reviewer) ⚠️ Review applies to c218381
Tobias Hartmann (@TobiHartmann - Reviewer) ⚠️ Review applies to c218381

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.java.net/jdk pull/7307/head:pull/7307
$ git checkout pull/7307

Update a local copy of the PR:
$ git checkout pull/7307
$ git pull https://git.openjdk.java.net/jdk pull/7307/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 7307

View PR using the GUI difftool:
$ git pr show -t 7307

Using diff file

Download this PR as a diff file:
https://git.openjdk.java.net/jdk/pull/7307.diff

bridgekeeper · 2022-02-01T15:04:11Z

👋 Welcome back roland! A progress list of the required criteria for merging this PR into master will be added to the body of your pull request. There are additional pull request commands available for use with this pull request.

openjdk · 2022-02-01T15:07:23Z

@rwestrel The following label will be automatically applied to this pull request:

hotspot-compiler

When this pull request is ready to be reviewed, an "RFR" email will be sent to the corresponding mailing list. If you would like to change these labels, use the /label pull request command.

mlbridge · 2022-02-01T15:10:37Z

Webrevs

01: Full - Incremental (a888a0fd)
00: Full (c218381f)

vnkozlov

Seems good. Will run testing before approval.

vnkozlov

Testing passed.

openjdk · 2022-02-02T02:45:16Z

@rwestrel This change now passes all automated pre-integration checks.

ℹ️ This project also has non-automated pre-integration requirements. Please see the file CONTRIBUTING.md for details.

After integration, the commit message for the final commit will be:

8280600: C2: assert(!had_error) failed: bad dominance

Reviewed-by: kvn, thartmann

You can use pull request commands such as /summary, /contributor and /issue to adjust it as needed.

At the time when this comment was updated there had been 52 new commits pushed to the master branch:

4304a77: 8279535: C2: Dead code in PhaseIdealLoop::create_loop_nest after JDK-8276116
ab63834: 8280885: Shenandoah: Some tests failed with "EA: missing allocation reference path"
48a32b5: 8280976: Incorrect encoding of avx512 vpsraq instruction with mask and constant shift.
97af323: 8280842: Access violation in ciTypeFlow::profiled_count
d32f99e: 8279219: [REDO] C2 crash when allocating array of size too large
85d839f: 8280601: ClhsdbThreadContext.java test is triggering codecache related asserts
9ca7ff3: 8281082: Improve javadoc references to JOSS
c74b8f4: 8275914: SHA3: changing java implementation to help C2 create high-performance code
a18beb4: 8280867: Cpuid1Ecx feature parsing is incorrect for AMD CPUs
fdd9ca7: 8280642: ObjectInputStream.readObject should throw InvalidClassException instead of IllegalAccessError
... and 42 more: https://git.openjdk.java.net/jdk/compare/973dda5ce0747a8ea67ec3a34c2ef2a0b2b6b140...master

As there are no conflicts, your changes will automatically be rebased on top of these commits when integrating. If you prefer to avoid this automatic rebasing, please check the documentation for the /integrate command for further details.

➡️ To integrate this PR with the above commit message to the master branch, type /integrate in a new comment.

TobiHartmann

Looks reasonable to me.

TobiHartmann · 2022-02-02T07:13:22Z

test/hotspot/jtreg/compiler/loopopts/TestCastIIMakesMainLoopPhiDead.java

+  public static void main(String[] k) {
+      TestCastIIMakesMainLoopPhiDead t = new TestCastIIMakesMainLoopPhiDead();
+      for (int i = 0; i < 3; i++) {
+        t.test();


Indentation is inconsistent (4 whitespaces for Java code).

rwestrel · 2022-02-02T07:45:05Z

@vnkozlov thanks for review and testing
@TobiHartmann thanks for the review. I will fix indentation before I push the change.

rwestrel · 2022-02-02T07:59:47Z

/integrate

openjdk · 2022-02-02T08:01:16Z

Going to push as commit de826ba.
Since your change was applied there have been 52 commits pushed to the master branch:

4304a77: 8279535: C2: Dead code in PhaseIdealLoop::create_loop_nest after JDK-8276116
ab63834: 8280885: Shenandoah: Some tests failed with "EA: missing allocation reference path"
48a32b5: 8280976: Incorrect encoding of avx512 vpsraq instruction with mask and constant shift.
97af323: 8280842: Access violation in ciTypeFlow::profiled_count
d32f99e: 8279219: [REDO] C2 crash when allocating array of size too large
85d839f: 8280601: ClhsdbThreadContext.java test is triggering codecache related asserts
9ca7ff3: 8281082: Improve javadoc references to JOSS
c74b8f4: 8275914: SHA3: changing java implementation to help C2 create high-performance code
a18beb4: 8280867: Cpuid1Ecx feature parsing is incorrect for AMD CPUs
fdd9ca7: 8280642: ObjectInputStream.readObject should throw InvalidClassException instead of IllegalAccessError
... and 42 more: https://git.openjdk.java.net/jdk/compare/973dda5ce0747a8ea67ec3a34c2ef2a0b2b6b140...master

Your commit was automatically rebased without conflicts.

openjdk · 2022-02-02T08:01:39Z

@rwestrel Pushed as commit de826ba.

💡 You may see a message that your pull request was closed with unmerged commits. This can be safely ignored.

rwestrel added 2 commits February 1, 2022 11:04

fix

cf87a80

test

c218381

openjdk bot added the rfr Pull request is ready for review label Feb 1, 2022

openjdk bot added the hotspot-compiler hotspot-compiler-dev@openjdk.org label Feb 1, 2022

vnkozlov reviewed Feb 1, 2022

View changes

vnkozlov approved these changes Feb 2, 2022

View changes

openjdk bot added the ready Pull request is ready to be integrated label Feb 2, 2022

TobiHartmann approved these changes Feb 2, 2022

View changes

indentation

a888a0f

openjdk bot added the integrated Pull request has been integrated label Feb 2, 2022

openjdk bot closed this Feb 2, 2022

openjdk bot removed ready Pull request is ready to be integrated rfr Pull request is ready for review labels Feb 2, 2022

eme64 mentioned this pull request Feb 2, 2022

8281117: Add regression test for JDK-8280587 #7323

Closed

3 tasks

8280600: C2: assert(!had_error) failed: bad dominance #7307

8280600: C2: assert(!had_error) failed: bad dominance #7307

rwestrel commented Feb 1, 2022 •

edited by openjdk bot

bridgekeeper bot commented Feb 1, 2022

openjdk bot commented Feb 1, 2022

mlbridge bot commented Feb 1, 2022 •

edited

vnkozlov left a comment

vnkozlov left a comment

openjdk bot commented Feb 2, 2022 •

edited

TobiHartmann left a comment

TobiHartmann Feb 2, 2022

rwestrel commented Feb 2, 2022

rwestrel commented Feb 2, 2022

openjdk bot commented Feb 2, 2022

openjdk bot commented Feb 2, 2022

8280600: C2: assert(!had_error) failed: bad dominance #7307

8280600: C2: assert(!had_error) failed: bad dominance #7307

Conversation

rwestrel commented Feb 1, 2022 • edited by openjdk bot

Progress

Issue

Reviewers

Reviewing

bridgekeeper bot commented Feb 1, 2022

openjdk bot commented Feb 1, 2022

mlbridge bot commented Feb 1, 2022 • edited

Webrevs

vnkozlov left a comment

Choose a reason for hiding this comment

vnkozlov left a comment

Choose a reason for hiding this comment

openjdk bot commented Feb 2, 2022 • edited

TobiHartmann left a comment

Choose a reason for hiding this comment

TobiHartmann Feb 2, 2022

Choose a reason for hiding this comment

rwestrel commented Feb 2, 2022

rwestrel commented Feb 2, 2022

openjdk bot commented Feb 2, 2022

openjdk bot commented Feb 2, 2022

rwestrel commented Feb 1, 2022 •

edited by openjdk bot

mlbridge bot commented Feb 1, 2022 •

edited

openjdk bot commented Feb 2, 2022 •

edited