Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

8264018: AArch64: NEON loadV2 and storeV2 addressing is wrong #3168

Closed
wants to merge 2 commits into from
Closed

8264018: AArch64: NEON loadV2 and storeV2 addressing is wrong #3168

wants to merge 2 commits into from
+76 −76

Conversation

nsjian
Copy link

@nsjian nsjian commented Mar 24, 2021 •
edited by openjdk bot
Loading

AArch64 c2 loadV2 and storeV2 are used for Vector API mask value load/store with size of halfword. Their load/store immediate offset should be similar to other vector load/store, e.g. loadV4/storeV4, valid only for unscaled signed offset of imm9 or scaled unsigned offset of imm12:scale. This patch also groups other loadVs/storeVs together in the ad file.

Verified with tier1 and jtreg vector api cases, though I cannot reproduce this issue.

Progress

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

Issue

  • JDK-8264018: AArch64: NEON loadV2 and storeV2 addressing is wrong

Reviewers

Download

To checkout this PR locally:
$ git fetch https://git.openjdk.java.net/jdk pull/3168/head:pull/3168
$ git checkout pull/3168

To update a local copy of the PR:
$ git checkout pull/3168
$ git pull https://git.openjdk.java.net/jdk pull/3168/head

Sorry, something went wrong.

Ningsheng Jian added 2 commits March 24, 2021 11:35
8264018: AArch64: NEON loadV2 and storeV2 addressing is wrong
ad93ccc 
AArch64 c2 loadV2 and storeV2 are used for Vector API mask value
load/store with size of halfword. Their load/store immediate offset
should be similar to other vector load/store, e.g. loadV4/storeV4,
valid only for unscaled signed offset of imm9 or scaled unsigned
offset of imm12:scale. Verified with tier1 and vector api cases,
though I cannot reproduce this issue.
Groups other loadVs/storeVs together in the ad file.
d05efb9
@bridgekeeper
Copy link

bridgekeeper bot commented Mar 24, 2021

👋 Welcome back njian! 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 openjdk bot added the rfr Pull request is ready for review label Mar 24, 2021
@openjdk
Copy link

openjdk bot commented Mar 24, 2021

@nsjian 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.

@openjdk openjdk bot added the hotspot-compiler hotspot-compiler-dev@openjdk.org label Mar 24, 2021
@mlbridge
Copy link

mlbridge bot commented Mar 24, 2021

Webrevs

@theRealAph
Copy link
Contributor

I admire your decision to tidy up the files, but I can't easily see what exactly has changed.

@adinn
Copy link
Contributor

adinn commented Mar 24, 2021

The change is to the memory operand of the first load and store rules. It was special cased to memory and is now vmem2 (just as the relocated rules used vmem4, vmem8 etc).

adinn
adinn approved these changes Mar 24, 2021
View reviewed changes
Copy link
Contributor

@adinn adinn left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good.

@openjdk
Copy link

openjdk bot commented Mar 24, 2021 •
edited
Loading

@nsjian 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:

8264018: AArch64: NEON loadV2 and storeV2 addressing is wrong

Reviewed-by: adinn, aph

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 11 new commits pushed to the master branch:

  • 6c0fbf7: 8254196: jshell infinite loops when startup script contains System.exit call
  • a79f095: 8263721: Unify oop casting
  • 329697b: 8263358: Update java.lang to use instanceof pattern variable
  • ae9af57: 8264001: JFR: Modernize implementation
  • fad8484: 8263411: Convert jshell tool to use Stream.toList()
  • 06d46d6: 8264008: Incorrect metaspace statistics after JEP 387 when UseCompressedClassPointers is off
  • 45e1bab: 8264091: Use the blessed modifier order in java.logging
  • cb776ed: 8263981: java.awt.image.ComponentSampleModel equals/hashcode use numBands twice
  • da512bf: 8264050: Remove unused field VM_HeapWalkOperation::_collecting_heap_roots
  • 6e3a158: 8263352: assert(use == polladr) failed: the use should be a safepoint polling
  • ... and 1 more: https://git.openjdk.java.net/jdk/compare/0b2aa1b6050eeeb4d13cbedfcfc86fb700e975a2...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.

@openjdk openjdk bot added the ready Pull request is ready to be integrated label Mar 24, 2021
@nsjian
Copy link
Author

nsjian commented Mar 24, 2021

I admire your decision to tidy up the files, but I can't easily see what exactly has changed.

I should explain the change in my PR comment better. I have two commits in this PR, the first commit is the main change to fix the issue, which adds vmem2 opclass for loadv2 and storev2, as their imm address ranges should be different from others. The second commit is the code tidy up.

@theRealAph
Copy link
Contributor

I admire your decision to tidy up the files, but I can't easily see what exactly has changed.

I should explain the change in my PR comment better. I have two commits in this PR, the first commit is the main change to fix the issue, which adds vmem2 opclass for loadv2 and storev2, as their imm address ranges should be different from others. The second commit is the code tidy up.

Ah, so it is. My mistake.

@nsjian
Copy link
Author

nsjian commented Mar 25, 2021

Thank you for the review! @adinn @theRealAph

@nsjian
Copy link
Author

nsjian commented Mar 25, 2021

/integrate

@openjdk openjdk bot closed this Mar 25, 2021
@openjdk openjdk bot added integrated Pull request has been integrated and removed ready Pull request is ready to be integrated rfr Pull request is ready for review labels Mar 25, 2021
@openjdk
Copy link

openjdk bot commented Mar 25, 2021

@nsjian Since your change was applied there have been 29 commits pushed to the master branch:

Your commit was automatically rebased without conflicts.

Pushed as commit 3e18330.

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

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@theRealAph theRealAph

@adinn adinn

Assignees
No one assigned
Labels
hotspot-compiler hotspot-compiler-dev@openjdk.org integrated Pull request has been integrated
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

None yet
3 participants
@nsjian @theRealAph @adinn