8288294: [vector] Add Identity/Ideal transformations for vector logic operations #9211

XiaohongGong · 2022-06-20T07:50:09Z

This patch adds the following transformations for vector logic operations such as "AndV, OrV, XorV", incuding:

  (AndV v (Replicate m1))   => v
  (AndV v (Replicate zero)) => Replicate zero
  (AndV v v)                => v

  (OrV v (Replicate m1))    => Replicate m1
  (OrV v (Replicate zero))  => v
  (OrV v v)                 => v

  (XorV v v)                => Replicate zero

where "m1" is the integer constant -1, together with the same optimizations for vector mask operations like "AndVMask, OrVMask, XorVMask".

Progress

Change must be properly reviewed (1 review required, with at least 1 Reviewer)
Change must not contain extraneous whitespace
Commit message must refer to an issue

Issue

JDK-8288294: [vector] Add Identity/Ideal transformations for vector logic operations

Reviewers

Vladimir Kozlov (@vnkozlov - Reviewer) ⚠️ Review applies to ce06b861
Jatin Bhateja (@jatin-bhateja - Committer) ⚠️ Review applies to ce06b861

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk pull/9211/head:pull/9211
$ git checkout pull/9211

Update a local copy of the PR:
$ git checkout pull/9211
$ git pull https://git.openjdk.org/jdk pull/9211/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 9211

View PR using the GUI difftool:
$ git pr show -t 9211

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk/pull/9211.diff

… operations

bridgekeeper · 2022-06-20T07:51:44Z

👋 Welcome back xgong! 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 · 2022-06-20T07:53:53Z

@XiaohongGong 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 · 2022-06-20T07:58:33Z

Webrevs

01: Full (325f3333)
00: Full (ce06b861)

XiaohongGong · 2022-06-27T01:38:59Z

Hi, could anyone please help to take a look at this simple patch? Thanks a lot for your time!

vnkozlov · 2022-06-30T02:30:31Z

src/hotspot/share/opto/vectornode.cpp

+  }
+  // (AndV (Replicate zero) src)   => (Replicate zero)
+  // (AndVMask (MaskAll zero) src) => (MaskAll zero)
+  if (VectorNode::is_all_zeros_vector(in(1))) {


Why you expect it to be in(1) instead of in(2) as in previous case? Do we create inputs in such order based on mask value?
At least add comment explaining it.

We also have the case that the all zeros vector is in(2) in the followed codes. Please see line 1817. The main reason to do the different handle is the consideration for the predicated vector operations in Vector API.

That is what confuses me. The comment there says: masked operation requires the unmasked lanes to save the same values in the first operand.

I'm interpreting it as mask should be in(2):
https://github.com/openjdk/jdk/blob/master/src/hotspot/cpu/aarch64/aarch64_sve.ad#L379

But here you check in(1).

The comment means: for masked operations, the result of non-masked lanes should be from in(1), and the masked lanes are from the operation results.
For "AndV" with zero, the results is zero. So if in(1) is all zeros vector which is the expected result, no matter whether the AndV is masked or not, the result is right (i.e. for masked AndV, the non-masked lanes should be from in(1), and the masked lanes should be from the operation result which is also in(1)). But if the all zeros vector is in(2), this transformation will the results not right for masked AndV. That's why I added !is_predicated_vector() limit for the second case while not for the first one.

The comment means: for masked operations, the result of non-masked lanes should be from in(1), and the masked lanes are from the operation results.

Okay, I got it finally. Thank you for explanation.

vnkozlov · 2022-06-30T02:32:44Z

src/hotspot/share/opto/vectornode.cpp

+  }
+  // (OrV src (Replicate zero))   => src
+  // (OrVMask src (MaskAll zero)) => src
+  if (VectorNode::is_all_zeros_vector(in(2))) {


The same question as for AndVNode.

The same as AndVNode.

vnkozlov · 2022-06-30T02:41:43Z

src/hotspot/share/opto/vectornode.cpp

+  // (XorV src src)      => (Replicate zero)
+  // (XorVMask src src)  => (MaskAll zero)
+  //


Do we really need this? Is Replicate asm instruction faster than XorV? I understand it may help reduce registers pressure.

Thanks for looking at this patch! I think the main benefit is "(Replicate zero)" is loop invariant which could be hoist outside of the loop.

vnkozlov · 2022-06-30T04:48:35Z

@jatin-bhateja can you look on this. It is shared code and x86 is also affected.

vnkozlov

Tobias already ran testing and results are good.

openjdk · 2022-06-30T04:56:16Z

@XiaohongGong 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:

8288294: [vector] Add Identity/Ideal transformations for vector logic operations

Reviewed-by: kvn, jbhateja

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 no new commits pushed to the master branch. If another commit should be pushed before you perform the /integrate command, your PR will be automatically rebased. If you prefer to avoid any potential 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.

jatin-bhateja · 2022-06-30T05:17:11Z

We can also handle following vector constant folding cases :-
AndV (Replicate Const1) (Replicate Const2) => Replicate (Con1 And Con2)
OrV (Replicate Const1) (Replicate Const2) => Replicate (Con1 Or Con2 )

jatin-bhateja · 2022-06-30T05:17:50Z

Other than above patch looks good to me.

XiaohongGong · 2022-06-30T07:35:08Z

Thanks so much for the advice @jatin-bhateja ! I basically agree with this idea! It seems the similar optimization can also be applied to other binary arithmetic vector operations like add, sub, mul, div, shift ? So do you think it's better we create another patch special to handle the constant folding for such vector nodes? We'd better find a better way to handle this while not add the same transformation for each node. WDYT?

XiaohongGong · 2022-06-30T07:36:37Z

Tobias already ran testing and results are good.

Thanks for the review and testing @vnkozlov !

jatin-bhateja · 2022-06-30T08:47:55Z

Thanks so much for the advice @jatin-bhateja ! I basically agree with this idea! It seems the similar optimization can also be applied to other binary arithmetic vector operations like add, sub, mul, div, shift ? So do you think it's better we create another patch special to handle the constant folding for such vector nodes? We'd better find a better way to handle this while not add the same transformation for each node. WDYT?

Agree, addressing it in subsequent PR along with other operations should be ok. Thanks.

LGTM.

XiaohongGong · 2022-06-30T08:50:01Z

Thanks for the reivew @jatin-bhateja !

XiaohongGong · 2022-07-01T01:18:04Z

/integrate

openjdk · 2022-07-01T01:19:32Z

Going to push as commit 124c63c.
Since your change was applied there have been 4 commits pushed to the master branch:

918068a: Merge
c20b3aa: 8289278: Suspend/ResumeAllVirtualThreads need both can_suspend and can_support_virtual_threads
00d06d4: 8289440: Remove vmTestbase/nsk/monitoring/MemoryPoolMBean/isCollectionUsageThresholdExceeded/isexceeded003 from ProblemList.txt
feb223a: 8288707: javax/swing/JToolBar/4529206/bug4529206.java: setFloating does not work correctly

Your commit was automatically rebased without conflicts.

openjdk · 2022-07-01T01:19:43Z

@XiaohongGong Pushed as commit 124c63c.

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

8288294: [vector] Add Identity/Ideal transformations for vector logic…

ce06b86

… operations

openjdk bot added the rfr Pull request is ready for review label Jun 20, 2022

openjdk bot added the hotspot-compiler hotspot-compiler-dev@openjdk.org label Jun 20, 2022

vnkozlov reviewed Jun 30, 2022

View reviewed changes

vnkozlov approved these changes Jun 30, 2022

View reviewed changes

openjdk bot added the ready Pull request is ready to be integrated label Jun 30, 2022

jatin-bhateja approved these changes Jun 30, 2022

View reviewed changes

Merge branch 'jdk:master' into JDK-8288294

325f333

openjdk bot added the integrated Pull request has been integrated label Jul 1, 2022

openjdk bot closed this Jul 1, 2022

openjdk bot removed ready Pull request is ready to be integrated rfr Pull request is ready for review labels Jul 1, 2022

XiaohongGong deleted the JDK-8288294 branch July 1, 2022 01:20

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

8288294: [vector] Add Identity/Ideal transformations for vector logic operations #9211

8288294: [vector] Add Identity/Ideal transformations for vector logic operations #9211

XiaohongGong commented Jun 20, 2022 •

edited by openjdk bot

Loading

bridgekeeper bot commented Jun 20, 2022

openjdk bot commented Jun 20, 2022

mlbridge bot commented Jun 20, 2022 •

edited

Loading

XiaohongGong commented Jun 27, 2022

vnkozlov Jun 30, 2022

XiaohongGong Jun 30, 2022

vnkozlov Jun 30, 2022

XiaohongGong Jun 30, 2022

vnkozlov Jun 30, 2022

vnkozlov Jun 30, 2022

XiaohongGong Jun 30, 2022

vnkozlov Jun 30, 2022

XiaohongGong Jun 30, 2022

vnkozlov Jun 30, 2022

vnkozlov commented Jun 30, 2022

vnkozlov left a comment

openjdk bot commented Jun 30, 2022 •

edited

Loading

jatin-bhateja commented Jun 30, 2022 •

edited

Loading

jatin-bhateja commented Jun 30, 2022

XiaohongGong commented Jun 30, 2022

XiaohongGong commented Jun 30, 2022

jatin-bhateja commented Jun 30, 2022 •

edited

Loading

XiaohongGong commented Jun 30, 2022

XiaohongGong commented Jul 1, 2022

openjdk bot commented Jul 1, 2022

openjdk bot commented Jul 1, 2022

8288294: [vector] Add Identity/Ideal transformations for vector logic operations #9211

8288294: [vector] Add Identity/Ideal transformations for vector logic operations #9211

Conversation

XiaohongGong commented Jun 20, 2022 • edited by openjdk bot Loading

Progress

Issue

Reviewers

Reviewing

bridgekeeper bot commented Jun 20, 2022

openjdk bot commented Jun 20, 2022

mlbridge bot commented Jun 20, 2022 • edited Loading

Webrevs

XiaohongGong commented Jun 27, 2022

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

vnkozlov commented Jun 30, 2022

vnkozlov left a comment

Choose a reason for hiding this comment

openjdk bot commented Jun 30, 2022 • edited Loading

jatin-bhateja commented Jun 30, 2022 • edited Loading

jatin-bhateja commented Jun 30, 2022

XiaohongGong commented Jun 30, 2022

XiaohongGong commented Jun 30, 2022

jatin-bhateja commented Jun 30, 2022 • edited Loading

XiaohongGong commented Jun 30, 2022

XiaohongGong commented Jul 1, 2022

openjdk bot commented Jul 1, 2022

openjdk bot commented Jul 1, 2022

XiaohongGong commented Jun 20, 2022 •

edited by openjdk bot

Loading

mlbridge bot commented Jun 20, 2022 •

edited

Loading

openjdk bot commented Jun 30, 2022 •

edited

Loading

jatin-bhateja commented Jun 30, 2022 •

edited

Loading

jatin-bhateja commented Jun 30, 2022 •

edited

Loading