8272315: Improve assert_different_registers #5083

kimbarrett · 2021-08-11T11:34:26Z

Please review this improvement to assert_different_register.

It reduces ~250 lines of repetitive code to ~10 lines, and improves the
error message on failure by printing a duplicated registers name.

Unfortunately, this isn't enough to eliminate AbstractRegister[Impl].
Although it seems to no longer be needed for the Oracle-supported platforms,
the linux-arm/ppc64le/s390 platforms all use the value() function provided
by AbstractRegisterImpl.

Testing:
tier1 on Oracle supported platforms.
built linux-arm/ppc64le/s390.

Locally changed a call to have a duplicate register and verified the error
message had a useful register name.

Progress

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

Issue

JDK-8272315: Improve assert_different_registers

Reviewers

Andrew Dinn (@adinn - Reviewer) ⚠️ Review applies to 839a5bb
Vladimir Ivanov (@iwanowww - Reviewer) ⚠️ Review applies to 839a5bb
Tobias Hartmann (@TobiHartmann - Reviewer) ⚠️ Review applies to 839a5bb

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.java.net/jdk pull/5083/head:pull/5083
$ git checkout pull/5083

Update a local copy of the PR:
$ git checkout pull/5083
$ git pull https://git.openjdk.java.net/jdk pull/5083/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 5083

View PR using the GUI difftool:
$ git pr show -t 5083

Using diff file

Download this PR as a diff file:
https://git.openjdk.java.net/jdk/pull/5083.diff

bridgekeeper · 2021-08-11T11:34:51Z

👋 Welcome back kbarrett! 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 · 2021-08-11T11:37:52Z

@kimbarrett 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 · 2021-08-11T11:41:37Z

Webrevs

adinn

Looks good.

openjdk · 2021-08-11T11:49:38Z

@kimbarrett 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:

8272315: Improve assert_different_registers

Reviewed-by: adinn, vlivanov, 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 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.

iwanowww

Looks nice!

kimbarrett · 2021-08-11T11:59:14Z

src/hotspot/share/asm/register.hpp

+  R regs[] = { first_register, more_registers... };
+  // Sort regs, so any equal entries are adjacent.
+  struct Compare { ptrdiff_t operator()(R x, R y) const { return x - y; } };
+  QuickSort::sort(regs, ARRAY_SIZE(regs), Compare(), false);


Sorting is kind of overkill. Maybe I should have just use the obvious nested loop? What do reviewers think?

Answering my own question, sorting imposes additional requirements on the type that might require adjustment for JDK-8270140. So I'm inclined to switch to the nested loop that only requires equality testing as before.

Well, a nested loop is order n^2 but n is always going to be pretty small (the old code allowed for 12 but cases with n > 6 are rare). So I see no great perf problem with using a nested loop.

... bearing in mind that quicksort is n.log(n) at best.

Sorting is kind of overkill. Maybe I should have just use the obvious nested loop? What do reviewers think?

No strong preference on my side. I'm fine with it either way.

Updated to use nested loops.

dean-long · 2021-08-11T21:20:57Z

Isn't this macro only used for integer registers? I have an alternative implementation for arm somewhere that uses RegisterSet, but you can just use an array of bytes or bits and check for duplicates as you go in O(n) time, right?

kimbarrett · 2021-08-12T03:52:12Z

Isn't this macro only used for integer registers? I have an alternative implementation for arm somewhere that uses RegisterSet, but you can just use an array of bytes or bits and check for duplicates as you go in O(n) time, right?

Probably, but that would require knowing the value range or dynamic allocation of the vector. And this is debug-only code, so keeping it simple would be nice. I don't think the N^2 behavior (which is what the old code had too) is a problem for the values of N we're dealing with.

TobiHartmann

Nice cleanup! Looks good to me.

adinn

Still good

kimbarrett · 2021-08-12T14:08:33Z

Thanks @TobiHartmann , @adinn , and @iwanowww for reviews.

kimbarrett · 2021-08-12T14:28:16Z

/integrate

openjdk · 2021-08-12T14:29:42Z

Going to push as commit 93cab7d.

openjdk · 2021-08-12T14:29:57Z

@kimbarrett Pushed as commit 93cab7d.

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

simplify assert_different_registers

5042f25

openjdk bot added the rfr Pull request is ready for review label Aug 11, 2021

openjdk bot added the hotspot-compiler hotspot-compiler-dev@openjdk.org label Aug 11, 2021

adinn approved these changes Aug 11, 2021

View reviewed changes

openjdk bot added the ready Pull request is ready to be integrated label Aug 11, 2021

iwanowww approved these changes Aug 11, 2021

View reviewed changes

kimbarrett commented Aug 11, 2021

View reviewed changes

nested loop instead of sort

839a5bb

TobiHartmann approved these changes Aug 12, 2021

View reviewed changes

adinn approved these changes Aug 12, 2021

View reviewed changes

iwanowww approved these changes Aug 12, 2021

View reviewed changes

Merge branch 'master' into reg_enum

2f8f773

openjdk bot closed this Aug 12, 2021

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 Aug 12, 2021

kimbarrett deleted the reg_enum branch August 12, 2021 14:33

8272315: Improve assert_different_registers #5083

8272315: Improve assert_different_registers #5083

kimbarrett commented Aug 11, 2021 •

edited by openjdk bot

bridgekeeper bot commented Aug 11, 2021

openjdk bot commented Aug 11, 2021

mlbridge bot commented Aug 11, 2021 •

edited

adinn left a comment

openjdk bot commented Aug 11, 2021 •

edited

iwanowww left a comment

kimbarrett Aug 11, 2021

kimbarrett Aug 11, 2021

adinn Aug 11, 2021

adinn Aug 11, 2021

iwanowww Aug 11, 2021

kimbarrett Aug 12, 2021

dean-long commented Aug 11, 2021

kimbarrett commented Aug 12, 2021

TobiHartmann left a comment

adinn left a comment

kimbarrett commented Aug 12, 2021

kimbarrett commented Aug 12, 2021

openjdk bot commented Aug 12, 2021

openjdk bot commented Aug 12, 2021

8272315: Improve assert_different_registers #5083

8272315: Improve assert_different_registers #5083

Conversation

kimbarrett commented Aug 11, 2021 • edited by openjdk bot

Progress

Issue

Reviewers

Reviewing

bridgekeeper bot commented Aug 11, 2021

openjdk bot commented Aug 11, 2021

mlbridge bot commented Aug 11, 2021 • edited

Webrevs

adinn left a comment

Choose a reason for hiding this comment

openjdk bot commented Aug 11, 2021 • edited

iwanowww left a comment

Choose a reason for hiding this comment

kimbarrett Aug 11, 2021

Choose a reason for hiding this comment

kimbarrett Aug 11, 2021

Choose a reason for hiding this comment

adinn Aug 11, 2021

Choose a reason for hiding this comment

adinn Aug 11, 2021

Choose a reason for hiding this comment

iwanowww Aug 11, 2021

Choose a reason for hiding this comment

kimbarrett Aug 12, 2021

Choose a reason for hiding this comment

dean-long commented Aug 11, 2021

kimbarrett commented Aug 12, 2021

TobiHartmann left a comment

Choose a reason for hiding this comment

adinn left a comment

Choose a reason for hiding this comment

kimbarrett commented Aug 12, 2021

kimbarrett commented Aug 12, 2021

openjdk bot commented Aug 12, 2021

openjdk bot commented Aug 12, 2021

kimbarrett commented Aug 11, 2021 •

edited by openjdk bot

mlbridge bot commented Aug 11, 2021 •

edited

openjdk bot commented Aug 11, 2021 •

edited