8282231: x86-32: runtime call to SharedRuntime::ldiv corrupts registers #7575

theRealAph · 2022-02-22T13:34:03Z

In x86_32.ad, a call to SharedRuntime::ldiv doesn't save all of the call-clobbered registers. Also SharedRuntime::lrem.
effect(CALL) is the simplest way to fix this, and AFAIK the most efficient too.

This is by no means the only bug in this area, but it's trivial to fix and allows a bootcycle build to complete on recent Linuxes.

Progress

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

Issue

JDK-8282231: x86-32: runtime call to SharedRuntime::ldiv corrupts registers

Reviewers

Aleksey Shipilev (@shipilev - Reviewer)
Jie Fu (@DamonFool - Reviewer)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.java.net/jdk pull/7575/head:pull/7575
$ git checkout pull/7575

Update a local copy of the PR:
$ git checkout pull/7575
$ git pull https://git.openjdk.java.net/jdk pull/7575/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 7575

View PR using the GUI difftool:
$ git pr show -t 7575

Using diff file

Download this PR as a diff file:
https://git.openjdk.java.net/jdk/pull/7575.diff

Co-authored-by: Aleksey Shipilev <shade@openjdk.org>

bridgekeeper · 2022-02-22T13:35:32Z

👋 Welcome back aph! 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 · 2022-02-22T13:37:25Z

⚠️ @theRealAph This pull request contains merges that bring in commits not present in the target repository. Since this is not a "merge style" pull request, these changes will be squashed when this pull request in integrated. If this is your intention, then please ignore this message. If you want to preserve the commit structure, you must change the title of this pull request to Merge <project>:<branch> where <project> is the name of another project in the OpenJDK organization (for example Merge jdk:master).

openjdk · 2022-02-22T13:37:59Z

@theRealAph 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 · 2022-02-22T13:41:18Z

Webrevs

shipilev

Yeah, I did a similar thing here: #7498 -- I think we can drop cr, cx, bx from match rule arguments too. Also drop spaces between parentheses, while we are at it. That change passed tier{1,2,3} on x86_32, I am sure this one passes it too.

theRealAph · 2022-02-22T14:23:38Z

Yeah, I did a similar thing here: #7498 -- I think we can drop cr, cx, bx from match rule arguments too. Also drop spaces between parentheses, while we are at it. That change passed tier{1,2,3} on x86_32, I am sure this one passes it too.

Yes, we seem to have collided. Suggest we go ahead with this one, OK? I've put you in as a co-author, so I guess someone else will have to review it.

cr, cx, and bx are call-clobbered, so we don't need them as an effect, I believe.

DamonFool · 2022-02-22T14:49:31Z

src/hotspot/cpu/x86/x86_32.ad

@@ -7827,7 +7827,7 @@ instruct divI_eReg(eAXRegI rax, eDXRegI rdx, eCXRegI div, eFlagsReg cr) %{
 // Divide Register Long
 instruct divL_eReg( eADXRegL dst, eRegL src1, eRegL src2, eFlagsReg cr, eCXRegI cx, eBXRegI bx ) %{


I agree with @shipilev that we can remove eFlagsReg cr, eCXRegI cx, eBXRegI bx here, right?

Done. I deleted them in the effect list but forgot to delete the corresponding args.

DamonFool · 2022-02-22T14:50:05Z

So how did you find this bug?
Is it possible to add a jtreg test for this fix?

Please also update the copyright year.
Thanks.

theRealAph · 2022-02-22T15:31:35Z

So how did you find this bug?

Found during testing of a new OS build. Fails bootstrap.

Is it possible to add a jtreg test for this fix?

No, because it corrupts registers in arbitrary places, causing seemingly-random crashes.
If you're lucky it'll corrupt a pointer value, in which case you're almost guaranteed a crash.
If you're unlucky it'll corrupt some data value somewhere. It's a Heisenbug.

Please also update the copyright year. Thanks.

OK.

DamonFool

Looks good to me.
Thanks for the update.

openjdk · 2022-02-22T23:22:31Z

@theRealAph 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:

8282231: x86-32: runtime call to SharedRuntime::ldiv corrupts registers

Reviewed-by: shade, jiefu

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 298 new commits pushed to the master branch:

e1060be: 8281615: Deadlock caused by jdwp agent
6f882de: 8280964: [Linux aarch64] : drawImage dithers TYPE_BYTE_INDEXED images incorrectly
2557ef8: 8282276: Problem list failing two Robot Screen Capture tests
6445ee4: 5041655: (ch) FileLock: negative param and overflow issues
7feabee: 8261407: ReflectionFactory.checkInitted() is not thread-safe
58e1882: 8282042: [testbug] FileEncodingTest.java depends on default encoding
3cb3867: 8281315: Unicode, (?i) flag and backreference throwing IndexOutOfBounds Exception
957dae0: 8280958: G1/Parallel: Unify marking code structure
e44d067: 8244593: Clean up GNM/NM after JEP 381
41355e2: 8276686: Malformed Javadoc inline tags in JDK source in /java/util/regex/Pattern.java
... and 288 more: https://git.openjdk.java.net/jdk/compare/f34f8d4d6a9b3e24a93a322b985c1413c27311cc...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.

theRealAph · 2022-02-23T10:14:00Z

/integrate

openjdk · 2022-02-23T10:15:47Z

Going to push as commit ecd85e6.
Since your change was applied there have been 299 commits pushed to the master branch:

378fa50: 8281962: Avoid unnecessary native calls in InflaterInputStream
e1060be: 8281615: Deadlock caused by jdwp agent
6f882de: 8280964: [Linux aarch64] : drawImage dithers TYPE_BYTE_INDEXED images incorrectly
2557ef8: 8282276: Problem list failing two Robot Screen Capture tests
6445ee4: 5041655: (ch) FileLock: negative param and overflow issues
7feabee: 8261407: ReflectionFactory.checkInitted() is not thread-safe
58e1882: 8282042: [testbug] FileEncodingTest.java depends on default encoding
3cb3867: 8281315: Unicode, (?i) flag and backreference throwing IndexOutOfBounds Exception
957dae0: 8280958: G1/Parallel: Unify marking code structure
e44d067: 8244593: Clean up GNM/NM after JEP 381
... and 289 more: https://git.openjdk.java.net/jdk/compare/f34f8d4d6a9b3e24a93a322b985c1413c27311cc...master

Your commit was automatically rebased without conflicts.

openjdk · 2022-02-23T10:16:20Z

@theRealAph Pushed as commit ecd85e6.

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

magicus and others added 11 commits October 13, 2021 01:49

Create hsdis backend using LLVM

effac9b

Merge https://github.com/openjdk/jdk

93537db

Merge https://github.com/openjdk/jdk

c08cf91

Merge branch 'master' of https://github.com/theRealAph/jdk

661026c

Merge

7f9794f

Merge branch 'master' of https://github.com/theRealAph/jdk

4042c89

Merge https://github.com/openjdk/jdk

8aed7a9

Merge https://github.com/openjdk/jdk

fae1e7e

Merge https://github.com/openjdk/jdk into master

eaf672b

Revert effac9b

1c14171

8282231: x86-32: runtime call to SharedRuntime::ldiv corrupts registers

fb02aa5

Co-authored-by: Aleksey Shipilev <shade@openjdk.org>

openjdk bot added the rfr Pull request is ready for review label Feb 22, 2022

openjdk bot added the hotspot-compiler hotspot-compiler-dev@openjdk.org label Feb 22, 2022

shipilev reviewed Feb 22, 2022

View reviewed changes

Whitespace

d912370

DamonFool reviewed Feb 22, 2022

View reviewed changes

Delete unused args

1382812

Copyright

82423ef

DamonFool approved these changes Feb 22, 2022

View reviewed changes

openjdk bot added the ready Pull request is ready to be integrated label Feb 22, 2022

shipilev approved these changes Feb 23, 2022

View reviewed changes

openjdk bot added the integrated Pull request has been integrated label Feb 23, 2022

openjdk bot closed this Feb 23, 2022

openjdk bot removed ready Pull request is ready to be integrated rfr Pull request is ready for review labels Feb 23, 2022

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

8282231: x86-32: runtime call to SharedRuntime::ldiv corrupts registers #7575

8282231: x86-32: runtime call to SharedRuntime::ldiv corrupts registers #7575

theRealAph commented Feb 22, 2022 •

edited by openjdk bot

Loading

bridgekeeper bot commented Feb 22, 2022

openjdk bot commented Feb 22, 2022

openjdk bot commented Feb 22, 2022

mlbridge bot commented Feb 22, 2022 •

edited

Loading

shipilev left a comment

theRealAph commented Feb 22, 2022

DamonFool Feb 22, 2022

theRealAph Feb 22, 2022

DamonFool commented Feb 22, 2022

theRealAph commented Feb 22, 2022 •

edited

Loading

DamonFool left a comment

openjdk bot commented Feb 22, 2022 •

edited

Loading

theRealAph commented Feb 23, 2022

openjdk bot commented Feb 23, 2022

openjdk bot commented Feb 23, 2022

		@@ -7827,7 +7827,7 @@ instruct divI_eReg(eAXRegI rax, eDXRegI rdx, eCXRegI div, eFlagsReg cr) %{
		// Divide Register Long
		instruct divL_eReg( eADXRegL dst, eRegL src1, eRegL src2, eFlagsReg cr, eCXRegI cx, eBXRegI bx ) %{

8282231: x86-32: runtime call to SharedRuntime::ldiv corrupts registers #7575

8282231: x86-32: runtime call to SharedRuntime::ldiv corrupts registers #7575

Conversation

theRealAph commented Feb 22, 2022 • edited by openjdk bot Loading

Progress

Issue

Reviewers

Reviewing

bridgekeeper bot commented Feb 22, 2022

openjdk bot commented Feb 22, 2022

openjdk bot commented Feb 22, 2022

mlbridge bot commented Feb 22, 2022 • edited Loading

Webrevs

shipilev left a comment

Choose a reason for hiding this comment

theRealAph commented Feb 22, 2022

DamonFool Feb 22, 2022

Choose a reason for hiding this comment

theRealAph Feb 22, 2022

Choose a reason for hiding this comment

DamonFool commented Feb 22, 2022

theRealAph commented Feb 22, 2022 • edited Loading

DamonFool left a comment

Choose a reason for hiding this comment

openjdk bot commented Feb 22, 2022 • edited Loading

theRealAph commented Feb 23, 2022

openjdk bot commented Feb 23, 2022

openjdk bot commented Feb 23, 2022

theRealAph commented Feb 22, 2022 •

edited by openjdk bot

Loading

mlbridge bot commented Feb 22, 2022 •

edited

Loading

theRealAph commented Feb 22, 2022 •

edited

Loading

openjdk bot commented Feb 22, 2022 •

edited

Loading