Bug 209538 – Bump boringssl version to M82

• Status: RESOLVED
• Resolution: FIXED
• Component: WebRTC

[youenn fablet]

Bump boringssl version to M82

[youenn fablet]

Created attachment 394485 [details]
Patch

[Eric Carlson]

Comment on attachment 394485 [details]
Patch

rs=me

[youenn fablet]

Created attachment 394587 [details]
Patch for landing

[EWS]

commit-queue failed to commit attachment 394587 [details] to WebKit repository.

[youenn fablet]

Created attachment 394592 [details]
Patch for landing

[EWS]

Committed r259042: <https://trac.webkit.org/changeset/259042>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 394592 [details].

[Radar WebKit Bug Importer]

<rdar://problem/60914881>

[WebKit Commit Bot]

Re-opened since this is blocked by bug 209602

[Alexey Proskuryakov]

Rolled back in http://trac.webkit.org/r259054

I noticed that the patch contained many changes without ChangeLogs, in WebCore and in LayoutTests. Seems like they were landed accidentally?

[youenn fablet]

Created attachment 394714 [details]
Patch

[youenn fablet]

(In reply to Alexey Proskuryakov from comment #9)
> Rolled back in http://trac.webkit.org/r259054
> 
> I noticed that the patch contained many changes without ChangeLogs, in
> WebCore and in LayoutTests. Seems like they were landed accidentally?

Right, I unexpectedly squashed these changes while doing a rebase.
This causes the test issues.

[EWS]

Committed r259111: <https://trac.webkit.org/changeset/259111>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 394714 [details].

[Aakash Jain]

(In reply to EWS from comment #12)
> Committed r259111: <https://trac.webkit.org/changeset/259111>
This seems to have broken webrtc/datachannel/multiple-connections.html on mac debug wk1.

History: https://results.webkit.org/?suite=layout-tests&test=webrtc%2Fdatachannel%2Fmultiple-connections.html&platform=mac&flavor=wk1

[youenn fablet]

(In reply to Aakash Jain from comment #13)
> (In reply to EWS from comment #12)
> > Committed r259111: <https://trac.webkit.org/changeset/259111>
> This seems to have broken webrtc/datachannel/multiple-connections.html on
> mac debug wk1.
> 
> History:
> https://results.webkit.org/?suite=layout-
> tests&test=webrtc%2Fdatachannel%2Fmultiple-connections.
> html&platform=mac&flavor=wk1

Hum, multiple-connections.html is a Slow test that tries to create connections like crazy. It might be that the new boringssl is much slower in Debug.

[youenn fablet]

Looking at https://build.webkit.org/results/Apple-Catalina-Debug-WK1-Tests/r259111%20(3416)/webrtc/datachannel/multiple-connections-stderr.txt, nothing wrong here, test seems to run normally.

[youenn fablet]

Test takes 22 seconds before boringssl bump.

[youenn fablet]

(In reply to youenn fablet from comment #16)
> Test takes 22 seconds before boringssl bump.

Test takes 31 seconds after boringssl bump but is timing out.
I would have expected a bigger time out here, 150000

[youenn fablet]

This does not repro locally when Tools/Scripts/run-webkit-tests --no-retry LayoutTests/webrtc/datachannel/multiple-connections.html -1 --iterations=50 -f --child-processes=5

This reproes with child-processes=20.