1633628 - WebRender causes driver resets on gen6 on startup

Status: RESOLVED FIXED

(Core :: Graphics: WebRender, defect, P3)

Description

[Jeff Muizelaar [:jrmuizel]]

It looks like we're sending work to the GPU that takes >7 seconds. This causes a timeout and the driver gets reset.

It seems like it might have to do with our initial allocations.

Comment 1

[Jeff Muizelaar [:jrmuizel]]

Disabling picture caching seems to improve things substantially

Comment 2

[Jeff Muizelaar [:jrmuizel]]

It looks like others see this as well:
https://www.reddit.com/r/firefox/comments/g9u0r0/call_for_webrender_testers_on_intel_sandybridge/foy3sq9/

I'm able to load Firefox captures using wrench without issue, so there's definitely something weird going on.

Comment 3

[Darkspirit]

(In reply to Jeff Muizelaar [:jrmuizel] from comment #0)

initial allocations.

Is this a regression from bug 1598380 comment 8? (bug 1631353 comment 11)

Comment 4

[Jeff Muizelaar [:jrmuizel]]

Could be, I'll try reverting it.

Comment 5

[Jeff Muizelaar [:jrmuizel]]

Further investigation has discovered that disabling dual source blending avoids the hangs.

Comment 6

[Jeff Muizelaar [:jrmuizel]]

We still have the mystery of why dual source blending seems to work in other situations (wrench/picture-caching disabled) without problem.

Comment 7

[Jeff Muizelaar [:jrmuizel]]

So we discovered some more yesterday. It looks like the combination of dual source blending and bug 1580063 caused this.
That change added scissored clears and our theory is that the scissored clear ends up going through the regular 3d path and if the previous draw used dual source blending that would cause a TDR.

I was able to get an apitrace recording of the problem and replaying it also causes TDR, I confirmed that it was a call to ClearView() that causes the TDR and that the blend state is set to use dual source blending prior to calling ClearView().

I tried resetting the blend state to the default before doing the ClearView() but that didn't seem to help. As Lee suggested, it could well be that the blend state doesn't end up set on the GPU unless you actually do a draw with it.

ANGLE has an alternative code path that avoids the use of ClearView(), forcing that path avoids the problem.

Comment 8

[Jeff Muizelaar [:jrmuizel]]

Attachment: Bug 1633628. Don't use ClearView if we previously used dual source blending on Intel gen6.

Comment 9

[Kelsey Gilbert [:jgilbert]]

Attachment: Bug 1633628 - Vender: Don't use ClearView if we previously used dual source blending on Intel gen6.

Comment 10

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/302c30b598b0
Vender: Don't use ClearView if we previously used dual source blending on Intel gen6. r=jrmuizel

Comment 11

[Cosmin Sabou [:CosminS]]

https://hg.mozilla.org/mozilla-central/rev/302c30b598b0

Comment 12

[Pulsebot]

Backout by [e-mail]:
https://hg.mozilla.org/mozilla-central/rev/293d13ecf47e
Backed out changeset 302c30b598b0 for causing bug 1636122 a=backout

Comment 13

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/754436d8b3a2
Vender: Don't use ClearView if we previously used dual source blending on Intel gen6. r=jrmuizel

Comment 14

[Cosmin Sabou [:CosminS]]

https://hg.mozilla.org/mozilla-central/rev/754436d8b3a2

Comment 15

[BugBot [:suhaib / :marco/ :calixte]]

The patch landed in nightly and beta is affected.
:jrmuizel, is this bug important enough to require an uplift?
If not please set status_beta to wontfix.
If yes, don't forget to request an uplift for the patches in the regression caused by this fix.

For more information, please visit auto_nag documentation.

Comment 16

[Jeff Muizelaar [:jrmuizel]]

Nope. We're not turning on gen6 in beta