WebGPU: release on stable on Windows π
Categories
(Core :: Graphics: WebGPU, enhancement, P1)
Tracking
()
People
(Reporter: ErichDonGubler, Assigned: ErichDonGubler, NeedInfo)
References
(Blocks 1 open bug)
Details
(Keywords: dev-doc-complete)
Attachments
(3 files, 3 obsolete files)
| Assignee | ||
Updated•5 months ago
|
| Assignee | ||
Comment 1•5 months ago
|
||
Updated•5 months ago
|
Comment 2•5 months ago
|
||
(will require an addition to our 141 release notes)
Comment 3•5 months ago
|
||
Applying D253924 on top of bug 1972918.
Comment 4•5 months ago
•
|
||
BTW we still have known crashes in WPT, are they somehow okay to be ignored for now? Otherwise I wonder it should be a release blocker. https://searchfox.org/mozilla-central/search?q=expected%3A+CRASH&path=webgpu&case=true®exp=false
And also, as I just pinged you in Matrix, do we want to ship webgpu for workers too here?
| Assignee | ||
Updated•5 months ago
|
| Assignee | ||
Comment 6•5 months ago
|
||
Answering for Jim: Yes, we definitely want to ship support for all worker types minus service workers, which do not yet have CTS runs (see bug 1942431 and its dependencies).
| Assignee | ||
Updated•5 months ago
|
| Assignee | ||
Comment 7•5 months ago
|
||
(In reply to Kagami Rosylight [:saschanaz] (they/them) from comment #4)
BTW we still have known crashes in WPT, are they somehow okay to be ignored for now? Otherwise I wonder it should be a release blocker. https://searchfox.org/mozilla-central/search?q=expected%3A+CRASH&path=webgpu&case=true®exp=false
These are risks we've accepted for now with shipping. Our team is focusing on removing any CRASHes we think are dangerous to end users for the next iteration, with uplifts to Beta being our current plan. Most of these crashes are panics on conditions not being validated in earlier parts of the WebGPU API but which are standardized to be checked, but not exploitable. Some others are not panics, and we need a fast follow on these.
Comment 8•5 months ago
|
||
(In reply to Erich Gubler [:ErichDonGubler] (he/him) from comment #6)
Answering for Jim: Yes, we definitely want to ship support for all worker types minus service workers, which do not yet have CTS runs (see bug 1942431 and its dependencies).
minus service worker sounds like we need an extra pref, because the current pref covers all workers?
(In reply to Erich Gubler [:ErichDonGubler] (he/him) from comment #7)
These are risks we've accepted for now with shipping. Our team is focusing on removing any
CRASHes we think are dangerous to end users for the next iteration, with uplifts to Beta being our current plan. Most of these crashes are panics on conditions not being validated in earlier parts of the WebGPU API but which are standardized to be checked, but not exploitable. Some others are not panics, and we need a fast follow on these.
I guess that decision is based on crash stats on beta/nightly, right?
Updated•5 months ago
|
| Assignee | ||
Comment 9•5 months ago
|
||
:saschanaz: I'm moving conversation from here into D253924; let's keep things in one place, and not fragmented between two patches and this bug's feed. π
| Assignee | ||
Comment 10•5 months ago
|
||
I guess that decision is based on crash stats on beta/nightly, right?
So far, yes!
| Assignee | ||
Comment 11•5 months ago
•
|
||
I guess that decision is based on crash stats on beta/nightly, right?
My position on this is further explained in the patch.
| Assignee | ||
Updated•5 months ago
|
| Assignee | ||
Comment 12•5 months ago
|
||
Updated•5 months ago
|
| Assignee | ||
Updated•5 months ago
|
Comment 13•5 months ago
|
||
Comment 14•5 months ago
|
||
Comment 15•5 months ago
|
||
Reverted this because it was causing mochitets failures in browser_aboutURLs.js.
- Revert link
- Push with failures
- Failure Log
- Failure line: TEST-UNEXPECTED-FAIL | browser/components/contextualidentity/test/browser/browser_aboutURLs.js | A promise chain failed to handle a rejection: Component returned failure code: 0x8000ffff (NS_ERROR_UNEXPECTED) [nsIPrefBranch.getBoolPref] - stack: refreshPage@chrome://global/content/aboutInference.js:609:20
| Assignee | ||
Updated•5 months ago
|
Comment 16•5 months ago
|
||
Comment 17•5 months ago
|
||
| bugherder | ||
| Assignee | ||
Comment 18•5 months ago
|
||
Next step: uplift to Beta! π
| Assignee | ||
Comment 19•5 months ago
|
||
Original Revision: https://phabricator.services.mozilla.com/D255381
Updated•5 months ago
|
| Assignee | ||
Comment 20•5 months ago
|
||
Make IDL available in all contexts. Limit adapters to content, dedicated
workers, and shared workers; basically everything but service workers,
which is blocked on bug 1942431.
Original Revision: https://phabricator.services.mozilla.com/D253924
Updated•5 months ago
|
Updated•5 months ago
|
Updated•5 months ago
|
Updated•5 months ago
|
Comment 21•5 months ago
|
||
Aggregates D255381 and D253924 for uplift to Beta.
Original Revision: https://phabricator.services.mozilla.com/D255649
Updated•5 months ago
|
Updated•5 months ago
|
Comment 22•5 months ago
|
||
firefox-beta Uplift Approval Request
- User impact if declined: WebGPU will not ship in Fx141
- Code covered by automated testing: yes
- Fix verified in Nightly: yes
- Needs manual QE test: no
- Steps to reproduce for manual QE testing: none
- Risk associated with taking this patch: WebGPU is a new DOM API.
- Explanation of risk level: low; it's been in Nightly for years
- String changes made/needed: none
- Is Android affected?: no
Updated•5 months ago
|
Comment 23•5 months ago
|
||
| uplift | ||
| Assignee | ||
Updated•5 months ago
|
| Assignee | ||
Updated•5 months ago
|
Comment 24•5 months ago
|
||
Could we get a release note for this change? Thanks
https://wiki.mozilla.org/Release_Management/Release_Notes_Nomination#Nomination_in_Bugzilla
Updated•5 months ago
|
| Assignee | ||
Comment 25•5 months ago
|
||
Jim has stated in team meetings that he intends to be the one to write a release note for this.
Updated•4 months ago
|
Comment 26•4 months ago
|
||
FF141 MDN work for this can be tracked in https://github.com/mdn/content/issues/40473
Description
•