New issue
Advanced search Search tips
Note: Color blocks (like █ or █) mean that a user may not be available. Tooltip shows the reason.

Issue 798174 link

Starred by 2 users
Status: Duplicate
Merged: issue 772653
Owner:
█ oetu...@nvidia.com
Email to this user bounced
Closed: Jan 2018
Cc:
█ sc00335...@techmahindra.com
Components:
EstimatedDays: ----
NextAction: ----
OS: Linux
Pri: 2
Type: Bug-Regression

Blocked on:
issue 772653


Sign in to add a comment

GLSL bug: s=vec2(1) , m=mat2(s,s) give wrong result

Reported by fabrice....@gmail.com, Dec 31 2017 
UserAgent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/63.0.3239.108 Safari/537.36

Example URL:
https://www.shadertoy.com/view/XlBBRD

Steps to reproduce the problem:
see URL:
s=vec2(1) , m=mat2(s,s)  give wrong m
s=vec2(1) ; m=mat2(s,s)  give right m

What is the expected behavior?
m=(1,1,1,1)

What went wrong?
m=(0,0,0,0)

Does it occur on multiple sites: Yes

Is it a problem with a plugin? No 

Did this work before? N/A 

Does this work in other browsers? Yes

Chrome version: 63.0.3239.108  Channel: stable
OS Version: 14.04
Flash Version: Shockwave Flash 28.0 r0

wrong on chrome, right on firefox.   (I'm linux)

might be caused again by scalarize_vec_and_mat_constructor_args

somebody please set flags Blink>WebGL , Internals>GPU ?
thanks !

Comment 1 by fabrice....@gmail.com, Dec 31 2017 

Version 63.0.3239.108 (Build officiel) (64 bits)

NVIDIA Corporation
GL_RENDERER Quadro K2100M/PCIe/SSE2
GL_VERSION 4.5.0 NVIDIA 384.90
aboutgpuHome-Ubuntu14b.html
74.3 KB View Download

Comment 2 by kbr@chromium.org, Dec 31 2017

Cc: oetu...@nvidia.com
Components: -Blink Blink>WebGL
Labels: GPU-NVidia
Unless there's a bug in ANGLE's shader translator, this is probably a bug in the driver which we'll need to work around.

Comment 3 by kbr@chromium.org, Dec 31 2017

Blockedon: 772653

Comment 4 by fabrice....@gmail.com, Jan 1 2018 

it does not occurs in firefox, so it look to me quite similar to https://bugs.chromium.org/p/chromium/issues/detail?id=772653 , that occurred to be a bug in scalarize_vec_and_mat_constructor_args Angle turnaround called by chrome but not by firefox.

Comment 5 by fabrice....@gmail.com, Jan 1 2018 

( uh, maybe it's what you meant by "Blockedon: 772653" :-) )

Comment 6 by krajshree@chromium.org, Jan 1 2018

Labels: Needs-Triage-M63

Comment 7 by oetu...@nvidia.com, Jan 2 2018 

Does this reproduce for you on Chrome Dev channel? To me it seems like the issue is already fixed by the work I did for  bug 772653 .

Comment 8 by sc00335...@techmahindra.com, Jan 3 2018

Cc: -oetu...@nvidia.com sc00335...@techmahindra.com
Labels: -Type-Bug -Pri-2 hasbisect-per-revision ReleaseBlock-Stable Triaged-ET M-63 Pri-1 Type-Bug-Regression
Owner: oetu...@nvidia.com
Status: Assigned (was: Unconfirmed)
Able to reproduce this issue on reported version 63.0.3239.108 stable using Ubuntu 14.04,mesa driver but got fixed on 64.0.3282.39 and 65.0.3310.0 dev with URL given in comment#0. Hence providing reverse bisect info.

Reverse Bisect Info:
=====================
Last Bad Build: 64.0.3280.0
First Good Build: 64.0.3281.0

You are probably looking for a change made after 520140 (known good), but no later than 520141 (first known bad).
CHANGELOG URL:
  https://chromium.googlesource.com/chromium/src/+log/f5d8ece50864fc3e7278c33cd1b450bf2cb01128..aba4bb1b5fda581c28c7ba798de4e6fa54662b0b

Reviewed-on: https://chromium-review.googlesource.com/790412

Suspecting above change from changelog.

@oetuaho: Please confirm the bug and help in re-assigning if it is not related to your change. Adding RB-Stable for M-63. Remove if not the case. Please merge to M-63 if required.

Comment 9 by oetu...@nvidia.com, Jan 3 2018

Labels: -Pri-1 -ReleaseBlock-Stable -Needs-Triage-M63 Pri-2
Mergedinto: 772653
Status: Duplicate (was: Assigned)
I don't think this bug should be release blocker for 63 stable, the bug only happens in corner cases and integrating ANGLE changes is tricky. Reducing priority accordingly and closing this as a duplicate of  bug 772653 . If you feel that integrating this change back to 63 should still be looked at, notify kbr@chromium.org.
►

Sign in to add a comment