1220091 - Weird wrapping of header/footer on GitHub source view

Status: RESOLVED FIXED

(Core :: Panning and Zooming, defect, P1)

Description

[(not currently active) Ted Mielczarek]

Attachment: Capture.PNG

Browsing some source on GitHub, I noticed that when I paged down to certain points the GitHub header and footer would repeat over the page content.

Comment 1

[(not currently active) Ted Mielczarek]

Graphics
--------

Adapter Description: NVIDIA Quadro K600
Adapter Drivers: nvd3dumx,nvwgf2umx,nvwgf2umx,nvwgf2umx nvd3dum,nvwgf2um,nvwgf2um,nvwgf2um
Adapter RAM: 1024
Asynchronous Pan/Zoom: wheel input enabled
ClearType Parameters: Gamma: 1800 Pixel Structure: R ClearType Level: 100 Enhanced Contrast: 50
Device ID: 0x0ffa
Direct2D Enabled: true
DirectWrite Enabled: true (10.0.10240.16430)
Driver Date: 7-22-2015
Driver Version: 10.18.13.5362
GPU #2 Active: false
GPU Accelerated Windows: 2/2 Direct3D 11 (OMTC)
Subsys ID: 094b10de
Supports Hardware H264 Decoding: Yes
Vendor ID: 0x10de
WebGL Renderer: Google Inc. -- ANGLE (NVIDIA Quadro K600 Direct3D11 vs_5_0 ps_5_0)
windowLayerManagerRemote: true
AzureCanvasBackend: direct2d 1.1
AzureContentBackend: direct2d 1.1
AzureFallbackCanvasBackend: cairo
AzureSkiaAccelerated: 0

Comment 2

[(not currently active) Ted Mielczarek]

I'm on the latest Windows Nightly build from yesterday (2015-10-29).

Comment 3

[Timothy Nikkel (:tnikkel)]

Does the problem go away if you disable the pref layers.async-pan-zoom.enabled (requires restart)?

Comment 4

[(not currently active) Ted Mielczarek]

Yes, toggling that pref makes it go away.

I also figured out my actual STR:
1) Load that URL
2) Scroll down with the mouse wheel for a bit

It doesn't reproduce scrolling with the keyboard only. Once the graphical glitch is rendered, however, I can scroll with the keyboard and it persists.

Comment 5

[Kartikaya Gupta (email:[e-mail])]

The file got moved in the repo, so updating URL. However I can't seem to repro the issue on my OS X nightly. I'll try on Windows.

Comment 6

[David Anderson [:dvander] - inactive, e-mail if emergency]

When I try this on Windows with Direct2D, we assert immediately. The displayport is clamped to the maximum texture size (16k), but scrollbars ignore the displayport. In this case the scrollbars are ~135,000 pixels high, so we fail to allocate a new frontbuffer.

This is probably bug 1214261. (Assigning to re-test this after that bug is fixed.)

Comment 7

[Benoit Girard (:BenWa)]

I think this is the same bug as bug 1225868. The screenshot looks very similar to it.

:ted can you try and see if the bug is correlated with a page zoom like bug 1225868 is?

Comment 8

[David Anderson [:dvander] - inactive, e-mail if emergency]

(In reply to Benoit Girard (:BenWa) from comment #7)
> I think this is the same bug as bug 1225868. The screenshot looks very
> similar to it.
> 
> :ted can you try and see if the bug is correlated with a page zoom like bug
> 1225868 is?

See comment #6, I'm fairly certain this is bug 1214261. Disabling the assert or using an opt build repros the problem, and clamping scrollbars to the displayport makes it go away.

Comment 9

[Kartikaya Gupta (email:[e-mail])]

It could be that bug 1225868 is also a scrollbar issue - it's true that the screenshot here looks very similar to what :jdm was seeing.

Comment 10

[Benoit Girard (:BenWa)]

Right. I saw comment 6 which I don't disagree with. Like Kats said perhaps these bugs could all be related. In any case if we fix one we should check the others.

Comment 11

[Kartikaya Gupta (email:[e-mail])]

Please retest on the Jan 11 nightly; this should be fixed now that bug 1214261 has landed.

Comment 12

[(not currently active) Ted Mielczarek]

This doesn't reproduce for me anymore. Thanks! (Sorry I never answered the NI here...)