Bug 87143 - Frontbuffer tracking doesn't track movements in Xterm
Summary: Frontbuffer tracking doesn't track movements in Xterm
Status: CLOSED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: unspecified
Hardware: All All
: medium normal
Assignee: Intel GFX Bugs mailing list
QA Contact: Intel GFX Bugs mailing list
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2014-12-09 08:33 UTC by Sonika
Modified: 2017-07-24 22:50 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
Frontbuffer tracking (111.38 KB, text/plain)
2014-12-09 08:33 UTC, Sonika
no flags Details

Description Sonika 2014-12-09 08:33:55 UTC
Created attachment 110609 [details]
Frontbuffer tracking

HW: BYT RVP

Boot from USB with:
1. 1st partition contains grub and kernel bootimage
2. 2nd partition contains rootfs.

Kernel: drm-intel-nightly

Bug description
---------------
After booting up kernel and mouting of rootfs when an Xterm is started using startx, frontbuffer updates are not tracked in intel_frontbuffer.c
Calls to intel_frontbuffer_flush and intel_fb_obj_invalidate are not made:
1. When moving from console mode to UI mode and vice-versa (F1 and F7)
2. When mouse movements or keystrokes are performed.

In the attached logs, control was moved from F1 to F7 and vice-versa many times but a flush was seen only for the first time.
Comment 1 Chris Wilson 2014-12-09 09:06:27 UTC
Can you also clarify whether by cursor you mean the mouse pointer on the screen, or the insertion point (the solid rectangle or rectangle outline) in the xterm?
Comment 2 Sonika 2014-12-09 09:16:39 UTC
For the first time when I startX, I don't see mouse pointer. But the cursor (rectangle) is present but in disabled state. So I am unable to enter anything from keyboard.
From next time onwards, I start seeing the mouse pointer and I can bring the rectangle pointer to enabled and then I enter text on the terminal.
I don't see frontbuffer flush and invalidate for any of the above.
Comment 3 Daniel Vetter 2014-12-15 08:38:25 UTC
Please test

https://freedesktop.org/patch/38393/
Comment 4 Sonika 2014-12-15 08:52:20 UTC
(In reply to Daniel Vetter from comment #3)
> Please test
> 
> https://freedesktop.org/patch/38393/

In -nightly, I don't see any third argument to intel_fb_obj_flush function which is used in the patch for ORIGIN_GPU/ORIGIN_GTT.
Please let me know the equivalent function for this.
Comment 5 Jani Nikula 2015-02-12 14:13:59 UTC
Please try first five patches of http://mid.gmane.org/1423500395-1787-1-git-send-email-przanoni@gmail.com
Comment 6 Daniel Vetter 2015-03-18 10:16:21 UTC
Patches should all have landed now, can you please retest latest -nightly?
Comment 7 Ander Conselvan de Oliveira 2015-06-04 14:24:48 UTC
(In reply to Daniel Vetter from comment #6)
> Patches should all have landed now, can you please retest latest -nightly?

No answer since March. Presuming fixed.


Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.