Bug 73514

Summary:

Slow performance with bochs

Product:

xorg

Reporter:

Richard Eames <freedesktop>

Component:

Driver/Radeon

Assignee:

xf86-video-ati maintainers <xorg-driver-ati>

Status:

RESOLVED MOVED

QA Contact:

Xorg Project Team <xorg-team>

Severity:

normal

Priority:

medium

Version:

git

Hardware:

x86-64 (AMD64)

OS:

Linux (All)

Whiteboard:

i915 platform:

i915 features:

Attachments:

Description	Flags
dmesg slow	none
Xorg log slow	none

Description Richard Eames 2014-01-12 02:04:17 UTC

I've been experiencing a performance regression when I'm using bochs, and using the latest git X and radeon. Browsing through boch's gui source, it looks like it's using a lot of bitmaps. I'm getting probably 1-2fps. What can I do to help narrow down what's causing this?

xserver-xorg-video-radeon:  Installed: 1:7.2.0+git1401090730.3213df~gd~s
linux-image-3.13.0-999-generic:  Installed: 3.13.0-999.201401020405

Comment 1 Alex Deucher 2014-01-12 23:29:56 UTC

What versions of the components were you using when the performance was better?  What hardware are you using?  Please attach your xorg log and dmesg output from both the well performing setup and the poorly performing setup.

Comment 2 Richard Eames 2014-01-14 02:38:23 UTC

Created attachment 92012 [details]
dmesg slow

Comment 3 Richard Eames 2014-01-14 02:38:43 UTC

Created attachment 92013 [details]
Xorg log slow

Comment 4 Richard Eames 2014-01-14 22:03:09 UTC

I've tried different builds of the kernel, mesa, and the radeon driver from over the past three weeks, and always get the performance issues. I think when I remember there not being an issue must have been when I was using fglrx. 

So, there's not really a "regression" per se, but there is a performance issue somewhere.

I forgot to mention in my first post, I'm using a duel card setup: Radeon R7 260x and a Radeon HD 4850.

Comment 5 Alex Deucher 2014-01-14 22:20:15 UTC

You might try the glamor patches in bug 64738, bug 68524, bug 71813.

Comment 6 Richard Eames 2014-03-16 00:00:32 UTC

Well, bug 64738 and bug 71813 didn't help, and in fact gtkperf shows they made things worse.

As for bug 68524, the current patch is still too slow to tell if it helps at all.

Comment 7 Michel Dänzer 2014-03-17 06:34:02 UTC

If the slowness is really related to bitmaps as you suggest, current glamor Git or the pending patches for glamor in the xserver tree might help.

Comment 8 Michel Dänzer 2015-04-23 01:51:25 UTC

With current xf86-video-ati Git master, does Option "ShadowPrimary" help? (I recommend enabling Option "TearFree" as well)

Comment 9 Martin Peres 2019-11-19 07:44:18 UTC

-- GitLab Migration Automatic Message --

This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/xorg/driver/xf86-video-ati/issues/90.

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.