Bug 94842

Summary: Ditching xf86-video-ati in favor of xf86-video-modesetting?
Product: xorg Reporter: N. W. <nw9165-3201>
Component: Driver/RadeonAssignee: xf86-video-ati maintainers <xorg-driver-ati>
Status: RESOLVED NOTABUG QA Contact: Xorg Project Team <xorg-team>
Severity: normal    
Priority: medium    
Version: unspecified   
Hardware: All   
OS: All   
Whiteboard:
i915 platform: i915 features:

Description N. W. 2016-04-06 12:19:49 UTC
Hello,

according to the discussion in the following Phoronix forum thread:

https://www.phoronix.com/forums/forum/phoronix/latest-phoronix-articles/863332-intel-s-unreleased-3-0-x-org-driver-gets-more-fixes-for-dri3-present

the generic xf86-video-modesetting DDX driver included in xorg-server seems to be better than most vendor specific xf86-video- drivers.

I am wondering:

Then why not ditch xf86-video-ati in favor of xf86-video-modesetting?

Regards
Comment 1 Alex Deucher 2016-04-06 14:47:32 UTC
Older chips don't have the necessary functionality to support glamor.  They still need the asic specific acceleration code.  R600 and newer could eventually move to -modesetting once all the features from -ati get ported to -modesetting.
Comment 2 N. W. 2016-04-07 17:15:42 UTC
(In reply to Alex Deucher from comment #1)
> R600 and newer could
> eventually move to -modesetting once all the features from -ati get ported
> to -modesetting.

Is this something that you are planning to do? I mean porting all the features from xf86-video-ati to xf86-video-modesetting?

Or are you merely hinting at the possibility that someone could do it, in which case you would think about moving to x86-video-modesetting?

Regards
Comment 3 N. W. 2016-04-23 12:09:08 UTC
(In reply to Alex Deucher from comment #1)
> Older chips don't have the necessary functionality to support glamor.  They
> still need the asic specific acceleration code.  R600 and newer could
> eventually move to -modesetting once all the features from -ati get ported
> to -modesetting.

You've made it to the news, see:

http://www.phoronix.com/scan.php?page=news_item&px=Radeon-Modesetting-DDX-Xenial

;D
Comment 4 Hi-Angel 2017-02-11 23:22:36 UTC
(In reply to N. W. from comment #0)
> Hello,
> 
> according to the discussion in the following Phoronix forum thread:
> 
> https://www.phoronix.com/forums/forum/phoronix/latest-phoronix-articles/
> 863332-intel-s-unreleased-3-0-x-org-driver-gets-more-fixes-for-dri3-present
> 
> the generic xf86-video-modesetting DDX driver included in xorg-server seems
> to be better than most vendor specific xf86-video- drivers.

Problems mentioned there are specifically about xf86-video-intel. Although I do agree — it'd be cool if everyone contributed to a single driver.

> I am wondering:
> 
> Then why not ditch xf86-video-ati in favor of xf86-video-modesetting?
> 
> Regards

I was just wondering alike, but about my local PC. So, just FTR, if anyone would query the internet for the same question: upon research I didn't find much difference between -ati and -modesetting. From the latest benchmarks I found (April 2016) they seem to have comparable performance. Then I looked an activity of both drivers, and found that -modesetting¹ has latest commit 3(!) years ago, whilst -ati² just 4 days ago. So, for end-users: unless you see problems, it's better to stick with -ati.

1: https://cgit.freedesktop.org/xorg/driver/xf86-video-modesetting/log/
2: https://cgit.freedesktop.org/xorg/driver/xf86-video-ati/log/
Comment 5 Hi-Angel 2017-02-12 07:34:40 UTC
(In reply to Hi-Angel from comment #4)
> (In reply to N. W. from comment #0)
> > Hello,
> > 
> > according to the discussion in the following Phoronix forum thread:
> > 
> > https://www.phoronix.com/forums/forum/phoronix/latest-phoronix-articles/
> > 863332-intel-s-unreleased-3-0-x-org-driver-gets-more-fixes-for-dri3-present
> > 
> > the generic xf86-video-modesetting DDX driver included in xorg-server seems
> > to be better than most vendor specific xf86-video- drivers.
> 
> Problems mentioned there are specifically about xf86-video-intel. Although I
> do agree — it'd be cool if everyone contributed to a single driver.
> 
> > I am wondering:
> > 
> > Then why not ditch xf86-video-ati in favor of xf86-video-modesetting?
> > 
> > Regards
> 
> I was just wondering alike, but about my local PC. So, just FTR, if anyone
> would query the internet for the same question: upon research I didn't find
> much difference between -ati and -modesetting. From the latest benchmarks I
> found (April 2016) they seem to have comparable performance. Then I looked
> an activity of both drivers, and found that -modesetting¹ has latest commit
> 3(!) years ago, whilst -ati² just 4 days ago. So, for end-users: unless you
> see problems, it's better to stick with -ati.
> 
> 1: https://cgit.freedesktop.org/xorg/driver/xf86-video-modesetting/log/
> 2: https://cgit.freedesktop.org/xorg/driver/xf86-video-ati/log/

A correction: there seem to be a sound confusion: the official documentation for Glamor¹ points to old location of the code. One can easily get to -modesetting driver code by editing the URL as in "s/glamor/xf86-video-modesetting". I do also know that -modesetting was included into Xserver, which seems to be okay with those URLs, because they located in "xorg" part of repository.

Now a plot twist: the documentation is outdated, and the repositories are not the place of the code nowadays. One can see the activity of xf86-video-modesetting here https://cgit.freedesktop.org/xorg/xserver/log/hw/xfree86/drivers/modesetting

I'm very curious to how to edit this "wiki", it's not the first time I see this site provides terribly outdated information.

1: https://www.freedesktop.org/wiki/Software/Glamor/

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.