dropping UMS - xf86-video-ati-7.0.0

Luc Verhaegen libv at skynet.be
Mon Jun 18 09:41:53 PDT 2012


On Mon, Jun 18, 2012 at 04:04:05PM +0100, Robert Swindells wrote:
> 
> Michael D?nzer wrote:
> >On Don, 2012-06-14 at 20:19 +0100, Dave Airlie wrote: 
> >> I'm seriously thinking of resurrecting the kms killing branch,
> >
> >I'm not opposed, it's becoming rather painful to keep UMS building, let
> >alone working.
> >
> >> can anyone give me a reason not too,
> >
> >Not sure a reason is enough, someone would need to step up and take care
> >of UMS vs. ongoing development.
> 
> If people were interested in working on UMS to use on non-linux systems
> how would they get the documentation on recent chips ?
> 
> Robert Swindells

The only answer would be: read the kernel source.

Here are the docs that are available:
http://www.x.org/docs/AMD/
http://developer.amd.com/documentation/guides/Pages/default.aspx#open_gpu

Afaict, there is no register documentation being made available 
by ATI anymore. All you get is the shader instruction sets from AMDs 
GPGPU people.

The deal with AMD at the time was that all basic hw and display 
information would be made public (these days, our proposal can be read 
at 
http://www.phoronix.com/scan.php?page=article&item=amd_radeonhd_four&num=1), 
and that we at SuSE would not get any documentation that was not meant 
to be made public, but that we would get it already while it was in the 
process of being cleared legally. While the documents that we did get 
initially (soon 5 years ago) were cleared during the time the SuSE/AMD 
contract was in place, are made public today, the documents that we got 
later on, like some ATI R7xx registers docs and R7xx AtomBIOS 
documentation (we never saw AtomBIOS documentation for R5xx or any R6xx 
type chip), were not cleared in that timeframe and were subsequently not 
released, despite Mr Bridgmans initial assurance that these docs would 
go out.

So all you have is the kernel source and Matthias Hopfs excellent 
AtomBIOS disassembler (in as far as the latter is still useful on 
recent AtomBIOSes). The disassembler is a real life saver when trying to 
figure out how the AtomBIOS interface changes, or with spotting bugs in 
the BIOS, but i have no idea how useful that becomes without any 
register information.

Luc Verhaegen.


More information about the xorg-driver-ati mailing list