Re: Visual artifacts on radeon 9500 dri

Alex Jackson <[email protected]> Thu, 20 Mar 2008 01:41:47 -0400
Newsgroups gmane.comp.video.dri.user
Message-ID <[email protected]>
--===============1187116071==
Content-Type: multipart/alternative;
	boundary="_4dcc4ffe-8cee-4718-bc9d-cdfbf35b5e17_"

--_4dcc4ffe-8cee-4718-bc9d-cdfbf35b5e17_
Content-Type: text/plain; charset="Windows-1252"
Content-Transfer-Encoding: quoted-printable


> Date: Wed, 19 Mar 2008 19:51:57 -0400
> From: [email protected]
> To: [email protected]
> Subject: Re: [Dri-users] Visual artifacts on radeon 9500 dri
> CC: [email protected]; [email protected]
>=20
> On Wed, Mar 19, 2008 at 7:29 PM, Alex Jackson <[email protected]> =
wrote:
> > Wouldn't it be better, now that we have those register docs, to get the
> > number of working pipes from the card itself (specifically GB_PIPE_SELE=
CT)
> > rather than using a list of PCI ids?
>=20
> Yes, I'm working on it :)
> However, r3xx chips don't have GB_PIPE_SELECT.  Only r4xx, rs4xx and r5xx=
 do.

Yeah, I just looked at the newly-dropped R300 doc and noticed that there's =
no sign of that register.  I guess that's why the 9500s were easy to "softm=
od" in Windows, and there's no choice for those cards but to use a list of =
PCI IDs...

According to the R500 doc, it looks like you can just set GB_PIPE_SELECT to=
 the maximum (16 pipes) on all R4xx and R5xx cards and the card will automa=
tically correct it to the actual number of working pipes.  Is this right or=
 am I misinterpreting the doc?

I guess we'll want to query the card anyway for the real number of working =
pipes when/if we start tweaking the bits in GB_TILE_CONFIG for performance =
like the binary drivers (presumably) do.  For example it looks like there's=
 a bit in that register that's specific to 12-pipe configurations, and sele=
cts one of two ways of splitting the pixels between the three quads, any id=
ea how that works?  Is it something like

[1][2]
[3]

versus

   [1]
[2][3]

??

--AWJ--

_________________________________________________________________
Enter today for your chance to win $1000 a day=97today until May 12th. Lear=
n more at SignInAndWIN.ca
http://g.msn.ca/ca55/215=

--_4dcc4ffe-8cee-4718-bc9d-cdfbf35b5e17_
Content-Type: text/html; charset="Windows-1252"
Content-Transfer-Encoding: quoted-printable

<html>
<head>
<style>
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
FONT-SIZE: 10pt;
FONT-FAMILY:Tahoma
}
</style>
</head>
<body class=3D'hmmessage'>
&gt; Date: Wed, 19 Mar 2008 19:51:57 -0400<br>&gt; From: alexdeucher@gmail.=
com<br>&gt; To: [email protected]<br>&gt; Subject: Re: [Dri-users] V=
isual artifacts on radeon 9500 dri<br>&gt; CC: [email protected]; dri-us=
[email protected]<br>&gt; <br>&gt; On Wed, Mar 19, 2008 at 7:29 PM,=
 Alex Jackson &lt;[email protected]&gt; wrote:<br>&gt; &gt; Wouldn't=
 it be better, now that we have those register docs, to get the<br>&gt; &gt=
; number of working pipes from the card itself (specifically GB_PIPE_SELECT=
)<br>&gt; &gt; rather than using a list of PCI ids?<br>&gt; <br>&gt; Yes, I=
'm working on it :)<br>&gt; However, r3xx chips don't have GB_PIPE_SELECT. =
 Only r4xx, rs4xx and r5xx do.<br><br>Yeah, I just looked at the newly-drop=
ped R300 doc and noticed that there's no sign of that register.&nbsp; I gue=
ss that's why the 9500s were easy to "softmod" in Windows, and there's no c=
hoice for those cards but to use a list of PCI IDs...<br><br>According to t=
he R500 doc, it looks like you can just set GB_PIPE_SELECT to the maximum (=
16 pipes) on all R4xx and R5xx cards and the card will automatically correc=
t it to the actual number of working pipes.&nbsp; Is this right or am I mis=
interpreting the doc?<br><br>I guess we'll want to query the card anyway fo=
r the real number of working pipes when/if we start tweaking the bits in GB=
_TILE_CONFIG for performance like the binary drivers (presumably) do.&nbsp;=
 For example it looks like there's a bit in that register that's specific t=
o 12-pipe configurations, and selects one of two ways of splitting the pixe=
ls between the three quads, any idea how that works?&nbsp; Is it something =
like<br><br>[1][2]<br>[3]<br><br>versus<br><br>&nbsp;&nbsp; [1]<br>[2][3]<b=
r><br>??<br><br>--AWJ--<br><br /><hr />Sign in to Windows Live Messenger, a=
nd enter for your chance to win $1000 a day=97today until May 12th. <a href=
=3D'http://g.msn.ca/ca55/210' target=3D'_new'>Visit SignInAndWIN.ca</a></bo=
dy>
</html>=

--_4dcc4ffe-8cee-4718-bc9d-cdfbf35b5e17_--


--===============1187116071==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
--===============1187116071==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

--
_______________________________________________
Dri-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/dri-users

--===============1187116071==--