RE: Radeon DRI support patch

RE: Radeon DRI support patch

[daRonin]

Hi,

I install benh kernel and applied your patches on my new Powerbook Titanium.

Seems to be working, it's nice to get more than 400FPS. Unfortunately it still seems to be a bit flaky.

It does freeze the machine randomly when running GLX apps.

daRonin


> Also...  you may need to replace your
> /usr/X11R6/lib/modules/drivers/radeon_drv.o with
> http://www.unixbox.com/~ajoshi/radeon_drv.o.bz2

> ani


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

RE: Radeon DRI support patch

[Derrik Pates]

On Thu, 24 Jan 2002, daRonin wrote:

> I install benh kernel and applied your patches on my new Powerbook Titanium.
>
> Seems to be working, it's nice to get more than 400FPS. Unfortunately it still seems to be a bit flaky.
>
> It does freeze the machine randomly when running GLX apps.

That's the ever-lovely UniNorth AGP adding joy to your day. Ben's still
working on getting the kinks out of supporting it - it's weird.

Derrik Pates      |   Sysadmin, Douglas School   |    #linuxOS on EFnet
dpates@dsdk12.net |     District (dsdk12.net)    |    #linuxOS on OPN


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: Radeon DRI support patch

[Michel Lanners]

On  24 Jan, this message from Derrik Pates echoed through cyberspace:
>
> On Thu, 24 Jan 2002, daRonin wrote:
>
>> I install benh kernel and applied your patches on my new Powerbook Titanium.
>>
>> Seems to be working, it's nice to get more than 400FPS. Unfortunately it still seems to be a bit flaky.
>>
>> It does freeze the machine randomly when running GLX apps.
>
> That's the ever-lovely UniNorth AGP adding joy to your day. Ben's still
> working on getting the kinks out of supporting it - it's weird.

By chance, this would not be something comparable to what has been
discovered for the athlon processor under linux recently? Namely cache
coherency problems between processor and AGP.

>From what I understood (which may be wrong ;-) AGP gart is non-coherent
(by design?) . So, if the processor has a _cacheable_ mapping for the
same memory as the AGP card also has, there can be coherency problems if
the (valid) data in the processor's cache is modified by the AGP card
before the processor writes the cache line back to memory.

It was also mentioned that the erronous behaviour was related to
speculative writes, which are also use on the PPC processors.

Just a shot in the dark...

Michel

-------------------------------------------------------------------------
Michel Lanners                 |  " Read Philosophy.  Study Art.
23, Rue Paul Henkes            |    Ask Questions.  Make Mistakes.
L-1710 Luxembourg              |
email   mlan@cpu.lu            |
http://www.cpu.lu/~mlan        |                     Learn Always. "


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: Radeon DRI support patch

[Tony 'Nicoya' Mantler]

At 10:59 PM -0600 1/24/02, Michel Lanners wrote:
[...]
>By chance, this would not be something comparable to what has been
>discovered for the athlon processor under linux recently? Namely cache
>coherency problems between processor and AGP.
>
>>From what I understood (which may be wrong ;-) AGP gart is non-coherent
>(by design?) . So, if the processor has a _cacheable_ mapping for the
>same memory as the AGP card also has, there can be coherency problems if
>the (valid) data in the processor's cache is modified by the AGP card
>before the processor writes the cache line back to memory.
>
>It was also mentioned that the erronous behaviour was related to
>speculative writes, which are also use on the PPC processors.
>
>Just a shot in the dark...

It's not the speculative write that kills the athlon, it's the fact that
the incomplete speculative write dirties the otherwise clean cache line,
which then gets written back unmodified somewhere down the road.

Caching non-coherent segments should be safe-ish so long as you:

1: Write a full cache line, and always a full cache line.
2: Always flush sequentially after writing.
3: Always invalidate sequentially before reading.

The athlon's speculative writes would violate rules #1 and #2. If the PPC
dirties the cache line on it's incomplete speculative writes, then it also
violates the same rules. (Does it? Anyone?)

Of course, if the code itself violates these rules, then all hell breaks
loose. :)

Indeed the simple solution is to turn off caching on AGP-mapped pages.


Cheers - Tony 'Nicoya' Mantler :)


--
Tony "Nicoya" Mantler - Renaissance Nerd Extraordinaire - nicoya@apia.dhs.org
Winnipeg, Manitoba, Canada           --           http://nicoya.feline.pp.se/


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

RE: Radeon DRI support patch

[daRonin]

On Thu, 24 Jan 2002, Derrik Pates wrote:

> On Thu, 24 Jan 2002, daRonin wrote:
>
> > I install benh kernel and applied your patches on my new Powerbook Titanium.
> >
> > Seems to be working, it's nice to get more than 400FPS. Unfortunately it still seems to be a bit flaky.
> >
> > It does freeze the machine randomly when running GLX apps.
>
> That's the ever-lovely UniNorth AGP adding joy to your day. Ben's still
> working on getting the kinks out of supporting it - it's weird.

  Yes, weird would describe it. After running it for a while, even without
it crashing the system, after a while the GLX libs just seem to disappear
from /usr/X11/lib/modules/extensions. No log or console messages, nothing.
Very weird...

daRonin


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Radeon DRI support patch

[Ani Joshi]

For those interested in testing Radeon DRI, I've placed a patch located at
http://www.unixbox.com/~ajoshi/radeon-dri.diff which whould patch cleanly
against recent Ben H.'s kernel.  You must turn on AGPgart support, PCIgart
isn't done yet.

Also, you will need to replace your
/usr/X11R6/lib/modules/dri/radeon_dri.so with

http://www.unixbox.com/~ajoshi/radeon_dri.so.bz2

This was tested on a G4 Dual800 w/ Radeon AGP 32MB board, your results may
vary, so please include machine specs if you reply.


ani

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: Radeon DRI support patch

[Ani Joshi]

Also...  you may need to replace your
/usr/X11R6/lib/modules/drivers/radeon_drv.o with

http://www.unixbox.com/~ajoshi/radeon_drv.o.bz2


ani


On Mon, 21 Jan 2002, Ani Joshi wrote:

> For those interested in testing Radeon DRI, I've placed a patch located at
> http://www.unixbox.com/~ajoshi/radeon-dri.diff which whould patch cleanly
> against recent Ben H.'s kernel.  You must turn on AGPgart support, PCIgart
> isn't done yet.
>
> Also, you will need to replace your
> /usr/X11R6/lib/modules/dri/radeon_dri.so with
>
> http://www.unixbox.com/~ajoshi/radeon_dri.so.bz2
>
> This was tested on a G4 Dual800 w/ Radeon AGP 32MB board, your results may
> vary, so please include machine specs if you reply.

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/