From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-yw0-x22f.google.com (mail-yw0-x22f.google.com [IPv6:2607:f8b0:4002:c05::22f]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by lists.ozlabs.org (Postfix) with ESMTPS id F18C31A0221 for ; Fri, 5 Feb 2016 13:47:27 +1100 (AEDT) Received: by mail-yw0-x22f.google.com with SMTP id g127so40751086ywf.2 for ; Thu, 04 Feb 2016 18:47:27 -0800 (PST) MIME-Version: 1.0 Date: Fri, 5 Feb 2016 02:47:24 +0000 Message-ID: Subject: PowerPC agpmode issues From: Mike To: Pranith Kumar Cc: "Aneesh Kumar K.V" , "open list:LINUX FOR POWERPC..." Content-Type: multipart/alternative; boundary=94eb2c081bb6ffb839052afcde8a List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , --94eb2c081bb6ffb839052afcde8a Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Hi. Managed to get the Radeon R300 running on mesa 11.1.1 with an old 2013 patch from Michel D=C3=A0nzer, next problem is of course enabling agpmode, running with pci-mode with radeon.agpmode=3D-1 works, but is of course slow= , and seems to load the cpu a lot. Upon initial investigation i could not initially believe agp could be this this broken for this long, until i found this. "committed with Ben Skeggs on Feb 26, 2013" https://github.com/DespairFactor/bullhead/commit/650e1203c11354ba84d69ba445= abc0efcfe3890a http://lxr.free-electrons.com/source/drivers/gpu/drm/nouveau/nouveau_agp.c?= v=3D4.2 #ifdef __powerpc__ /* Disable AGP by default on all PowerPC machines for * now -- At least some UniNorth-2 AGP bridges are * known to be broken: DMA from the host to the card * works just fine, but writeback from the card to the * host goes straight to memory untranslated bypassing * the GATT somehow, making them quite painful to deal * with... */ if (nouveau_agpmode =3D=3D -1) return false; #endif and now later this: https://github.com/torvalds/linux/blob/master/drivers/gpu/drm/nouveau/nvkm/= subdev/pci/agp.c #ifdef __powerpc__ /* Disable AGP by default on all PowerPC machines for now -- At * least some UniNorth-2 AGP bridges are known to be broken: * DMA from the host to the card works just fine, but writeback * from the card to the host goes straight to memory * untranslated bypassing that GATT somehow, making them quite * painful to deal with... */ mode =3D 0; #endif All seems to point to serious issues had around the time of change to ums to kms and a serious regression hitting the linux kernel? No? Cheers -Mike --94eb2c081bb6ffb839052afcde8a Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
Hi.= =C2=A0
Managed to get the Radeon R300 runni= ng on mesa 11.1.1 with an old 2013 patch from Michel D=C3=A0nzer, next prob= lem is of course enabling agpmode, running with pci-mode with radeon.agpmod= e=3D-1 works, but is of course slow, and seems to load the cpu a lot.
=

Upon initia= l investigation i could not initially believe agp could be this this broken= for this long, until i found this.
=C2=A0"committed with Ben Skeggs on Feb 26, 2013&q= uot;
https://git= hub.com/DespairFactor/bullhead/commit/650e1203c11354ba84d69ba445abc0efcfe38= 90a
http://lx= r.free-electrons.com/source/drivers/gpu/drm/nouveau/nouveau_agp.c?v=3D4.2
#ifdef __powerpc__
/* Disable A= GP by default on all PowerPC machines for
<= span class=3D"" style=3D"white-space:pre"> * now -- At least some = UniNorth-2 AGP bridges are
* known to be broken: DMA from the hos= t to the card
* works just fine, but writeback from the card to t= he
* host goes straight to memory untranslated bypassing
* the GATT somehow, making them quite painful to deal
* with...=
*/
if (nouveau_agpmode =3D=3D -1)
return fals= e;
#endif
= =C2=A0
=C2=A0and now later this:=C2=A0
https://github.com= /torvalds/linux/blob/master/drivers/gpu/drm/nouveau/nvkm/subdev/pci/agp.c
#ifdef __powerpc__
/* Disable AG= P by default on all PowerPC machines for now -- At
* least some Un= iNorth-2 AGP bridges are known to be broken:
* DMA from the host t= o the card works just fine, but writeback
<= span class=3D"" style=3D"white-space:pre"> * from the card to the h= ost goes straight to memory
* untranslated bypassing that GATT som= ehow, making them quite
* painful to deal with...
*/
mode =3D 0;
#endif

All seems to point to ser= ious issues had around the time of change to ums to kms and a serious regre= ssion hitting the linux kernel? No?

Cheers
-Mike<= /div>
--94eb2c081bb6ffb839052afcde8a--