Below is a message I posted to x11@ regarding an apparent problem I am having with a new mobo and getting DRI support. At the very least, it appears that my AGP is not being probed correctly, and I hope that is where my DRI problem lies. The mobo is an ECS 755-A2 (Athlon 64 socket 754, but I am running i386, not amd64, and this is a fairly recent 5.3-STABLE) using the SiS 755 chipset. I do not find this listed in src/sys/pci/agp_sis.c It appears a guess is being made that this is a SiS 650 chipset, based on the output from pciconf -lv: device = 'SiSM650? Host-to-PCI Bridge' .
The device ID for that chipset (taken from agp_sis.c) is 0x06501039, but
pciconf shows mine to be 0x07551039. Presumably I can add another condition
to the case statement the generates that information and it will at least ID
correctly, but I've never done that and I want to make sure I do it correctly
before trying it. I would think something along the lines of:
case 0x07551039:
return ("SiS 755 host to AGP bridge");
but is all that going to accomplish is fixing the ID or is there more that
needs to be done?
---------- Forwarded Message ----------
Subject: Radeon DRI problems after system upgrade
Date: Monday 10 January 2005 07:45
From: Jonathan Fosburgh <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED]
I recently (this weekend) upgraded my aging P-III system to an AMD Athlon 64
(still running the i386 version of FreeBSD) and DRI has stopped working. I
am running a recent 5.3-STABLE with Xorg 6.8.1 and the current graphics/dri
from ports. AGP is compiled into the kernel. The chipset is SiS 755. I
receive the following error now when drm0 loads:
Jan 9 04:27:17 gw kernel: agp0: binding memory at bad offset 0
Jan 9 04:27:17 gw kernel: error: [drm:pid688:radeon_cp_init] *ERROR*
radeon_cp_init called without lock held
Jan 9 04:27:17 gw kernel: error: [drm:pid688:radeon_unlock] *ERROR* Process
688 using kernel context 0
From postings to mailing lists, this indicates that AGP is not loaded, but I
made sure to build it statically into the kernel when I first noticed the
problem. Here are some particulars:
Jan 9 04:27:17 gw kernel: drm0: <ATI Radeon QY RV100 7000/VE> port
0xd000-0xd0ff mem 0xe8020000-0xe802ffff,0xe0000000-0xe7ffffff irq 16 at
device 0.0 on pci1
Jan 9 04:27:17 gw kernel: info: [drm] AGP at 0xd8000000 128MB
Jan 9 04:27:17 gw kernel: info: [drm] Initialized radeon 1.11.0 20020828 on
minor 0
pciconf -lv reports the following for AGP:
[EMAIL PROTECTED]:0:0: class=0x060000 card=0x18911019 chip=0x07551039 rev=0x01
hdr=0x00
vendor = 'Silicon Integrated Systems (SiS)'
device = 'SiSM650? Host-to-PCI Bridge'
class = bridge
subclass = HOST-PCI
and for drm0:
[EMAIL PROTECTED]:0:0: class=0x030000 card=0x2081148c chip=0x51591002 rev=0x00
hdr=0x00
vendor = 'ATI Technologies Inc.'
device = 'RV100 Radeon 7000 / Radeon VE'
class = display
subclass = VGA
DRI used to work with this release of Xorg and the same -STABLE sources.
From old syslogs:
Dec 27 21:09:44 gw kernel: agp0: <VIA 82C691 (Apollo Pro) host to PCI bridge>
mem 0xe4000000-0xe7ffffff at device 0.0 on pci0
Dec 27 21:10:00 gw kernel: drm0: <ATI Radeon QY RV100 7000/VE> port
0xd800-0xd8ff mem 0xd7000000-0xd700ffff,0xd8000000-0xdfffffff irq 11 at
device 0.0 on pci1
Dec 27 21:10:00 gw kernel: info: [drm] AGP at 0xe4000000 64MB
Dec 27 21:10:00 gw kernel: info: [drm] Initialized radeon 1.11.0 20020828 on
minor 0
My only guess is that this has something to do with AGP support for this
chipset. Does anyone have a suggestion?
--
Jonathan Fosburgh
AIX and Storage Administrator
UT MD Anderson Cancer Center
Houston, TX
-------------------------------------------------------
--
Jonathan Fosburgh
AIX and Storage Administrator
UT MD Anderson Cancer Center
Houston, TX
pgp6lCQBqWPpu.pgp
Description: PGP signature
