AGP serverworks chipset not being initializated properly?

From: Diego Calleja
Date: Sun Feb 12 2006 - 20:26:10 EST


For a while, I've been having this on my dmesg:

agpgart: Xorg tried to set rate=x12. Setting to AGP3 x8 mode.
agpgart: Putting AGP V2 device at 0000:00:00.1 into 2x mode
agpgart: Putting AGP V2 device at 0000:01:00.0 into 2x mode
[drm] Loading R200 Microcode

This is really weird since my motherboard (CNB20HE chipset)
only supports 2x AGP cards and my xorg.conf file has this:

Section "Device"
Identifier "RADEON9200SE"
Driver "radeon"
Option "AGPMode" "2"
EndSection


Looking at the code I found out that the code path being run to get
this printk is:
------------------------------------------------------------------
/* Check to see if we are operating in 3.0 mode */
if (agp_bridge->mode & AGPSTAT_MODE_3_0)
agp_v3_parse_one(&requested_mode, &bridge_agpstat, &vga_agpstat); <----------
else
agp_v2_parse_one(&requested_mode, &bridge_agpstat, &vga_agpstat);
------------------------------------------------------------------


so it looks like agp_brige->mode is not being initializated properly, it is
sworks-agp.c:serverworks_configure() who initializates it in my box so
something has to be wrong with one of these 2 lines?:
------------------------------------------------------------------
agp_bridge->capndx = pci_find_capability(serverworks_private.svrwrks_dev, PCI_CAP_ID_AGP);

/* Fill in the mode register */
pci_read_config_dword(serverworks_private.svrwrks_dev,
agp_bridge->capndx+PCI_AGP_STATUS, &agp_bridge->mode);
------------------------------------------------------------------


What is going on, is this a know bug, should I just ignore it (it doesn't
harms anything), file a bugzilla bug? My dmesg says:
Linux agpgart interface v0.101 (c) Dave Jones
agpgart: AGP aperture is 128M @ 0xd0000000
agpgart: Detected ServerWorks CNB20HE chipset: No AGP present.
agpgart: Detected ServerWorks CNB20HE chipset: No AGP present.

Full lspci output is included.



Attachment: lspci
Description: Binary data