Okay, more investigation on this error.
I noticed the same message (****INVALID IO ALLOCATION****) in the logfile
for another chipset (ATI Technologies Inc RS480 [Radeon Xpress 200G Series])
on a Nec VL350 computer. I noticed a beep when starting XFree86, which this
time doesn't fail. This beep probably comes from the BIOS. I tested again
on the Dell GX270 and Dell GX280 and, indeed, they emit the beep too (but, as said,
XFree86 fails on the computers).
I looked deeper into the patch 4.6.99.17-4.6.99.18.diff.bz2 and found that this
file modified by the patch is the culprit :
xc/programs/Xserver/hw/xfree86/os-support/bus/linuxPci.c (1.14)
bugfixes :
111. Various PCI-related changes (Marc La France):
[...]
So it has to be reworked, I think.
Loïc, Toulouse, France
| loic.mahe@xxxxxxxxxxxxxx
Envoyé par : devel-admin@xxxxxxxxxxx 18/06/2007 08:58
|
|
Hello,
I'm building XFree86 4.6+patches for a selection of computers with various video chipsets.
The kernel is a 2.4.34.
I found that the patch 4.6.99.17-4.6.99.18.diff.bz2 breaks the i810 driver for the Intel
graphic chipset on these two Dell computers :
- Dell Optiplex GX270 / i865G
- Dell Optiplex GX280 / i915G
But note that this computer works well : Dell Optiplex GX260 / i845G
The syndrome : as soon as X startx, the screen becomes dark and the text consoles become unusable too
(but the system doesn't crash). I can see this in /var/log/XFree86.0.log (this example is for the i915G) :
[...]
(--) Chipset 915G found
(II) resource ranges after xf86ClaimFixedResources() call:
[...]
(II) Found conflict at: 0x3df
(II) Found conflict at: 0x3bb
(WW) ****INVALID IO ALLOCATION**** b: 0xe898 e: 0xe89f correcting^G
(II) window:
[0] -1 0 0x00000000 - 0x0000ffff (0x10000) IX[B]
(II) resSize:
[0] -1 0 0x00000000 - 0xffffffff (0x0) IX[B]
(II) window fixed:
[0] -1 0 0x00000000 - 0x0000ffff (0x10000) IX[B]
and the log file stops here.
Moreover, I have the same graphic chipsets (i865G and i915G) on non-Dell motherboards
(MSI or Intel motherboards) and they work perfectly here. So I suspect that the problem comes from
the Dell BIOS. Dell BIOSes are mostly not configurable, as you may know, and I didn't find a parameter
which could be related to the problem.
There are some pci-related files which are altered by the patch, the error probably comes from there.
Loïc, Toulouse, France
