simple pnp bios io resources bug makes system unusable
my one line patch just skips an io registration with a simple sanity check. never once have i heard a device with an ioport of 0x0. question is why it happens and only once.
tested with
ISA Plug and Play:U.S. Robotics Sportster 33600 FAX/Voice Int
Creative ViBRA16C PnP
Crystal Codec
Host/PCI Bridge:
VIA Technologies, In VT82C585VP [Apollo V
VIA Technologies, In VT82C586/A/B PCI-to-
VIA Technologies, In VT82C586/B/686A/B PI
VIA Technologies, In USB
VIA Technologies, In VT82C586B ACPI
(also noticing cutoff in /sys/devices/pci0/*/name)
Quoted 25 lines
Yes, I also ran into this problem, Adam is was also working on a fix although I wonder if it's the same fix?
Shawn S.
Quoted 20 lines
List: linux-kernel Subject: simple pnp bios io resources bug makes system unusable From: CarlosRomero <caberome () bellsouth ! net> Date: 2003-07-01 3:38:17cat /sys/devices/pnp0/00\:0c/name
Reserved Motherboard Resourcescat /sys/devices/pnp0/00\:0c/resources
state = active
io 0x4d0-0x4d1
io 0xcf8-0xcff
io 0x3f7-0x3f7
io 0x401-0x407
io 0x298-0x298
io 0x00000000-0xffffffff
mem 0xfffe0000-0xffffffff
mem 0x100000-0x7fffffffixup: check for null io base, other devices are now able to initialize.
Quoted 2 lines · CarlosRomero
my one line patch just skips an io registration with a simple sanity check. never once have i heard a device with an ioport of
0x0.
AT DMA controller as stated in my other message, is one example.
Quoted 1 line
question is why it happens and only once.
It only happens once because in 99% of cases the entire device is disabled, not just an individual range.
Quoted 1 line
(also noticing cutoff in
/sys/devices/pci0/*/name)
There is a hard limit for device name length, see include/linux/device.h. That is probably the cause.
Thanks,
Adam