Handling of bogus PCI bus numbering

Handling of bogus PCI bus numbering

[Stephen Mollett]

The 2.4 PCI subsystem seems not to handle bogus BIOS-assigned PCI bus numbers 
well. (OK, it shouldn't be necessary for the OS to have to tweak the 
numbering, but alas, it is.)

I've got an IBM Thinkpad 240 and the BIOS incorrectly assigns bus number 0 to 
the CardBus (obviously, bus 0 is the primary internal PCI bus). Because of 
this, it's impossible to use CardBus devices - when one is inserted, its 
device number (00:00.0) collides with the 82443BX Northbridge and Card 
Services understandably gets rather confused, thinking I've just plugged in a 
440BX chip... ahem.

I've emailed Martin Mares (listed as the PCI maintainer in the PCI-driver 
source) a while back but I think he must be busy at the moment. I had a look 
at the source myself with a view to trying to write a patch but had to get 
myself a stiff drink after about 5 minutes to help me recover from the 
shell-shock :o)

(I'm using stock 2.4.17 on the machine at the moment.)

Here's /proc/pci without any CardBus devices inserted:

PCI devices found:
  Bus  0, device   0, function  0:
    Host bridge: Intel Corp. 440BX/ZX - 82443BX/ZX Host bridge (AGP disabled) 
(rev 3).
      Master Capable.  Latency=64.  
      Prefetchable 32 bit memory at 0xf8000000 [0xfbffffff].
  Bus  0, device   7, function  0:
    Bridge: Intel Corp. 82371AB PIIX4 ISA (rev 2).
  Bus  0, device   7, function  1:
    IDE interface: Intel Corp. 82371AB PIIX4 IDE (rev 1).
      Master Capable.  Latency=64.  
      I/O at 0x1040 [0x104f].
  Bus  0, device   7, function  2:
    USB Controller: Intel Corp. 82371AB PIIX4 USB (rev 1).
      IRQ 10.
      Master Capable.  Latency=64.  
      I/O at 0x1060 [0x107f].
  Bus  0, device   7, function  3:
    Bridge: Intel Corp. 82371AB PIIX4 ACPI (rev 2).
      IRQ 9.
  Bus  0, device   9, function  0:
    VGA compatible controller: Neomagic Corporation NM2160 [MagicGraph 128XD] 
(rev 1).
      Master Capable.  Latency=128.  Min Gnt=16.Max Lat=255.
      Prefetchable 32 bit memory at 0xf5000000 [0xf5ffffff].
      Non-prefetchable 32 bit memory at 0xf4000000 [0xf41fffff].
      Non-prefetchable 32 bit memory at 0xf4200000 [0xf42fffff].
  Bus  0, device  10, function  0:
    CardBus bridge: Texas Instruments PCI1211 (rev 0).
      IRQ 11.
      Master Capable.  Latency=168.  Min Gnt=192.Max Lat=7.
      Non-prefetchable 32 bit memory at 0x10000000 [0x10000fff].
  Bus  0, device  11, function  0:
    Multimedia audio controller: ESS Technology ES1969 Solo-1 Audiodrive (rev 
2).
      IRQ 3.
      Master Capable.  Latency=64.  Min Gnt=2.Max Lat=24.
      I/O at 0x1000 [0x103f].
      I/O at 0x1090 [0x109f].
      I/O at 0x1080 [0x108f].
      I/O at 0x10a4 [0x10a7].
      I/O at 0x10a0 [0x10a3].
  Bus  0, device  12, function  0:
    Communication controller: Lucent Microelectronics WinModem 56k (rev 1).
      IRQ 10.
      Master Capable.  No bursts.  Min Gnt=252.Max Lat=14.
      Non-prefetchable 32 bit memory at 0xf4300000 [0xf43000ff].
      I/O at 0x10a8 [0x10af].
      I/O at 0x1400 [0x14ff].

With a 3Com 3CCFE575CT CardBus 100Mbit NIC inserted, /proc/pci is identical 
except for one extra device at the end of the listing:

  Bus  0, device   0, function  0:
    Host bridge:  (rev 3).
      Master Capable.  Latency=64.  

and the entry for the CardBus bridge, which shows Min Gnt=128 instead of 192.

Card Services (version 3.1.31, from pcmcia-cs.sf.net) reports:

Mar  2 12:10:08 rodin cardmgr[47]: unsupported card in socket 0
Mar  2 12:10:08 rodin kernel: cs: cb_alloc(bus 0): vendor 0x8086, device 
0x7192
Mar  2 12:10:08 rodin cardmgr[47]:   no product info available
Mar  2 12:10:08 rodin cardmgr[47]:   PCI id: 0x8086, 0x7192

when I insert the NIC.

Please cc any responses to me - I can't take the lkml traffic!

Thanks
Stephen

Re: Handling of bogus PCI bus numbering

[Kai Germaschewski]

On Sat, 2 Mar 2002, Stephen Mollett wrote:

> I've got an IBM Thinkpad 240 and the BIOS incorrectly assigns bus number 0 to 
> the CardBus (obviously, bus 0 is the primary internal PCI bus). Because of 
> this, it's impossible to use CardBus devices - when one is inserted, its 
> device number (00:00.0) collides with the 82443BX Northbridge and Card 
> Services understandably gets rather confused, thinking I've just plugged in a 
> 440BX chip... ahem.

Could you replace "#undef DEBUG" with "#define DEBUG" in drivers/pci/pci.c 
and arch/i386/kernel/pci-i386.h, rebuild the kernel and send "dmesg" 
output after rebooting/inserting the card?

--Kai

Re: Handling of bogus PCI bus numbering

[Stephen Mollett]

On Saturday 02 Mar 2002 15:15, Kai Germaschewski wrote:
> On Sat, 2 Mar 2002, Stephen Mollett wrote:
>> I've got an IBM Thinkpad 240 and the BIOS incorrectly assigns bus number
>> 0 to the CardBus ...
> Could you replace "#undef DEBUG" with "#define DEBUG" in drivers/pci/pci.c
> and arch/i386/kernel/pci-i386.h, rebuild the kernel and send "dmesg"
> output after rebooting/inserting the card?

Here's the full output. The card was inserted immediately after the boot 
completed - the insertion shows up as the last line of the messages.

Stephen

Linux version 2.4.17-pcidebug (root@rodin) (gcc version 2.95.3 20010315 
(release)) #6 Sun Mar 3 11:26:37 GMT 2002
BIOS-provided physical RAM map:
 BIOS-e820: 0000000000000000 - 000000000009fc00 (usable)
 BIOS-e820: 000000000009fc00 - 00000000000a0000 (reserved)
 BIOS-e820: 00000000000e0000 - 0000000000100000 (reserved)
 BIOS-e820: 0000000000100000 - 000000000bff0000 (usable)
 BIOS-e820: 000000000bff0000 - 000000000bfffc00 (ACPI data)
 BIOS-e820: 000000000bfffc00 - 000000000c000000 (ACPI NVS)
 BIOS-e820: 00000000fff80000 - 0000000100000000 (reserved)
On node 0 totalpages: 49136
zone(0): 4096 pages.
zone(1): 45040 pages.
zone(2): 0 pages.
Local APIC disabled by BIOS -- reenabling.
Could not enable APIC!
Kernel command line: BOOT_IMAGE=Linux-Backup ro root=302
Initializing CPU#0
Detected 298.650 MHz processor.
Console: colour VGA+ 80x30
Calibrating delay loop... 596.37 BogoMIPS
Memory: 191648k/196544k available (829k kernel code, 4508k reserved, 195k 
data, 188k init, 0k highmem)
Dentry-cache hash table entries: 32768 (order: 6, 262144 bytes)
Inode-cache hash table entries: 16384 (order: 5, 131072 bytes)
Mount-cache hash table entries: 4096 (order: 3, 32768 bytes)
Buffer-cache hash table entries: 8192 (order: 3, 32768 bytes)
Page-cache hash table entries: 65536 (order: 6, 262144 bytes)
CPU: Before vendor init, caps: 0183f9ff 00000000 00000000, vendor = 0
CPU: L1 I cache: 16K, L1 D cache: 16K
CPU: L2 cache: 128K
CPU: After vendor init, caps: 0183f9ff 00000000 00000000 00000000
Intel machine check architecture supported.
Intel machine check reporting enabled on CPU#0.
CPU:     After generic, caps: 0183f9ff 00000000 00000000 00000000
CPU:             Common caps: 0183f9ff 00000000 00000000 00000000
CPU: Intel Celeron (Mendocino) stepping 0a
Enabling fast FPU save and restore... done.
Checking 'hlt' instruction... OK.
POSIX conformance testing by UNIFIX
mtrr: v1.40 (20010327) Richard Gooch (rgooch@atnf.csiro.au)
mtrr: detected mtrr type: Intel
PCI: BIOS32 Service Directory structure at 0xc00f6b30
PCI: BIOS32 Service Directory entry at 0xfd7c0
PCI: BIOS probe returned s=00 hw=01 ver=02.10 l=00
PCI: PCI BIOS revision 2.10 entry at 0xfd9e5, last bus=0
PCI: Using configuration type 1
PCI: Probing PCI hardware
Scanning bus 00
Found 00:00 [8086/7192] 000600 00
Found 00:38 [8086/7110] 000680 00
Found 00:39 [8086/7111] 000101 00
PCI: IDE base address fixup for 00:07.1
Found 00:3a [8086/7112] 000c03 00
Found 00:3b [8086/7113] 000680 00
Found 00:48 [10c8/0004] 000300 00
Found 00:50 [104c/ac1e] 000607 02
Found 00:58 [125d/1969] 000401 00
Found 00:60 [11c1/0449] 000780 00
Fixups for bus 00
PCI: Scanning for ghost devices on bus 0
Scanning behind PCI bridge 00:0a.0, config 010000, pass 0
Scanning behind PCI bridge 00:0a.0, config 010000, pass 1
Bus scan for 00 returning with max=01
PCI: IRQ init
PCI: Interrupt Routing Table found at 0xc00fdf60
00:00 slot=00 0:60/def8 1:61/def8 2:62/def8 3:63/def8
00:07 slot=00 0:60/def8 1:61/def8 2:62/def8 3:63/8eb8
00:09 slot=00 0:00/def8 1:00/def8 2:00/def8 3:00/def8
00:0a slot=00 0:61/8eb8 1:00/def8 2:00/def8 3:00/def8
00:0b slot=00 0:62/8eb8 1:00/def8 2:00/def8 3:00/def8
00:0c slot=00 0:63/8eb8 1:00/def8 2:00/def8 3:00/def8
PCI: Attempting to find IRQ router for 8086:122e
PCI: Using IRQ router PIIX [8086/7110] at 00:07.0
PCI: IRQ fixup
00:09.0: ignoring bogus IRQ 255
IRQ for 00:09.0:0 -> not routed
PCI: Allocating resources
PCI: Resource f8000000-fbffffff (f=1208, d=0, p=0)
PCI: Resource 00001040-0000104f (f=101, d=0, p=0)
PCI: Resource 00001060-0000107f (f=101, d=0, p=0)
PCI: Resource f5000000-f5ffffff (f=1208, d=0, p=0)
PCI: Resource f4000000-f41fffff (f=200, d=0, p=0)
PCI: Resource f4200000-f42fffff (f=200, d=0, p=0)
PCI: Resource 00001000-0000103f (f=101, d=0, p=0)
PCI: Resource 00001090-0000109f (f=101, d=0, p=0)
PCI: Resource 00001080-0000108f (f=101, d=0, p=0)
PCI: Resource 000010a4-000010a7 (f=105, d=0, p=0)
PCI: Resource 000010a0-000010a3 (f=101, d=0, p=0)
PCI: Resource f4300000-f43000ff (f=200, d=0, p=0)
PCI: Resource 000010a8-000010af (f=109, d=0, p=0)
PCI: Resource 00001400-000014ff (f=101, d=0, p=0)
PCI: Sorting device list...
Limiting direct PCI/PCI transfers.
Linux NET4.0 for Linux 2.4
Based upon Swansea University Computer Society NET3.039
Initializing RT netlink socket
IBM machine detected. Enabling interrupts during APM calls.
apm: BIOS version 1.2 Flags 0x03 (Driver version 1.15)
Starting kswapd
pty: 256 Unix98 ptys configured
block: 128 slots per queue, batch=32
Uniform Multi-Platform E-IDE driver Revision: 6.31
ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
PIIX4: IDE controller on PCI bus 00 dev 39
PIIX4: chipset revision 1
PIIX4: not 100% native mode: will probe irqs later
    ide0: BM-DMA at 0x1040-0x1047, BIOS settings: hda:DMA, hdb:pio
    ide1: BM-DMA at 0x1048-0x104f, BIOS settings: hdc:pio, hdd:pio
hda: IBM-DBCA-206480, ATA DISK drive
ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
hda: 12685680 sectors (6495 MB) w/420KiB Cache, CHS=789/255/63, UDMA(33)
Partition check:
 hda: hda1 hda2 hda3 hda4
NET4: Linux TCP/IP 1.0 for NET4.0
IP Protocols: ICMP, UDP, TCP
IP: routing cache hash table of 1024 buckets, 8Kbytes
TCP: Hash tables configured (established 16384 bind 16384)
NET4: Unix domain sockets 1.0/SMP for Linux NET4.0.
reiserfs: checking transaction log (device 03:02) ...
Using tea hash to sort names
ReiserFS version 3.6.25
VFS: Mounted root (reiserfs filesystem) readonly.
Freeing unused kernel memory: 188k freed
reiserfs: checking transaction log (device 03:03) ...
Using tea hash to sort names
ReiserFS version 3.6.25
reiserfs: checking transaction log (device 03:04) ...
Using tea hash to sort names
ReiserFS version 3.6.25
Real Time Clock Driver v1.10e
Linux PCMCIA Card Services 3.1.31
  kernel build: 2.4.17-pcidebug #6 Sun Mar 3 11:26:37 GMT 2002
  options:  [pci] [cardbus] [apm] [pnp]
PnP: PNP BIOS installation structure at 0xc00f6b80
PnP: PNP BIOS version 1.0, entry at f0000:8440, dseg at 400
Intel ISA/PCI/CardBus PCIC probe:
IRQ for 00:0a.0:0 -> PIRQ 61, mask 8eb8, excl 0000 -> newirq=11 -> got IRQ 11
PCI: Found IRQ 11 for device 00:0a.0
  TI 1211 rev 00 PCI-to-CardBus at slot 00:0a, mem 0x10000000
    host opts [0]: [ring] [pci + serial irq] [pci irq 11] [lat 168/176] [bus 
0/1]
    ISA irqs (scanned) = 5,15 PCI status changes
parport0: PC-style at 0x3bc (0x7bc), irq 7, using FIFO 
[PCSPP,TRISTATE,COMPAT,ECP]
lp0: using parport0 (interrupt-driven).
lp0: compatibility mode
lp0: compatibility mode
Serial driver version 5.05c (2001-07-08) with MANY_PORTS SHARE_IRQ SERIAL_PCI 
enabled
ttyS00 at 0x03f8 (irq = 4) is a 16550A
ttyS02 at 0x03e8 (irq = 4) is a 16550A
cs: cb_alloc(bus 0): vendor 0x8086, device 0x7192

Re: Handling of bogus PCI bus numbering

[Alan Cox]

> The 2.4 PCI subsystem seems not to handle bogus BIOS-assigned PCI bus numbers 
> well. (OK, it shouldn't be necessary for the OS to have to tweak the 
> numbering, but alas, it is.)

Does the pci=assign-busses option help ?

Re: Handling of bogus PCI bus numbering - case closed

[Stephen Mollett]

On Sunday 03 Mar 2002 14:02, Alan Cox wrote:
>> The 2.4 PCI subsystem seems not to handle bogus BIOS-assigned PCI bus
>> numbers well...
> Does the pci=assign-busses option help ?

Yes :)

The CardBus gets assigned a real bus number at last.

I must have been so brain-fried when I was scouring the documentation and 
source that I overlooked that one.

Thanks

Stephen

Re: Handling of bogus PCI bus numbering - case closed

[Alan Cox]

> >> numbers well...
> > Does the pci=assign-busses option help ?
> 
> Yes :)
> The CardBus gets assigned a real bus number at last.
> 
> I must have been so brain-fried when I was scouring the documentation and 
> source that I overlooked that one.

Its not terribly well documented. Can you run dmidecode on your laptop
and hopefully we can catch that laptop and automtically do bus assignment
for it

Re: Handling of bogus PCI bus numbering - case closed

[Stephen Mollett]

On Sunday 03 Mar 2002 15:00, Alan Cox wrote:
> ... Can you run dmidecode on your laptop
> and hopefully we can catch that laptop and automtically do bus assignment
> for it

dmidecode output follows:

SMBIOS 2.3 present.
DMI 2.3 present.
31 structures occupying 876 bytes.
DMI table at 0x000EABC0.
Handle 0x0000
	DMI type 0, 20 bytes.
	BIOS Information Block
		Vendor: IBM
		Version: IRET75WW
		Release: 11/30/1999
		BIOS base: 0xEA720
		ROM size: 448K
		Capabilities:
			Flags: 0x0000000000011F90
Handle 0x0001
	DMI type 1, 25 bytes.
	System Information Block
		Vendor: IBM
		Product: 260921G
		Version: Not Available
		Serial Number: BAZAPVF
Handle 0x0002
	DMI type 2, 8 bytes.
	Board Information Block
		Vendor: IBM
		Product: 2609BS1
		Version: Not Available
		Serial Number: PLR-SERIAL#
Handle 0x0003
	DMI type 3, 17 bytes.
	Chassis Information Block
		Vendor: IBM
		Chassis Type: Notebook
		Version: Not Available
		Serial Number: Not Available
		Asset Tag: Not Available
Handle 0x0004
	DMI type 4, 32 bytes.
	Processor
		Socket Designation: MMC 2
		Processor Type: Central Processor
		Processor Family: Celeron processor
		Processor Manufacturer: GenuineIntel
		Processor Version: Celeron(TM)
Handle 0x0005
	DMI type 5, 20 bytes.
	Memory Controller
Handle 0x0006
	DMI type 6, 12 bytes.
	Memory Bank
		Socket: CN24
		Banks: 0 1
		Speed: 15nS
		Type: 
		Installed Size: 64Mbyte (Double sided)
		Enabled Size: 64Mbyte (Double sided)
Handle 0x0007
	DMI type 6, 12 bytes.
	Memory Bank
		Socket: CN25
		Banks: 2 3
		Speed: 15nS
		Type: 
		Installed Size: 128Mbyte (Double sided)
		Enabled Size: 128Mbyte (Double sided)
Handle 0x0008
	DMI type 7, 19 bytes.
	Cache
		Socket: L1 Cache
		L1 socketed Internal Cache: write-back
		L1 Cache Size: 32K
		L1 Cache Maximum: 32K
		L1 Cache Type: Asynchronous 
Handle 0x0009
	DMI type 7, 19 bytes.
	Cache
		Socket: L2 Cache
		L2 socketed External Cache: write-back
		L2 Cache Size: 256K
		L2 Cache Maximum: 128K
		L2 Cache Type: Unknown 
Handle 0x000A
	DMI type 8, 9 bytes.
	Port Connector
		Internal Designator: CN7
		Internal Connector Type: 9 Pin Dual Inline (pin 10 cut)
		External Designator: COM 1
		External Connector Type: DB-9 pin male
		Port Type: Serial Port 16650A Compatible
Handle 0x000B
	DMI type 8, 9 bytes.
	Port Connector
		Internal Designator: CN1
		Internal Connector Type: 25 Pin Dual Inline (pin 26 cut)
		External Designator: Parallel
		External Connector Type: DB-25 pin female
		Port Type: Parallel Port ECP/EPP
Handle 0x000C
	DMI type 8, 9 bytes.
	Port Connector
		Internal Designator: CN14
		Internal Connector Type: None
		External Designator: Keyboard
		External Connector Type: Circular DIN-8 male
		Port Type: Keyboard Port
Handle 0x000D
	DMI type 8, 9 bytes.
	Port Connector
		Internal Designator: CN14
		Internal Connector Type: None
		External Designator: PS/2 Mouse
		External Connector Type: Circular DIN-8 male
		Port Type: Mouse Port
Handle 0x000E
	DMI type 8, 9 bytes.
	Port Connector
		Internal Designator: CN4
		Internal Connector Type: None
		External Designator: USB
		External Connector Type: Access Bus (USB)
		Port Type: USB
Handle 0x000F
	DMI type 8, 9 bytes.
	Port Connector
		Internal Designator: CN13
		Internal Connector Type: None
		External Designator: Infrared
		External Connector Type: Infrared
		Port Type: Other
Handle 0x0010
	DMI type 8, 9 bytes.
	Port Connector
		Internal Designator: SM1
		Internal Connector Type: None
		External Designator: Lucent Modem
		External Connector Type: RJ-11
		Port Type: Modem Port
Handle 0x0011
	DMI type 9, 13 bytes.
	Card Slot
		Slot: CardBus Slot 1
		Type: 32bit PCMCIA VLB 
		Status: Available.
		Slot Features: 5v 3.3v PCCard16 CardBus Zoom-Video 
Handle 0x0012
	DMI type 10, 8 bytes.
	On Board Devices Information
		Description: Video (NeoMagic NM2160) : Disabled
		Type: 
		Description: ESS 1946 : Disabled
		Type: 
Handle 0x0013
	DMI type 11, 5 bytes.
	OEM Data
		Intel 440BX NoteBook
		IBM ThinkPad
Handle 0x0014
	DMI type 12, 5 bytes.
	Configuration Information
Handle 0x0015
	DMI type 16, 15 bytes.
	Physical Memory Array
Handle 0x0016
	DMI type 17, 23 bytes.
	Memory Device
Handle 0x0017
	DMI type 17, 23 bytes.
	Memory Device
Handle 0x0018
	DMI type 19, 15 bytes.
	Memory Array Mapped Address
Handle 0x0019
	DMI type 20, 19 bytes.
	Memory Device Mapped Address
Handle 0x001A
	DMI type 21, 7 bytes.
03 04 02                                        	...             
Handle 0x001B
	DMI type 31, 28 bytes.
00 00 00 00 4e 61 bc 00 4e 61 bc 00 00 00 00 00 	....Na..Na......
00 00 00 00 00 00 00 00                         	........        
Handle 0x001C
	DMI type 32, 20 bytes.
	System Boot Information
Handle 0x001D
	DMI type 126, 4 bytes.
	Inactive
Handle 0x001E
	DMI type 127, 4 bytes.
	End-of-Table