Pb1500 PCI problems

DISCUSSION Chris Zimman 4 participants 7 messages Feb 25, 2003 – Feb 26, 2003 MIPS

Share & tools

Chris Zimman

author

2003-02-25 08:01 UTC #0 lore raw

I've seen some strange stuff in the PCI code in 2.4.19-rc1 and 2.4.20 from the CVS tree.

Neither trees compile out of the box for the PB1500, both having errors in one place or another. 2.4.20 blows up during boot in:

...

00:10.0 Class 2000: 0356:2000 (rev 56)
Mem unavailable -- skipping
Mem unavailable -- skipping
Mem unavailable -- skipping
Mem unavailable -- skipping
Mem unavailable -- skipping
Mem unavailable -- skipping
00:11.0 Class 0000: 0000:0000

        Mem at 0x40000000 [size=0xffd0]
        Mem at 0x4000ffd0 [size=0xffd0]

Reserved instruction in kernel code in traps.c::do_ri, line 650:
$0 : 00000000 1000fc00 c0000000 c0000000 00000001 c0000000 00000000 1000fc00 $8 : 810fa7f0 00000000 ffffffbf ffffffff fffffff8 ffffffff 00000010 00000003 $16: 00000000 00000000 8034be88 00000000 00000000 00000098 00000000 00000000
$24: 8034bd53 00000000 8034a000 8034be48 00000004 80274c80
Hi : 00000000
Lo : 000000c0
epc : 80274ca0 Not tainted
Status: 1000fc02
Cause : 00800028
Process swapper (pid: 1, stackpage=8034a000)

Stack:    00000000 00000098 8027fe6f 80274c80 00000400 0000000d 00000000
 8034bef0 00000000 8030e108 80274e0c 00000088 00000000 00000000 802b4588
 802dbc68 8027fe58 4000ffd0 1000fc01 00000090 801e73d8 4000ffd0 ffff0036
 00000000 00000000 00000088 00000000 801084f0 00000000 00000098 00000000
 00000000 802b49d0 802b4b34 8027fea0 00000000 00000001 00000001 00000000
 8034bef0 ...
Call Trace:   [<8027fe6f>] [<80274c80>] [<80274e0c>] [<8027fe58>] [<801e73d8>]
 [<801084f0>] [<8027fea0>] [<8010078c>] [<8027ff6c>] [<8012cf90>] [<8010078c>]
 [<80101f24>] [<801e9040>] [<8010078c>] [<8010079c>] [<801022d4>] [<80100780>]
 [<801022c4>]

Code: 24040001  12640023  00431825 <8c620000> ae420000  0000000f  3c03000d  34631b72  3c02802d 
Kernel panic: Attempted to kill init!

2.4.19-rc1 fares a little better, but has strange problems as well:

chris@au1500:~$ lspci -vv
00:01.0 PIC: Unknown device bad7:0800 (rev db) (prog-if ba)
        Control: I/O- Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
        Status: Cap- 66Mhz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
        Interrupt: pin ? routed to IRQ 255
        Region 0: Memory at <ignored> (type 3, prefetchable) [disabled]
        Region 2: I/O ports at <ignored> [disabled]
        Region 4: Memory at <ignored> (low-1M, prefetchable) [disabled]
        Expansion ROM at 0800b800 [disabled] [size=2K]

00:05.0 Class 1060: Unknown device 0007:1040 (rev 0d)
!!! Invalid class 1060 for header type 02
Subsystem: Unknown device 4054:0800
Control: I/O- Mem+ BusMaster- SpecCycle+ MemWINV+ VGASnoop+ ParErr- Stepping- SERR- FastB2B-
Status: Cap- 66Mhz- UDF- FastB2B- ParErr- DEVSEL=slow >TAbort- <TAbort- <MAbort+ >SERR- <PERR-
Interrupt: pin ? routed to IRQ 255
Region 0: I/O ports at <ignored> [disabled]
Bus: primary=55, secondary=40, subordinate=00, sec-latency=8
BridgeCtl: Parity+ SERR- ISA+ VGA- MAbort- >Reset- 16bInt- PostWrite+
16-bit legacy interface ports at 0006

00:0a.0 Class 8e10: Unknown device f809:0040 (rev 10)
Subsystem: Unknown device 0008:03e0
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop+ ParErr- Stepping- SERR- FastB2B-
Status: Cap- 66Mhz- UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
Latency: 128 (1000ns min, 3000ns max), cache line size da
Interrupt: pin Region 5: Memory at <invalid-64bit-slot> (64-bit, non-prefetchable)
Expansion ROM at 27bd0000 [disabled] [size=2K]

00:10.0 Class 2000: Unknown device 0356:2000 (rev 56) (prog-if 03)

        Subsystem: Unknown device 0356:2000
        Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV+ VGASnoop- ParErr+ Stepping- SERR+ FastB2B+
        Status: Cap- 66Mhz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort+ >SERR- <PERR-
        Latency: 3 (8000ns max), cache line size 56
        Interrupt: pin C routed to IRQ 255
        Region 0: Memory at 20000350 (type 3, non-prefetchable) [size=16]
        Region 1: Memory at 20000350 (type 3, non-prefetchable) [size=16]
        Region 2: Memory at 20000350 (type 3, non-prefetchable) [size=16]
        Region 3: Memory at 20000350 (type 3, non-prefetchable) [size=16]
        Region 4: Memory at 20000350 (type 3, non-prefetchable) [size=16]
        Region 5: Memory at 20000350 (type 3, non-prefetchable) [size=16]
        Expansion ROM at 20000000 [disabled] [size=512M]

Before I go digging too much, I'd like it if someone else with a PB1500 or similar can confirm what I'm seeing.

The kernel was built with defconfig-pb1500, and using GCC 3.2.2 and bintutils 2.13

The 2.4.19-rc1 kernel seems to work fine otherwise, BTW

Thanks

--Chris

Pete Popov

2003-02-25 16:12 UTC #1 lore raw

Quoted 5 lines · Chris Zimman

I've seen some strange stuff in the PCI code in 2.4.19-rc1 and 2.4.20 from the CVS tree.
Neither trees compile out of the box for the PB1500, both having errors in one place or another. 2.4.20 blows up during boot in:

I probably need to put this in the main FAQ or something.

There are some required patches that you need to apply which are not in the tree yet. Take a look at ftp.linux-mips.org:/pub/linux/mips/people/ppopov. There is a README there that describes the patches. At the very least you need to apply the 36 bit patch.

Pete

Quoted 97 lines

...
00:10.0 Class 2000: 0356:2000 (rev 56)
Mem unavailable -- skipping
Mem unavailable -- skipping
Mem unavailable -- skipping
Mem unavailable -- skipping
Mem unavailable -- skipping
Mem unavailable -- skipping
00:11.0 Class 0000: 0000:0000
        Mem at 0x40000000 [size=0xffd0]
        Mem at 0x4000ffd0 [size=0xffd0]
Reserved instruction in kernel code in traps.c::do_ri, line 650:
$0 : 00000000 1000fc00 c0000000 c0000000 00000001 c0000000 00000000 1000fc00 $8 : 810fa7f0 00000000 ffffffbf ffffffff fffffff8 ffffffff 00000010 00000003 $16: 00000000 00000000 8034be88 00000000 00000000 00000098 00000000 00000000
$24: 8034bd53 00000000 8034a000 8034be48 00000004 80274c80
Hi : 00000000
Lo : 000000c0
epc : 80274ca0 Not tainted
Status: 1000fc02
Cause : 00800028
Process swapper (pid: 1, stackpage=8034a000)
Stack:    00000000 00000098 8027fe6f 80274c80 00000400 0000000d 00000000
 8034bef0 00000000 8030e108 80274e0c 00000088 00000000 00000000 802b4588
 802dbc68 8027fe58 4000ffd0 1000fc01 00000090 801e73d8 4000ffd0 ffff0036
 00000000 00000000 00000088 00000000 801084f0 00000000 00000098 00000000
 00000000 802b49d0 802b4b34 8027fea0 00000000 00000001 00000001 00000000
 8034bef0 ...
Call Trace:   [<8027fe6f>] [<80274c80>] [<80274e0c>] [<8027fe58>] [<801e73d8>]
 [<801084f0>] [<8027fea0>] [<8010078c>] [<8027ff6c>] [<8012cf90>] [<8010078c>]
 [<80101f24>] [<801e9040>] [<8010078c>] [<8010079c>] [<801022d4>] [<80100780>]
 [<801022c4>]

Code: 24040001  12640023  00431825 <8c620000> ae420000  0000000f  3c03000d  34631b72  3c02802d 
Kernel panic: Attempted to kill init!
2.4.19-rc1 fares a little better, but has strange problems as well:
chris@au1500:~$ lspci -vv
00:01.0 PIC: Unknown device bad7:0800 (rev db) (prog-if ba)
        Control: I/O- Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
        Status: Cap- 66Mhz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
        Interrupt: pin ? routed to IRQ 255
        Region 0: Memory at <ignored> (type 3, prefetchable) [disabled]
        Region 2: I/O ports at <ignored> [disabled]
        Region 4: Memory at <ignored> (low-1M, prefetchable) [disabled]
        Expansion ROM at 0800b800 [disabled] [size=2K]
00:05.0 Class 1060: Unknown device 0007:1040 (rev 0d)
!!! Invalid class 1060 for header type 02
Subsystem: Unknown device 4054:0800
Control: I/O- Mem+ BusMaster- SpecCycle+ MemWINV+ VGASnoop+ ParErr- Stepping- SERR- FastB2B-
Status: Cap- 66Mhz- UDF- FastB2B- ParErr- DEVSEL=slow >TAbort- <TAbort- <MAbort+ >SERR- <PERR-
Interrupt: pin ? routed to IRQ 255
Region 0: I/O ports at <ignored> [disabled]
Bus: primary=55, secondary=40, subordinate=00, sec-latency=8
BridgeCtl: Parity+ SERR- ISA+ VGA- MAbort- >Reset- 16bInt- PostWrite+
16-bit legacy interface ports at 0006
00:0a.0 Class 8e10: Unknown device f809:0040 (rev 10)
Subsystem: Unknown device 0008:03e0
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop+ ParErr- Stepping- SERR- FastB2B-
Status: Cap- 66Mhz- UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
Latency: 128 (1000ns min, 3000ns max), cache line size da
Interrupt: pin Region 5: Memory at <invalid-64bit-slot> (64-bit, non-prefetchable)
Expansion ROM at 27bd0000 [disabled] [size=2K]
00:10.0 Class 2000: Unknown device 0356:2000 (rev 56) (prog-if 03)
        Subsystem: Unknown device 0356:2000
        Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV+ VGASnoop- ParErr+ Stepping- SERR+ FastB2B+
        Status: Cap- 66Mhz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort+ >SERR- <PERR-
        Latency: 3 (8000ns max), cache line size 56
        Interrupt: pin C routed to IRQ 255
        Region 0: Memory at 20000350 (type 3, non-prefetchable) [size=16]
        Region 1: Memory at 20000350 (type 3, non-prefetchable) [size=16]
        Region 2: Memory at 20000350 (type 3, non-prefetchable) [size=16]
        Region 3: Memory at 20000350 (type 3, non-prefetchable) [size=16]
        Region 4: Memory at 20000350 (type 3, non-prefetchable) [size=16]
        Region 5: Memory at 20000350 (type 3, non-prefetchable) [size=16]
        Expansion ROM at 20000000 [disabled] [size=512M]
Before I go digging too much, I'd like it if someone else with a PB1500 or similar can confirm what I'm seeing.
The kernel was built with defconfig-pb1500, and using GCC 3.2.2 and bintutils 2.13
The 2.4.19-rc1 kernel seems to work fine otherwise, BTW
Thanks
--Chris

Jiahan Chen

2003-02-25 16:49 UTC #2 lore raw

Kernel Source Tree & Rebuild for Mips

Quoted 4 lines · --- Pete Popov

Quoted 3 lines · Chris Zimman
I've seen some strange stuff in the PCI code in 2.4.19-rc1 and 2.4.20 from the CVS tree.

Where and how can I get CVS source tree to build customized Linux kernel for Mips?

Recently, I downloaded linux-2.4.18.tar.gz, patch-2.4.19.bz2, patch-2.4.20.bz2 from www.kernel.org, used cross-compiler mipsel-linux-gcc, mips-linux-ld
on Redhat 7.3 PC envoronment, and got quite a few errors from compiling and ld. Can you or someone give me help?

Thanks,

Jiahan

__________________________________________________
Do you Yahoo!?
Yahoo! Tax Center - forms, calculators, tips, more
http://taxes.yahoo.com/

Geert Uytterhoeven

2003-02-25 17:05 UTC #3 lore raw

Re: Kernel Source Tree & Rebuild for Mips

Quoted 8 lines · Jiahan Chen

Quoted 4 lines · --- Pete Popov
Quoted 3 lines · Chris Zimman
I've seen some strange stuff in the PCI code in 2.4.19-rc1 and 2.4.20 from the CVS tree.
Where and how can I get CVS source tree to build customized Linux kernel for Mips?

http://www.google.com/search?q=Linux+MIPS+CVS

Gr{oetje,eeting}s,

Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds

Jiahan Chen

2003-02-26 03:06 UTC #4 lore raw

CVS Usage and Kernel Build

Quoted 9 lines

Quoted 4 lines
Where and how can I get CVS source tree to build customized Linux kernel for Mips?
http://www.google.com/search?q=Linux+MIPS+CVS
Gr{oetje,eeting}s,

Geert Uytterhoeven

2003-02-26 09:28 UTC #5 lore raw

Re: CVS Usage and Kernel Build

Quoted 10 lines · Jiahan Chen

Quoted 9 lines
Quoted 2 lines
Where and how can I get CVS source tree to build customized Linux kernel for Mips?
http://www.google.com/search?q=Linux+MIPS+CVS
Gr{oetje,eeting}s,
From Mips web-site, I read:

Quoted 7 lines

cvs -d :pserver:[email protected]:/home/cvs login
(Only needed the first time you use anonymous CVS, the password is "cvs") cvs -d :pserver:[email protected]:/home/cvs co <repository>
I have a few questions:
1. There should be a client "cvs" in my linux PC, then to use above command to get CVS source files INDIVIDUALLY?

cvs -d :pserver:[email protected]:/home/cvs co linux/path/to/file.

Quoted 3 lines

2. After get everything from ftp site as above, do we use the similar procedure to re-build linux kernel for MIPS, such as make config; make dep; make vmlinux

Yes.

Quoted 1 line

3. Does this source tree support R3000 (CPU) and USB?

Yes.

Quoted 2 lines

4. In order to add a new USB device driver, do I need update drivers/usb/Config.In and drivers/usb/Makefile manully?

Yes.

Quoted 5 lines

Currently, I am in the initial phase for development, the Network card is not available and Winmoden doesn't work with Linux, so I have no ftp connection from my Linux box to get CVS. In this case, is there any alternative to get CVS source tree?

Use CVS on another box.

Gr{oetje,eeting}s,

Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds

Pete Popov

2003-02-25 17:33 UTC #6 lore raw

Re: Kernel Source Tree & Rebuild for Mips

In reply to Jiahan Chen

Quoted 6 lines · Jiahan Chen

Quoted 4 lines · --- Pete Popov
Quoted 3 lines · Chris Zimman
I've seen some strange stuff in the PCI code in 2.4.19-rc1 and 2.4.20 from the CVS tree.

Quoted 8 lines

Where and how can I get CVS source tree to build customized Linux kernel for Mips?
Recently, I downloaded linux-2.4.18.tar.gz, patch-2.4.19.bz2, patch-2.4.20.bz2 from www.kernel.org, used cross-compiler mipsel-linux-gcc, mips-linux-ld
on Redhat 7.3 PC envoronment, and got quite a few errors from compiling and ld. Can you or someone give me help?

The mips linux port is hosted on linux-mips.org. Take a look at the documentation on www.linux-mips.org and go from there.

Pete