Re: [parisc-linux] C240 and Linux 2.6.x

Re: [parisc-linux] C240 and Linux 2.6.x

[Joel Soete]

Hello Alexander,

Alexander Bartolich wrote:
> I have a C240 satisfyingly running Debian Testing and kernels
> from 2.4.17 to 2.4.27-pa4, both stock and custom compiled.
> 
> However, no variation of 2.6.x ever booted. Since the last message
> I see is that infamous »you may need to switch your console« I
> eventually organized a serial cable, a second machine running
> minicom and got the same result:
> 
[...]
> 
> Where is this FAQ item mentioned above?
Simply at <http://www.parisc-linux.org/> :-)

> And how is this ment, anyway?
> Just switching consoles within HP's »BIOS«,
Two possibility:
     1) disconect all kdb from builtin interfaces (on my c110 there are ps2 and hp-hil) then pdc (hp's 'bios') will switch 
automaticaly to serial port 1 as console (then booting kernel will also use this)
     2) at pdc boot prompt (Main Menu: Enter command or menu > ) (this is the one of my c110) choose submenu "COnfiguration menu" in 
which option "PAth [PRI|ALT] [<path>]" would help you to select your console (see: help path) (e.g. to set serial1 as console: "path 
con serial_1"

> or passing some arguments via /etc/palo.conf,
only requested for some k model iirc (to force kernel to use ttyB0 for the mux)

> or compiling the kernel with different settings?
> 
At a first glance I would try first the b180 config (a quick look at <http://www.openpa.net/index.html>, seems to me the nearest):
prebuild kernel for different model are available at <http://cvs.parisc-linux.org/download/linux-2.6/> (see instruction at the top 
of page to install package :-)

or build your own: in the subdir arch/parisc/configs/ you would find b180_defconfig which you copy as .config in the top of your 
kernel tree then "make oldconfig; make vmlinux" ;-)

> I'd really like to help more, but without any diagnostics of 2.6.x ...
> Well, here is the start of dmesg when booting with a custom compiled
> kernel:
> 
Not trivial exercise, but you can try to build an equivalent one for 2.6 (simply copy the 2.4 .config into 2.6 was not enough for my 
c110, I realy have to check option by option)

[...]
> STI device: HPA4554A
> stifb: Unsupported gfx card id 0x2fc1066b
A few gfx card are supported (lack of doc) :-(

Even thought some foreign pci card (some ATI and Vodoo iirc) are reported to works (not using stifb obviously because couldn't be 
detected by pdc) for X (see m-l: <http://www.parisc-linux.org/mailing-lists/index.html> )

Hth,
	Joel
_______________________________________________
parisc-linux mailing list
parisc-linux@lists.parisc-linux.org
http://lists.parisc-linux.org/mailman/listinfo/parisc-linux

Re: [parisc-linux] C240 and Linux 2.6.x

[Alexander Bartolich]

[-- Attachment #1.1: Type: text/plain, Size: 4745 bytes --]

On Fri, Dec 31, 2004 at 02:23:30PM -0700, Grant Grundler wrote:
> Yes, STI should continue to work with 2.6.x kernels.
> This sounds like a bug.

I fetched palinux-b180-defconfig-2.6.10-pa2_0-2_all.deb from
http://cvs.parisc-linux.org/download/linux-2.6/autobuild/b180_defconfig/
and build a nice vmlinux-2.6.10-pa2.

This gives me nice console output on booting, so there is definitely
something wrong with Debian's kernel-image-2.6.8-1-32.

Howver, now comes the bad news.

=======================================================================
Command line for kernel: 'root=/dev/sda2 HOME=/ max_loop=64 console=ttyS0 TERM=vt102 palo_kernel=2/boot/vmlinux'
Selected kernel: /boot/vmlinux from partition 2
ELF32 executable
Entry 00100000 first 00100000 n 1
Segment 0 load 00100000 size 3514501 mediaptr 0x1000
Branching to kernel entry point 0x00100000.  If this is the last
message you see, you may need to switch your console.  This is
a common symptom -- search the FAQ and mailing list at parisc-linux.org

CPU'
CPU(s): 1 x PA8200 (PCX-U+) at 236.000000 MHz
Setting cache flush threshold to 40 (1 CPUs online)
Lasi version 0 at 0xffd00000 found.
Dino [8/0] version 3.1 found at 0xf2000000
Cujo version 2.0 found at 0xf1000000
SCSI subsystem initialized
unwind_init: start = 0x1034aea0, end = 0x1036ed70, entries = 9197
Installing knfsd (copyright (C) 1996 okir@monad.swb.de).
Initializing Cryptographic API
Soft power switch enabled, polling @ 0xf0140000.
STI GSC/PCI core graphics driver Version 0.9a
STI PCI graphic ROM found at f1e00000 (2048 kB), fb at f6000000 (32 MB)
    id 2fc1066b-9a02587, conforms to spec rev. 8.09
    graphics card name: HPA4554A
sticon: Initializing STI text console.
Console: switching to colour STI console 160x64
stifb: 'HPA4554A' (id: 0x2fc1066b) not supported.
lp: driver loaded but no devices found
Generic RTC Driver v1.07
serio: GSC PS/2 keyboard port at 0xffd08000 irq 69 @ 8:16:7
serio: GSC PS/2 mouse port at 0xffd08100 irq 69 @ 8:16:8
Serial: 8250/16550 driver $Revision: 1.90 $ 13 ports, IRQ sharing enabled
ttyS0 at MMIO 0xffd05800 (irq = 90) is a 16550A
ttyS1 at MMIO 0xf2003800 (irq = 106) is a 16550A
parport_init_chip: initialize bidirectional-mode.
parport0: PC-style at 0xffd02800, irq 88 [PCSPP,TRISTATE]
lp0: using parport0 (interrupt-driven).
io scheduler noop registered
io scheduler anticipatory registered
io scheduler deadline registered
io scheduler cfq registered
loop: loaded (max 64 devices)
PPP generic driver version 2.4.2
Linux Tulip driver version 1.1.13 (May 11, 2002)
Backtrace:
 [<10202ea4>] pci_device_probe_static+0x60/0x78
 [<10202ed8>] __pci_device_probe+0x1c/0x34
 [<10202f1c>] pci_device_probe+0x2c/0x58
 [<1023dbe0>] driver_probe_device+0x48/0x90
 [<1023dd5c>] driver_attach+0x80/0xa8
 [<1023e2e0>] bus_add_driver+0xb8/0xe0
 [<1023e878>] driver_register+0x44/0x50
 [<102031f8>] pci_register_driver+0x54/0x60
 [<103d189c>] do_initcalls+0x8c/0x138
 [<10103760>] init+0x38/0x104
 [<1010bc5c>] ret_from_kernel_thread+0x1c/0x24


Kernel Fault: Code=26 regs=11474480 (Addr=00000004)

     YZrvWESTHLNXBCVMcbcbcbcbOGFRQPDI
PSW: 00000000000001001111111100001111 Not tainted
r00-03  00000000 10460010 103eb100 0000000c
r04-07  103915cc 10391590 ffffffed 11750000
r08-11  11462800 00000003 10313800 11750220
r12-15  10332800 10391810 80000080 11462848
r16-19  f000167c 00000060 f00001f0 00000000
r20-23  00000030 0000000e f2000000 00000020
r24-27  11750230 00000a00 11462848 1036f010
r28-31  00000000 11474488 11474480 1020119c
sr0-3   00000000 00000000 00000000 00000000
sr4-7   00000000 00000000 00000000 00000000

IASQ: 00000000 00000000 IAOQ: 103eb120 103eb124
 IIR: 0e681096    ISR: 00000000  IOR: 00000004
 CPU:        0   CR30: 11474000 CR31: 103c8000
 ORIG_R28: 1048a1a0
 IAOQ[0]: tulip_init_one+0x23c/0xdc0
 IAOQ[1]: tulip_init_one+0x240/0xdc0
 RP(r2): tulip_init_one+0x21c/0xdc0
Kernel panic - not syncing: Kernel Fault
=======================================================================

2.4.27-pa4 runs fine with the Tulip NIC.

$ grep -i tulip /var/log/dmesg 
Linux Tulip driver version 0.9.15-pre12 (Aug 9, 2002)
tulip0: no phy info, aborting mtable build
tulip0:  MII transceiver #1 config 1000 status 782d advertising 0061.
eth0: Digital DS21143 Tulip rev 48 at 0xff00, 00:60:B0:EA:C8:DE, IRQ 96.

$ /usr/bin/lspci
0000:00:13.0 SCSI storage controller: LSI Logic / Symbios Logic 53c875 (rev 04)
0000:00:14.0 Ethernet controller: Digital Equipment Corporation DECchip 21142/43 (rev 30)
0000:01:04.0 Display controller: Hewlett-Packard Company Visualize FX4 (rev 02)

Alright. Where is the right place to send this? 

-- 
This ain't no sig.

[-- Attachment #1.2: Type: application/pgp-signature, Size: 307 bytes --]

[-- Attachment #2: Type: text/plain, Size: 169 bytes --]

_______________________________________________
parisc-linux mailing list
parisc-linux@lists.parisc-linux.org
http://lists.parisc-linux.org/mailman/listinfo/parisc-linux

Re: [parisc-linux] C240 and Linux 2.6.x

[Grant Grundler]

On Mon, Jan 03, 2005 at 07:13:34PM +0100, Alexander Bartolich wrote:
> This gives me nice console output on booting, so there is definitely
> something wrong with Debian's kernel-image-2.6.8-1-32.

Good.

> Howver, now comes the bad news.

Thanks for capturing and sending the console log.

...
> Linux Tulip driver version 1.1.13 (May 11, 2002)
> Backtrace:

[ hrm...IOAQ is missing from the backtrace ... maybe that's intentional ]

>  [<10202ea4>] pci_device_probe_static+0x60/0x78
>  [<10202ed8>] __pci_device_probe+0x1c/0x34
>  [<10202f1c>] pci_device_probe+0x2c/0x58
>  [<1023dbe0>] driver_probe_device+0x48/0x90
...
> Kernel Fault: Code=26 regs=11474480 (Addr=00000004)

I don't see this with -pa2 or -pa3 on a500 or j6000.
Likely some problem with dino or ccio support.

>  IAOQ[0]: tulip_init_one+0x23c/0xdc0
>  IAOQ[1]: tulip_init_one+0x240/0xdc0
>  RP(r2): tulip_init_one+0x21c/0xdc0

Just have to figure out where in tulip_init_one() +0x240 is pointing.
I'll try to track this down after plowing through the
back-from-2-weeks-off email backlog.

> Alright. Where is the right place to send this? 

Normally just to parisc-linux@lists.p-l.o mailing list.
I've dropped debian-hppa since the bits aren't in debian (yet).

thanks,
grant
_______________________________________________
parisc-linux mailing list
parisc-linux@lists.parisc-linux.org
http://lists.parisc-linux.org/mailman/listinfo/parisc-linux

Re: [parisc-linux] C240 and Linux 2.6.x

[Grant Grundler]

On Mon, Jan 03, 2005 at 07:13:34PM +0100, Alexander Bartolich wrote:
> I fetched palinux-b180-defconfig-2.6.10-pa2_0-2_all.deb from
> http://cvs.parisc-linux.org/download/linux-2.6/autobuild/b180_defconfig/
> and build a nice vmlinux-2.6.10-pa2.
...

> Linux Tulip driver version 1.1.13 (May 11, 2002)
...
> Kernel Fault: Code=26 regs=11474480 (Addr=00000004)
...
>  IAOQ[0]: tulip_init_one+0x23c/0xdc0
>  IAOQ[1]: tulip_init_one+0x240/0xdc0
>  RP(r2): tulip_init_one+0x21c/0xdc0

The code in question is:
...
103eb100:       34 17 00 40     ldi 20,r23
103eb104:       2b 78 e0 00     addil f1000,dp,%r1
103eb108:       34 1a 00 00     ldi 0,r26
103eb10c:       48 33 01 60     ldw b0(,r1),r19
103eb110:       34 f8 04 60     ldo 230(r7),r24
103eb114:       91 00 20 00     cmpiclr,= 0,r8,r0
103eb118:       08 0f 02 5a     copy r15,r26
103eb11c:       34 19 14 00     ldi a00,r25
103eb120:       0e 68 10 96     ldw 4(,r19),r22		<== Fault
103eb124:       e6 c0 20 00     be,l 0(sr4,r22),%sr0,%r31
103eb128:       08 1f 02 42     copy r31,rp
103eb12c:       87 80 2c 50     cmpib,= 0,ret0,103eb75c <tulip_init_one+0x878>
103eb130:       0d 7c 12 90     stw ret0,8(,r11)
103eb134:       49 73 00 20     ldw 10(,r11),r19
103eb138:       37 94 10 00     ldo 800(ret0),r20

"103eb124: be,l" is an indirect function call.
The fault occurs when the function pointer is loaded.
This means the function table pointer (R19) hasn't been initialized.

I'm still trying to track down which indirection function this is.

grant
_______________________________________________
parisc-linux mailing list
parisc-linux@lists.parisc-linux.org
http://lists.parisc-linux.org/mailman/listinfo/parisc-linux

Re: [parisc-linux] C240 and Linux 2.6.x

[Grant Grundler]

On Mon, Jan 03, 2005 at 07:13:34PM +0100, Alexander Bartolich wrote:
> http://cvs.parisc-linux.org/download/linux-2.6/autobuild/b180_defconfig/
> and build a nice vmlinux-2.6.10-pa2.
...
> CPU(s): 1 x PA8200 (PCX-U+) at 236.000000 MHz

Sorry - I should have recognized this sooner.
The B180 defconfig will NOT run on C-class machines:
#
# Bus options (PCI, PCMCIA, EISA, GSC, ISA)
# 
CONFIG_GSC=y
# CONFIG_HPPB is not set
# CONFIG_IOMMU_CCIO is not set
...

CONFIG_IOMMU_CCIO needs to be enabled.
The failing function call was probably dma_ops table not being initialized.

Can you try this "32" bit kernel?
http://cvs.parisc-linux.org/download/linux-2.6/autobuild/32/palinux-32-2.6.10-pa5_0-2_all.deb

thanks,
grant
_______________________________________________
parisc-linux mailing list
parisc-linux@lists.parisc-linux.org
http://lists.parisc-linux.org/mailman/listinfo/parisc-linux