2.2.17pre20-ben3 + pmud-0.7-1: still kernel panic upon wakeup

2.2.17pre20-ben3 + pmud-0.7-1: still kernel panic upon wakeup

[Albrecht Dress]

I upgraded the pmud package to the latest version (pmud-0.7-1), but
unfortunately I still get a kernel panic when I try to wake up my Lombard
again (see message dump below). Does anyone have an idea what goes wrong
here? Or does anybody have a working pmud/2.2.17pre20-ben3 combination?

Thanks for any help,

Albrecht.

pmud: PMU version 11: 1999 G3 Lombard
pmud: pmud startup succeeded
pmud[1734]: running /etc/power/pwrctl maximum ac
[...]
pmud[1734]: initiating user requested sleep
pmud[1734]: running /etc/power/pwrctl sleep ac
pmud[1734]: going to sleep
kernel: media bay 0 contains a CD-ROM drive
kernel: keybdev.c: Removing keyboard: input1
kernel: ADB keyboard at 2, handler 1
kernel: Detected ADB keyboard, type ISO, swapping keys.
kernel: keybdev.c: Adding keyboard: input1
kernel: event1: Event device for input1
kernel: input1: ADB HID on ID 2:2.c4
kernel: event2: Event device for input2
kernel: input2: ADB HID on ID 7:7.1f
kernel: ADB mouse at 3, handler set to 4 (trackpad)
kernel: mouse1: PS/2 mouse device for input3
kernel: event3: Event device for input3
kernel: input3: ADB HID on ID 3:3.01
kernel: hda: Enabling MultiWord DMA 2
kernel: Unable to handle kernel NULL pointer dereference at virtual
address 00000024 (error 40000000)
kernel: NIP: C00B74C8 XER: 00000000 LR: C00B7C04 REGS: c92f1c80 TRAP: 0300
kernel: MSR: 00009032 [EEIRDRME]
kernel: TASK = c92f0000[1734] 'pmud' mm->pgd c9f91000 Last syscall: 54
kernel: last math d2366000
kernel: GPR00: C00B7C04 C92F1D30 C92F0000 C01CBD70 00000000 C014A490
00004000 C01C07F0
kernel: GPR08: 00000000 00000000 80800020 C0270A40 44000024 1001E8BC
00000000 100A3310
kernel: GPR16: 10040000 10040000 00000000 00000000 00009032 092F1E80
C01CC14C C01CBBE8
kernel: GPR24: C0190000 C01CC148 00000002 00418570 00000008 00000001
00000000 C01CBD70
kernel: Call backtrace:
kernel: C017E6D0 C00B7C04 C00E88A8 C01898F4 C00E8CA0 C003C178 C000392C
kernel: 0FF8D9A4 1000339C 10001610 0FF0B69C 00000000
kernel: Kernel panic: kernel access of bad area pc c00b74c8 lr c00b7c04
address 24 tsk pmud/1734


+-----------------------------------------------------------------------------+
| Dr.-Ing. Albrecht Dre\ss                                     """"           |
| Max-Planck-Institut f\"ur Radioastronomie   |\       /      /o  o\          |
| Abteilung f\"ur Infrarot-Interferometrie    |  \    /      |  /   |         |
| Auf dem H\"ugel 69                          |    \ |        \ ---/          |
| D-53121 Bonn (Germany)          ------------+------+-------------------     |
| Phone (+49) 228 525 319                     |    / |                        |
| Fax   (+49) 228 525 411                     |  /  /                         |
| Mail  ad@mpifr-bonn.mpg.de                  |/   /                          |
| http://www.mpifr-bonn.mpg.de/div/iri                                        |
+-------------- electrical engineers do it with less resistance --------------+


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: 2.2.17pre20-ben3 + pmud-0.7-1: still kernel panic upon wakeup

[Olaf Hering]

On Thu, Oct 12, Albrecht Dress wrote:

>
> I upgraded the pmud package to the latest version (pmud-0.7-1), but
> unfortunately I still get a kernel panic when I try to wake up my Lombard
> again (see message dump below). Does anyone have an idea what goes wrong
> here? Or does anybody have a working pmud/2.2.17pre20-ben3 combination?

Can you try Benhs latest?
http://penguinppc.org/~olaf/vmlinux-benh.gz


Gruss Olaf

--
 $ man clone

BUGS
       Main feature not yet implemented...

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: 2.2.17pre20-ben3 + pmud-0.7-1: still kernel panic upon wakeup

[Michael Schmitz]

>
> I upgraded the pmud package to the latest version (pmud-0.7-1), but
> unfortunately I still get a kernel panic when I try to wake up my Lombard
> again (see message dump below). Does anyone have an idea what goes wrong
> here? Or does anybody have a working pmud/2.2.17pre20-ben3 combination?

Was the CDROM mounted when you put the machine to sleep? I don't see this
with either 2.2.17pre20-ben1 nor with 2.2.18pre14-ben1 (one of which may
be patched in ide-pmac.c from the CDROM sleep debugging sessions).

	Michael


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: 2.2.17pre20-ben3 + pmud-0.7-1: still kernel panic upon wakeup

[Albrecht Dress]

On Thu, 12 Oct 2000, Olaf Hering wrote:
> Can you try Benhs latest?
> http://penguinppc.org/~olaf/vmlinux-benh.gz

Got it, this one does not panic. I'll rsync from Ben's tree asap, thanks a
lot for your help! However, I recognised a quite odd behaviour of this
kernel: using the trackpad, the x and y coordinates are exchanged (moving
the trackpad up and down moves the cursor left and right and vice versa).
Not sure if this was caused by sleeping the machine (I reverted to my old
kernel as I doid not want to download the modules, too).

Thanks again, Albrecht.


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: 2.2.17pre20-ben3 + pmud-0.7-1: still kernel panic upon wakeup

[Michael Schmitz]

> > Can you try Benhs latest?
> > http://penguinppc.org/~olaf/vmlinux-benh.gz
>
> Got it, this one does not panic. I'll rsync from Ben's tree asap, thanks a

Good.

> lot for your help! However, I recognised a quite odd behaviour of this
> kernel: using the trackpad, the x and y coordinates are exchanged (moving
> the trackpad up and down moves the cursor left and right and vice versa).

Fixed in the pmud-0.7-2 Debian package. All you need to do is patch
pmud.c:

--- pmud-0.7.orig/pmud.c
+++ pmud-0.7/pmud.c
@@ -1502,7 +1502,7 @@
 {
 	int nr;
 	int tries = 5;
-	unsigned char buf[9];
+	unsigned char buf[10];

 	if(adb_program(fd, id, 1) < 0)
 		return 0;
@@ -1522,7 +1522,8 @@
 		buf[2] = settings & TPS_DRAG ? 0x94 : 0x14;
 		buf[4] = settings & TPS_LOCK ? 0xff : 0xb2;

-		if(adbwrite(fd, id, 2, ADBlisten, buf, nr) <= 0)
+		/* buf+1 because the read buffer contains command byte at buf[0] */
+		if(adbwrite(fd, id, 2, ADBlisten, buf+1, nr) <= 0)
 		{
 			syslog(LOG_ERR, "set.set adbwrite: %m");
 			continue;

	Michael


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/