Re: [PATCH] ide-cd: fix kernel crash on hppa regression

[Helge Deller <deller@gmx.de>]

Hi Borislav,

Borislav Petkov wrote:
> Hi,
> 
> On Thu, Apr 16, 2009 at 12:45:31AM +0200, Helge Deller wrote:
>> With 2.6.30-rc2 I face a kernel crash on the 32bit hppa architecture 
>> due to ide-cd when udev creates the device nodes at startup:
>>
>> Kernel Fault: Code=26 regs=8ed34c40 (Addr=00000024)
>> IASQ: 00000000 00000000 IAOQ: 1034b5ac 1034b5b0
>>  IIR: 4ab30048    ISR: 00000000  IOR: 00000024
>>  CPU:        0   CR30: 8ed34000 CR31: ffff55ff
>>  ORIG_R28: 00000000
>>  IAOQ[0]: ide_complete_rq+0x2c/0x70
>>  IAOQ[1]: ide_complete_rq+0x30/0x70
>>  RP(r2): cdrom_newpc_intr+0x178/0x46c
>> Backtrace:
>>  [<1035c608>] cdrom_newpc_intr+0x178/0x46c
>>  [<1034c494>] ide_intr+0x1b0/0x214
>>  [<1016d284>] handle_IRQ_event+0x70/0x150
>>  [<1016d4b0>] __do_IRQ+0x14c/0x1cc
>>  [<102f7864>] superio_interrupt+0x88/0xbc
>>  [<1016d284>] handle_IRQ_event+0x70/0x150
>>  [<1016d4b0>] __do_IRQ+0x14c/0x1cc
>>  [<10112efc>] do_cpu_irq_mask+0x9c/0xd0
>>  [<10116068>] intr_return+0x0/0x4
>>
>> This crash seems to happen due to an uninitialized variable "rc".
>> The compiler even warns about that:
>>   CC      drivers/ide/ide-cd.o                                                         
>> /mnt/sda4/home/cvs/parisc/git-kernel/linus-linux-2.6/drivers/ide/ide-cd.c: In function `cdrom_newpc_intr':
>> /mnt/sda4/home/cvs/parisc/git-kernel/linus-linux-2.6/drivers/ide/ide-cd.c:612: warning: `rc' might be used uninitialized in this function
> 
> can you do
> 
> gdb ./vmlinux
> 
> and then
> 
> (gdb) l *(cdrom_newpc_intr+0x178)
> 
> I think I have a suspect but I want to be sure. In order to do that
> you'll need a debug kernel (set CONFIG_DEBUG_INFO to 'y' in your
> .config). In case the oops above is not from a debug kernel, can you
> catch it again and _then_ do the gdb thing since the offset of 0x178 is
> most probably going to change. In that case, you have to do
> 
> (gdb) l *(cdrom_newpc_intr+NEW_OFFSET_FROM_THE_OOPS)

I tried to build a kernel with CONFIG_DEBUG_INFO.
There are two problem with that, because I'm cross-compiling:
a) I don't have a cross-gdb (but I could use addr2line instead though)
b) the resulting kernel is becoming too big and won't link/boot:
...
  SYSMAP  System.map
  SYSMAP  .tmp_System.map
/sbin/palo -f /mnt/sda4/home/cvs/parisc/git-kernel/linus-linux-2.6-32bit/palo.conf
palo version 1.14 deller@p100 Fri Jul 28 22:55:11 CEST 2006
ELF32 executable
ipl: addr 2048 size 36864 entry 0x0
 ko 0x0 ksz 0 k64o 0x0 k64sz 0 rdo 0 rdsz 0
<>
Your 32-bit-kernel is too big for your F0 partition
make[3]: *** [palo] Error 2

I could now just try to look into the assembly by hand and try
to find which line produces the problem. But reading hppa assembly code
for such a big function is not very easy though.

So, my question would be:
If you have an idea where the problem is, maybe you have a patch for me to try?
That would simplify the whole thing for me a lot, as just compiling/run-testing is much easier for me.

Best regards,
Helge

BTW, even with my patch ATAPI CDROM accesses don't work with the current kernel...