On Tue, Oct 07, 2008 at 09:27:33PM +0200, Thomas Pfaff wrote:
| Hi.
|
| In light of the recent thread about bigmem support, I ran a test (bigmem=1)
| on my system with not-so-great results. Following is a dmesg excerpt:
|
| ...
| uhci4: host controller process error
| uhci4: host controller halted
| ehci0: unrecoverable error, controller halted
| ehci0: blocking intrs 0x10
| ...
| mtrr: Pentium Pro MTRR support
| uhci_freex: xfer=0xffffffe8002301400 not busy, 0x4f4e5155
| ...
Have you confirmed that this only occurs with bigmem=1 ? Try a GENERIC
kernel (so with bigmem=0 and from the same source) and make sure this
doesn't happen in that case too.
| and here it stops. Unfortunately I don't have a serial console so I don't
| have the full dmesg available to you. If you want to, I can pull out my
| camera and do a quick photo session with dmesg in various postures.
Are you sure there's no dmesg data in the buffer if you boot another
kernel ? It's quite common that the data in the dmesg is retained
across reboots (can be *extremely* useful, exactly for cases such as
this).
If you do the photo session, the best way is to take good quality
pictures and then you copy the text from the pictures in an e-mail.
Then put the pictures online somewhere and refer to them in your mail.
| Anything I can do to help you debug this?
You can also try disabling uhci in UKC (boot -c; disable uhci; quit)
and see if that gets you any further.
Good luck,
Paul 'WEiRD' de Weerd
--
>++++++++[<++++++++++>-]<+++++++.>+++[<------>-]<.>+++[<+
+++++++++++>-]<.>++[<------------>-]<+.--------------.[-]
http://www.weirdnet.nl/
