2.6.0-test5: Hang in i8042_init

[Nick Piggin]

I'm still getting these hangs in 2.6.0-test5.


Nick Piggin wrote:

> Andrew Morton wrote:
>
> > Nick Piggin <piggin@xxxxxxxxxxxxxxx> wrote:
> >
> > > Felipe Alfaro Solana wrote:
> > >
> > >
> > > > On Fri, 2003-08-29 at 06:53, Nick Piggin wrote:
> > > >
> > > >
> > > > > Is what I am getting. Last line is something like input: PC Speaker
> > > > > (followed by the initcall).
> > > > >
> > > > > dmseg and lspci from a working kernel attached. Let me know if I can
> > > > > do more.
> > > > Could it be something related with
> > > > http://bugzilla.kernel.org/show_bug.cgi?id=1123?
> > > Yes it seems quite likely. Further poking reveals that the
> > > box still locks with a PS2 mouse _and_ the USB mouse. A PS2
> > > mouse on its own allows the system to boot, although
> > > interrupt 10 (eth0, usb) is not working. Booting with
> > > acpi=off allows the system to boot normally with the USB
> > > mouse, and interrupt 10 works.
> >
> > Nick, please send a full report.  There's been quite some breakage 
> > from the
> > recent ACPI updates and it'd be nice to work out what happened.
> >
> > Here's the recipe:
> >
> >  The dmesg output of the failing case is really helpful, As is the
> >  output of acpidmp to examine the ACPI tables on the system.  (Red Hat
> >  includes both of these in their severn beta1, acpidmp is also in 
> > pmtools
> >  on intel's ACPI web page) dmidecode output is useful to identify the 
> > BIOS
> >  version.
> >
> >  Of course the 1st thing to check with ACPI failures is that the BIOS
> >  version shown by dmidmp is the latest provided by the vendor...  
> > Plus, if
> >  we determine the BIOS is toast, DMI provides what we need to add the
> >  system to the DMI or acpi blacklists.
> >
> >  We're seeting the most problems on VIA chip-sets with no IO-APIC.  The
> >  one below is unusual because it is a 2-way system with 3 IO-APICs.
> >
> >  The latest code in linus' tree includes ACPICA 20030813, which is
> >  slightly newer than the one below, it might be a good idea to try that
> >  with CONFIG_ACPI_DEBUG.  Note that it will spit out the DMI info 
> > upon the
> >  mount root failure automatically.
> >
> > Thanks.
>
> OK, well the computer is a 1 CPU Pentium 4 Celeron. I can't get the
> failing dmesg because I can't get a serial console onto it (it isn't
> my computer). The last line of the dmesg is: "input: PC Speaker", and
> the last initcall is i8042_init.
>
> The computer has a PS2 keyboard and a USB mouse. Booting with a PS2
> mouse attached as well still causes a hang. Booting without the USB
> mouse works, however the USB interrupt (2 ohci-hcd's and eth0 share it)
> no longer works, so eth0 doesn't work, and plugging in the USB mouse
> does nothing.
>
> Attached are dmesg, interrupts, .config, acpidmp, dmidecode, etc.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/