Hi there,
I've got some strange problems here, since 2.4.12 (2.4.10 was ok, I
never tried .11)
The System boots fine, I can login on a text console, but when I log out
again, the system partly gets unusable. The load rises, and the text
console where I logout hangs.
I've got seveal processes hanging in "D" state, especially devfsd. I
think something with devfs and/or devfsd is broken here. If I kill
devfsd before, the problem does not arise (but I need devfsd :-)
Now, It gets even more strange: The problem does only exist if I
deactivate ACPI! I tried with a ACPI enabled kernel with the command
line "acpi=off", I tried on a machine which is too old for ACPI, and I
even compiled a kernel without ACPI at all. Everywhere the same problem.
When I boot with activated ACPI, there's no problem.
I don't understand what's happening here. I tried to debug a bit, but
couldn't find something sense- or useful.
Any hints? If someone could tell me what I should try or which debug
info could be useful, please let me know!
TIA, Michael
--
netWorks Vox: +43 316 698260
Michael Reinelt Fax: +43 316 692343
Geisslergasse 4 GSM: +43 676 3079941
A-8045 Graz, Austria e-mail: [email protected]
Michael Reinelt <[email protected]> wrote:
> Hi there,
>
> I've got some strange problems here, since 2.4.12 (2.4.10 was ok, I
> never tried .11)
2.4.10 was the first kernel that broke for me.
> I've got seveal processes hanging in "D" state, especially devfsd. I
> think something with devfs and/or devfsd is broken here. If I kill
> devfsd before, the problem does not arise (but I need devfsd :-)
I've had this problem as well - it seems there's a deadlock in devfs.
I have sent a mail to Richard Gooch about this some days ago, but he
hasn't responded yet.
I will resend my bug report to the list then.
> Now, It gets even more strange: The problem does only exist if I
> deactivate ACPI! I tried with a ACPI enabled kernel with the command
> line "acpi=off", I tried on a machine which is too old for ACPI, and I
> even compiled a kernel without ACPI at all. Everywhere the same problem.
> When I boot with activated ACPI, there's no problem.
I don't know about this.
> Any hints? If someone could tell me what I should try or which debug
> info could be useful, please let me know!
>
> TIA, Michael
>
--
Johannes Kloos
"Walk a mile on these paws and call me a liar."
-- Gaspode the wonder dog (Terry Pratchett, Moving Pictures)
Johannes Kloos wrote:
>
> Michael Reinelt <[email protected]> wrote:
> > I've got seveal processes hanging in "D" state, especially devfsd. I
> > think something with devfs and/or devfsd is broken here. If I kill
> > devfsd before, the problem does not arise (but I need devfsd :-)
> I've had this problem as well - it seems there's a deadlock in devfs.
> I have sent a mail to Richard Gooch about this some days ago, but he
> hasn't responded yet.
> I will resend my bug report to the list then.
Mybe you could CC me the reports? I'm not a regular reader of lkml.
Thanks...
> > Now, It gets even more strange: The problem does only exist if I
> > deactivate ACPI! I tried with a ACPI enabled kernel with the command
> > line "acpi=off", I tried on a machine which is too old for ACPI, and I
> > even compiled a kernel without ACPI at all. Everywhere the same problem.
> > When I boot with activated ACPI, there's no problem.
> I don't know about this.
Well, I've just had the same problem with ACPI. So forget about the last
paragraph.
But I found another strange thing: If I deactivate nscd, the problem
does not occur when logging in and out.
bye, Michael
--
netWorks Vox: +43 316 698260
Michael Reinelt Fax: +43 316 692343
Geisslergasse 4 GSM: +43 676 3079941
A-8045 Graz, Austria e-mail: [email protected]