Update/new info (was: [ltp] X30 + 2.6.13 + APM == broken suspend/resume... help?)

[Karl Klashinsky]

For those who might be interested...

Adding mm patch-set, and bringing the kernel up to 2.6.13-mm2 resolved
the problem.

Based on the IRQ/device messages coming from the kernel during the
resume (but prior to the x30 locking up), it appeared that the machine
was freezing up near the end of resuming the graphics card/display, or
just as the PCMCIA subsystem was resuming.  I decided to try the mm2
patch-set after noticing these two comments in the mm2 changelog:

- Large revamp of pcmcia suspend handling
- Lots of framebuffer driver updates

Anyway, with 2.6.13-mm2, everything seems to be fine (but I've only been
using it for about 12hrs so far).

On Fri, 2005-09-02 at 09:52 -0700, Karl Klashinsky wrote:
> Note, my X30 is one of the earlier models, with an i830 graphics chip
> (i.e., no radeon chipset in this one).
> 
> I'm sure I'm not running ACPI... I disabled ACPI during the kernel
> build, and there is no /proc/acpi after boot.
> 
> The suspend works fine, box goes nicely to sleep (fans, disk, screen,
> etc, all power down), then my little "moon" light comes on.
> 
> The resume starts out ok, the screen comes back to life, disk spins up,
> screen gets redrawn... then it freezes.  No response to keyboard, etc. 
> I've let it sit for up to 30mins to be sure it wasn't just a
> hiccup/pause.   Only recourse is the "hold power button for 10 secs"
> ploy.
> 
> The odd part is that swsusp (suspend to swap disk) works just fine. 
> Also odd is that basic suspend/resume has worked fine in every other
> kernel I've used on this machine, incl. my most recent kernel, 2.6.12.3.
> 
> I've also tried switching to the text console and initiating the suspend
> manually via "apm --suspend", then watching what happens on resume
> (i.e., watching the "dmesg" style diag's about which devices are being
> restarted, etc).
> 
> On resume, I only see one "set" of those diag's, the ones for my
> graphics card, before things lock up.  That hints to me that the freeze
> is happening either right at the end of initializing the graphics card,
> or as it starts to initialize the next device.
> 
> Any clues?  
> 
> -- 
> klash        |       |        Mitchell's Law of Committees: 
>             .|.     .|.       Any simple problem can be made
>            .|||.   .|||.      insoluble if enough meetings 
>          .:::::::.:::::::..   are held to discuss it.
-- 
klash        |       |      Galbraith's Law of Human Nature: Faced 
            .|.     .|.     with the choice between changing one's
           .|||.   .|||.    mind and proving that there is no need
         .:::::::.:::::::.  to do so, almost everybody gets busy
                            on the proof.