[ltp] Fedora Core 5 & T43: Suspend & resume?

[Clark Williams]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Matthias Runge wrote:
> Thomas Roessler wrote:
>> Hello,
>>
>> I've upgraded my Thinkpad T43 to FC5.  Things continue to work quite
>> nicely; ACPI suspend and resume works when I use this kernel:
>>
>>   Linux raktajino 2.6.16-1.2115_FC4 #1 Mon Jun 5 14:45:53 EDT 2006
i686 i686 i386 GNU/Linux
>>
>> Relevant kernel command line parameters:
>>
>>   acpi_sleep=s3_bios acpi_serialize
>>
>> However, the more recent kernels that come with FC5 kill the X
>> server as soon as the machine wakes up; this is, of course, not
>> acceptable.
>>
>> I wonder if this is one of the usual suspend/resume problems from
>> the "they may fix it in half a year" category, or whether there is
>> some configuration trick, additional driver, or whatever else that
>> would help fix this.
>>
>> Also, what do people here use to suspend and successfully resume
>> their Thinkpads?
>>
>> Thanks,
> Hello,
>
> this seems to be a known problem. (All 2.6.17.*-kernels from Redhat are
> broken). I solved this issue by compiling a new kernel on my own.
> But still there are other issues, e.g. wakeup from works, but
> the systems goes instantly to suspend. After the third wake up,
> the system stays on. Even this is known and documented in RH bugzilla.
>
> Cheers,

Well, I wouldn't say that all the 2.6.17.x kernels are broken. I was
running FC5 on my T41 up to two weeks ago when I moved to "rawhide"
(the FC6 pre-release). I exclusively use suspend to RAM when moving
around (i.e. I close the lid to suspend). The 2.6.17-1.2543.fc6 kernel
suspends and resumes just fine. While I've had other problems with
2.6.17.x kernels, suspend/resume hasn't been one of them.

I should say that I'm using XFCE as my desktop environment, so I had
to go in to /etc/acpi/events and create a lid.conf that looks like this:

event=button/lid
action=/usr/sbin/pm-suspend

Since a large percentage of RH developers work on Thinkpads (mostly
T4[123]'s) I would be surprised if problems like this take a long time
to fix (we had the radeon force-sleep patch in the FC kernel for a
long time, until the sleep issues were fixed in the radeon driver).

What did you do when you compiled your own kernel that fixed this problem?

Clark

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iD8DBQFE4JEnHyuj/+TTEp0RAmWwAKDk0Le+MeQ0xvozevGHJBQEtayBhQCgmotg
IzK9qf6bjrv1g5ZCsMV4NmI=
=8EwA
-----END PGP SIGNATURE-----