[Pkg-acpi-devel] Bug#522756: Same problem on ThinkPad X40

[Soenke]

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

It seems that I have the same problem (or at least a similar one) on my
ThinkPad X40.
After resume, acpid sometimes crashes. Normally, this happens in
conjunction with USB hub errors. I have an external keyboard and mouse
attached by USB.
Also, acpid crashes in much the same way when I unplug my USB hub (the
one with the KB/mouse on it) during normal operation, making the error
nicely reproducible. ;)

Output of some commands when unplugging the USB hub:

- ------------------------------------------------------
acpi -dl:

acpid: input layer read error: No such device (19)
acpid: input layer read error: No such device (19)
acpid: input layer read error: No such device (19)
acpid: input layer read error: No such device (19)
acpid: input layer read error: No such device (19)
acpid: too many errors reading input layer - aborting

The same appears in the syslog.
- ------------------------------------------------------

- ------------------------------------------------------
strace -p $(pidof acpid):

select(12, [3 4 5 6 7 8 9 10 11], NULL, NULL, NULL) = 2 (in [8 9])
read(8, 0xbfae2a50, 16)                 = -1 ENODEV (No such device)
time(NULL)                              = 1240423440
writev(2, [{"acpid: input layer read error: No"..., 51}], 1) = 51
send(13, "<27>Apr 22 20:04:00 acpid: input "..., 71, MSG_NOSIGNAL) = 71
read(9, 0xbfae2a50, 16)                 = -1 ENODEV (No such device)
time(NULL)                              = 1240423440
writev(2, [{"acpid: input layer read error: No"..., 51}], 1) = 51
send(13, "<27>Apr 22 20:04:00 acpid: input "..., 71, MSG_NOSIGNAL) = 71
select(12, [3 4 5 6 7 8 9 10 11], NULL, NULL, NULL) = 2 (in [8 9])
read(8, 0xbfae2a50, 16)                 = -1 ENODEV (No such device)
time(NULL)                              = 1240423440
writev(2, [{"acpid: input layer read error: No"..., 51}], 1) = 51
send(13, "<27>Apr 22 20:04:00 acpid: input "..., 71, MSG_NOSIGNAL) = 71
read(9, 0xbfae2a50, 16)                 = -1 ENODEV (No such device)
time(NULL)                              = 1240423440
writev(2, [{"acpid: input layer read error: No"..., 51}], 1) = 51
send(13, "<27>Apr 22 20:04:00 acpid: input "..., 71, MSG_NOSIGNAL) = 71
select(12, [3 4 5 6 7 8 9 10 11], NULL, NULL, NULL) = 2 (in [8 9])
read(8, 0xbfae2a50, 16)                 = -1 ENODEV (No such device)
time(NULL)                              = 1240423440
writev(2, [{"acpid: input layer read error: No"..., 51}], 1) = 51
send(13, "<27>Apr 22 20:04:00 acpid: input "..., 71, MSG_NOSIGNAL) = 71
time(NULL)                              = 1240423440
writev(2, [{"acpid: too many errors reading in"..., 54}], 1) = 54
send(13, "<27>Apr 22 20:04:00 acpid: too ma"..., 74, MSG_NOSIGNAL) = 74
exit_group(1)                           = ?
Process detached
- ------------------------------------------------------

I recently upgraded to kernel version 2.6.29-1-686 (from 2.6.26-2-686)
and the new Xorg in unstable. Could there be a connection with the evdev
input driver? I was using kbd and mouse Xorg drivers before without trouble.
Please tell me if I can provide more information to help in solving this.

Regards,

Soenke

Versions of possibly relevant packages:
acpid 1.0.8-8
linux-image-2.6.29-1-686
xserver-xorg-core-2:1.6.1-1
xserver-xorg-input-evdev-1:2.2.1-1
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)

iD8DBQFJ72BF094f/H6wEuMRAmNyAKCgJelVp/1zujJ/q8ysoLNDzpeqfACgl59m
eHsIgbBDO78HmV+Eku+X6LQ=
=pgUI
-----END PGP SIGNATURE-----