Re: [Cooker] X mouse: depends on how X started...

2002-04-06 Thread Pierre Fortin

On Thu, 4 Apr 2002 07:39:20 -0500 (EST) Pierre Fortin
[EMAIL PROTECTED] wrote:

No-one has responded yet, and I'm still digging...

If I start X via reboot or init 5 -- no mouse.

However, while trying to debug this, if I issue:

  init 5; sleep 7; ps h -C X -o pid= | xargs strace -tto X.strace -p 

In this case, X *does* connect to the mouse, so the strace output is not
very useful possibly except for some error messages:

# while logging the startup
12:43:19.300661 write(0, (**) , 5)= 5
12:43:19.301448 write(0, Option \Device\, 15) = 15
12:43:19.302244 write(0,  \/dev/psaux\, 13) = 13
12:43:19.303023 write(0, \n, 1)   = 1
12:43:19.303760 open(/dev/psaux, O_RDWR|O_NONBLOCK|O_EXCL) = 7
12:43:19.304756 ioctl(7, SNDCTL_TMR_TIMEBASE, 0xbae0) = -1 ENOTTY
(Inappropriate ioctl for device)
12:43:19.305540 close(7)= 0

# after accessing: /usr/X11R6/lib/X11/xkb/compiled/server-0.xkm
12:43:21.627157 open(/dev/psaux, O_RDWR|O_NONBLOCK|O_EXCL) = 7
12:43:21.628012 ioctl(7, SNDCTL_TMR_TIMEBASE, 0xbac0) = -1 ENOTTY
(Inappropriate ioctl for device)
12:43:21.628720 ioctl(7, SNDCTL_TMR_TIMEBASE, 0xba50) = -1 ENOTTY
(Inappropriate ioctl for device)
12:43:21.629402 write(7, \364, 1) = 1
12:43:21.631133 nanosleep({0, 3000}, NULL) = 0
12:43:21.665253 ioctl(7, 0x540b, 0) = -1 ENOTTY
(Inappropriate ioctl for device)
12:43:21.666046 select(1024, [7], NULL, NULL, {0, 0}) = 1
(in [7], left {0, 0})
12:43:21.668206 read(7, \372, 4)  = 1
12:43:21.668938 select(1024, [7], NULL, NULL, {0, 0}) = 0 (Timeout)
12:43:21.670278 write(7, \346, 1) = 1
12:43:21.672004 write(7, \364, 1) = 1
12:43:21.675894 write(7, \363d, 2)= 2
12:43:21.682674 write(7, \350\2, 2)   = 2
12:43:21.689536 nanosleep({0, 3000}, NULL) = 0
12:43:21.725137 ioctl(7, 0x540b, 0) = -1 ENOTTY
(Inappropriate ioctl for device)

There are other instances where the mouse is opened/closed; but no
additional hints as to what may be preventing the mouse from being seen on
bootup or init 5...

Any ideas for tracking this down would be appreciated... unfortunately,
trying to monitor the startup appears to alter timings...  

Pierre

 
 On LM8.2, I've been chasing a strange problem...
 
 The mouse worked after 8.2 install; but now, I did what I hate most
 (reboot) after taking my laptop along on a short trip...
 
 Under the following  conditions, the mouse *works in VC mode* (gdm);
 but *not* under X:
 - booting with or without autologin
 - init 3 -- init 5
 
 # lsof | grep psaux  # *should* give something like:
 X1905  root  5u  CHR10,1  84 /dev/misc/psaux
 but I get nothing when the problem exists...
 
 From a VC, as a regular user, startx and mouse works.
 
 Anyone have a hint...?
 
 Still looking,
 Pierre
 
 




[Cooker] X mouse: depends on how X started...

2002-04-04 Thread Pierre Fortin


On LM8.2, I've been chasing a strange problem...

The mouse worked after 8.2 install; but now, I did what I hate most
(reboot) after taking my laptop along on a short trip...

Under the following  conditions, the mouse *works in VC mode* (gdm);
but *not* under X:
- booting with or without autologin
- init 3 -- init 5

# lsof | grep psaux  # *should* give something like:
X1905  root  5u  CHR10,1  84 /dev/misc/psaux
but I get nothing when the problem exists...

From a VC, as a regular user, startx and mouse works.

Anyone have a hint...?

Still looking,
Pierre