Trackpoint (psmouse) dead after suspend/resume (S3)

DISCUSSION [email protected] 4 participants 6 messages Jul 19, 2004 – Aug 12, 2004 4 messages not in the archive ACPI

Share & tools

[email protected]

author

2004-07-19 22:43 UTC #0 lore raw

On 2.6.7, the suspend/resume cycle kills the trackpoint on my Fujitsu P2120. This definitely worked before, but I can't figure out what the critical change is. Here is the approximate history:

2.5.xx: keyboard and trackpoint dead, reloading psaux module fixes both 2.6.4: both devices work, no reloading necessary
2.6.7: keyboard works, trackpoint dead, reloading psmouse doesn't help

I've poked around a bit and it seems that all writes to the AUX-port times out, this causes the psmouse to disconnect.

Any suggestions appreciated, Ole

-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=4721&alloc_id=10040&op=click

A message that is not in the archive (sent off-list or not yet received): yzollhfbou5.fsf-bi+AKbBUZKY6gyzm1THtWbp2dZbC/[email protected]

Vojtech Pavlik

2004-07-27 13:52 UTC #2 lore raw

Quoted 10 lines · On Tue, Jul 20, 2004 at 12:43:14AM +0200…

On 2.6.7, the suspend/resume cycle kills the trackpoint on my Fujitsu P2120. This definitely worked before, but I can't figure out what the critical change is. Here is the approximate history:
2.5.xx: keyboard and trackpoint dead, reloading psaux module fixes both 2.6.4: both devices work, no reloading necessary
2.6.7: keyboard works, trackpoint dead, reloading psmouse doesn't help
I've poked around a bit and it seems that all writes to the AUX-port times out, this causes the psmouse to disconnect.

There's a bug in the s3 code in 2.6.7 that doesn't call the proper callbacks upon resume. IIRC this was fixed recently.

Signature · 9 lines

Vojtech Pavlik
SuSE Labs, SuSE CR


-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=4721&alloc_id=10040&op=click

A message that is not in the archive (sent off-list or not yet received): [email protected]

[email protected]

author

2004-07-27 14:55 UTC #4 lore raw

Vojtech> There's a bug in the s3 code in 2.6.7 that doesn't call the Vojtech> proper callbacks upon resume. IIRC this was fixed recently.

Thanks for letting me know!

For the Fujitsu P2120 I ended up having to change the DSDT for it to work with 2.6.7. I'll try and see if the stock DSDT works with 2.6.8 when it comes out.

Regards, Ole

Pavel Machek

2004-08-11 13:18 UTC #5 lore raw

In reply to a message not in the archive

Hi!

Quoted 8 lines

2.5.xx: keyboard and trackpoint dead, reloading psaux module fixes both 2.6.4: both devices work, no reloading necessary
2.6.7: keyboard works, trackpoint dead, reloading psmouse doesn't help
I've poked around a bit and it seems that all writes to the AUX-port times out, this causes the psmouse to disconnect.
Any suggestions appreciated, Ole

See if psmouse.proto=bare helps?

Signature · 9 lines

64 bytes from 195.113.31.123: icmp_seq=28 ttl=51 time=448769.1 ms         



-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285

A message that is not in the archive (sent off-list or not yet received): 20040811131848.GA1120-u08AdweFZfgxtPtxi4kahqVXKuFTiq87@public.gmane.org

Nathan Bryant

2004-08-11 20:06 UTC #7 lore raw

Quoted 7 lines · Pavel Machek

Hi!
Quoted 4 lines
2.5.xx: keyboard and trackpoint dead, reloading psaux module fixes both 2.6.4: both devices work, no reloading necessary
2.6.7: keyboard works, trackpoint dead, reloading psmouse doesn't help

Quoted 7 lines

Quoted 4 lines
I've poked around a bit and it seems that all writes to the AUX-port times out, this causes the psmouse to disconnect.
Any suggestions appreciated, Ole
See if psmouse.proto=bare helps?

maybe also i8042.reset?

Ole, is it losing interrupts or is it completely dead? I know that serio only has knowledge of the generic i8042 device; in most cases it doesn't know a lot about initializing multi-function controller chips, such as the Winbond. The Winbond documentation is a little obscure on this subject - OK, a lot obscure - depending on which chip docs you look at, it only lists the register maps and doesn't describe the I/O port accesses needed to set the registers. However, if you look at parport_pc.c, you can see how to access the registers.

In any case, psmouse is just a shim between input and serio, not the driver itself, so as you have seen, reloading it doesn't help in most cases.

-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media 100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33 Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285

A message that is not in the archive (sent off-list or not yet received): [email protected]

[email protected]

author

2004-08-12 05:13 UTC #9 lore raw

Ole> 2.5.xx: keyboard and trackpoint dead, reloading psaux module fixes both Ole> 2.6.4: both devices work, no reloading necessary
Ole> 2.6.7: keyboard works, trackpoint dead, reloading psmouse doesn't help

Pavel> See if psmouse.proto=bare helps?

Nathan> maybe also i8042.reset?

I had concluded that the track-point hardware was in fact not powered up and I was able to work around the problem by telling BIOS I'm running some Windows. This makes the S3 resume sequence take notably longer - it used to wake up almost promptly but now it spends a few seconds flashing various symbols ("LEDs") before it comes back to C.

Strangely, it seems that I have to use a custom DSDT where I explicitly initialize "Name (TOSS, 0x01)", I'm not able to achieve the same effect by means of acpi_osi.

Strangely as well, 2.6.4 worked (and had the "long" resume sequence) with the stock DSDT.

Thanks, Ole