Skip to contentSkip to conversation

Reboot & ACPI suspend Laptop display initialization

DISCUSSION Mike Mohr 2 participants 6 messages Sep 16, 2005 – Sep 22, 2005 1 message not in the archive

lore
Share & tools
lore

Mike Mohr

author
#0 lore raw

There seems to be an issue with the VESA driver for my laptop's video card. I can install any version of the kernel (tested with 2.4.{23,24}, 2.4.31, and various 2.6.XX > 2.6.8) and the system turns on and boots fine. However, if I go into suspend to ram (which I think is ACPI S3) where the display is turned off -or- I reboot without shutting down first, the display is not re-initialized.

The system seems to come back on into a semi-usable state (i.e. if I reboot the kernel seems to boot and the system comes up), but the display remains solid black, so I have not been able to verify this. Come to think of it, I could probably connect the laptop up to my home network and SSH into it after this happens, but I haven't yet done so.

My laptop is a Toshiba Satellite 1800 S207 with a Trident XPAi1 video card. If I use the tridentfb driver (at least as of kernel 2.4) at boot time, the screen turns various shades of gold immediately after the kernel loads. Can someone help me debug this issue so that a fix can be written? The problem happens even with a pristine installation of Slackware 10.2 with the default bare 2.4.31 kernel.

TIA

Mike

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

author
#2 lore raw

This isn't just a problem with ACPI's suspend. If I reboot the machine, without without power management, the screen isn't properly reinitialized. The problem with the suspend is certainly an issue with ACPI; however, the reboot issue probably isn't. Unless someone can enlighten me?

Quoted 13 lines · Michal Jaegermann
Quoted 3 lines · Mike Mohr

However, if I go into suspend to ram (which I
think is ACPI S3) where the display is turned off -or- I reboot without shutting down first, the display is not re-initialized.

Did you try with 'acpi_sleep=s3_bios' kernel parameter as documentation suggests? I have the same problem with video on Acer Travelmate and this helps. Remember that with ACPI (or APM) you are to a great extent at a mercy of your BIOS supplier and that part is often seriously buggy. APM is simpler so it often used to be in a better shape.

Michal

#3 lore raw

Hi!

Quoted 5 lines

This isn't just a problem with ACPI's suspend. If I reboot the machine, without without power management, the screen isn't properly reinitialized. The problem with the suspend is certainly an issue with ACPI; however, the reboot issue probably isn't. Unless someone can enlighten me?

Well, the reboot issue is broken BIOS... It should reset the hardware when you are rebooting.

Read Doc*/power/video.txt for suspend issues.
Pavel

Signature · 1 line
if you have sharp zaurus hardware you don't need... you know my address
author
#4 lore raw

Hmm. It seems unlikely to me that Toshiba will bother fixing a bug like this (even as serious as this one is) for a 3-yr-old laptop on my request. I checked, and I already have the latest BIOS released by them on 9-24-2002. The fact that they let a bug as serious as this through their QA really shocks me, esp. since my wife works for Panasonic QA and I know the procedures they go through. I will never buy from Toshiba again.

Thanks for the responses guys!

Mike

Quoted 16 lines · Pavel Machek

Hi!

Quoted 5 lines

This isn't just a problem with ACPI's suspend. If I reboot the machine, without without power management, the screen isn't properly reinitialized. The problem with the suspend is certainly an issue with ACPI; however, the reboot issue probably isn't. Unless someone can enlighten me?

Well, the reboot issue is broken BIOS... It should reset the hardware when you are rebooting.

Read Doc*/power/video.txt for suspend issues.
Pavel
--
if you have sharp zaurus hardware you don't need... you know my address

#5 lore raw

Hi!

Don't top-post.

Quoted 5 lines

Hmm. It seems unlikely to me that Toshiba will bother fixing a bug like this (even as serious as this one is) for a 3-yr-old laptop on my request. I checked, and I already have the latest BIOS released by them on 9-24-2002. The fact that they let a bug as serious as this through their QA really shocks me, esp. since my wife works for

Actually I'm not surprised. Embedded controller locked up because Linux did something unexpected. That's still hardware problem, but quite tricky to catch.

You can still try to find "what us it that upsets EC", and work around it in Linux. printk/mdelay is your friend.

Pavel

Signature · 1 line
64 bytes from 195.113.31.123: icmp_seq=28 ttl=51 time=448769.1 ms         
#6 lore raw
In reply to Mike Mohr

Hi!

Quoted 7 lines

Hmm. It seems unlikely to me that Toshiba will bother fixing a bug like this (even as serious as this one is) for a 3-yr-old laptop on my request. I checked, and I already have the latest BIOS released by them on 9-24-2002. The fact that they let a bug as serious as this through their QA really shocks me, esp. since my wife works for Panasonic QA and I know the procedures they go through. I will never buy from Toshiba again.

Okay, sorry, it is not EC; I was replying to wrong mail. Chance of linux workaround still exists.

Signature · 1 line
64 bytes from 195.113.31.123: icmp_seq=28 ttl=51 time=448769.1 ms