Toshiba M45

Hi MajorCedar;

I'm not sure this'll fix your problem, but give it a chance;

I've had some similar problems with my laptop, so I added this line in my /boot/grub/menu.lst: acpi=off.

So now it should look something like this:
title MEPIS 6.0.1
root (hd1,1)
kernel /boot/vmlinuz-2.6.15-26-386 root=/dev/hdc2 nomce quiet vga=791 acpi=off
initrd /boot/initrd.img-2.6.15-26-386
boot

In my case it stopped the mouse-cursor from having any hick-ups, and didn't slow-down any other process from that point...

If it doesn't do the job, you can just delete this line.

Ruud
=====
Compaq N1050v, Pentium 4 1.8 GHz, 1024Mb Ram, MEPISLite, MEPIS 6 (acpi=off ;D).
Shuttle SN41G2, AMD Athlon XP 3000+, 512Mb RAM Mepis 6.0.
Linuxuserreg # 417292 Tiocfaidh ár lá
==

It didn't work. I lost connectivity after I made the change. Plus, the machine locked up when I tried to suspend to disk. I ended up having to reboot, which is what I feel like I do a lot of, especially since I cannot suspend to disk. Thanks for the suggestion. Hopefully someone out here can help me uderstand this problem.

By the way, what is the function of "ACPI?" Too, someone told me to update my BIOS, which I have not done. I get nervous doing things like that. I don't want to end up with a cool looking paper weight, if you know what I mean. Does that make a difference?

Thanks.

Can anyone out here help me configure my system so that I can suspend to disk/ram? When I do try to suspend to disk (ram does not work at all and ends in a hang up), at startup, I get a small white square at the bottom of my screen and my mouse (touchpad on laptop) does not function. Frustrating! I have to continually shutdown/up. Thanks.

http://www.cantrip.org/toshiba-m45.html?seenIEPage=1

Make sure you read carefully the whole section on getting suspend/hibernate/sleep to work and follow it carefully -- I think it will do what you need.

If that post doesn't help, you'll probably need to learn to compile the kernel (and guess what to change) -- I think there are modules for suspend which may not be in the standard Mepis kernel -- there's one called suspend2, for example -- but I think that post will be all you need if you follow it carefully ...

Mike

I keep hoping someone else knows about this -- you may want to search http://www.ubuntuforums.org
something like "dapper suspend"
or "suspend dapper M45" or "suspend dapper toshiba"
M40 and M40X are very similar to M45 (european models)
I suggest putting dapper in the search as Mepis is based Ubuntu dapper version

I read through the tutorial you sent me. Wow, I hope I'm not over my head on this one. (I work on this in a day or two; I have a big test coming up.) I searched the Ubuntu forum and found, what I think might be helpful. The writer subtly promises a fix for most laptop suspend/hibernate problems. However, since I'm a newbie, a green one at that, I wonder if someone might take a look at see what they think. I have a Toshiba M45-S169. Thanks.

Here's the link: http://www.linux.com/article.pl?sid=06/05/24/1716222.

I try the easy, simple things -- often a lot of them.

I'll admit it -- I haven't yet compiled a kernel, though I've been gluing the easy parts of Linux together for over 2 years now, and I'm an experienced programmer and dba.

I've heard it said that an expert has made all the possible mistakes ( hopefully only once for each mistake -- I'm not there yet for Linux ).

There are complex things which have important returns -- some folks like to fix the BIOS on their own (with DSDT) and so on.

Suspend and resume just don't happen to be useful for me right now -- apparently they're very useful for you.

You'll be likely to get additional help at http://www.mepislovers.com

Mike

Okay, I'm ready to make a go and fix my sleep mode. I feel like I need it because, it being a laptop, I'm always on the go, which, without the ability to sleep, means I'm constantly shutting down. I read somewhere that this is actually bad for linux (not sure why????).

Following the advice from the link above, I began with this instruction:

Getting sleep mode working was a bit of a chore. It turns out the "hibernate" package is obsolete (and didn't work too well); all the work is done by the ACPI apparatus (acpi, acpid, acpitool, and particularly acpi-support). To make ACPI work right, it turns out, you have to add "acpi_irq_balance" to the boot parameters listed in /boot/grub/menu.lst; be sure to add it to every line that contains "silent". I edited /etc/default/acpi-support, too, uncommenting "ACPI_SLEEP=true" and "ENABLE_LAPTOP_MODE=true". I had to rename /etc/acpi/lid.sh and then copy /etc/acpi/sleep.sh in its place so closing the lid would suspend it properly. The scripts use cardctl, from package pcmcia-cs rather than pccardctl from pcmciautils, the right tool for a 2.6.15 kernel -- not that it matters so much, since PCMCIA doesn't work anyway; however, suspend/resume doesn't work right, yet, without pcmcia-cs installed.

Problem, is however, when I go to /boot/grub/menu.lst, I don't have "silent" in any of my lines. I have this:
timeout 15
color cyan/blue white/blue
foreground ffffff
background 0639a1

gfxmenu /boot/grub/message

title Windows at hda1
rootnoverify (hd0,0)
chainloader +1

title MEPIS at hda7, kernel 2.6.15-26-386
root
kernel /boot/vmlinuz-2.6.15-26-386 root=/dev/hda7 nomce quiet vga=791
boot

title MEMTEST
kernel /boot/memtest86+.bin

Can someone please help? No joke, if I can get this sleep/suspend bit fixed, I'm sold on Mepis!

add after vga=791

so it will look like this

kernel /boot/vmlinuz-2.6.15-26-386 root=/dev/hda7 nomce quiet vga=791 acpi_irq_balance

You only need to change one line in this part -- not sure why the author makes it sound like there is more than one line to be changed

Mike

Not sure I followed the rest of it, tho

Here's what I got. I made the change, rebooted, and the tried to "suspend to disk." When I did, I got this message:

The resume partition is not set up. Probably you need to add a 'resume=...' option to your kernel command line and reboot. Suspend to disk and resume is not possible without a resume partition, please consult the documentation. You can skip this check by setting SUSPEND2DISK_SKIP_RESUME_CHECK to 'yes' in the sleep configuration file.

Now what?

Here's the rest of that article I sent you that concerns sleep and resume:

I edited /etc/default/acpi-support, too, uncommenting "ACPI_SLEEP=true" and "ENABLE_LAPTOP_MODE=true". I had to rename /etc/acpi/lid.sh and then copy /etc/acpi/sleep.sh in its place so closing the lid would suspend it properly. The scripts use cardctl, from package pcmcia-cs rather than pccardctl from pcmciautils, the right tool for a 2.6.15 kernel -- not that it matters so much, since PCMCIA doesn't work anyway; however, suspend/resume doesn't work right, yet, without pcmcia-cs installed.

After suspend/resume, unfortunately, the DPMS setting (turn off screen backlight) seems to be ignored, so I added "xset dpms 0 0 1800" at the end of the /etc/acpi/resume.sh script. The display driver forgets the correct contrast setting during suspend, too, so I also added "xvattr -a XV_CONTRAST -v 64". Experiment with gxvattr if you're not happy with that value.

I added ",noatime" to all the disk partition entries in /etc/fstab. The default settings in /etc/laptop-mode/laptop-mode.conf seemed OK but I turned on verbose mode, for now, so I can see what decisions it's making.

Also unfortunately, when it comes back after a suspend, the network drivers (sky2 for the ethernet, ipw2200 for the wireless) get mixed up. On boot they're eth0 for wireless and eth2 for wired; after resume, they're eth0 for wired and eth1 for wireless (or something like that). I copied
/usr/share/doc/ifupdown/examples/get-mac-address.sh
to /etc/network/ and made it executable. I changed the "mapping" section of /etc/network/interfaces to
mapping eth*
script /etc/network/get-mac-address.sh
map 00:a0:d1:xx:xx:xx lan
map 00:13:ce:xx:xx:xx wlan
(use "ifconfig -a" to find out your own MAC addresses) to tie those names to physical interfaces, and changed the "iface" stanza labels to "lan" and "wlan"; then I also edited /etc/iftab the same way (although I still don't know what reads this file), i.e.:
# Assign persistent names to network interfaces. See iftab(5).
lan mac 00:a0:d1:xx:xx:xx
wlan mac 00:13:ce:xx:xx:xx
Next, I added "lan" and "wlan" to INTERFACES in /etc/default/ifplugd, and just "wlan" in /etc/default/waproamd. (Note that the lan driver, sky2, sucks rocks at the moment, which is to say it fails at (attempted) high throughput rates. That might improve soon; in the meantime I turned it off entirely. The ipw2200 connection seems to max out at ~15 Mbps on a 54 Mbps channel; that might improve soon, too.)

Probably you'll need to figure it out or get help to figure it out -- just piece it out very slowly and experiment

Mike

I don't have these files: /etc/acpi/lid.sh and then copy /etc/acpi/sleep.sh, well, I don't have them where the writer says I will.

Boo! I'm not having too much luck with this suspend. HELP!

Does everything look right in this file?

## Path: System/Powermanagement/Powersave/Events
## Description: In this section you can define what actions to perform \
## if a powersaved event occurs. Read the documentation in \
## /usr/share/doc/packages/powersave/ for more informations on how to \
## add customized actions. You can specify multiple actions for a given \
## event, just separate them with whitespace.\n \
## Note: Your system might not generate all of the events.

## Path: System/Powermanagement/Powersave/Events
## Type: string
# Events executed at daemon start.
EVENT_DAEMON_START="ignore"

## Path: System/Powermanagement/Powersave/Events
## Type: string
# Events executed at a scheme change. The scheme change event is also thrown
# if the machine is e.g. unplugged from AC or the daemon is started.
EVENT_DAEMON_SCHEME_CHANGE="set_disk_settings"

## Path: System/Powermanagement/Powersave/Events
## Type: string
#
# What events should be processed if the
# powersave daemon terminates?
EVENT_DAEMON_TERMINATE="dethrottle"

## Path: System/Powermanagement/Powersave/Events
## Type: string
EVENT_ACADAPTER_ONLINE="reread_cpu_capabilities"
EVENT_ACADAPTER_OFFLINE="reread_cpu_capabilities"

## Path: System/Powermanagement/Powersave/Events
## Type: string(notify,wm_shutdown,wm_logout,standby,suspend_to_ram,suspend_disk,screen_saver,ignore)
#
# The specified values(actions) will take effect, when the
# battery state changes from "normal" to "warning".
EVENT_BATTERY_WARNING="notify"

## Path: System/Powermanagement/Powersave/Events
## Type: string(notify,wm_shutdown,wm_logout,standby,suspend_to_disk,suspend_to_ram,screen_saver,ignore)
#
# The specified values(actions) will take effect, when the
# battery state changes from "warning" to "low".
EVENT_BATTERY_LOW="notify"

## Path: System/Powermanagement/Powersave/Events
## Type: string(notify,wm_shutdown,wm_logout,standby,suspend_to_disk,suspend_to_ram,screen_saver,ignore)
#
# The specified values(actions) will take effect, when the
# battery state changes from "low" to "critical".
EVENT_BATTERY_CRITICAL="wm_shutdown"

# This event is never reached
EVENT_BATTERY_NORMAL="ignore"

## Path: System/Powermanagement/Powersave/Events
## Type: string(notify,wm_shutdown,wm_logout,standby,suspend_to_disk,suspend_to_ram,screen_saver,switch_vt,ignore)
#
# Button events. What functions of the proxy should be
# triggered if a ACPI button is pressed?
#
# Hint: switch_vt could be useful for the LID_OPEN event (replace ignore)
# some graphic cards might need switching to a virtual terminal
# and back to let the X-server recover properly
EVENT_BUTTON_POWER="wm_shutdown"
EVENT_BUTTON_SLEEP="suspend_to_disk"
EVENT_BUTTON_LID_OPEN="ignore"
EVENT_BUTTON_LID_CLOSED="screen_saver"

## Path: System/Powermanagement/Powersave/Events
## Type: string
#
# Gets processed if the powersave tool requests
# a CPU frequency policy change. (e.g. powersave --performance)
# These values are thought for later enhancements.
# You may want to get your own function invoked to enable advanced
# power saving functionalities.
# However it would be easier to modify the binary that should
# be invoked for the event in /etc/powersave.conf.
EVENT_PROCESSOR_PERFORMANCE="ignore"
EVENT_PROCESSOR_POWERSAVE="ignore"
EVENT_PROCESSOR_DYNAMIC="ignore"

## Path: System/Powermanagement/Powersave/Events
## Type: list(notify,wm_shutdown,wm_logout,standby,suspend_to_disk,suspend_to_ram,power_off,screen_saver)
#
# What should happen if a temperature limit is reached?
EVENT_TEMPERATURE_CRITICAL="notify"
EVENT_TEMPERATURE_HOT="notify"
EVENT_TEMPERATURE_PASSIVE="ignore"
EVENT_TEMPERATURE_ACTIVE="ignore"
EVENT_TEMPERATURE_OK="ignore"

## Path: System/Powermanagement/Powersave/Events
## Type: string
#
# This event gets thrown if a unknown event is encountered by the daemon.
# Unknown events include e.g. ACPI hotkeys on some machines.
# A description of this mechanism is available in
# /usr/share/doc/packages/powersave/README.events.
EVENT_OTHER="hotkey_handler"

## Path: System/Powermanagement/Powersave/Events
## Type: string
#
# The kernel may inform that changes on CPU frequencies or throttling
# states capabilites have been taken place.
# In this case a cpu_notify event is generated by the powersave daemon and
# the internal function reread_cpu_capabilities should be processed
EVENT_PROCESSOR_NOTIFY="reread_cpu_capabilities"

## Path: System/Powermanagement/Powersave/Events
## Type: string
#
# After the processor is idle (CPU_LOW) for a specific time
# see a processor.idle event is thrown.
# If the cpu usage goes beyond the CPU_HIGH value
# a processor.busy event is thrown at once.
# There will never be one event thrown twice in a row, they
# alternate! This is a perfect place to add additional
# power saving functionalities.
EVENT_PROCESSOR_BUSY="dethrottle"
EVENT_PROCESSOR_IDLE="throttle"

## Path: System/Powermanagement/Powersave/Events
## Type: string
#
# The specified functions that get processed
# when an suspend/standby is requested should not be modified.
# See System/Powermanagement/Powersave/Suspend or Standby
# for possible modifications if you encounter problems
# with these.
# You should change those only if you know what you are doing,
# if you remove e.g. the "screen_saver" event here, graphical
# clients like kpowersave will no longer lock the screen on
# suspend, no matter what. You have been warned
EVENT_GLOBAL_SUSPEND2DISK="prepare_suspend_to_disk screen_saver do_suspend_to_disk"
EVENT_GLOBAL_SUSPEND2RAM="prepare_suspend_to_ram screen_saver do_suspend_to_ram"
EVENT_GLOBAL_STANDBY="prepare_standby screen_saver do_standby"

EVENT_GLOBAL_RESUME_SUSPEND2DISK="restore_after_suspend_to_disk"
EVENT_GLOBAL_RESUME_SUSPEND2RAM="restore_after_suspend_to_ram"
EVENT_GLOBAL_RESUME_STANDBY="restore_after_standby"

## Path: System/Powermanagement/Powersave/Events
## Type: string
#
# This event gets executed on suspend to disk if a suspend2
# enabled kernel is found (/proc/suspend2 exists)
EVENT_GLOBAL_SUSPEND2DISK_OTHER="suspend2"

MajorCedar, the "resume=" thing is a reference to your swap partition. On my system that happens to be /dev/hda5. So my /boot/grub/menu.lst file has these lines:

title MEPIS at hda1, kernel 2.6.15-26-386
root (hd0,0)
kernel /boot/vmlinuz-2.6.15-26-386 root=/dev/hda1 resume=/dev/hda5 nomce quiet vga=791
boot

Not that I have been able to get suspend and all the other goodies working on my laptop. But that is not a big priority to me at the moment
But the suspend to disk does work on my box (sort of).

Jon

SimplyMEPIS Value CD: $17.95
Amount saved not buying Windows Vista Ultimate: $399
Experience gained learning Mepis and helping others: Priceless