Linux Hibernate!

[khurrum1990]

Does hibernate work for anyone? Why can't Linux developers solve this problem? Microsoft solved their ages back, even OS X running on top of freeBSD kernel can hiberate. What is Apple doing that we don't, I know OS X runs on fewer hardware, but there method of hibernate is different then ours and apparently it works.

[tageiru]

Does hibernate work for anyone?
Yes

Why can't Linux developers solve this problem?
http://www.osnews.com/story.php/17689/Bill-Gates-on-Making-ACPI-Not-Work-with-Linux/

Not the whole answer, but an interesting fact none the less.

even OS X running on top of freeBSD kernel

False

[NovaAesa]

Hibernate kind of works. When I turn on again, my internet doesn't work though which is a deal breaker for me.

I would have really liked to hibernate then boot into Windows (for when I feel the urge to do some serious gaming) then restart and unhibernate back into Ubuntu but alas it looks like I'm stuck with reboots.

[x0as]

Never worked for me, which is why I've kept xp my laptop.

[aimran]

IMO screw hibernate. Fast bootups ftw.

[khurrum1990]

First of all I don't get what Bill Gates could have possibly done thats stops us from developing a method of hibernate that works. Second if Linux doesn't improve this one little thing then even though people may install Linux on desktops where hibernate doesn't matter, but on laptops people r only going to use Windows.

[pelle.k]

Hibernate works for me. I think you phrased it badly, since hibernate probably fails on less computer than it works on. Nevertheless, it's a pain in the *** when it doesn't!
As usual, it's the industry ACPI standards and specifications that is to blame, for the most part. Often, bad drivers can screw it up too, but many linux drivers are reverse engineered, because the specs aren't available to the kernel developers.

microsoft has the advantage of being *the* testbed for most manufacturers, and so hibernate and suspend often work as a result of that.

[khurrum1990]

Quote:
Originally Posted by khurrum1990:
"even OS X running on top of freeBSD kernel"
False

What is false about this? Is it that OS X does not run on top of freeBSD kernel or does OS X does have problems with sleep?

[Mateo]

It's a major problem, almost a deal breaker for laptops.

[tageiru]

Quote:
Originally Posted by khurrum1990:
"even OS X running on top of freeBSD kernel"
False

What is false about this? Is it that OS X does not run on top of freeBSD kernel or does OS X does have problems with sleep?

OS X uses the Darwin kernel, not the FreeBSD kernel. You probably confused the kernel with userland as OS X uses the FreeBSD userland.

As for how Microsoft affects Linux hibernation it seems like many implementations of ACPI doesn't follow the the public specification. Whether that is because of lazy manufacturers or fowl play is not clear, but the mail from Gates suggests the latter.

[khurrum1990]

OS X uses the Darwin kernel, not the freebsd kernel. It uses the freebsd userland however.

Yeah u were right I kept forgetting, but still OS X has no problems with sleep even on Intel Macs. What r they doing right? Linux has hibernate troubles on Intel Macs as well.

[xeth_delta]

First of all I don't get what Bill Gates could have possibly done thats stops us from developing a method of hibernate that works. Second if Linux doesn't improve this one little thing then even though people may install Linux on desktops where hibernate doesn't matter, but on laptops people r only going to use Windows.

This is of course my personal choice, but I would not use windows even if hibernate did not work on my laptop.

That being said, Hibernate and software suspend both work on my laptop. It is Core 2 Duo Macbook, second generation. I run Ubuntu Feisty Fawn with a Gutsy Kernel.

These problems, for instance hibetnation are solved as time goes on. The Feisty kernel did not allow me to suspend or hibernate. I recompiled a newer version and later just used the gutsy kernel, in both cases it worked.

Most of the times these problems are associated with several drivers / modules not supporting suspend/hibernation properly, and in many cases work-arounds are possible until a permanent fix is done.

A way to identify the affected module is to disable them and then try to suspend. If that doesn't work, then start without the graphical interface and suspend from the comand line, eliminating one by one possible problematic modules. There can be some candidates such as modules for network / wireless cards, sound cards, modem.

It might help if we know what modules your system uses. Can you please post the specs of your machine? Please, also post, between code tags, the output of:


lsmod
lspci
lsusb
dmesg
cat /var/log/messages

Xeth

[khurrum1990]

Even though I really wasn't asking for help, but hey why not?
I am on a Macbook right now the machine I am talking about is:
A Gigabyte motherboard Intel P4 2.7ghz
1.5gb of ram
realtek ac'97 sound card
lan
lg dvd-rw
80gb hard disk
XFX Geforce 6200 512 mb AGP
I am running kernel 2.6.22-12 on that machine I think and not the latest 2.6.22-14 cause on that nothing works and there has been a bug report.

lsmod:
Module Size Used by
xt_limit 3584 8
ipt_LOG 7552 8
ipt_MASQUERADE 4608 0
ipt_TOS 3200 0
ipt_REJECT 5760 1
nf_conntrack_irc 8088 0
nf_conntrack_ftp 11136 0
xt_state 3456 6
xt_TCPMSS 5888 0
xt_tcpmss 3200 0
xt_tcpudp 4224 10
binfmt_misc 12936 1
rfcomm 42136 4
l2cap 26240 11 rfcomm
vboxdrv 60080 0
ppdev 10244 0
pppoe 15680 2
pppox 4872 1 pppoe
speedstep_lib 6404 0
cpufreq_userspace 5280 0
cpufreq_powersave 2688 0
cpufreq_conservative 8072 0
cpufreq_ondemand 9612 0
cpufreq_stats 7232 0
freq_table 5792 2 cpufreq_ondemand,cpufreq_stats
video 18060 0
sbs 19592 0
container 5504 0
button 8976 0
ac 6148 0
dock 10656 0
battery 11012 0
af_packet 24840 2
ppp_generic 29332 6 pppoe,pppox
slhc 7552 1 ppp_generic
lp 12580 0
snd_intel8x0 34972 1
snd_ac97_codec 100644 1 snd_intel8x0
ac97_bus 3200 1 snd_ac97_codec
snd_pcm_oss 44672 0
snd_mixer_oss 17664 1 snd_pcm_oss
snd_mpu401 9640 0
gspca 608336 0
snd_pcm 80388 3 snd_intel8x0,snd_ac97_codec,snd_pcm_oss
snd_seq_dummy 4740 0
snd_seq_oss 33152 0
snd_mpu401_uart 9600 1 snd_mpu401
videodev 29312 1 gspca
v4l2_common 18432 1 videodev
snd_seq_midi 9600 0
snd_seq_midi_event 8448 2 snd_seq_oss,snd_seq_midi
analog 13344 0
snd_rawmidi 25728 2 snd_mpu401_uart,snd_seq_midi
snd_seq 53232 6 snd_seq_dummy,snd_seq_oss,snd_seq_midi,snd_seq_mid i_event
hci_usb 18332 2
gameport 16776 1 analog
bluetooth 57060 7 rfcomm,l2cap,hci_usb
v4l1_compat 15364 1 videodev
parport_pc 37412 1
parport 37448 3 ppdev,lp,parport_pc
snd_timer 24324 2 snd_pcm,snd_seq
snd_seq_device 9228 5 snd_seq_dummy,snd_seq_oss,snd_seq_midi,snd_rawmidi ,snd_seq
nvidia 6221648 36
serio_raw 8068 0
pcspkr 4224 0
psmouse 39952 0
agpgart 35016 1 nvidia
snd 54660 14 snd_intel8x0,snd_ac97_codec,snd_pcm_oss,snd_mixer_ oss,snd_mpu401,snd_pcm,snd_seq_oss,snd_mpu401_uart ,snd_rawmidi,snd_seq,snd_timer,snd_seq_device
soundcore 8800 1 snd
snd_page_alloc 10888 2 snd_intel8x0,snd_pcm
i2c_sis96x 6404 0
i2c_core 26112 2 nvidia,i2c_sis96x
shpchp 34580 0
pci_hotplug 32704 1 shpchp
ipv6 274020 10
evdev 11136 4
iptable_nat 8708 0
nf_nat 20140 2 ipt_MASQUERADE,iptable_nat
nf_conntrack_ipv4 19724 8 iptable_nat
nf_conntrack 65288 7 ipt_MASQUERADE,nf_conntrack_irc,nf_conntrack_ftp,x t_state,iptable_nat,nf_nat,nf_conntrack_ipv4
nfnetlink 6936 3 nf_nat,nf_conntrack_ipv4,nf_conntrack
iptable_mangle 3840 0
iptable_filter 3968 1
ip_tables 13924 3 iptable_nat,iptable_mangle,iptable_filter
x_tables 16260 11 xt_limit,ipt_LOG,ipt_MASQUERADE,ipt_TOS,ipt_REJECT ,xt_state,xt_TCPMSS,xt_tcpmss,xt_tcpudp,iptable_na t,ip_tables
ext3 133768 2
jbd 60456 1 ext3
mbcache 9732 1 ext3
8139too 27776 0
sg 36764 0
sr_mod 17828 0
cdrom 37536 1 sr_mod
sd_mod 30336 4
floppy 60004 0
8139cp 25088 0
mii 6528 2 8139too,8139cp
ehci_hcd 36108 0
ohci_hcd 22916 0
usbcore 138248 5 gspca,hci_usb,ehci_hcd,ohci_hcd
pata_sis 15236 3
ata_generic 8452 0
libata 124528 2 pata_sis,ata_generic
scsi_mod 147084 4 sg,sr_mod,sd_mod,libata
thermal 14344 0
processor 31944 1 thermal
fan 5764 0
fuse 47124 1
apparmor 40600 0
commoncap 8320 1 apparmor

lspci:
00:00.0 Host bridge: Silicon Integrated Systems [SiS] 650/M650 Host (rev 80)
00:01.0 PCI bridge: Silicon Integrated Systems [SiS] Virtual PCI-to-PCI bridge (AGP)
00:02.0 ISA bridge: Silicon Integrated Systems [SiS] SiS962 [MuTIOL Media IO] (rev 25)
00:02.1 SMBus: Silicon Integrated Systems [SiS] SiS961/2 SMBus Controller
00:02.5 IDE interface: Silicon Integrated Systems [SiS] 5513 [IDE]
00:02.7 Multimedia audio controller: Silicon Integrated Systems [SiS] AC'97 Sound Controller (rev a0)
00:03.0 USB Controller: Silicon Integrated Systems [SiS] USB 1.0 Controller (rev 0f)
00:03.1 USB Controller: Silicon Integrated Systems [SiS] USB 1.0 Controller (rev 0f)
00:03.3 USB Controller: Silicon Integrated Systems [SiS] USB 2.0 Controller
00:0a.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139/8139C/8139C+ (rev 10)
01:00.0 VGA compatible controller: nVidia Corporation NV44A [GeForce 6200] (rev a1)

lsusb:
Bus 003 Device 001: ID 0000:0000
Bus 002 Device 002: ID 1131:1001 Integrated System Solution Corp. KY-BT100 Bluetooth Adapter
Bus 002 Device 001: ID 0000:0000
Bus 001 Device 003: ID 093a:2471 Pixart Imaging, Inc.
Bus 001 Device 001: ID 0000:0000

Dmesg is messed up cause that computer just started firestarter.

[xeth_delta]

Ok, I will have more time to look at it tomorrow evening, but I guess it might be possible to get it working after some tinkering.

Of course, the usual recommendation of making a back-up of important information, just in case.

As a starting point you could try starting without the graphical interface, in the most basic state possible. There should be less modules loaded, which you can check with lsmod.

Try to suspend from the command line:

sudo echo 3 > /proc/acpi/sleep
or

sudo echo -n mem > /sys/power/state

Try to wake up (keys, bower button, etc), if it does not you can start unload modules with rmmod, for instance let's say you want to unload nvidia


sudo rmmod nvidia

Try again. This might be laborious, but in the end, you could end up identifying the problematic module.

Have a look at this thread, might be useful, even though it concerns a thinkpad: http://ubuntuforums.org/showthread.php?t=200082&highlight=echo+%2Fproc%2F+suspend

If you want to try suspend to disk, too, I think this is what you have to do:

echo 4 > /proc/acpi/sleep
http://ubuntuforums.org/showthread.php?t=65457&highlight=echo+%2Fproc%2F+suspend

After a failed suspend - wake up, information should be written to /var/log/messages. Can you please post the output, but please, do use the code tags ( # button in the posting page ), it makes it easier to read.

Xeth

[khurrum1990]

Hey wow thanks, I will try this as soon as my family leaves the computer alone.

[Sporkman]

fowl play

Don't duck the issue.

[23meg]

Does hibernate work for anyone? Why can't Linux developers solve this problem?

ACPI, lack of drivers, lack of specifications.


What is Apple doing that we don't,

Manufacturing their own hardware that they can and do test extensively.

[Palmyra]

Nope, doesn't work for me. Neither suspend or hibernate work for me, but I recall a time they did (probably with Dapper). But then again, Oceana was, at that time, still at war with Eurasia. I am not quite sure if it is now at war with Eurasia or Eustasia.

[thisllub]

I modify GRUB to add ACPI=off
Then my laptop boots.
After it boots I find the hald storage polling process and kill it.
That stops the 20 second hang every few minutes.

[beercz]

I have an HP Compaq nx7400 laptop running gutsy.

I have got hibernate to work well (very useful for me).

I have not got suspend to work though (suspend not that useful for me so I haven't really bothered)

When I was running feisty on the same machine I had both suspend and hibernate working, using uswsp (or whatever it was - can't quite remember) - which I think has been removed in gutsy.

Really could do with better power management though.

[ssam]

when linux devs have good access to hardware info they can do amazing things. look up the power management systems on the OLPC

[xeth_delta]

when linux devs have good access to hardware info they can do amazing things. look up the power management systems on the OLPC

Really good point ssam