http://brainstorm.ubuntu.com/item/94/
[7253 votes] Solution #1: With proprietary drivers


[23 votes] Solution #3: Mix of improvements + Solution #1 (letting the user choose)


[-4 votes] Solution #4: Add an idle option


]]> en-us Thu, 28 Feb 2008 17:22:21 +0000 Fri, 23 Oct 2009 06:34:47 +0000 QAPoll module http://brainstorm.ubuntu.com/idea/94/
PS. it breaks quite a bit...]]> Thu, 28 Feb 2008 19:16:00 +0000 Thu, 28 Feb 2008 19:25:27 +0000 Thu, 28 Feb 2008 20:02:38 +0000

GNU/Linux Rules........!!!!]]> Thu, 28 Feb 2008 20:43:28 +0000 Fri, 29 Feb 2008 00:03:08 +0000 The number of Laptops is growing and will continue to grow so good working stand by and hibernate are really needed.
]]> Fri, 29 Feb 2008 00:09:23 +0000 Fri, 29 Feb 2008 00:22:45 +0000 Fri, 29 Feb 2008 00:25:02 +0000 http://ubuntuforums.org/showpost.php?p=4346978&postcount=7

I'm not sure why that isn't the default though. ]]> Fri, 29 Feb 2008 00:35:46 +0000 Fri, 29 Feb 2008 00:37:37 +0000
This is a high-value subject!]]> Fri, 29 Feb 2008 00:44:29 +0000
I have way more problems than with suspend and hibernate.

1. Suspend usually works, but not always. Often the mouse won't move after coming out of suspend.
2. Every time I've tried hibernate, my computer doesn't come back to life.
3. When I connect to a VGA (external monitor), horrible things happen. I have to change resolutions manually (something that ought to be done automatically, really). I don't know if this is a television problem, but when I connect VGA to my television, everything but the eject button stops working until it's unplugged and plugged in again (probably is an Insignia television issue).]]> Fri, 29 Feb 2008 00:44:46 +0000 Fri, 29 Feb 2008 01:11:02 +0000
This could be fixed by switching to suspend 2, when I was using this on my debian box, suspend and wakeup was an order of magnitude faster.]]> Fri, 29 Feb 2008 01:13:51 +0000
Can anyone point to an effort to improve suspend/hibernate? (or at least improve what hardware is supported?)]]> Fri, 29 Feb 2008 01:18:43 +0000 Fri, 29 Feb 2008 01:24:58 +0000 Fri, 29 Feb 2008 02:06:24 +0000 This is a must have just to be in the race.]]> Fri, 29 Feb 2008 02:37:31 +0000 Fri, 29 Feb 2008 02:37:44 +0000
Suspend/Hibernate is a fundamental feature that should be supported where at all possible. With this working, there is also much less pressure to improve boot times. ]]> Fri, 29 Feb 2008 02:57:07 +0000 In that case, I've no choice but to press the power button to force reboot.]]> Fri, 29 Feb 2008 03:01:29 +0000
Suspend/resume worked with Fiesty, but is a crap shoot on Gutsy. On my machine every thing works on resume but the usb-keyboard until I restart the x server using my usb-mouse. My machine is a Dell XPS 410 with proprietary nvida drivers. ]]> Fri, 29 Feb 2008 03:14:00 +0000 Fri, 29 Feb 2008 03:44:28 +0000
The Ubuntu responses to the endless posts about this issue on the bug-tracking system are not acceptable. This issue is a showstopper and must be fixed, no matter how painful. If the distro won't run on mainstream, business hardware used by millions around the world - well, it's dead at the starting gate.

Please - speak for us end-users and get in the face of ATI (now AMD) and others to get this fixed.

Many, many thanks for creating a truly impressive distro.]]> Fri, 29 Feb 2008 04:44:35 +0000 Fri, 29 Feb 2008 05:09:48 +0000
It failed to work at all pre-Feisty, worked with additional hacks in Feisty, and is broken again in Gutsy.


Dell e1505 w/ATI x1400]]> Fri, 29 Feb 2008 05:12:09 +0000
This is the fault of INTEL for refusing to open the BIOS or to provide sufficient documentation for the chipset and processor.

The reason that Intel doesn't open up the BIOS is that it has agreed with Phoenix Technologies and American Megatrends to protect their markets from competition from open source.

I know this because I own a company that manufacturers computers. I tried to get from Intel technical documentation to solve this problem. Intel's engineers candidly told me that they were prohibited from releasing the information because the company had agreed to protect Phoenix.

Happy Trails,

Loye Young
Isaac & Young Computer Company
Laredo, Texas
http://www.iycc.biz]]> Fri, 29 Feb 2008 05:44:11 +0000
1. sudo gedit /boot/grub/menu.lst
2. find the grub menu line that is the ubuntu default .. the first one that appears, that will look like this:
title Ubuntu 7.10, kernel 2.6.22-14-generic
root (hd0,0)
kernel ....... etc

3. Add the following to the end of the "kernel" line:

acpi_sleep s3_bios

Here is my "kernel" line.
kernel /boot/vmlinuz-2.6.22-14-generic root=UUID=b6e33b4a-d2ec-4107-9aaa-8956ed669d82 ro quiet splash acpi_sleep s3_bios

4. reboot computer

Now resume works every time.
]]> Fri, 29 Feb 2008 06:21:51 +0000 Fri, 29 Feb 2008 06:42:16 +0000 Fri, 29 Feb 2008 06:47:55 +0000 Fri, 29 Feb 2008 07:44:15 +0000 i don't know any notebook user who does not have problems with this in ubuntu]]> Fri, 29 Feb 2008 07:55:56 +0000 Fri, 29 Feb 2008 08:25:24 +0000 For this reason I don't think it is as big an issue as others that are out there.]]> Fri, 29 Feb 2008 08:34:37 +0000 Fri, 29 Feb 2008 08:54:03 +0000
However, even with hibernate problems, I'm running Ubuntu on fglrx with compiz-fusion full as it gives me an awesome user experience. Guess that kills off my battery life, so I'm stuck to a/c power.]]> Fri, 29 Feb 2008 09:03:11 +0000
It has nothing to do with proprietary drivers for video cards: because I have this problem on two systems with Intel video cards, which run on the usual open source drivers.

Please fix it.]]> Fri, 29 Feb 2008 09:29:33 +0000 Fri, 29 Feb 2008 09:34:17 +0000 Fri, 29 Feb 2008 09:43:19 +0000 Fri, 29 Feb 2008 09:45:33 +0000
Definitely needs to be fixed]]> Fri, 29 Feb 2008 09:56:55 +0000 Fri, 29 Feb 2008 10:00:57 +0000
Regards,
Johan Louwers.]]> Fri, 29 Feb 2008 10:12:33 +0000
but unfortunately I'm sure that it is not only kernel or ubuntu problem. I have an ATI X1300 Mobility card in my HP nc6400 and it doesn't work at all. ATI shoud commit it's patches as well.

HP 6510b and 6710b (both with intel x3100) suspend&hibernation doesn't work neither :(

Pleeeeeeeeeeeeeeeeeeease solve this issue :)]]> Fri, 29 Feb 2008 10:48:41 +0000
Maybe we should start a team that processes related bugs and tries to test things on a massive schale :)]]> Fri, 29 Feb 2008 10:48:42 +0000 Fri, 29 Feb 2008 12:31:06 +0000 Fri, 29 Feb 2008 12:59:48 +0000
why is this happening? why is it related to video drivers? honestly, i don't care if a proprietary driver is necessary or not - this just needs to be solved.]]> Fri, 29 Feb 2008 13:01:09 +0000
(But I still love Ubuntu! Thanks.)]]> Fri, 29 Feb 2008 13:02:59 +0000 Fri, 29 Feb 2008 13:03:34 +0000
My first experience with Linux a few years ago was trying to install it on an IBM T20, which I found out in the course of my frustration and research, was an exception and did not play nice with Linux at all on the issue of suspend and resume while being one of the most popular laptops of its day. It seemed to me at the time that there was something fundamentally wrong with Linux because even Windows 98 could suspend a T20. This left me with a negative impression that has lasted for some time even though I am a Linux fan and dual boot every machine I own. Since then I've learned more but nearly every computer I've tried to use it on has had an issue with it not working without having to address some issue. To use a MS phrase, it leaves a bad "OOBE" (out of box experience). The failure of Linux to do something fundamental like that produces its own "FUD" due to the logic that "there's never just one cockroach" ]]> Fri, 29 Feb 2008 13:35:44 +0000 Fri, 29 Feb 2008 13:52:55 +0000 Fri, 29 Feb 2008 13:58:56 +0000 Fri, 29 Feb 2008 14:12:59 +0000
And an important step is to make the hibernate and suspend thing impec!]]> Fri, 29 Feb 2008 14:31:50 +0000
At least adding it to the repos would be a good place to start]]> Fri, 29 Feb 2008 14:47:48 +0000
At least adding it to the repos would be a good place to start]]> Fri, 29 Feb 2008 14:47:49 +0000
I just realize this issue is about nvidia (restrict) driver. After active this driver hibernate does not works.

What happens:
- When pressed hibernate button the system is successful hibernated, but... when resuming the screen got stuck on a "white" panel filling all the screen. After a few seconds it show some "blurs" on the screen and the system freezes.
]]> Fri, 29 Feb 2008 15:11:25 +0000 Thanks for your comment. The Ubuntu environment will then for sure help us saving the _real_ environment.
And no, I'm not an activist. At all.]]> Fri, 29 Feb 2008 15:20:57 +0000
One thing, though. In this year fiddling around I found out that to fix or at least improve the problems with closed drivers it was sufficient to add some lines to the acpi, saying to leave some modules in the kernel or stop some services. My question then becomes, why when I install one of these drivers does the system NOT take care of this work which I have to do manually? I do not think it would be that hard to add some minor script so that restricted manager takes care of this, using the huge feedback already existing to create the base database of necessary hacks to the acpi-support (or any other file, I am not knowledgeable enough to tell where else one must act).

It seems to me that Ubuntu has mainly targeted the Desktop pcs. There Ubuntu rocks (ever put a desktop pc into suspend? yeah, me neither). But as laptops are continuously spreading in the user-base, IMHO Ubuntu should focus more on those to be a runner in the mid to long term.

Just my 2¢ :)]]> Fri, 29 Feb 2008 15:32:48 +0000
I had tried so hard to find solutions to fix this only got to nowhere.

And this Suspend and Hibernate have never been stable in term of one day I can get it working; usually after system was setup the way I wanted it, then failed the next day. So sometimes I would decide to reinstall the OS again only to lose all the settings and software I would have installed.

I am using a HP DV9xxx laptop. And what I have found so far, is that when I need to use special parameters at the boot time, like apm=off noapic and etc, then this function usually fail.

At one time, I found i386 edition can be loaded without using any of these parameters and the AMD64 one would always required.

But in summary, please have this function to be fixed perhaps in the near release.

Thanks!]]> Fri, 29 Feb 2008 15:54:04 +0000 http://brainstorm.ubuntu.com/idea/1544/

It suggests to include TuxOnIce into Ubuntu.]]> Fri, 29 Feb 2008 16:16:54 +0000 Fri, 29 Feb 2008 16:39:47 +0000
Jeremy]]> Fri, 29 Feb 2008 16:50:43 +0000 suspend is broken! Using Open source driver ati.

This my only headache with K(u)buntu!

Please fix this.

Thanks a lot

konrad]]> Fri, 29 Feb 2008 17:07:21 +0000 Fri, 29 Feb 2008 17:35:06 +0000 Fri, 29 Feb 2008 17:42:16 +0000 Suspend which I use(d) daily worked perfectly on Feisty with my Lenovo v100. This was the one thing that Gutsy broke on my system. Fixing for Hardy is an absolute priority.]]> Fri, 29 Feb 2008 17:44:54 +0000
]]> Fri, 29 Feb 2008 18:13:00 +0000 Fri, 29 Feb 2008 18:28:37 +0000 Fri, 29 Feb 2008 19:07:47 +0000 Fri, 29 Feb 2008 19:09:20 +0000 Fri, 29 Feb 2008 19:12:59 +0000 Fri, 29 Feb 2008 19:19:07 +0000 Fri, 29 Feb 2008 19:36:53 +0000 Fri, 29 Feb 2008 20:45:01 +0000 Fri, 29 Feb 2008 20:57:52 +0000
What would really be required to "fix suspend and hibernate" are three things. First of all, the rare breed: C wizards that are able to work with Linux kernel. Second, hardware donations (motherboards and/or complete systems) to the developers. This could however be substituted with giving SSH access to your computer - major trust issue but sometimes doable. Third, support from the vendors. (docs etc)]]> Fri, 29 Feb 2008 21:17:47 +0000 Fri, 29 Feb 2008 22:06:04 +0000 Fri, 29 Feb 2008 22:22:22 +0000
For a user new to Linux, hibernating bringing up a screen full of techo-babble is a Bad Thing, and they may never hibernate again.

With that said, the core issue is making sure hibernate works at all. Making it user-friendly is just icing on the cake.]]> Fri, 29 Feb 2008 22:57:21 +0000 Sat, 01 Mar 2008 01:51:08 +0000 Sat, 01 Mar 2008 01:56:13 +0000
* Promote a standard test procedure and related tools to probe the hardware and try out the different methods of suspending.

* Develop a reporting procedure where people can report the behaviour of their systems to the different tests

* Perhaps a suspend-test package that will automate some of this, and try out the suspend options, sending back the debugging information to a database.

The test/debug tools package would probably need to change regularly (and require re-testing) as developers gain new ideas about how to fix specific computers, and what different things to probe.

Basically, rather than (or in addition to) developers getting donation PCs or SSH access, if the testing procedure is automated, they can do the tests in parallel on thousands of PCs at once, and all developers can benefit from the tests others have made.
]]> Sat, 01 Mar 2008 02:06:51 +0000
Also annoying, I tell power manager to leave the PC on when plugged in but it goes to sleep anyway.

]]> Sat, 01 Mar 2008 02:42:34 +0000
I've never gotten hibernate to work.]]> Sat, 01 Mar 2008 05:15:38 +0000
My EEE PC works great!

My older Dell that started with Windows XP doesn't hibernate, and doesn't wake up from suspend in 7.10.

Here's to supporting even more hardware in our quest to stomp out bug #1.]]> Sat, 01 Mar 2008 06:28:51 +0000 Sat, 01 Mar 2008 07:30:39 +0000 Sat, 01 Mar 2008 07:45:21 +0000 Sat, 01 Mar 2008 07:55:57 +0000 Sat, 01 Mar 2008 09:43:04 +0000 Sat, 01 Mar 2008 09:52:50 +0000 Sat, 01 Mar 2008 10:04:09 +0000 Sat, 01 Mar 2008 11:43:20 +0000
[URL=http://brainstorm.ubuntu.com/idea/2065/][IMG]http://brainstorm.ubuntu.com/idea/2065/image/2/[/IMG][/URL]]]> Sat, 01 Mar 2008 12:23:59 +0000 Sat, 01 Mar 2008 15:53:45 +0000
For me, when using the opensource ATi driver, suspend works like a charm. Using the closed source driver, it doesn't work at all. I bet the same goes for 90% of the people talking about this 'bug'. The issue here isn't suspend resume per-say, but rather the closed source drivers. Get those fixed, and it's all fixed.]]> Sat, 01 Mar 2008 16:47:06 +0000
Testing, testing, testing.

It needs to be made easier to test, report, and diagnose suspend / hibernate problems.]]> Sat, 01 Mar 2008 17:17:48 +0000 Sat, 01 Mar 2008 17:50:28 +0000 I can tell suspend2 is light-years fastest than default suspend and hibernate shipped with ubuntu...

Why don't you think to include that in future releases?

i think suspend and hibernate should be also quick in going and coming back, default system takes more time than a normal reboot... otherwise they're kinda useless.]]> Sat, 01 Mar 2008 19:56:32 +0000 world ecologically talking ...

P.D: Sorry about my english :-(]]> Sat, 01 Mar 2008 22:05:30 +0000
http://brainstorm.ubuntu.com/idea/2367/]]> Sun, 02 Mar 2008 02:11:06 +0000 Sun, 02 Mar 2008 02:50:19 +0000
... it defends a perfect suspend/resume using open source kernel code, that works everytime, even if the user is using some proprietary drivers (like nvidia) that have problems sleeping nowadays.

It might be just me, but it seems a lot of people have read this entry wrong.]]> Sun, 02 Mar 2008 04:40:25 +0000 I don't think it should be considered a Ubuntu problem more of an general Linux problem.

I guess Ubuntu developers can do something about it to a certain extend. What I see as the complex issue is not only what should be done to fix, but how much time Ubuntu should spend on it.

I am not sure how Ubuntu and other distribution / kernel people keep record of what is broken / what works when it come to Suspend.
What I can see is Ubuntu as a big active community, if there was a big Suspend issue database or the-like that would open up, I am sure quite a lot of people to help to provide useful information.
Well it's just one idea... I don't know what the developer need the more for fixing those suspend issue.]]> Sun, 02 Mar 2008 07:44:21 +0000 Sun, 02 Mar 2008 09:25:24 +0000
There is also Suspend2, tuxonice, and other suspend methods. However, currently there isn't much current documentation about what is available, and the differences between those.

A package could be developed for people willing to fiddle more that simply click "hibernate" and observe it does/doesn't work. - instead trying the different options, with/without X, with different X driver etc..
]]> Sun, 02 Mar 2008 13:31:17 +0000 Sun, 02 Mar 2008 13:32:53 +0000
It should me much more easy to implement scripts running when suspending or hibernating, and waking up. This is a problem for several users.
See this german thread
http://forum.ubuntuusers.de/topic/153082/?p=1258820#1258820

thanks for fixing it, showing how to to it or making it more easy.

Its very important, for example to unmount truecrypt containers, or restarting a special daemon]]> Sun, 02 Mar 2008 16:25:01 +0000
Suspend/standy works when I select it manually from the menu. Even though klaptopdmanager/daemon is set up to suspend when closing the lid, this only works occasionally. Most of the times, it just starts the screensaver.

Hibernation works, but takes extremely long when comparing eg to Windows.

In both cases, when coming back up, it happens quite often that I see my desktop and then it goes into suspend mode again, so I have to start it one more time before I can actually use the laptop.

No big issues but it would be nice if this could be better.]]> Sun, 02 Mar 2008 17:40:50 +0000 Sun, 02 Mar 2008 17:49:48 +0000 Sun, 02 Mar 2008 21:04:15 +0000
00:02.0 VGA compatible controller: Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (rev 0c)
00:02.1 Display controller: Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (rev 0c)

xserver-xorg-video-intel drivers

2.6.24-10-generic (ubuntu 8.04)

and I have problems suspending to ram, mainly the major problem is that the screen goes blank and it doesn't come back. I wish there was a per laptop model wiki because of the way each one has different settings for propper suspend.]]> Sun, 02 Mar 2008 21:05:31 +0000 Sun, 02 Mar 2008 22:39:05 +0000 Sun, 02 Mar 2008 23:21:47 +0000
Suspend/Hibernate are not only important for laptop users.

But also people who want to be enviromentally friendly, but not wait for their computer to boot each time they want to use it.]]> Mon, 03 Mar 2008 00:02:42 +0000
So what needs to happen? The Linux community needs to forward a hardware standard for which Linux is guaranteed to work. Make a "Guaranteed to work with Linux" logo and certification process. This is bigger than Ubu/Canonical and probably needs to be addressed by the Linux Foundation and broader community. Unfortunately, the Linux Foundation has limited pockets and at least one of their major employees (rhymes with highness) appears not to use a laptop (or at least has stated he's only really interested in desktop hibernate).

Short of that, the only bandaid solution is to have laptop-model specific packages that are maintained by the community. Installing these packages would do a "best practices" tweaking of the system.]]> Mon, 03 Mar 2008 00:17:49 +0000 Mon, 03 Mar 2008 02:06:01 +0000
uswsusp
suspend2-userui (just userui???)
hibernate

Maybe some of these may be able to help some situations. Apparently there are some kernel patches to be done to use tuxonice (suspend2) - but I haven't seen how this is done on Ubuntu yet.
]]> Mon, 03 Mar 2008 02:08:25 +0000 However gutsy would not work on lappy. Hardy does I'm pleased to say :).

It's a Acer Aspire 5633 WLMi.]]> Mon, 03 Mar 2008 02:28:20 +0000 Mon, 03 Mar 2008 02:30:28 +0000
At the very least, if it doesn't reliably work, they should be removed from the menu.]]> Mon, 03 Mar 2008 05:49:23 +0000
I'm using a Lenovo R60 with Gutsy. It doesn't work.]]> Mon, 03 Mar 2008 08:25:08 +0000
In feisty on my laptop, a Gateway MT3707, wireless and suspend and hibernate worked out of box. There was sound as well.

Now, in gutsy, wireless causes a kernel panic upon connect (so I'm forced to use ndiswrapper), sound doesn't work and I needed to compile Alsa from source and apply a patch that isn't going to go upstream, and suspend and hibernate doesn't work.

Interestingly enough, every time I fix sound, after every kernel update they push, I have to recompile alsa.

I really just want suspend and hibernate to work. The other problems I can deal with, but this is ridiculous. I'm actually thinking of switching to another distro. ]]> Mon, 03 Mar 2008 12:49:01 +0000 ]]> Mon, 03 Mar 2008 13:21:49 +0000
Focus on speeding up booting and shut down times. Remove hibernate and suspend buttons for default.

Hibernation takes ages to be done when you have a lot of RAM anyways. And suspend does not really save energy.

If booting and shut down times could be halved we wouldn't even need these things.

If you really want to keep hibernate and suspend, improve them but please, avoid proprietary drivers, I don't want to see people's freedom reduced for such unnecessary features.]]> Mon, 03 Mar 2008 16:33:07 +0000 Improving hibernate should be a priority.

I see a big need for a documentation-package for suspend and hibernate.]]> Mon, 03 Mar 2008 18:16:07 +0000
If suspend and hibernate were working, then we wouldn't need to worry so much about boot times.

The idea is not talking about using proprietary drivers to do it. It is talking about making it work even when people are using proprietary video drivers. If anything, if you wanted to remove proprietary driver dependency, you would be better off encouraging suspend - because it is one of the things that is driving people to the open source video drivers. ]]> Mon, 03 Mar 2008 21:43:37 +0000 Mon, 03 Mar 2008 22:14:53 +0000
1. It is very complex to get working because it depends on every peripheral driver on the system behaving correctly. And sometimes there are just buggy drivers.
2. The single biggest culprit preventing suspend/hibernate is the video driver.
3. With binary drivers (mainly ATI and Nvidia), we do not have _any_ control over making these drivers behave - so users are left to choose between better 3D support or hibernate.
4. Suspend2 suffers from lack of support upstream (Linus') tree. Coupled with the fact that the patch is very intrusive, it makes maintenance of the patch harder for Ubuntu.
5. As has been shown recently on LKML and various talks [1], both the swsusp and suspend2 (tuxonice) suffer from significant design drawbacks, so that they may never work correctly. There is new work ongoing on using a kexec-based approach that is being currently seen as the successor to the current design.
6. Using open drivers has a better chance of success. If that doesn't work, I urge you to file bugs on launchpad.net so that we might try to fix those.
7. References [2] and [3] are pointers to further debugging your suspend/hibernate issues. Help us in fixing your machines.

Regards,
Amit

[1] http://linux.conf.au/programme/detail?TalkID=139
[2] https://wiki.ubuntu.com/DebuggingKernelSuspend
[3] https://wiki.ubuntu.com/UnderstandingSuspend]]> Mon, 03 Mar 2008 22:34:09 +0000 Tue, 04 Mar 2008 16:26:28 +0000 Tue, 04 Mar 2008 22:15:55 +0000
I am willing to hear suggestions on how we can get proprietary drivers vendors to co-operate with us. Most of the Linux and *BSD communities would be interested in finding a solution to this problem.

One solution that is encouraged by these communities is to ask users to vote with their wallets - buy hardware that ships with open drivers (or atleast open specifications). Most Intel graphics chips[1] and many ATI ones[2][3] now have open drivers or are working towards open drivers. So the expectation is that these problems will go away, but it will take some time.

In the meanwhile, if there is something that works on another Linux distros but not on Ubuntu, we want to hear about it.

Regards,
Amit

[1] intellinuxgraphics.org
[2] http://www.desktoplinux.com/news/NS2568321546.html
[3] http://airlied.livejournal.com/50613.html]]> Wed, 05 Mar 2008 00:31:09 +0000 Wed, 05 Mar 2008 09:59:13 +0000
Hardware:
Samsung P35 notebook, ATI Radeon 9700 Mobility (r300) with free drivers, ipw2200 WLAN

Before:
Gutsy, s2ram (suspend) works like a charm (needs only vbe restore)

Now:
Hardy, freeze and black screen during resume

I have to go deep into the suspend/resume process and debug it.
https://wiki.ubuntu.com/DebuggingGNOMEPowerManager will hopefully help ....

*suggestion*:
Wouldn't it be possible to help the user migrating a working setup to the new pm-tools based infrastructure? Like asking
- "Does suspend/resume work for you?" Y/n
- "Shall I try to migrate the old settings?" Y/n
Additionaly after a resume the user is asked, if the resume worked as expected. In case of a failure either the default settings may be loaded or some probing script may be started to evaluate a working setup AND publish the results.


]]> Wed, 05 Mar 2008 12:56:32 +0000 Wed, 05 Mar 2008 13:23:13 +0000
http://support.microsoft.com/search/default.aspx?query=hibernation&catalog=LCID%3D1033&spid=&mode=r&lsc=0]]> Wed, 05 Mar 2008 16:47:50 +0000 Thu, 06 Mar 2008 02:09:13 +0000 Thu, 06 Mar 2008 04:51:17 +0000 Thu, 06 Mar 2008 23:16:09 +0000 full AK

this comes down to ACPI support, better broken (from vendor side) ACPI support]]> Thu, 06 Mar 2008 23:19:00 +0000 Thu, 06 Mar 2008 23:20:19 +0000 Thu, 06 Mar 2008 23:20:49 +0000
From a usability perspective, this is better than hibernate. Current huge amounts of RAM (up to 4 GB) take almost as much time to write to disk then plain old rebooting.

The only 'drawback' of Macbook sleep being that you can't swap batteries, because it would purge RAM. But why not find an adapter instead? I have heard reports of Macbooks that can be in the sleep state for days and still be used. ]]> Fri, 07 Mar 2008 15:03:42 +0000
S3 Suspend is good on desktops too, where the current draw is typically very little more than the "off" leakage current.

S3 Suspend is theoretically implemented in Ubuntu, but the problem is there is just so much hardware that doesn't work properly with it, and there isn't much coherent documentation about how to get it going when it doesn't "just work" .. hence the demand here for it to be fixed.]]> Sat, 08 Mar 2008 22:43:36 +0000 Mon, 10 Mar 2008 18:20:06 +0000
http://www.ubuntugeek.com/fix-for-suspend-and-hibernation-problem-for-laptops.html]]> Tue, 11 Mar 2008 01:17:04 +0000
On the other hand, for me the following works on my Dell D400,

sudo apt-get install pm-utils

now the pm-suspend does s3 suspend correctly, and it seems to also work through the graphical option on the shutdown menu
]]> Wed, 12 Mar 2008 03:55:06 +0000 1. Ask the user "Is this a Laptop?" during installation.
2. If yes as user "What brand laptop is it?"
3. Taylor installation of APM to fit the laptop
4. If it is an unrecognized brand inform the user that suspend/hibernate may have problems]]> Wed, 12 Mar 2008 15:18:41 +0000
For each method, it's also an issue of knowing what is the right way to do it for each model. That is the hard part. Once that's known, then the model that the user has can pretty much be auto-detected.

I think what really needs better support are the users with some technical know-how to do some level of debugging. That's less than 1% of users, who can then do directed debugging on their own hardware to better gather and incorporate that information. The results of that would then flow on to all users. The problem for that 1% is that the documentation is rather haphazard, and much is out of date. It would be good if there were an official ubuntu wiki that kept up to date with how things should be done for those users prepared to do a bit of debugging and information-gathering.
]]> Wed, 12 Mar 2008 22:43:34 +0000
http://video.google.com/videoplay?docid=7625509273628599358]]> Fri, 14 Mar 2008 02:33:21 +0000
But then neither uswsusp or pm-utils packages are installed by default it Gutsy ... why's that?]]> Fri, 14 Mar 2008 21:58:04 +0000 Sat, 15 Mar 2008 20:16:41 +0000 https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/164974]]> Mon, 17 Mar 2008 06:15:57 +0000 Mon, 17 Mar 2008 19:31:36 +0000
@jblackhall -- my sister's T61 has the X3100 chipset and both hibernate and suspend work beautifully, although there was one issue with the X3100. See: http://www.thinkwiki.org/wiki/Installing_Ubuntu_7.10_(Gutsy_Gibbon)_on_a_ThinkPad_T61#Hibernate.2FSuspend (I believe that it's a bug in the X3100 driver, but at least the issue is being tracked)]]> Tue, 18 Mar 2008 06:34:27 +0000
http://cryopid.berlios.de/]]> Wed, 19 Mar 2008 00:02:30 +0000
Use open source drivers (with minimal/bare 3D accel, think old ATi cards, intel cards etc) and have suspend/hibernate working pretty much fine, but expect slow 3D performance (on my 9600 mobile the open source drivers work well enough for wobbly windows etc, but obviously can't do heavy 3D). Or use closed source drivers, but don't expect suspend/hibernate to work. It's a choice atm you'll have to make.

You cannot blame ubuntu for this, it's driver makers fault. Some are doing there best (intel for example) some are changing their old ways (ATi). And as mentioned before, even windows has plenty of problems with suspend/hibernate. Though I absolutely agree, Suspend/hibernate is very important to us all.]]> Wed, 19 Mar 2008 00:13:06 +0000
"Its not Ubuntus fault..."

So what? It still must be fixed!]]> Wed, 19 Mar 2008 02:52:37 +0000
But if it does not work correct, Ubuntu needs to care about it. Otherwise users will run away from Ubuntu. I don't know where exactly the problem is, but maybe the Ubuntu developers need to collaborate more with Novell and the radeonhd developers (radeonhd is sponsored by Novell, afaik)]]> Sat, 22 Mar 2008 19:03:06 +0000 If suspend and hibernate were working, then we wouldn't need to worry so much about boot times. "

But then, if (re)boot was made faster then we wouldn't need to worry about suspend and hibernate either.

From the devs' point of view it looks like working on speeding those up is more viable than hoping for everyone to use hardware with open source 3d graphics cards.

As of your clarification about proprietary drivers, I stand corrected.]]> Sun, 23 Mar 2008 22:19:17 +0000
And I want a pony.

"It is required to work out of the box like it does in any modern respectable OS."

You mean, like in no OS? Even windows doesn't offer that out of the box and has issues.

"I used ubuntu with proprietary Graphic driver because these are the one that allow me to use my graphic adapter to its full potential."

Buy one that doesn't require that sort of driver.

"Using my graphic adapter to its full potential is the main reason why I did b*u*y a graphic adapter, so it should not be a reason for not having the possibility to suspend and resume."

You are just saying it is not a valid reason because you don't agree with it.


"it is much more important imho than having a desktop cube."

As long as the kernel devs work in compiz, this statement is relevant, else it isn't.

"It can mean 150$/year in economie when used overnight instead of leaving it always on."

Ubuntu does allow you to turn the computer off.


"This is for me the most anoying bug that there is in ubuntu. "

That's a matter of opinion.

-

I insist, remove those buttons, speed up (re)boot and just pretend these things never existed, if you want you computer to go green, turning it off saves more energy...]]> Sun, 23 Mar 2008 22:24:51 +0000 Sun, 23 Mar 2008 22:30:26 +0000 Mon, 24 Mar 2008 11:42:38 +0000
My point is, just because you don't think it useful, doesn't mean others shouldn't use it.]]> Mon, 24 Mar 2008 12:04:33 +0000 I think ubuntu should do updates with more care and not in a crazy way.]]> Tue, 25 Mar 2008 17:23:15 +0000 Tue, 25 Mar 2008 23:18:35 +0000
The point is being able to turn your laptop off and on quickly. NO ONE in everyday life of a laptop thinks "Hey I want to wait 5 minutes between on/off every time to save a teeny bit of power". Thats simply absurd. Someone forgot to add that last line to their resume, or whatever, they need to quickly open/close their laptop to do quick things.

Fact is people will not use Ubuntu if it cant handle a basic task like suspend/resume. I have seen the frustration first hand and I'm willing to bet Ubuntu has lost countless users to this unfortunate messup. Windows provides suspend/resume, so they return to what works.]]> Wed, 26 Mar 2008 18:05:05 +0000

I insist, speed up boot and reboot, imagine if it took 15 seconds to do either of them (which seems possible since right now it is taking 45 seconds) , for the very same reason you have stated you like to use suspend, you would be able to power off the computer that fast. And it will save energy as well.

So, after reading your statement, I would say that if boot/reboot took much less time, you would agree with the idea of removing suspend and hibernate, and that's what I am saying, since , according to the devs it is currently unfixable, the best approach would be to begin working on improving boot/shutdown times to make them a viable alternative to suspend and hibernate.

]]> Thu, 27 Mar 2008 15:36:42 +0000 Sat, 29 Mar 2008 22:10:49 +0000
If we can get full boot, including gnome desktop, and sessions started plus files loaded for emacs, firefox, and Matlab all within 5 seconds, then I'd say there's no need for suspend.

I agree that hibernate isn't such a big thing to get excited over though.
]]> Mon, 31 Mar 2008 10:09:05 +0000 Wed, 02 Apr 2008 05:32:29 +0000
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/185703

in the hope that it can be addressed before Hardy is released. I swear it got even slower between Gutsy and Hardy beta!
]]> Wed, 02 Apr 2008 15:53:29 +0000
Suspend/Resume does not even work out of the box! According to Dell this is Ubuntu's fault... if so, why would they sell a defective laptop in the first place?

This is incredibly embarrassing, one of the fundamental features of laptops, doesn't work with a pre-equipped Ubuntu machine.]]> Thu, 03 Apr 2008 14:09:22 +0000 Sat, 05 Apr 2008 16:04:49 +0000
Couple with that the fact that its a moving target, and for me for a given piece of hardware, it has generally gone backwards. My old laptop used to suspend fine all the time a few years ago, but then around dapper or so it stopped. Then my desktop used to suspend reliably, until a kernel update that shipped somewhere in fiesty .. and since then I've never been able to get it to do it.

The problem that Ubuntu can address is getting the right information more readily available, rather than relying on hunting through out-of-date forum posts to find out (a) whether some hardware works, and (b) the right methods to try in order to get it working. ]]> Sun, 06 Apr 2008 12:31:05 +0000 It is not the task of Ubuntu to support all the hardware in the world.

Err, it is if they ever want to make a dent in bug #1: https://bugs.launchpad.net/ubuntu/+bug/1]]> Sun, 06 Apr 2008 20:29:08 +0000
Also, to say that Ubuntu should get rid of suspend/resume and replace it with a quicker reboot, is not solving the problem, but trying to cover it up. The fact is that Suspend/Resume is a fundamental feature of laptops.]]> Mon, 07 Apr 2008 17:22:56 +0000
Kernel 2.6.24 and suspend and hibernate work for me, amd64.]]> Mon, 07 Apr 2008 22:37:08 +0000 Wed, 09 Apr 2008 03:33:25 +0000 Thu, 10 Apr 2008 14:41:29 +0000
this guy nails the problem. this whole suspend/hibernate issue needs fixing, I agree with the author in saying that it is mindboggling why this still doesn't work right, it just can't be.]]> Sat, 12 Apr 2008 13:05:03 +0000
It's also very annoying to hear people insist that things work fine on Windows, therefore Windows is doing a better job. That's complete BS. Even if nVidia cards suspend/resume perfectly in Windows (a very debatable point), that says nothing about Windows. Any reason it's working on Windows is because nVidia made their proprietary drivers work right specifically for Windows, but they didn't take the time or effort to do the same for Linux AND they won't make their specifications available so that we can do it ourselves.]]> Mon, 14 Apr 2008 01:02:37 +0000
There are also problems in the packaging and configuration of the available suspend tools. I found my laptop started suspending successfully after installing an additional package. I only knew to look for that package after seeing it mentioned in an obscure reply to a obscure post in an obscure forum somewhere.

Then there's documentation - It's hard to get authoritative current information about the way it's supposed to be done and what alternatives can be tried. Even for people willing to debug different hardware, its hard to find where to start. Doing a search on ubuntu suspend yields a whole lot of older pages with various attempts and anecdotes, many of which are outdated and don't represent the "right" way of doing things now - and also apply only for some unknown types of hardware

Then also yes, it does make sense to compare to Windows. If Windows will suspend and Ubuntu won't then Windows (as an operating system) is doing a better job in that aspect. Sure, it may be harder for Ubuntu to do it due to its development model and popularity, but that doesn't change the facts about which is actually working better for the user. ]]> Mon, 14 Apr 2008 02:28:30 +0000 "even with proprietary drivers."

Sorry but this is not the solution. Work is being done upstream with free software (the kernel, HAL,...), why would we want to short-circuit them and despise their work by using proprietary drivers when we could work with them and help them fix this issue only with free software ?]]> Mon, 14 Apr 2008 06:49:55 +0000
So it isn't about short-circuiting the open-source developers. At worst, its making life harder for them by getting them to test against systems with binary video drivers installed. Perhaps there's also some pressure there to get more dialogue going on with the video card companies to get their driver bugs with suspend sorted out.]]> Tue, 15 Apr 2008 12:07:30 +0000 That makes a lot more sense :)

The only problem with that is that it will be kind of difficult to ask the devs of "suspend / resume" to support those binary drivers.

In fact with FOS drivers, the devs of "suspend / resume" can say "hey, your driver is broken so suspend / resume won't work. You should fix it this way, here is a beginning of a patch. Thanks in advance for fixing it."

With binary drivers, those same devs have the choice between:
1. "We do all the work to try and make those binary work, without even knowing how they work, without knowing if their next release won't break all that we did."
2. "Suspend / resume won't work with those drivers"

I kinda understand them... :-\]]> Wed, 16 Apr 2008 06:48:54 +0000
We should go on hoping for the best.]]> Wed, 16 Apr 2008 07:03:27 +0000 it's just that when you claim "for human beings" you can't exclude the human beings who like to use laptop computers ;-) I love Ubuntu, but the popularity of this issue on Brainstorm proves how dissatisfied lots of people are about the state of laptop support in Ubuntu. I just hope all those cool programmers are working on it, while I hoped that they were working on it ever since I first tried it 2 years ago. 2 years is a long time to get something done that WinDumb has done almost since the inception of XP....]]> Wed, 16 Apr 2008 23:14:31 +0000
Suspend and Hibernate are the *most* important features for mobile computing, I really hope that Ubuntu team can pay a serious attention to these issues and play a key role to improve the situation, either by collaborating with the manufactures, or developing their own robust ACPI support. Wish these painful experiences with suspend/hibernate go away, forever.]]> Thu, 17 Apr 2008 05:52:09 +0000 Fri, 18 Apr 2008 17:30:33 +0000
I hope hardy or at least intrepid ibex fixes this!]]> Fri, 18 Apr 2008 22:06:57 +0000
It's hard for Ubuntu to do much about those sort of issues, but where they can help as packagers is to make the most of what code and options are available, and harness the power of their userbase for debugging and dissemination of up-to-date methods for getting it working. ]]> Sat, 19 Apr 2008 08:47:46 +0000
Btw, how come nobody's claiming to do anything about this issue; it has been at #1 position for weeks now.]]> Wed, 23 Apr 2008 09:09:23 +0000 Wed, 23 Apr 2008 23:25:35 +0000 Sat, 26 Apr 2008 00:54:17 +0000 Tue, 29 Apr 2008 03:59:57 +0000 Fri, 02 May 2008 00:46:05 +0000 Fri, 02 May 2008 01:07:15 +0000 Fri, 02 May 2008 21:38:22 +0000
HOWEVER sometimes when i suspend the machine and i try to wake it up again, the screen remains black. This seems to happen randomly and not very often. These are the problems that need to be fixed under windows it never like this.]]> Thu, 08 May 2008 11:55:05 +0000 Sun, 11 May 2008 12:48:31 +0000 Mon, 12 May 2008 22:19:09 +0000 I own an HP tx1215nr laptop and suspend doesn't work at all. I really need it when I am in between classes. It saves a lot of time when you don't really need to fully shut down your computer. For example, in between classes or when you leave school and take a short bus trip home and you know you'll use it again when you get home.
Cheerfully, hibernate does work... sometimes.]]> Thu, 15 May 2008 15:47:51 +0000 Tue, 20 May 2008 16:44:48 +0000 http://www.ubuntu.com/partners/hardwareprogramme/freesoftware]]> Thu, 22 May 2008 19:16:20 +0000 http://www.ubuntu.com/partners/hardwareprogramme/freesoftware]]> Thu, 22 May 2008 19:20:57 +0000
I think its embarrassing that even Dell laptops preloaded with Ubuntu have broken suspend/resume.

This issue needs to be taken seriously by Canonical, this should be #1 or very high priority for the 8.10 release! PLEASE!!!]]> Fri, 23 May 2008 14:55:10 +0000
Put some kind of wrapper around it so that it thinks it is booting. That seems to me like the solution to these proprietary driver problems. Just a thought!]]> Fri, 23 May 2008 15:04:19 +0000
I don't think that explains anything.]]> Tue, 27 May 2008 15:36:02 +0000
MAINTAINERS | 7 +
arch/x86/mm/fault.c | 22 +++
crypto/Kconfig | 8 +
crypto/Makefile | 1
drivers/macintosh/via-pmu.c | 1
drivers/md/md.c | 2
fs/buffer.c | 87 +++++++++++++
fs/drop_caches.c | 2
fs/fuse/control.c | 1
fs/fuse/dev.c | 7 +
fs/fuse/dir.c | 35 ++++-
fs/fuse/file.c | 14 ++
fs/fuse/inode.c | 4
fs/ioctl.c | 1
fs/namei.c | 2
include/asm-powerpc/suspend.h | 3
include/asm-ppc/suspend.h | 3
include/asm-x86/suspend_32.h | 3
include/asm-x86/suspend_64.h | 3
include/linux/Kbuild | 1
include/linux/buffer_head.h | 5
include/linux/freezer.h | 15 ++
include/linux/fs.h | 10 +
include/linux/kernel.h | 2
include/linux/mm.h | 1
include/linux/netlink.h | 2
include/linux/suspend.h | 65 ++++++++++
include/linux/swap.h | 8 +
init/do_mounts.c | 2
init/do_mounts_initrd.c | 6
init/main.c | 3
kernel/power/Kconfig | 264 ++++++++++++++++++++++++++++++++++++++++++
kernel/power/Makefile | 31 ++++
kernel/power/disk.c | 41 ++++--
kernel/power/power.h | 42 ++++++
kernel/power/process.c | 50 ++++++-
kernel/power/snapshot.c | 51 ++++----
kernel/printk.c | 22 ++-
kernel/timer.c | 56 ++++++++
lib/vsprintf.c | 23 +++
mm/Makefile | 2
mm/memory_hotplug.c | 2
mm/page_alloc.c | 29 ++++
mm/vmscan.c | 211 +++++++++++++++++++++++++++++++--

Most of that is adding support for freezing fuse filesystems, exporting swsusp functions so we can use them and adding an improved (finer grained) method of freeing memory.

TuxOnIce generally applies cleanly (or at worst with minor rejects) against Ubuntu kernels. Git trees with it already applied to Feisty, Gutsy and Hardy are available from kernel.ubuntu.com.

That said, I agree that I should be seeking to get it merged into vanilla. Unfortunately I don't have a lot of time to work on that now, and it seems that too many people have swallowed FUD to even consider it anyway.

Nigel]]> Wed, 28 May 2008 04:08:14 +0000 Fri, 30 May 2008 14:28:44 +0000 Sun, 01 Jun 2008 23:38:38 +0000
1) It reduces complication to the user, which increases work efficiency and makes computing a pleasure for the user

2) It saves EVEN MORE power because now ABSOLUTELY NO power is used when the computer is off. Sure, it takes power to turn on the computer, but first we can make boot up extremely fast and also make it power-efficient or even have an energy-saver boot up style in the power options that should be coming in a future release of Ubuntu.


To reiterate, scrap suspend, speed up boot and shutdown, and make hibernate really good.]]> Mon, 09 Jun 2008 22:00:18 +0000 On: 68W
S3 Suspend: 3.3W
Off: 3.1W

So the power needed for S3 suspend is negligable compared to the power supply leakage current. For modern computers, the only way to get absolutely no power draw is to pull the power plug out, and most people don't do that.

I don't see how having to re-start applications and find context again increases work efficiency and makes computing a pleasure for the user.

S3 Suspend is just expected to be available these days, and is strongly demanded. Having it working poorly in Ubuntu is an embarrassment, and the lack of debugging documentation is a shame.]]> Sat, 14 Jun 2008 05:52:43 +0000 Suspend and hibernate MUST ABSOLUTELY work. On my system it does not since one year and an half !!!
It does not matter who's fault it is. I used to have MS W and it worked, now with Ubuntu it does NOT work.
The final word is the quality of the service rendered to the users.
Ubuntu (and other Linux distro) is failing miserably on that topic, and as an old software GM, my answer is always the same, the developers (we) are not good enough, sorry guys...
But this poor performance is costing us 2 Nuclear Reactors of 1.000 MWatt/h each per year....On the other hand, I thrown away my Win and I am ecstatic about Ubuntu in general.]]> Tue, 17 Jun 2008 14:59:36 +0000 ]]> Wed, 18 Jun 2008 12:05:12 +0000 Fri, 20 Jun 2008 23:01:09 +0000
The time requiered to suspend is the catch.]]> Sat, 21 Jun 2008 20:42:06 +0000 Sat, 21 Jun 2008 21:27:23 +0000 This is not an idea its a bug.]]> Mon, 23 Jun 2008 19:53:12 +0000 Wed, 25 Jun 2008 03:43:49 +0000
Admittedly it's old, but it's really common, suspend/hibernate should work for it. I reverted to the 7.10 kernel, which has both suspend and hibernate support for it.

Even my 2 year old custom built system that has always worked with suspend (never tried hibernate on it) Does not work with 8.04 now. I guess I'll have to try reverting it's kernel as well.

But this sad... why exactly is the ability of Ubuntu to suspend/resume getting WORSE instead of better?

Neither one works with with or without using proprietary drivers.]]> Wed, 25 Jun 2008 18:59:05 +0000
Its been broken for two years, and is not even a priority for the future version (Ibex). Simply embarrassing.]]> Mon, 30 Jun 2008 18:29:20 +0000 Mon, 30 Jun 2008 20:33:12 +0000
Post ideas on how to fix this Suspend and Hibernate problem and I will be glad to vote for them.

This "idea" (problem) should be marked as "not an idea" before everyone starts posting their pet peeves and not including ideas to solve their pet peeves.

We are looking for ideas to solve the Suspend and Hibernate problems. Does anyone have any ideas on how to solve that?

Here's few:

1. Provide an experimental kernel with the debugging features enabled and the latest suspend and hibernate patches to make it easier for mass testing.
2. Patch Apport to mark the bug reports that a certain application crashed after a suspend or hibernate.
3. Create a wiki that list known working and non working suspend and hibernate hardware.
4. If we need hardware specs, petition the manufacturer.
5. If the bugs have been identified and no one is working on them, create a bounty.

These are just at the top of my head and that is what we should be voting on as well.
]]> Tue, 01 Jul 2008 18:26:29 +0000 Wed, 02 Jul 2008 15:42:45 +0000
I'd go as far and say every other issue/feature out there for intrepid isn't even half as important as this one..... maybe the Ubuntu makers, and I applaud them because I really like Ubuntu a lot and so much is happening, really, aside from my ranting, Ubuntu is really great, but maybe the Ubuntu makers don't suspend/hibernate enough to see how broken this is on many laptops? just an assumption. or people involved in making Ubuntu pick an Ubuntu-compatible laptop in the first place?]]> Sat, 05 Jul 2008 23:02:28 +0000 Sun, 06 Jul 2008 16:18:25 +0000 In my PC, Suspend is OK, though a little slow but acceptable but works most of the time but Hibernate does NOT work half the time.
Fixing Suspend and Hibernate should be of topmost priority.]]> Sun, 06 Jul 2008 19:26:36 +0000 "If you won't release an open source driver Ubuntu won't support display green colors with a nvidia card detected, when Ubuntu reached 20 % market share."]]> Tue, 08 Jul 2008 12:35:47 +0000
And if we go around blackmailing companies, they sure wont help us... And I guarantee many linux users would leave.

And I frankly find the suggestion semi offensive. We shouldn't associate linux with blackmail]]> Tue, 08 Jul 2008 12:55:27 +0000 Wed, 09 Jul 2008 16:07:33 +0000 Mon, 14 Jul 2008 13:12:09 +0000 Mon, 14 Jul 2008 22:28:22 +0000 Wed, 16 Jul 2008 12:47:36 +0000
Except, the drivers at the time we produced wouldn't have been as good... Not nearly as good. And their drivers showed many users they were serious about linux. And in turn, those drivers were used to prove to users, that Linux was good, which spurred a whole new segment of development.

If the nvidia drivers are causing problems, it may be a result of too much competition actually. Both ATI and Nvidia have to sacrifice some stability in the drivers department for much better optimisation of drivers (like doing dodgy clipping in benchmarks or lowering graphics quality). The best thing we could do is improve the bug reporting systems, so that we can report back exact systems that don't suspend/hibernate properly, so we can isolate which hardware configurations have issues so they can fix it (this would be the case normally anyway!)

Like it or not, consistent driver support from Nvidia is a beating pulse in our Linux timeline, and we should respect the effort they put in. Whilst something we coded might be a bit more stable, they probably wouldn't have been near as fast as Nvidia's official drivers.

But anyway, due to the changing environment (AMD swallowed ATI), this might change soon anyway. After all, Nvidia's purchase of 3DFX, probably contributed significantly towards their fight with ATI for so long. ]]> Wed, 16 Jul 2008 13:39:53 +0000 Thu, 24 Jul 2008 04:10:02 +0000
I'm using drivers version 173.14.09 on Ubuntu Hardy i386.]]> Sun, 27 Jul 2008 02:55:58 +0000 http://devresources.linux-foundation.org/dev/robustmutexes/src/fusyn.hg/Documentation/power/video.txt

Will have to try that out some time along with the links given by amitk.

My laptop, if anybody already knows of a working solution:
Compaq Presario 2500
]]> Wed, 30 Jul 2008 23:51:57 +0000
@ rorymccann, purpleidea, gcc and others
So "free software" should not allow me to choose proprietary drivers over open source and have them work? Doesn't sound so "free" to me. I reserve the right to choose which ever drivers I want to use, not you or anyone else. This is freedom of choice.

Anyway that's not what this is about. Its about the problems lots of people are having with suspend and hibernate irrespective of what being used to drive it.

The Ubuntu developers can't do much about proprietary drivers which are the responsibility of the driver developer, but they can assist where possible. ]]> Fri, 08 Aug 2008 13:00:45 +0000
Sure, it would be nice to have them, but... if Ubuntu manages to make them work, they've boldly gone where no (wo)man has gone before... +1]]> Sun, 24 Aug 2008 11:56:40 +0000
I routinely see suspend and hibernate work properly around here, on various flavors of Windoze and Linux.

In fact, the only os I cannot get hibernate to work right is on Hardy using a Dell Latitude D610. I seem to be in good company, as there is no shortage of hibernate threads for Hardy. Actually, it appears to hibernate correctly, it just cannot resume from the image and merely reboots.]]> Mon, 01 Sep 2008 22:33:32 +0000
Getting it to work reliably would be fantastic. Booting takes to long, and leaving it on is just not an environmentally friendly solution when these computers have these nice low power standby modes which work fine with Windows.]]> Wed, 03 Sep 2008 05:03:07 +0000 Sat, 06 Sep 2008 01:02:51 +0000 I have a Fujitsu P7150D. The hibernate works fine with 7.10 but broken in 8.04. Not sure how to fix it. Major issue for me.]]> Fri, 12 Sep 2008 17:30:59 +0000 Sat, 13 Sep 2008 00:04:18 +0000 Wed, 17 Sep 2008 00:23:57 +0000
this is on a new pc wich as a asus m2n-e motherboard and a asus nvidia 8600gt vid card using the proprietary drivers.

i am very surprised at that successfull result and even more that i did not even need to do anything to be able to achieve it.

Wow good job devellopers this is another step closer to the absolute perfect OS for me.]]> Thu, 18 Sep 2008 23:07:06 +0000

almost fried my laptop today >]]> Thu, 25 Sep 2008 05:41:26 +0000 Thu, 02 Oct 2008 16:47:42 +0000
Graphical explication

1. Delete de windows networks icon
http://img90.imageshack.us/my.php?image=85330472bu1.png


2. In the new Network:// detected the protocols in host
http://img137.imageshack.us/my.php?image=95376211hx7.png

3 Show the host detectes in local network
http://img90.imageshack.us/my.php?image=83103217gu9.png


4 Clic in host 1 for more details os the procotols avaibles
http://img259.imageshack.us/my.php?image=46306862qe0.png


5. Clic in protocos nfs for share the folder in the local host
http://img259.imageshack.us/my.php?image=11026040at8.png



And the p2p protocol share files local ]]> Thu, 02 Oct 2008 22:32:13 +0000
I just changed 3 lines in 2 config files, and my suspend works for the first time ever on this PC!!

I have been annoyed ever since I bought this motherboard 5 years ago...

I just thought my motherboard was buggy and didn't support suspend...

Stumbled across bug 34043 and the https://help.ubuntu.com/community/NvidiaLaptopBinaryDriverSuspend pages...

So I thought... what the hell, suspend would be nice, lets give it another go!

My setup is Hardy 8.04.1 with updates and backports and proposed and several PPAs and envy-ng with nvidia drivers: 173.14.12
Abit NF7-S mobo (NForce 2 Ultra)
Geforce 5900XT AGP
Never suspended correctly before in Windows XP

So I set the option in xorg.conf in Section "Device"
Option "NvAGP" "1"

And I changed the 2 options in /etc/default/acpi-support
SAVE_VBE_STATE=false
POST_VIDEO=false

AND guess what!! it F@#$#@king works for the first time ever!!! I'm so happy I could cry!!

Previously it would sleep ok.. then it would just beep continously with a blank screen until reset when you tried to wake up.

Any disadvantage to NvAGP "1" I should know?
If not, this should seriously be the default!

Thanks to all the hackers that discovered this!

References to:
https://bugs.launchpad.net/ubuntu/+source/acpi-support/+bug/34043
and
https://help.ubuntu.com/community/NvidiaLaptopBinaryDriverSuspend

Note: I personally didn't need to blacklist any modules... just changed the 3 options I mentioned before.
]]> Mon, 06 Oct 2008 14:17:54 +0000 I'm so happy :)]]> Sat, 18 Oct 2008 22:19:28 +0000
Option "NvAGP" "1"
was trying to load the NvAGP part of the nvidia driver, and failing, falling back to PCI mode, which made it equivalent to Option "NvAGP" "0" which I'm using now... it's a bit slower then actually having AGP support, but suspend is better then no suspend, I save so much time booting it's worth it... especially since the nvidia driver is so slow in 2D acceleration anyway.]]> Wed, 22 Oct 2008 00:47:43 +0000 Wed, 22 Oct 2008 21:05:34 +0000
i voted for this idea a few months ago (hardy). now i switched to intrepid ibex and hibernate works out of the box - i love it and want to thank you very much for your work on this issue!

Flo]]> Wed, 05 Nov 2008 12:12:19 +0000
I need a fix!!! :)]]> Tue, 25 Nov 2008 02:22:42 +0000 Sun, 30 Nov 2008 10:51:52 +0000 ]]> Sun, 07 Dec 2008 07:42:01 +0000
Linus and the guys just dived into this long standing problem, and fixed it.]]> Mon, 08 Dec 2008 17:56:22 +0000
Some of my machines won't suspend at all, some will just freeze when trying to wake up, and the remainder will lose at least some essential functionality after waking up, like Bluetooth, the touchpad, the keyboard, display backlight, or two or more of these in varying combinations.

Please do something about this and make suspend/hibernate one more area where Ubuntu is far superior to Windows!
]]> Mon, 29 Dec 2008 19:45:43 +0000 http://torvalds-family.blogspot.com/2008/12/life-is-good-again.html

Linus and the guys just dived into this long standing problem, and fixed it.

i'm kind of confused, if the blog post is really related to this very huge general bug in the suspend/hibernate functionality, isn't that the news of the year for the linux community? ]]> Tue, 13 Jan 2009 23:26:51 +0000 My motherboard is MSI P35 Platinum.

Now on ubuntu 8.10 I have sometimes problems with shutdown.
]]> Fri, 23 Jan 2009 21:59:35 +0000 Fri, 23 Jan 2009 22:20:48 +0000 Fri, 06 Feb 2009 18:33:02 +0000
Unfortunately it's broken in many ways- permissions, drivers, poorly written acpi, etc. It works fine for example on my laptop running Xubuntu, by my desktop will just return me to the GDM login screen. Why? I have no idea.

We need better tools to understand why suspend does not work, and how to fix it.

PowerTOP for suspend would be nice.

]]> Wed, 25 Feb 2009 07:51:09 +0000 Mon, 06 Apr 2009 14:22:27 +0000 Mon, 13 Apr 2009 19:59:26 +0000 Tue, 14 Apr 2009 16:25:17 +0000
I don't know "why" this works, but it does for my Thinkpad X60s. I suspect its something that will work for others, since its a bios level thing, and that relates to the previous post.

1. sudo gedit /boot/grub/menu.lst
2. find the grub menu line that is the ubuntu default .. the first one that appears, that will look like this:
title Ubuntu 7.10, kernel 2.6.22-14-generic
root (hd0,0)
kernel ....... etc

3. Add the following to the end of the "kernel" line:

acpi_sleep s3_bios

Here is my "kernel" line.
kernel /boot/vmlinuz-2.6.22-14-generic root=UUID=b6e33b4a-d2ec-4107-9aaa-8956ed669d82 ro quiet splash acpi_sleep s3_bios

4. reboot computer

Now resume works every time. ]]> Tue, 28 Apr 2009 23:43:06 +0000
Nvidia driver: 180.44
Geforce Go 7600

Hp dv9000 series; Model dv9207us
kernel: 2.6.28-11-generic (#42-Ubuntu SMP Fri Apr 17 01:57:59 UTC 2009)]]> Sun, 10 May 2009 05:38:29 +0000 Thu, 27 Aug 2009 15:30:21 +0000