[EMAIL PROTECTED] said:
>> [EMAIL PROTECTED]:~$ sudo s2ram -i
>> This machine can be identified by:
>> sys_vendor = "System manufacturer"
>> sys_product = "System Product Name"
>> sys_version = "System Version"
>> bios_version = "ASUS A8N-SLI DELUXE ACPI BIOS Revision 1015"
>> S
sure, here's an updated patch.
Jason
---
Add a pointer to hibernate scripts to the ususpend README.
---
README | 13 +
1 file changed, 13 insertions(+)
Index: suspend/README
===
--- suspend.orig/README
+++ suspend/
On Mon, Nov 06, 2006 at 08:08:59AM +0100, Stefan Seyfried wrote:
> Well, the common upstream solution will soon be pm-utils, so maybe
> the hibernate users should start looking at that, so that it does
> what they want :-)
Why wouldn't the pm-utils authors start with hibernate as a base? It
looks
On Sun, Nov 05, 2006 at 11:19:52PM +0100, Pavel Machek wrote:
> > configuration issues. I think suspend.sf.net ought to point users to
> > hibernate just as suspend2 does.
>
> I guess we can take README patch ;-).
OK, how about this?
Jason
---
README | 13 +
1 file changed, 13 in
On Sun, Nov 05, 2006 at 10:22:24PM +0100, Pavel Machek wrote:
> Well, having some default scripts in suspend.sf.net package... would
> not be that bad. Not all the stuff is done best in C.
hibernate supports swsusp, ususpend, and suspend2 equally. It isn't an
integrated part of any of them. So all
On Sun, Nov 05, 2006 at 06:18:34PM +0100, Pavel Machek wrote:
> Okay, why not. If you resend the patch, and noone dislikes it, I can
> apply it..
Here it is again. Just to be clear, the patch is for Bernard's hibernate
scripts; I just CC'd suspend-devel because it may be of general interest
to usu
On Mon, Oct 30, 2006 at 12:14:30PM +, Matej Cepl wrote:
> Why not for swap? OK, except that suspend cannot work :-). Does it
> mean that I should switch to inserting password for swap?
yes, that's necessary. luks is nice for managing the passwords.
Jason
-
On Mon, Oct 16, 2006 at 01:26:16PM +0200, Matej Cepl wrote:
> is something like this possible? I have here notebook with Fedora Core
> 5 with swap encrypted using cryptsetup create. Is it possible to
> suspend/resume such notebook with ususpend?
Yes, it works just fine. I've been running a laptop
The advantage to using s2both is that if the battery runs out while
suspended to ram, the system can still resume from disk (after power is
restored). The disadvantage is that it's slower, and most of the time
the disk image is never used.
This implementation of USuspendMethodAuto takes two integ
On Mon, Sep 25, 2006 at 08:22:09AM +0200, Stefan Seyfried wrote:
> - What happens if i suspend with a "timing enabled" s2disk and resume with a
> "pre-timing-enabled" resume? (Somebody updates the package but forgets to
> run mkinitrd).
> Will it just fail (invalid header) or will it cause tota
On Sun, Sep 24, 2006 at 08:28:21PM +0200, Pavel Machek wrote:
> If you want to do it _right_, write timing info onto disk, then read
> it from there. That will not slow the user down, while still telling
> them nice statistics when resume is done. Hmmm?
where? there's no fs mounted. Is there a pla
On Sun, Sep 24, 2006 at 07:17:46PM +0200, Rafael J. Wysocki wrote:
> But then you probably won't be using this switch any more. ;-)
I will when the async i/o support comes in 2.6.19.
> I think it would be better to make it possible to compile out the
> benchmarking code.
Maybe eventually. Right
On Sun, Sep 24, 2006 at 10:11:17AM +0200, Pavel Machek wrote:
> I'm not sure I like this one. It is little too specialized...
without this, I can't see the i/o timings at all. Is there another way
to do it?
I guess it could be a command-line option, or an #ifdef used only by
someone benchmarking.
Make it possible to set the s2disk shutdown method in the config file.
Jason
---
HOWTO |6 ++
resume.c |5 +
suspend.c | 14 +-
swsusp.h |2 +-
4 files changed, 25 insertions(+), 2 deletions(-)
Index: suspend/suspend.c
===
Add an option to allow the user to wait for keypress during the resume
process so that image i/o stats may be read.
Jason
---
HOWTO |4
resume.c | 13 +
suspend.c |5 +
swsusp.h |2 +-
4 files changed, 23 insertions(+), 1 deletion(-)
Index: suspend/resu
On Fri, Sep 22, 2006 at 03:38:39PM +0200, Pavel Machek wrote:
> Ok, I'm not applying this one, but command line switch selecting
> between reboot/platform/shutdown/S3 would be welcome.
OK. The patch was obviously only meant for testing. I'll send patches
that make the reboot-on-s2disk and pause-on
On Fri, Sep 22, 2006 at 03:48:20PM +0200, Pavel Machek wrote:
> Well, it did not apply (reject) and did not compile when applied
I was working against the debian uswsusp 0.2 package - i'll use cvs next
time. sorry.
> (missing swsusp.h changes). I fixed it up, but please check it after
> me...
I
Make it easier to benchmark a lot of s2disk cycles.
Reboot the machine instead of powering off, and wait for keypress after
printing timings during resume.
---
resume.c |1 +
suspend.c |1 +
2 files changed, 2 insertions(+)
Index: suspend-0.2/suspend.c
The following two patches add suspend/resume benchmarking.
The first is the more important one. During suspend, it times the image
writeout + fsync(), and stores the result in the swsusp header. During
resume, it times image load, then prints writeout, readin, and total
timings and speeds.
The s
Print s2disk benchmarks. The image writeout time is saved in the
swsusp_info header during suspend, then statistics are printed during
resume about elapsed time and image load/store speed for writeout,
readin, and total.
---
resume.c | 24 +++-
suspend.c | 11 +++
[EMAIL PROTECTED] said:
> Seems okay to me... if 45% image is significantly faster than 50%
> image, lets do it...
agreed. I also found that limiting the image (even a little) produces
faster writeout and much more consistent writeout times.
I'm still fooling around with my patch that prints benc
[EMAIL PROTECTED] said:
> Hmm.. we can probably do that. Is smoothly-running progressbar really
> all that important?
it is to me. That's why I wrote the patch, because I got tired of
looking at inaccurate progress output. And I don't use splash.
I agree about correctness, but this *is* correct.
On Mon, Sep 18, 2006 at 01:10:45PM +0200, Tim Dijkstra wrote:
> > It would be nice to actually benchmark this, and hardcode the option that
> > provides acceptable performance...
>
> Sure, maybe Jason can provide some numbers.
yeah, i'll whip up a writeout-timing patch tonight. That might be use
Increase the granularity of calls to start_writeout() when
early_writeout is used. This results in a smooth progress display during
writeout, rather than a noticeable pause each 20% of the way through.
I think performance shouldn't be hurt, since we're still batching pages
1% at a time, down from
On Thu, Aug 10, 2006 at 11:55:09AM +0200, Sheer El-Showk wrote:
> piix9092 0 [permanent]
> ide_core 115900 3 ide_cd,piix,ide_disk
>
> Your patch was at the more general level of the ide drivers so do you
> have any idea why it didn't work on this particular drive
On Thu, Aug 10, 2006 at 12:54:19AM +0200, Sheer El-Showk wrote:
> Is this a similar problem to what you were having on your laptop
> before the patch? Would you expect your patch to work for me or do
> you think its an unrelated problem? Whatever it is I definately have
> an issue with the disk t
On Wed, Aug 09, 2006 at 10:12:10PM +0200, Rafael J. Wysocki wrote:
> AFAIK, the AHCI suspend to RAM support is on its way, but does anyone know
> about the other SATA/PATA drivers?
AFAICT, the situation with drivers/ide is that it happens to work (by
luck) well enough for most people that it's bee
On Wed, Aug 09, 2006 at 06:27:13PM +0200, Sheer El-Showk wrote:
> So, might this general IDE fix solve my problem? I compiled
> 2.6.18-rc3-mm2 and tried 's2ram -f' but it just crashed while
> suspending so I couldn't test this.
Maybe. Which ide driver are you using? There's plenty of other stuff
On Wed, Aug 09, 2006 at 01:42:17PM +0200, Rafael J. Wysocki wrote:
> The fix was for AMD/NVidia IDE, but Alan vetoed it on the basis that it should
> be done at the general level. Everyone agreed, but I don't know if anyone
> took care of that. Jason?
I redid the patch at the general IDE level.
[EMAIL PROTECTED] said:
> I've tried running s2ram with all of the suggested arguments from
> init=/bin/bash, with and without noapic, running kernel 2.6.18-rc3, I
> believe. This was configured with *minimal* support -- no modules, sound,
> usb, IDE, SMP, preempt, cpufreq, mce etc. (I've done simi
[EMAIL PROTECTED] said:
> (aside from your fierce opposition to the competition, is there any rational
> reason why I should not have suspend2 in my kernel?)
On more than one occasion, the suspend2 patches have broken both
in-kernel swsusp and uswsusp on my laptop.
Come to think of it, I wonder h
[EMAIL PROTECTED] said:
> Would that kill keyboard leds?
the opposite, actually. After about 30 seconds or so hung, my capslock
LED would start blinking on and off
Jason
-
Take Surveys. Earn Cash. Influence the Future of IT
[EMAIL PROTECTED] said:
> I guess it's an IDE-related problem similar to the one which Jason Lunz's
> patch fixed, or a (P)SATA-related one.
If it's like what I saw, most of the time the hang will keep the hard
drive LED on solid. But not always.
Jason
--
On Mon, Jul 10, 2006 at 08:31:13AM +0200, Stefan Seyfried wrote:
> Ok, try the following (not with s2ram, but with plain vbetool. If it works,
> it will be in s2ram real soon ;-):
>
> ---
> #!/bin/bash
> X=`vbetool vbemode get`
> echo mem > /sys/
On Wed, Jul 05, 2006 at 07:42:46PM +0200, Luca Tettamanti wrote:
> I'm not sure that the setting is preserved. Anyway the bus is locked
> and it seems that it's caused by IDE.
> I saw patches on lkml that added support for running ACPI methods to
> bring IDE back to life. But it seems that it's onl
[EMAIL PROTECTED] said:
> Any suggestions? (Even if only how to figure out what pnp device
> 00:07 and 00:08 are?)
I found this info in dmesg output after boot, where the pnp code
initializes and probes hardware.
Jason
Using Tomcat but need to do more? Need to support web services, security?
G
On Wed, Jul 05, 2006 at 10:16:39AM +0200, Stefan Seyfried wrote:
> The "Get video state buffer size failed" adn "base is 0x0" looks suspicious
> to me.
yeah. The program should probably exit when that fails - in my case it
then calls LRMI_alloc_real() with size 0, so it's not surprising it
crashes
On Wed, Jul 05, 2006 at 01:29:30AM +0200, Pavel Machek wrote:
> But we are running them under emulator, so that should not be a
> problem... and they are probably 16-bit x86 instructions, anyway.
yes, I see. That emulator seems be causing the segfaults when I use -s:
# gdb s2ram
GNU gdb 6.4.90-de
On Wed, Jul 05, 2006 at 01:21:11AM +0200, Pavel Machek wrote:
> > Further investigation shows that it works within X, whether or
> > not the boot console is using framebuffer. I'm not unloading any
> > modules, or doing anything other than running "s2ram -f -p".
> > After resume int
On Tue, Jul 04, 2006 at 06:34:00PM +, Pavel Machek wrote:
> I'd prefer not to apply this. It is better to just fix the kernel.
> Whitelist is for video problems, not for random kernel bugs.
I agree, I'm not about to switch to a 32-bit kernel. But giving a
warning like this could save people a
On Tue, Jul 04, 2006 at 06:40:02PM +, Pavel Machek wrote:
> Disable MCE checking and see what happens?
ha! very funny. I'll try that. But I think it won't help - by the time
the mce check fires, the machine has been hung for 20-30 seconds.
Jason
Using Tomcat but need to do more? Need to supp
On Tue, Jul 04, 2006 at 10:53:27PM +0200, Luca Tettamanti wrote:
> The blinking is done by the kernel panic handler.
yes.
> >bcm43xx: set security called enabled = 1 encrypt = 1
> >hda: dma_timer_expiry: dma status == 0x21
> >hda: DMA timeout error
> >
> >HARDWARE ERROR
> >CPU 0: Machine Check Ex
On Mon, Jul 03, 2006 at 11:48:08PM +0200, Pavel Machek wrote:
> Important info is "did kernel get control and crashed, or did it lock
> up in firmware". Beeping patch can be used to debug that, as can be
> Linus's RTC hack present in very new kernels.
I tried out netconsole, and it works very well
nel.\n");
+
if (test_mode) {
machine_known(id);
goto out;
Index: suspend-0.2/whitelist.c
===
--- suspend-0.2.orig/whitelist.c
+++ suspend-0.2/whitelist.c
@@ -57,7 +57,7 @@
/* The full mat
full match is:
+* { "Hewlett-Packard ", "Pavilion zv5000 (DP523AV) ", "F.35",
"F.35", VBE_POST|NOFB },
+* but I don't see any reason to match that tightly. - Jason Lunz */
+ { "Hewlett-Packard
[EMAIL PROTECTED] said:
> Interesting idea. I just booted knoppix 5.0.1 on my amd64, and tried a
> 32-bit "echo mem > /sys/power/state" from a running X desktop. It
> suspended successfully, and it still failed to resume, but this time at
> least the hard drive isn't on solid. Otherwise the same, t
[EMAIL PROTECTED] said:
> Interesting datapoint would be "does it work ok with 32-bit kernel"?
> Some live-CD might help answer that question.
Interesting idea. I just booted knoppix 5.0.1 on my amd64, and tried a
32-bit "echo mem > /sys/power/state" from a running X desktop. It
suspended successf
On Mon, Jul 03, 2006 at 08:34:58PM +0200, Rafael J. Wysocki wrote:
> Great. I'd like to create a HOWTO patch describing this, but I have no
> experience with initramfs and/or yaird. Would you please describe
> shortly all steps needed to make it work?
It's not bad. I really think that getting us
[EMAIL PROTECTED] said:
> - screen blank and black
> - no keyboard -- caps lock doesn't do anything and typing things like
> "ls", "find /" doesn't do anything;
> - no network -- cannot ssh in
> Seem to be able to get out only with the three-seconds-power-button push.
same for me. HD light on s
on debian unstable, this is all I needed to get yaird to generate an
initramfs image with ususpend support.
Jason
--- Templates.cfg.orig 2006-06-14 23:33:34.0 -0400
+++ Templates.cfg 2006-06-14 23:39:58.0 -0400
@@ -400,6 +400,8 @@
#
TEMPLATE resume
[EMAIL PROTECTED] said:
> I'm familiar with hibernate script, but I see a couple of minor issues:
Thanks for looking at it. Here's a version that's less half-assed; the
last one was a quick search-and-replace job.
> suspend may fail, yet you don't propagate the return value. Is this
> correct?
I
[EMAIL PROTECTED] said:
> [EMAIL PROTECTED] said:
>> Or if somebody wants to make a scriptlet to support it (use the
>> sysfs_power_state scriptlet as a guide), I'll happily merge it for
>> the next release.
>
> That's exactly what I did. I'll post it tonight when my laptop's on.
> It's completely
[EMAIL PROTECTED] said:
> Or if somebody wants to make a scriptlet to support it (use the
> sysfs_power_state scriptlet as a guide), I'll happily merge it for
> the next release.
That's exactly what I did. I'll post it tonight when my laptop's on.
It's completely straightforward.
Jason
___
53 matches
Mail list logo
