Linux Archive

Linux Archive (http://www.linux-archive.org/)
-   Gentoo User (http://www.linux-archive.org/gentoo-user/)
-   -   Update to newer kernel completely hoses suspend (http://www.linux-archive.org/gentoo-user/703357-update-newer-kernel-completely-hoses-suspend.html)

Canek Peláez Valdés 09-13-2012 12:59 AM

Update to newer kernel completely hoses suspend
 
On Wed, Sep 12, 2012 at 7:18 PM, Daniel Frey <djqfrey@gmail.com> wrote:
> So about a month ago I decided to update my kernel to the dreaded 3.x
> series. My old 2.6.x kernel was working fine, but of course I decided
> to try to update it anyway, knowing there were problems with suspend
> and a few other things.
>
> I've always used gentoo-sources. So I tried 3.3.8.
>
> Hrm. Suspend doesn't work. I tried 3.4.5, 3.4.9 and 3.0.35 (older
> versions are no longer available.) If I'd known it would completely
> kill my suspend and make it useless, I wouldn't have bothered.
>
> Here's the problem:
>
> I can suspend fine. It appears to work. It powers off and goes into
> its suspend state. I press the space bar. Nothing. So, then I
> discovered that as of 3.2 USB wakeup had completely changed in the
> kernel, and you need to set hubs and devices in /proc/acpi/wakeup
> (which is normally done for you) *and* in /sys/devices. No biggie, I
> wrote a script to do just that at
> http://forums.gentoo.org/viewtopic-t-933934.html. So now I can wake
> with the keyboard as before. Or can I?
>
> If I suspend and wake up the PC within about 10 minutes it works.
> After that, all hell breaks loose. The PC is dead. Completely. Waking
> up no longer works, not with the keyboard, or even the power button.
> The *only* way is to pull the power plug and leave it unplugged for a
> few seconds. Then the PC comes to life.
>
> I've never seen an issue quite like this one...
>
> I use mdraid in my kernel with IMSM to dual boot Windows. I've been
> using it for a long time, so that's not it. The only thing that's
> changed are the kernel versions I've tried. So far, every 3.x kernel
> has done this. Now, this could very well be a kernel problem, heres my
> ACPI config:
>
> # Power management and ACPI options
> CONFIG_ACPI=y
> CONFIG_ACPI_SLEEP=y
> # CONFIG_ACPI_PROCFS is not set
> # CONFIG_ACPI_PROCFS_POWER is not set
> # CONFIG_ACPI_EC_DEBUGFS is not set
> CONFIG_ACPI_PROC_EVENT=y
> CONFIG_ACPI_AC=y
> CONFIG_ACPI_BATTERY=y
> CONFIG_ACPI_BUTTON=y
> CONFIG_ACPI_FAN=y
> # CONFIG_ACPI_DOCK is not set
> CONFIG_ACPI_PROCESSOR=y
> CONFIG_ACPI_HOTPLUG_CPU=y
> # CONFIG_ACPI_PROCESSOR_AGGREGATOR is not set
> CONFIG_ACPI_THERMAL=y
> # CONFIG_ACPI_CUSTOM_DSDT is not set
> CONFIG_ACPI_BLACKLIST_YEAR=0
> # CONFIG_ACPI_DEBUG is not set
> # CONFIG_ACPI_PCI_SLOT is not set
> CONFIG_ACPI_CONTAINER=y
> # CONFIG_ACPI_SBS is not set
> # CONFIG_ACPI_HED is not set
> # CONFIG_ACPI_APEI is not set
> CONFIG_X86_ACPI_CPUFREQ=y
> CONFIG_PNPACPI=y
> CONFIG_ATA_ACPI=y
> # CONFIG_PATA_ACPI is not set
> # ACPI drivers
> # ACPI drivers
> # CONFIG_SENSORS_ACPI_POWER is not set
>
> Suspend stuff:
> CONFIG_ARCH_SUSPEND_POSSIBLE=y
> CONFIG_SUSPEND=y
> CONFIG_SUSPEND_FREEZER=y
> CONFIG_USB_SUSPEND=y
>
>
> Here's output from my script (usbwakeup -l): (USB4 is where my keyboard is)
> ~ # usbwakeup -l
> Listing USB hubs/devices and their wakeup status...
>
> USB ID :: Device* :: Status :: Device Description
> ----------------------------------------------------
> 1d6b:0001 :: usb3 :: enabled :: UHCI Host Controller
> 051d:0002 :: 3-1 :: disabled :: Back-UPS RS 1200 FW:8.g1 .D USB FW:g1
> 1d6b:0001 :: usb4 :: enabled :: UHCI Host Controller
> 046d:c508 :: 4-1 :: disabled :: USB Receiver
> 046d:c221 :: 4-2.1 :: enabled :: Gaming Keyboard
> 1d6b:0001 :: usb5 :: enabled :: UHCI Host Controller
> 1d6b:0002 :: usb1 :: enabled :: EHCI Host Controller
> 1d6b:0001 :: usb6 :: enabled :: UHCI Host Controller
> 1d6b:0001 :: usb7 :: enabled :: UHCI Host Controller
> 1d6b:0001 :: usb8 :: enabled :: UHCI Host Controller
> 1d6b:0002 :: usb2 :: enabled :: EHCI Host Controller
>
> *Use the Device column to identify hubs/devices to be toggled.
>
> 11 USB hubs/devices listed.
>
> Output from acpitool -w:
> osoikaze ~ # acpitool -w
> Device S-state Status Sysfs node
> ---------------------------------------
> 1. P0P1 S3 *disabled pci:0000:00:01.0
> 2. UAR1 S3 *disabled pnp:00:03
> 3. P0P2 S4 *disabled pci:0000:00:1e.0
> 4. USB0 S3 *disabled pci:0000:00:1d.0
> 5. USB1 S3 *disabled pci:0000:00:1d.1
> 6. USB2 S3 *disabled pci:0000:00:1d.2
> 7. USB5 S3 *disabled
> 8. USB6 S3 *disabled pci:0000:00:1a.2
> 9. EUSB S3 *disabled pci:0000:00:1d.7
> 10. USB3 S3 *disabled pci:0000:00:1a.0
> 11. USB4 S3 *enabled pci:0000:00:1a.1
> 12. USBE S3 *disabled pci:0000:00:1a.7
> 13. PEX0 S4 *disabled pci:0000:00:1c.0
> 14. PEX1 S4 *disabled pci:0000:00:1c.1
> 15. PEX2 S4 *disabled pci:0000:00:1c.2
> 16. PEX3 S4 *disabled pci:0000:00:1c.3
> 17. PEX4 S4 *disabled pci:0000:00:1c.4
> 18. PEX5 S4 *disabled pci:0000:00:1c.5
> 19. SLPB S4 *enabled
> 20. PWRB S3 *enabled
>
> Does anyone have any idea what could be wrong here? This is driving me
> crazy, I hate shutting down my PC when I'm not using it. I could live
> without the keyboard if the damn power button would work, but even if
> I don't set the USB wakeup (through /proc/acpi/wakeup or my script) it
> still gets stuck in the 'eternal' sleep. The whole kernel config is at
> http://pastebin.com/2G9vWD0R
>
> The only thing I haven't tried yet is installing something like Ubuntu
> and see if it has the same problem.

I switched to 3.0 more than a year ago (I use vanilla-sources). Never
had a problem with suspend and/or hibernate; I'm now running kernel
3.5.3.

You didn't specify how do you suspend. pm-utils? dbus-send to upower?
echo mem > /sys/power/state?

I would recommend you to shut down X, and try pm-suspend from the
console. It may tell you more info.

Regards.
--
Canek Peláez Valdés
Posgrado en Ciencia e Ingeniería de la Computación
Universidad Nacional Autónoma de México

Chris Stankevitz 09-13-2012 04:49 AM

Update to newer kernel completely hoses suspend
 
On Wed, Sep 12, 2012 at 5:18 PM, Daniel Frey <djqfrey@gmail.com> wrote:
> So about a month ago I decided to update my kernel to the dreaded 3.x
> series. My old 2.6.x kernel ...

FYI Linus Torvalds says there was no change between 2.6 and 3.0. A quote:

So what are the big changes? NOTHING. Absolutely nothing. Sure, we
have the usual two thirds driver
changes, and a lot of random fixes, but the point is that 3.0 is
*just* about renumbering, we are very much *not* doing a KDE-4 or a
Gnome-3 here. No breakage, no special scary new features, nothing at
all like that.

You can read his entire letter here:
https://lkml.org/lkml/2011/5/29/204

Chris

Volker Armin Hemmann 09-13-2012 05:37 PM

Update to newer kernel completely hoses suspend
 
Am Mittwoch, 12. September 2012, 17:18:38 schrieb Daniel Frey:
> So about a month ago I decided to update my kernel to the dreaded 3.x
> series. My old 2.6.x kernel was working fine, but of course I decided
> to try to update it anyway, knowing there were problems with suspend
> and a few other things.
>
> I've always used gentoo-sources. So I tried 3.3.8.
>
> Hrm. Suspend doesn't work. I tried 3.4.5, 3.4.9 and 3.0.35 (older
> versions are no longer available.) If I'd known it would completely
> kill my suspend and make it useless, I wouldn't have bothered.
>
> Here's the problem:
>
> I can suspend fine. It appears to work. It powers off and goes into
> its suspend state. I press the space bar. Nothing. So, then I
> discovered that as of 3.2 USB wakeup had completely changed in the
> kernel, and you need to set hubs and devices in /proc/acpi/wakeup

I don't have to do that.

> (which is normally done for you) *and* in /sys/devices. No biggie, I
> wrote a script to do just that at

neither that. In fact, I have done nothing. It just works, with fglrx.

uname -a
Linux energy 3.4.10 #1 SMP Sun Sep 9 23:01:01 CEST 2012 x86_64 AMD Phenom(tm)
II X4 955 Processor AuthenticAMD GNU/Linux

and 3.2 before that

and 3.0 before that.

--
#163933

Daniel Frey 09-14-2012 02:15 AM

Update to newer kernel completely hoses suspend
 
On 09/13/2012 10:37 AM, Volker Armin Hemmann wrote:
> Am Mittwoch, 12. September 2012, 17:18:38 schrieb Daniel Frey:
>> So about a month ago I decided to update my kernel to the dreaded 3.x
>> series. My old 2.6.x kernel was working fine, but of course I decided
>> to try to update it anyway, knowing there were problems with suspend
>> and a few other things.
>>
>> I've always used gentoo-sources. So I tried 3.3.8.
>>
>> Hrm. Suspend doesn't work. I tried 3.4.5, 3.4.9 and 3.0.35 (older
>> versions are no longer available.) If I'd known it would completely
>> kill my suspend and make it useless, I wouldn't have bothered.
>>
>> Here's the problem:
>>
>> I can suspend fine. It appears to work. It powers off and goes into
>> its suspend state. I press the space bar. Nothing. So, then I
>> discovered that as of 3.2 USB wakeup had completely changed in the
>> kernel, and you need to set hubs and devices in /proc/acpi/wakeup
> I don't have to do that.

You're lucky then - I've googled the issue and it has hit a lot of users
from all sorts of distros, it's not specific to gentoo. Usually when the
keyboard won't wake up the PC you can use the power button, which didn't
work on my machine either.

I'm kind of suspecting the PSU now. It's getting worse, and it happens
on several kernel versions (even on 3.0.x which I had for about 3 weeks
before going to something > 3.2. When I powered up my machine today
(after being off all night) X bombed with no screens found. I restarted
xdm and it worked... I'm thinking if the PSU is off maybe it's taking a
while to warm up to be reliable - my video card uses one of those 6 port
extra power plugs. I'll bet after being on for a few minutes it was
fine. I hate diagnosing weird & intermittent hardware problems. The PSU
in here is way too much for what I've got in here anyway. Come to think
of it, I think I have a spare in my closet, although I can't remember
the wattage.


>
>> (which is normally done for you) *and* in /sys/devices. No biggie, I
>> wrote a script to do just that at
> neither that. In fact, I have done nothing. It just works, with fglrx.
>
> uname -a
> Linux energy 3.4.10 #1 SMP Sun Sep 9 23:01:01 CEST 2012 x86_64 AMD Phenom(tm)
> II X4 955 Processor AuthenticAMD GNU/Linux
>
> and 3.2 before that
>
> and 3.0 before that.
>
It worked with 3.0.x for a couple of weeks, all hell broke loose after I
upgraded to 3.3.8 (I think.) I've also tried 3.4.5 and 3.4.9, nothing
works. Sigh...

I'd better figure out a way to rule out the PSU. It's possible the
emerge world that I did at the same time I upgraded to 3.3.8 stressed
the PSU.

Dan

Daniel Frey 09-14-2012 02:17 AM

Update to newer kernel completely hoses suspend
 
On 09/12/2012 05:59 PM, Canek Peláez Valdés wrote:
> I switched to 3.0 more than a year ago (I use vanilla-sources). Never
> had a problem with suspend and/or hibernate; I'm now running kernel
> 3.5.3. You didn't specify how do you suspend. pm-utils? dbus-send to
> upower? echo mem > /sys/power/state? I would recommend you to shut
> down X, and try pm-suspend from the console. It may tell you more
> info. Regards.

I use KDE and use it to suspend. It's worked flawlessly for years now,
although I had to enable usb wakeup on one of the USB hubs for the
keyboard to wake.

The thing is, it goes to sleep fine. It goes into low power state &
shuts fans etc. off. The problem is, after leaving it for a while, you
can't wake it. I've tried waking immediately after suspending, and it
works. Leave it for a duration (like overnight) and neither the keyboard
or power button wakes it (as in absolutely nothing happens, it won't
even turn on/spin up fans etc.)

I think I might have a hardware issue now. I shut down last night and
this morning X wouldn't start right away - I turned it on and walked
away from it. Restarting xdm made it start, but it had been running for
> 5 minutes at that point.

Dan

Daniel Frey 09-14-2012 02:20 AM

Update to newer kernel completely hoses suspend
 
On 09/12/2012 09:49 PM, Chris Stankevitz wrote:
> On Wed, Sep 12, 2012 at 5:18 PM, Daniel Frey <djqfrey@gmail.com> wrote:
>> So about a month ago I decided to update my kernel to the dreaded 3.x
>> series. My old 2.6.x kernel ...
> FYI Linus Torvalds says there was no change between 2.6 and 3.0. A quote:
>
> So what are the big changes? NOTHING. Absolutely nothing. Sure, we
> have the usual two thirds driver
> changes, and a lot of random fixes, but the point is that 3.0 is
> *just* about renumbering, we are very much *not* doing a KDE-4 or a
> Gnome-3 here. No breakage, no special scary new features, nothing at
> all like that.
>
> You can read his entire letter here:
> https://lkml.org/lkml/2011/5/29/204
>
> Chris
When I updated, I knew about changes in 3.2 that affected USB keyboard
wake in suspend (& mostly how it deals with acpi. Most of the stuff
moved to /sys/devices, the normal /proc/acpi/wakeup didn't really do
anything.) This affected many users over many distros.

It also changed how lirc works, although that happened around 2.6.38??,
so my htpc frontend is still on 2.6.32. When I tried updating that
machine to 3.0, nothing worked and I spent about a day troubleshooting
it before I put the image I took of it before I upgraded it back on.

Dan

Daniel Frey 09-15-2012 04:28 PM

Update to newer kernel completely hoses suspend
 
Well, it turns out it was my PSU. The voltage drop on the 5V line was
4.08, but it would slowly warm up to 4.95V, then the PC would behave
normally. I opened the PSU and there was a ruptured cap.

I've replaced it and the problems are all gone.

I guess it was not really a coincidence that the failure happened after
a major update. This isn't the first time an `emerge -pvuDN world`
killed my computer. :-)

Dan

On 09/13/2012 07:20 PM, Daniel Frey wrote:
> On 09/12/2012 09:49 PM, Chris Stankevitz wrote:
>> On Wed, Sep 12, 2012 at 5:18 PM, Daniel Frey <djqfrey@gmail.com> wrote:
>>> So about a month ago I decided to update my kernel to the dreaded 3.x
>>> series. My old 2.6.x kernel ...
>> FYI Linus Torvalds says there was no change between 2.6 and 3.0. A quote:
>>
>> So what are the big changes? NOTHING. Absolutely nothing. Sure, we
>> have the usual two thirds driver
>> changes, and a lot of random fixes, but the point is that 3.0 is
>> *just* about renumbering, we are very much *not* doing a KDE-4 or a
>> Gnome-3 here. No breakage, no special scary new features, nothing at
>> all like that.
>>
>> You can read his entire letter here:
>> https://lkml.org/lkml/2011/5/29/204
>>
>> Chris
> When I updated, I knew about changes in 3.2 that affected USB keyboard
> wake in suspend (& mostly how it deals with acpi. Most of the stuff
> moved to /sys/devices, the normal /proc/acpi/wakeup didn't really do
> anything.) This affected many users over many distros.
>
> It also changed how lirc works, although that happened around 2.6.38??,
> so my htpc frontend is still on 2.6.32. When I tried updating that
> machine to 3.0, nothing worked and I spent about a day troubleshooting
> it before I put the image I took of it before I upgraded it back on.
>
> Dan

Mick 09-15-2012 10:26 PM

Update to newer kernel completely hoses suspend
 
On Saturday 15 Sep 2012 17:28:26 Daniel Frey wrote:
> Well, it turns out it was my PSU. The voltage drop on the 5V line was
> 4.08, but it would slowly warm up to 4.95V, then the PC would behave
> normally. I opened the PSU and there was a ruptured cap.
>
> I've replaced it and the problems are all gone.
>
> I guess it was not really a coincidence that the failure happened after
> a major update. This isn't the first time an `emerge -pvuDN world`
> killed my computer. :-)
>
> Dan
>
> On 09/13/2012 07:20 PM, Daniel Frey wrote:
> > On 09/12/2012 09:49 PM, Chris Stankevitz wrote:
> >> On Wed, Sep 12, 2012 at 5:18 PM, Daniel Frey <djqfrey@gmail.com> wrote:
> >>> So about a month ago I decided to update my kernel to the dreaded 3.x
> >>> series. My old 2.6.x kernel ...
> >>
> >> FYI Linus Torvalds says there was no change between 2.6 and 3.0. A
> >> quote:
> >>
> >> So what are the big changes? NOTHING. Absolutely nothing. Sure, we
> >> have the usual two thirds driver
> >> changes, and a lot of random fixes, but the point is that 3.0 is
> >> *just* about renumbering, we are very much *not* doing a KDE-4 or a
> >> Gnome-3 here. No breakage, no special scary new features, nothing at
> >> all like that.
> >>
> >> You can read his entire letter here:
> >> https://lkml.org/lkml/2011/5/29/204
> >>
> >> Chris
> >
> > When I updated, I knew about changes in 3.2 that affected USB keyboard
> > wake in suspend (& mostly how it deals with acpi. Most of the stuff
> > moved to /sys/devices, the normal /proc/acpi/wakeup didn't really do
> > anything.) This affected many users over many distros.
> >
> > It also changed how lirc works, although that happened around 2.6.38??,
> > so my htpc frontend is still on 2.6.32. When I tried updating that
> > machine to 3.0, nothing worked and I spent about a day troubleshooting
> > it before I put the image I took of it before I upgraded it back on.
> >
> > Dan

I was also replacing capacitors last weekend. It is a good idea to upgrade
them if there are alternatives of a higher maximum temperature as they will
probably last longer. A belts & braces approach is to add another/larger case
fan to keep the in-case temperatures lower.
--
Regards,
Mick

Dale 09-15-2012 10:29 PM

Update to newer kernel completely hoses suspend
 
Daniel Frey wrote:
> Well, it turns out it was my PSU. The voltage drop on the 5V line was
> 4.08, but it would slowly warm up to 4.95V, then the PC would behave
> normally. I opened the PSU and there was a ruptured cap.
>
> I've replaced it and the problems are all gone.
>
> I guess it was not really a coincidence that the failure happened after
> a major update. This isn't the first time an `emerge -pvuDN world`
> killed my computer. :-)
>
> Dan
>
>

*cough cough* Maybe you need a better or more powerful power supply?
If that cap went bad, you could have some others that are ready for the
same problem. I'd at least be on the look out for a new P/S. The next
one could go out and take a mobo or something with it. That would be
bad for sure.

Just a thought.

Dale

:-) :-)

--
I am only responsible for what I said ... Not for what you understood or how you interpreted my words!

Daniel Frey 09-16-2012 06:22 PM

Update to newer kernel completely hoses suspend
 
On 09/15/2012 03:26 PM, Mick wrote:
> I was also replacing capacitors last weekend. It is a good idea to upgrade
> them if there are alternatives of a higher maximum temperature as they will
> probably last longer. A belts & braces approach is to add another/larger case
> fan to keep the in-case temperatures lower.

Well, after I replaced the cap, I decided to get a better power supply
anyway. The one that died was an OCZ 1010 W supply, which is way too
much for this machine. I used to have 12 hard drives in here
(Thermaltake Full Armor), but a few years ago I built a server to do
those tasks. I replaced it with a platinum 650 W rated supply. The new
PSU looks like it's built a lot better.

I already have three 12 cm fans that are controlled by the BIOS. I've
had really poor luck with OCZ anything (to the point that I won't buy
anything they make anymore.)

Dan


All times are GMT. The time now is 10:14 AM.

VBulletin, Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO ©2007, Crawlability, Inc.