[Bug 500525] New: A key has to be pressed after hibernation in order to activate X.org server (Intel 945GM)
http://bugzilla.novell.com/show_bug.cgi?id=500525 Summary: A key has to be pressed after hibernation in order to activate X.org server (Intel 945GM) Classification: openSUSE Product: openSUSE 11.1 Version: Final Platform: x86-64 OS/Version: openSUSE 11.1 Status: NEW Severity: Normal Priority: P5 - None Component: X.Org AssignedTo: bnc-team-xorg-bugs@forge.provo.novell.com ReportedBy: nice@titanic.nyme.hu QAContact: xorg-maintainer-bugs@forge.provo.novell.com Found By: --- User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; hu-HU; rv:1.9.0.9) Gecko/2009041500 SUSE/3.0.9-0.1.1 Firefox/3.0.9 After resuming from a suspend to disk, the display is totally black (but not turned of), it doesn't show the screen saver, and is unaware of any mouse movements. I have to press a key on the keyboard, then suddenly the text console appears for a moment, and the my X.org server resumes to normal operation. It only happens with s2disk, never with s2ram. BTW, I'm using KDE 4.2.2 and compiz as display manager. Reproducible: Always Steps to Reproduce: 1. Initate a suspend to disk in KDE4's power devil applet 2. Wait for shutting down 3. Turn on the machine Actual Results: X.org won't work after a key is pressed. Expected Results: It should work immediately. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=500525
User nice@titanic.nyme.hu added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c1
--- Comment #1 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=500525
User nice@titanic.nyme.hu added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c2
--- Comment #2 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=500525
Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=500525
Mark Gray
http://bugzilla.novell.com/show_bug.cgi?id=500525
User sndirsch@novell.com added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c3
Stefan Dirsch
http://bugzilla.novell.com/show_bug.cgi?id=500525
User nice@titanic.nyme.hu added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c4
Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=500525
User sndirsch@novell.com added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c5
Stefan Dirsch
http://bugzilla.novell.com/show_bug.cgi?id=500525
User nice@titanic.nyme.hu added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c6
--- Comment #6 from Tamás Németh
Ah. mouse movement does not wake up the screen. Sorry, I didn't read your initial commment carefully. Then it's unlikely that it is DPMS stepping in.
In fact, the display backlight is ON, but the display shows a blank, black screen. Instead of this, it should show the screensaver IMMEDIATELY, as it does when returning from a suspend to RAM (As I've said this problem only happens after suspend to disk). Also it's annoying, that I can start the X server by the keyboard, while it doesn't seem to be able to sense the mouse. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=500525
User nice@titanic.nyme.hu added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c7
--- Comment #7 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=500525
User sndirsch@novell.com added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c8
Stefan Dirsch
http://bugzilla.novell.com/show_bug.cgi?id=500525
User nice@titanic.nyme.hu added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c9
--- Comment #9 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=500525
User seife@novell.com added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c10
Stefan Seyfried
http://bugzilla.novell.com/show_bug.cgi?id=500525
User nice@titanic.nyme.hu added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c11
--- Comment #11 from Tamás Németh
http://bugzilla.novell.com/show_bug.cgi?id=500525
User rjw@novell.com added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c12
--- Comment #12 from Rafael Wysocki
Rafael: do we still have nohz problems so that we need to press a key on resume to make the machine continue?
I believe we do on some (mostly firmware-broken) systems. It could be worth trying to force a specific clock source with the clocksource= kernel command line option and see if that changes the symptoms.
Holger: does the libdirectfb stuff trigger anything for you? Have you seen something like that already? Maybe it's time to runs s2disk under valgrind ;)
I think it would be easier to just try 'echo disk > /sys/power/state' on the affected box and see if the problem can be reproduced. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=500525
User rjw@novell.com added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c13
--- Comment #13 from Rafael Wysocki
Now I realized that on Xen kernel neither s2dsik no s2ram does work.
I don't really think they are supported with Xen. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=500525
User seife@novell.com added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c14
--- Comment #14 from Stefan Seyfried
Holger: does the libdirectfb stuff trigger anything for you? Have you seen something like that already? Maybe it's time to runs s2disk under valgrind ;)
I think it would be easier to just try 'echo disk > /sys/power/state' on the affected box and see if the problem can be reproduced.
The libdirectfb stuff comes from splashy, so no, this won't happen with in-kernel-suspend ;) -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=500525
User nice@titanic.nyme.hu added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c15
--- Comment #15 from Tamás Németh
(In reply to comment #11)
Now I realized that on Xen kernel neither s2dsik no s2ram does work.
I don't really think they are supported with Xen.
Yes, you may be right. Although Xen documentation seems to be outdated or nonexistent, it seems to me that there is no suspend/hibernate support in dom0: http://wiki.xensource.com/xenwiki/XenFaq#head-cb7861c9865fc8fbae3bb472f04c5d... But why down then KDE think that it will be able to do an s2ram? Is it a KDE or a kernel bug? I adore your efforts and this last kernel seems to be very cool, but maybe somehow misguides KDE? -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=500525
User rjw@novell.com added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c16
--- Comment #16 from Rafael Wysocki
(In reply to comment #12)
Holger: does the libdirectfb stuff trigger anything for you? Have you seen something like that already? Maybe it's time to runs s2disk under valgrind ;)
I think it would be easier to just try 'echo disk > /sys/power/state' on the affected box and see if the problem can be reproduced.
The libdirectfb stuff comes from splashy, so no, this won't happen with in-kernel-suspend ;)
Well, that's my point. :-) Why do you think the problem is with libdirectfb and not the kernel's internal VT switching? -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=500525
User rjw@novell.com added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c17
--- Comment #17 from Rafael Wysocki
(In reply to comment #13)
(In reply to comment #11)
Now I realized that on Xen kernel neither s2dsik no s2ram does work.
I don't really think they are supported with Xen.
Yes, you may be right. Although Xen documentation seems to be outdated or nonexistent, it seems to me that there is no suspend/hibernate support in dom0:
http://wiki.xensource.com/xenwiki/XenFaq#head-cb7861c9865fc8fbae3bb472f04c5d...
From what I can tell, the support for suspend to RAM for Xen dom0 is being worked on, so there's a chance it will be there in future kernel versions.
But why down then KDE think that it will be able to do an s2ram? Is it a KDE or a kernel bug? I adore your efforts and this last kernel seems to be very cool, but maybe somehow misguides KDE?
I don't really know what method is used by KDE to check whether or not suspend to RAM is supported. It is likely, though, that the kernel reports support for suspend to RAM via /sys/power/state and that's what makes KDE think S2RAM is supported. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=500525
User nice@titanic.nyme.hu added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c18
--- Comment #18 from Tamás Németh
(In reply to comment #10)
Rafael: do we still have nohz problems so that we need to press a key on resume to make the machine continue?
I think it would be easier to just try 'echo disk > /sys/power/state' on the affected box and see if the problem can be reproduced.
/sys/power/state' worked. (I tried both methods only once.) I didn't observe
I've tried that, and worked. Both 'echo disk > /sys/power/state' and 'echo mem this actual problem (a key has to be pressed...) when down s2disk by kernel, but: -It was a bit annoying that I saw no feedback doring the suspend process -It was a bit scary to see that upon restart I was able to choose any entry from the grub menu. What if I chose a different kernel? Or will even the kernel be overwritten by the saved image? -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=500525
User nice@titanic.nyme.hu added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c19
--- Comment #19 from Tamás Németh
(In reply to comment #15)
(In reply to comment #13)
(In reply to comment #11)
Now I realized that on Xen kernel neither s2dsik no s2ram does work.
I don't really think they are supported with Xen.
Yes, you may be right. Although Xen documentation seems to be outdated or nonexistent, it seems to me that there is no suspend/hibernate support in dom0:
http://wiki.xensource.com/xenwiki/XenFaq#head-cb7861c9865fc8fbae3bb472f04c5d...
From what I can tell, the support for suspend to RAM for Xen dom0 is being worked on, so there's a chance it will be there in future kernel versions.
But why down then KDE think that it will be able to do an s2ram? Is it a KDE or a kernel bug? I adore your efforts and this last kernel seems to be very cool, but maybe somehow misguides KDE?
I don't really know what method is used by KDE to check whether or not suspend to RAM is supported. It is likely, though, that the kernel reports support for suspend to RAM via /sys/power/state and that's what makes KDE think S2RAM is supported.
I've reported it here: https://bugzilla.novell.com/show_bug.cgi?id=502193 -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=500525
User nice@titanic.nyme.hu added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c20
--- Comment #20 from Tamás Németh
(In reply to comment #14)
(In reply to comment #12)
Holger: does the libdirectfb stuff trigger anything for you? Have you seen something like that already? Maybe it's time to runs s2disk under valgrind ;)
I think it would be easier to just try 'echo disk > /sys/power/state' on the affected box and see if the problem can be reproduced.
The libdirectfb stuff comes from splashy, so no, this won't happen with in-kernel-suspend ;)
Well, that's my point. :-) Why do you think the problem is with libdirectfb and not the kernel's internal VT switching?
VT switching works well otherwise except this: https://bugs.freedesktop.org/show_bug.cgi?id=21592 -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=500525
User hmacht@novell.com added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c21
Holger Macht
Holger: does the libdirectfb stuff trigger anything for you? Have you seen
It seems to be segfaulting or worse :-(
something like that already? Maybe it's time to runs s2disk under valgrind ;)
No, please don't ;-) Tamás, just to make sure, please set "splash = n" in /etc/suspend.conf and try again with s2disk. You shouldn't see this issue. If so, it's a splashy bug. But to warn you in advance, I guess we won't have the time to fix splashy for this issue for openSUSE 11.1 when a workaround exists. Sorry. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=500525
User nice@titanic.nyme.hu added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c22
Tamás Németh
set "splash = n" in /etc/suspend.conf and try again with s2disk. You shouldn't see this issue. If so, it's a splashy bug.
It WORKED!
But to warn you in advance, I guess we won't have the time to fix splashy for this issue for openSUSE 11.1 when a workaround exists. Sorry.
No problem. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
http://bugzilla.novell.com/show_bug.cgi?id=500525
User hmacht@novell.com added comment
http://bugzilla.novell.com/show_bug.cgi?id=500525#c23
Holger Macht
participants (1)
-
bugzilla_noreply@novell.com