2008-11-05 22:04:47

by Bernhard Schmidt

[permalink] [raw]
Subject: [2.6.28-rc3] regression when switching TTY->X, input related?

Hi,

I have a pretty weird issue with 2.6.28-rc3 that I originally thought to
be related with suspend2ram, then with intel's xorg driver and now it
looks like something related to the input subsystem.

I have a Dell Latitude D630 (Intel GM965 graphics) running Ubuntu 8.10
amd64. Since the included (heavily patched by Ubuntu) 2.6.27 is
extremely unstable in iwlagn I've been trying my luck with 2.6.28-rc3.

One thing that did not work anymore was suspend2ram and resume. It would
suspend just fine and resume as well, but then I would sit in front of a
black screen with an unmovable mouse cursor and not be able to do
anything, not even switch vtys or kill X with ctrl-alt-backspace. Magic
Sysrq to reboot did usually work though.

I've been following a few threads about similar issues in here, among
others this one: http://lkml.org/lkml/2008/11/5/217 . However, the
workaround mentioned in there (switch to a textmode console prior to
suspend) didn't work, in fact when switching back to X without
suspending I had the same problem. After sitting there and scratching my
head for a couple of minutes the screensaver unlock dialog suddenly
appeared and I could enter my password and "work", but the touchpad
didn't work. I verified that a couple of times and it is always the
same, after switching to a textmode vty and back to X I have a black
screen with my mousecursor, I can neither move the cursor with the
touchpad nor any keypress has any effect, but the box isn't locked up.
There is a bit of disk activity, I can login using the serial console
just fine and immediately kill and restart X and it comes back fine, in
seldom cases without a working touchpad, but that might be a different
issue. After about 30-120s the situation resolves itself and I get my
normal X, but again sometimes without touchpad.

I think this is an input issue because I get the following messages in
/var/log/Xorg.0.log when switching away:

(II) AIGLX: Suspending AIGLX clients for VT switch
(II) intel(0): xf86UnbindGARTMemory: unbind key 0
(II) intel(0): xf86UnbindGARTMemory: unbind key 1

when switching back

(II) Open ACPI successful (/var/run/acpid.socket)
(II) AIGLX: Resuming AIGLX clients after VT switch
(II) intel(0): xf86BindGARTMemory: bind key 0 at 0x0c6b0000 (pgoffset 50864)
(II) intel(0): xf86BindGARTMemory: bind key 1 at 0x0e000000 (pgoffset 57344)
(II) intel(0): Fixed memory allocation layout:
(II) intel(0): 0x00000000-0x00000fff: power context (4 kB)
(II) intel(0): 0x0077f000: end of stolen memory
(II) intel(0): 0x0077f000-0x0c6affff: DRI memory manager (195780 kB)
(II) intel(0): 0x0c6b0000-0x0dffffff: exa offscreen (25920 kB)
(II) intel(0): 0x0e000000-0x0fffffff: classic textures (32768 kB)
(II) intel(0): 0x10000000: end of aperture
(II) intel(0): BO memory allocation layout:
(II) intel(0): 0x0077f000: start of memory manager
(II) intel(0): 0x0079f000-0x0100efff: depth buffer (8640 kB) Y tiled
(II) intel(0): 0x0179f000-0x0200efff: back buffer (8640 kB) X tiled
(II) intel(0): 0x02800000-0x0306ffff: front buffer (8640 kB) X tiled
(II) intel(0): 0x0279f000-0x0279ffff: overlay registers (4 kB)
(II) intel(0): 0x027a0000-0x027b5fff: exa G965 state buffer (88 kB)
(II) intel(0): 0x027c0000-0x027c7fff: logical 3D context (32 kB)
(II) intel(0): 0x027c8000-0x027d1fff: HW cursors (40 kB)
(II) intel(0): 0x0c6b0000: end of memory manager
(WW) intel(0): ESR is 0x00000001
(WW) intel(0): PRB0_CTL (0x0001f001) indicates ring buffer enabled
(WW) intel(0): Existing errors found in hardware state.
(II) intel(0): using SSC reference clock of 100 MHz
(II) intel(0): Selecting standard 18 bit TMDS pixel format.
(II) intel(0): Output configuration:
(II) intel(0): Pipe A is off
(II) intel(0): Display plane A is now disabled and connected to pipe A.
(II) intel(0): Pipe B is on
(II) intel(0): Display plane B is now enabled and connected to pipe B.
(II) intel(0): Output VGA is connected to pipe none
(II) intel(0): Output LVDS is connected to pipe B
(II) intel(0): Output TMDS-1 is connected to pipe none
(II) intel(0): Output TV is connected to pipe none
(II) intel(0): [drm] mapped front buffer at 0xe2800000, handle = 0x2fff8000
(II) Configured Mouse: x-axis range 0 - 1023
(II) Configured Mouse: y-axis range 0 - 767
(--) Configured Mouse touchpad found
(II) AlpsPS/2 ALPS GlidePoint: x-axis range 0 - 1023
(II) AlpsPS/2 ALPS GlidePoint: y-axis range 0 - 767
(WW) AlpsPS/2 ALPS GlidePoint can't grab event device, errno=16
(--) AlpsPS/2 ALPS GlidePoint touchpad found

then it hangs some seconds/minutes with the usual symptoms. When I move
my finger on the touchpad in this stage I get my xorg log flooded with

[mi] EQ overflowing. The server is probably stuck in an infinite loop.
[mi] mieqEnequeue: out-of-order valuator event; dropping.

And then it suddenly shows me the login screen, and at the same time the
Xorg logfile gets the following lines added (at once, so there is a long
delay right before the following line)

(II) PS/2 Mouse: Device reopened after 10 attempts.
(**) PS/2 Mouse: YAxisMapping: buttons 4 and 5
(**) PS/2 Mouse: EmulateWheelButton: 4, EmulateWheelInertia: 10, EmulateWheelTimeout: 200
(II) AT Translated Set 2 keyboard: Device reopened after 10 attempts.
(II) Video Bus: Device reopened after 10 attempts.
(II) Video Bus: Device reopened after 10 attempts.
(II) Video Bus: Device reopened after 10 attempts.
(II) intel(0): EDID vendor "SEC", prod id 17495
(II) intel(0): Printing DDC gathered Modelines:
(II) intel(0): Modeline "1440x900"x0.0 105.06 1440 1488 1616 1920 900 906 909 912 -hsync -vsync (54.7 kHz)

I think this is a regression because it used to work with the 2.6.27
kernel from Ubuntu. I'll try with a vanilla 2.6.27 but I had been
running ext4dev which I need to work a bit to mount on 2.6.27 (it got
flagged as ext4 without test_fs flag the first time 2.6.28-rc was
booting). It's not really related to suspend, because it happens on VTY
switching as well. It happens with both Intel xorg driver 2.4.1 and
2.5.0. I think it's related to input because the first message after it
starts working again is always related to the mouse.

Anyone having an idea where to look? dmesg and config attached.

Bernhard


Attachments:
(No filename) (6.34 kB)
dmesg.txt (45.02 kB)
config (77.06 kB)
Download all attachments

2008-11-06 04:15:34

by Andrey Borzenkov

[permalink] [raw]
Subject: Re: [2.6.28-rc3] regression when switching TTY->X, input related?

Bernhard Schmidt wrote:

> Hi,
>
> I have a pretty weird issue with 2.6.28-rc3 that I originally thought to
> be related with suspend2ram, then with intel's xorg driver and now it
> looks like something related to the input subsystem.
>
> I have a Dell Latitude D630 (Intel GM965 graphics) running Ubuntu 8.10
> amd64. Since the included (heavily patched by Ubuntu) 2.6.27 is
> extremely unstable in iwlagn I've been trying my luck with 2.6.28-rc3.
>
> One thing that did not work anymore was suspend2ram and resume. It would
> suspend just fine and resume as well, but then I would sit in front of a
> black screen with an unmovable mouse cursor and not be able to do
> anything, not even switch vtys or kill X with ctrl-alt-backspace. Magic
> Sysrq to reboot did usually work though.
>
> I've been following a few threads about similar issues in here, among
> others this one: http://lkml.org/lkml/2008/11/5/217 . However, the
> workaround mentioned in there (switch to a textmode console prior to
> suspend) didn't work, in fact when switching back to X without
> suspending I had the same problem. After sitting there and scratching my
> head for a couple of minutes the screensaver unlock dialog suddenly
> appeared and I could enter my password and "work", but the touchpad
> didn't work.

Cannot comment on input stuff, but I noticed that you are using driver
that requests firmware:

[ 39.735183] iwlagn 0000:0c:00.0: PCI INT A -> GSI 17 (level, low) -> IRQ
17
[ 39.735272] iwlagn 0000:0c:00.0: restoring config space at offset 0x1
(was 0x100102, writing 0x100106)
[ 39.735379] iwlagn 0000:0c:00.0: irq 42 for MSI/MSI-X
[ 39.735448] iwlagn 0000:0c:00.0: firmware: requesting
iwlwifi-4965-2.ucode

If it tries to do it during resume in ->resume method, this won't work and
will timeout; symptoms are exactly as you describe (sitting there for 180
seconds and suddenly start to work). Do you have WiFi connection on resume?

> I verified that a couple of times and it is always the
> same, after switching to a textmode vty and back to X I have a black
> screen with my mousecursor, I can neither move the cursor with the
> touchpad nor any keypress has any effect, but the box isn't locked up.
> There is a bit of disk activity, I can login using the serial console
> just fine and immediately kill and restart X and it comes back fine, in
> seldom cases without a working touchpad, but that might be a different
> issue. After about 30-120s the situation resolves itself and I get my
> normal X, but again sometimes without touchpad.
>
> I think this is an input issue because I get the following messages in
> /var/log/Xorg.0.log when switching away:
>
> (II) AIGLX: Suspending AIGLX clients for VT switch
> (II) intel(0): xf86UnbindGARTMemory: unbind key 0
> (II) intel(0): xf86UnbindGARTMemory: unbind key 1
>
> when switching back
>
> (II) Open ACPI successful (/var/run/acpid.socket)
> (II) AIGLX: Resuming AIGLX clients after VT switch
> (II) intel(0): xf86BindGARTMemory: bind key 0 at 0x0c6b0000 (pgoffset
> 50864) (II) intel(0): xf86BindGARTMemory: bind key 1 at 0x0e000000
> (pgoffset 57344) (II) intel(0): Fixed memory allocation layout:
> (II) intel(0): 0x00000000-0x00000fff: power context (4 kB)
> (II) intel(0): 0x0077f000: end of stolen memory
> (II) intel(0): 0x0077f000-0x0c6affff: DRI memory manager (195780 kB)
> (II) intel(0): 0x0c6b0000-0x0dffffff: exa offscreen (25920 kB)
> (II) intel(0): 0x0e000000-0x0fffffff: classic textures (32768 kB)
> (II) intel(0): 0x10000000: end of aperture
> (II) intel(0): BO memory allocation layout:
> (II) intel(0): 0x0077f000: start of memory manager
> (II) intel(0): 0x0079f000-0x0100efff: depth buffer (8640 kB) Y tiled
> (II) intel(0): 0x0179f000-0x0200efff: back buffer (8640 kB) X tiled
> (II) intel(0): 0x02800000-0x0306ffff: front buffer (8640 kB) X tiled
> (II) intel(0): 0x0279f000-0x0279ffff: overlay registers (4 kB)
> (II) intel(0): 0x027a0000-0x027b5fff: exa G965 state buffer (88 kB)
> (II) intel(0): 0x027c0000-0x027c7fff: logical 3D context (32 kB)
> (II) intel(0): 0x027c8000-0x027d1fff: HW cursors (40 kB)
> (II) intel(0): 0x0c6b0000: end of memory manager
> (WW) intel(0): ESR is 0x00000001
> (WW) intel(0): PRB0_CTL (0x0001f001) indicates ring buffer enabled
> (WW) intel(0): Existing errors found in hardware state.
> (II) intel(0): using SSC reference clock of 100 MHz
> (II) intel(0): Selecting standard 18 bit TMDS pixel format.
> (II) intel(0): Output configuration:
> (II) intel(0): Pipe A is off
> (II) intel(0): Display plane A is now disabled and connected to pipe A.
> (II) intel(0): Pipe B is on
> (II) intel(0): Display plane B is now enabled and connected to pipe B.
> (II) intel(0): Output VGA is connected to pipe none
> (II) intel(0): Output LVDS is connected to pipe B
> (II) intel(0): Output TMDS-1 is connected to pipe none
> (II) intel(0): Output TV is connected to pipe none
> (II) intel(0): [drm] mapped front buffer at 0xe2800000, handle =
> 0x2fff8000 (II) Configured Mouse: x-axis range 0 - 1023
> (II) Configured Mouse: y-axis range 0 - 767
> (--) Configured Mouse touchpad found
> (II) AlpsPS/2 ALPS GlidePoint: x-axis range 0 - 1023
> (II) AlpsPS/2 ALPS GlidePoint: y-axis range 0 - 767
> (WW) AlpsPS/2 ALPS GlidePoint can't grab event device, errno=16
> (--) AlpsPS/2 ALPS GlidePoint touchpad found
>
> then it hangs some seconds/minutes with the usual symptoms. When I move
> my finger on the touchpad in this stage I get my xorg log flooded with
>
> [mi] EQ overflowing. The server is probably stuck in an infinite loop.
> [mi] mieqEnequeue: out-of-order valuator event; dropping.
>
> And then it suddenly shows me the login screen, and at the same time the
> Xorg logfile gets the following lines added (at once, so there is a long
> delay right before the following line)
>
> (II) PS/2 Mouse: Device reopened after 10 attempts.
> (**) PS/2 Mouse: YAxisMapping: buttons 4 and 5
> (**) PS/2 Mouse: EmulateWheelButton: 4, EmulateWheelInertia: 10,
> EmulateWheelTimeout: 200 (II) AT Translated Set 2 keyboard: Device
> reopened after 10 attempts. (II) Video Bus: Device reopened after 10
> attempts. (II) Video Bus: Device reopened after 10 attempts.
> (II) Video Bus: Device reopened after 10 attempts.
> (II) intel(0): EDID vendor "SEC", prod id 17495
> (II) intel(0): Printing DDC gathered Modelines:
> (II) intel(0): Modeline "1440x900"x0.0 105.06 1440 1488 1616 1920 900
> 906 909 912 -hsync -vsync (54.7 kHz)
>
> I think this is a regression because it used to work with the 2.6.27
> kernel from Ubuntu. I'll try with a vanilla 2.6.27 but I had been
> running ext4dev which I need to work a bit to mount on 2.6.27 (it got
> flagged as ext4 without test_fs flag the first time 2.6.28-rc was
> booting). It's not really related to suspend, because it happens on VTY
> switching as well. It happens with both Intel xorg driver 2.4.1 and
> 2.5.0. I think it's related to input because the first message after it
> starts working again is always related to the mouse.
>
> Anyone having an idea where to look? dmesg and config attached.
>
> Bernhard

2008-11-06 11:34:29

by Bernhard Schmidt

[permalink] [raw]
Subject: Re: [2.6.28-rc3] regression when switching TTY->X, input related?

On Thu, Nov 06, 2008 at 07:15:11AM +0300, Andrey Borzenkov wrote:

> > I have a pretty weird issue with 2.6.28-rc3 that I originally thought to
> > be related with suspend2ram, then with intel's xorg driver and now it
> > looks like something related to the input subsystem.
> >
> > I have a Dell Latitude D630 (Intel GM965 graphics) running Ubuntu 8.10
> > amd64. Since the included (heavily patched by Ubuntu) 2.6.27 is
> > extremely unstable in iwlagn I've been trying my luck with 2.6.28-rc3.
> >
> > One thing that did not work anymore was suspend2ram and resume. It would
> > suspend just fine and resume as well, but then I would sit in front of a
> > black screen with an unmovable mouse cursor and not be able to do
> > anything, not even switch vtys or kill X with ctrl-alt-backspace. Magic
> > Sysrq to reboot did usually work though.
> >
> > I've been following a few threads about similar issues in here, among
> > others this one: http://lkml.org/lkml/2008/11/5/217 . However, the
> > workaround mentioned in there (switch to a textmode console prior to
> > suspend) didn't work, in fact when switching back to X without
> > suspending I had the same problem. After sitting there and scratching my
> > head for a couple of minutes the screensaver unlock dialog suddenly
> > appeared and I could enter my password and "work", but the touchpad
> > didn't work.
>
> Cannot comment on input stuff, but I noticed that you are using driver
> that requests firmware:
>
> [ 39.735183] iwlagn 0000:0c:00.0: PCI INT A -> GSI 17 (level, low) -> IRQ
> 17
> [ 39.735272] iwlagn 0000:0c:00.0: restoring config space at offset 0x1
> (was 0x100102, writing 0x100106)
> [ 39.735379] iwlagn 0000:0c:00.0: irq 42 for MSI/MSI-X
> [ 39.735448] iwlagn 0000:0c:00.0: firmware: requesting
> iwlwifi-4965-2.ucode
>
> If it tries to do it during resume in ->resume method, this won't work and
> will timeout; symptoms are exactly as you describe (sitting there for 180
> seconds and suddenly start to work). Do you have WiFi connection on resume?

Yes. In fact the WiFi connection is reestablished immediately after
resume (with NetworkManager) and I can ping the box through WiFI, while
X hangs and waits for something.

Bernhard

2008-11-09 14:01:56

by Pavel Machek

[permalink] [raw]
Subject: Re: [2.6.28-rc3] regression when switching TTY->X, input related?

Hi!


> I think this is a regression because it used to work with the 2.6.27
> kernel from Ubuntu. I'll try with a vanilla 2.6.27 but I had been

Yes, please try that. It is possible that ubuntu has some patches in
and that their userland depens on them...


--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

2008-11-09 20:35:36

by Bernhard Schmidt

[permalink] [raw]
Subject: Re: [2.6.28-rc3] regression when switching TTY->X, input related?

On Sun, Nov 09, 2008 at 02:51:46PM +0100, Pavel Machek wrote:

> > I think this is a regression because it used to work with the 2.6.27
> > kernel from Ubuntu. I'll try with a vanilla 2.6.27 but I had been
> Yes, please try that. It is possible that ubuntu has some patches in
> and that their userland depens on them...

Will do, it might take me a while to figure out how to boot that ext4
filesystem with ext4dev again :-)

The (most probably) very same bug has been reported in
http://bugzilla.kernel.org/show_bug.cgi?id=11947 and says
"Latest working kernel version: 2.6.27.4".

Bernhard

2008-11-09 21:07:28

by Bernhard Schmidt

[permalink] [raw]
Subject: Re: [2.6.28-rc3] regression when switching TTY->X, input related?

On Sun, Nov 09, 2008 at 09:35:21PM +0100, Bernhard Schmidt wrote:

> > > I think this is a regression because it used to work with the 2.6.27
> > > kernel from Ubuntu. I'll try with a vanilla 2.6.27 but I had been
> > Yes, please try that. It is possible that ubuntu has some patches in
> > and that their userland depens on them...
> Will do, it might take me a while to figure out how to boot that ext4
> filesystem with ext4dev again :-)

Confirmed, vanilla 2.6.27.4 works just fine, both for switching and the
full suspend/resume cycle, so it is really a regression in 2.6.28-rc

The other bugzilla entry suggested something related to the vblank
events.

Bernhard