Bug#832484: xserver-xorg-core: modesetting on some Intel hw causes hangs on xscreensaver (and on VT switches)

2016-08-01 Thread Francesco Poli
Control: reassign -1 libdrm-intel1 2.4.68-1


On Thu, 28 Jul 2016 23:06:47 +0200 Francesco Poli wrote:

[...]
> Hence, I think it's definitely a Linux kernel bug.
> I am consequently reassigning the bug report.
[...]

Update: after today's upgrade, I tried to boot again with the current
Debian testing Linux kernel:

  $ uname -srvmo
  Linux 4.6.0-1-amd64 #1 SMP Debian 4.6.4-1 (2016-07-18) x86_64 GNU/Linux

The issues I reported (kern.log traces caused by VT switches, non-working
OpenGL applications) seem to have vanished.

By looking at /var/log/aptitude, I think that the upgrade that fixed
things up is:

  [UPGRADE] libdrm-intel1:amd64 2.4.68-1 -> 2.4.70-1

I am consequently reassigning the bug report to package libdrm-intel1,
version 2.4.68-1 .

Please feel free to close this bug report as fixed in version
2.4.70-1 of libdrm-intel1 ...
If nobody objects in the next 30 days, I will close it myself.

Thanks for your time.
Bye.


-- 
 http://www.inventati.org/frx/
 There's not a second to spare! To the laboratory!
. Francesco Poli .
 GnuPG key fpr == CA01 1147 9CD2 EFDF FB82  3925 3E1C 27E1 1F69 BFFE


pgpqQ_GGECsE2.pgp
Description: PGP signature


Bug#832484: xserver-xorg-core: modesetting on some Intel hw causes hangs on xscreensaver (and on VT switches)

2016-07-28 Thread Francesco Poli
Control: reassign -1 linux 4.6.4-1


On Tue, 26 Jul 2016 18:47:40 +0200 Francesco Poli wrote:

[...]
> After restarting the X session, I found out that the same issues happen
> even with the "intel" X video driver.

Update: I cannot understand why the behavior changed, but I am now
completely unable run OpenGL applications (with either modesetting or
intel X video driver).
Any program that is supposed to display something via OpenGL (such as
GL screensavers from package xscreensaver-gl or even glxgears from
package mesa-utils) simply hangs and displays nothing or a static image.

> Could these issues depend on the kernel upgrade that was performed on
> the box along with the xserver-xorg-core upgrade?
> 
>   [UPGRADE] linux-image-4.6.0-1-amd64:amd64 4.6.3-1 -> 4.6.4-1
> 
> I am Cc-ing the Debian Kernel Team about this.
> 
> Dear Kernel Team, should bug #832484 be reassigned to source package
> linux/4.6.4-1 ?

I tried to boot with a previous kernel:

  $ uname -srvmo
  Linux 4.5.0-2-amd64 #1 SMP Debian 4.5.5-1 (2016-05-29) x86_64 GNU/Linux

and all the issues vanished (no more traces on kern.log, no more OpenGL
application hangs).

Hence, I think it's definitely a Linux kernel bug.
I am consequently reassigning the bug report.

Dear Kernel Team, please investigate this bug and fix it and/or forward
my bug report upstream, as appropriate.

Thanks for your time.


-- 
 http://www.inventati.org/frx/
 There's not a second to spare! To the laboratory!
. Francesco Poli .
 GnuPG key fpr == CA01 1147 9CD2 EFDF FB82  3925 3E1C 27E1 1F69 BFFE


pgpIaU8gHSDPk.pgp
Description: PGP signature


Bug#832484: xserver-xorg-core: modesetting on some Intel hw causes hangs on xscreensaver (and on VT switches)

2016-07-26 Thread Francesco Poli
On Mon, 25 Jul 2016 23:55:49 +0200 Francesco Poli (wintermute) wrote:

[...]
> I found a number of kernel traces in /var/log/kern.log (attached).
> The times showed that the first traces were experienced well before
> I came back to the box, while the last ones were obtained during
> my attempts to re-gain control of my X session...
> 
> I also noticed that switching VT back and forth (without a running
> screensaver or a locked display) generates new traces in
> /var/log/kern.log ...
[...]

Update: I tried to force the use of the "intel" X video driver with

  $ cat /etc/X11/xorg.conf 
  Section "Device"
  Identifier "Intel"
  Driver "intel"
  #   Option "AccelMethod" "uxa"
  EndSection

After restarting the X session, I found out that the same issues happen
even with the "intel" X video driver.
Could these issues depend on the kernel upgrade that was performed on
the box along with the xserver-xorg-core upgrade?

  [UPGRADE] linux-image-4.6.0-1-amd64:amd64 4.6.3-1 -> 4.6.4-1

I am Cc-ing the Debian Kernel Team about this.

Dear Kernel Team, should bug #832484 be reassigned to source package
linux/4.6.4-1 ?

Please let me know, thanks for your time.


-- 
 http://www.inventati.org/frx/
 There's not a second to spare! To the laboratory!
. Francesco Poli .
 GnuPG key fpr == CA01 1147 9CD2 EFDF FB82  3925 3E1C 27E1 1F69 BFFE


pgpLeFEn_ENsA.pgp
Description: PGP signature


Bug#832484: xserver-xorg-core: modesetting on some Intel hw causes hangs on xscreensaver (and on VT switches)

2016-07-25 Thread Francesco Poli (wintermute)
Package: xserver-xorg-core
Version: 2:1.18.4-1
Severity: normal

Hello X Strike Force,
thanks for taking the step to use the modesetting X video driver by default
on most Intel graphics hardware. I hope this move will help in solving
the issues that many users have been facing in recent times with the Intel
X video driver.


I have been giving this modesetting driver a try since this morning.
Unfortunately, I am experiencing some new issues on one box (these
issues do not seem to show on another box with more recent Intel
graphics hardware...).

After leaving the display locked (with xscreensaver) for the lunch
break, I came back to the box and found the screen in dpmsOff state
(as usual), I pressed [Shift] on the keyboard in order to wake up
the display (as usual) and found a hanged screensaver (as has often
happened in the last year or two, with the Intel X video driver).

By "hanged screensaver" I mean that you see a static image (one of the
frames generated by the screensaver under consideration) which is
completely frozen and, if you hit a key, you do not get the usual
login dialog window, but you instead see the mouse pointer which
you can move inside an invisible rectangle (exactly where the login
dialog window is supposed to be displayed).

For a long time, my usual workaround has been the following:
  * switch to another virtual terminal (VT) with [Ctrl+Alt+F2]
  * switch back to the original VT with [Ctrl+Alt+F1] (I use startx
from tty1, with no display manager)
After this workaround, everything used to come back to normal (another
screensaver running, login dialog window, entering password unlocks the
display, and so forth...).

A very different behavior was exhibited by the modesetting driver
on the box under consideration.

The VT switch did not produce any visible result. My attempt to switch
back to the original VT produced a completely frozen screensaver:
static image, no login dialog, no mouse pointer, no sign of life.

After a worrying wait of many seconds (perhaps tens of seconds,
I haven't measured), during which I desperately tried to enter my
password anyway, everything came back to normal: the display was
unlocked and my X session was working again.

I found a number of kernel traces in /var/log/kern.log (attached).
The times showed that the first traces were experienced well before
I came back to the box, while the last ones were obtained during
my attempts to re-gain control of my X session...

I also noticed that switching VT back and forth (without a running
screensaver or a locked display) generates new traces in
/var/log/kern.log ...


Please investigate this bug and fix it and/or forward my bug report
upstream, as appropriate.

Thanks for your time.



-- Package-specific info:
X server symlink status:

lrwxrwxrwx 1 root root 13 Nov  6  2009 /etc/X11/X -> /usr/bin/Xorg
-rwxr-xr-x 1 root root 274 Jul 20 05:00 /usr/bin/Xorg

VGA-compatible devices on PCI bus:
--
00:02.0 VGA compatible controller [0300]: Intel Corporation 4 Series Chipset 
Integrated Graphics Controller [8086:2e22] (rev 03)

/etc/X11/xorg.conf does not exist.

/etc/X11/xorg.conf.d does not exist.

KMS configuration files:

/etc/modprobe.d/i915-kms.conf:
  options i915 modeset=1
/etc/modprobe.d/radeon-kms.conf:
  options radeon modeset=1

Kernel version (/proc/version):
---
Linux version 4.6.0-1-amd64 (debian-ker...@lists.debian.org) (gcc version 5.4.0 
20160609 (Debian 5.4.0-6) ) #1 SMP Debian 4.6.4-1 (2016-07-18)

No Xorg X server log files found.

udev information:
-
P: /devices/LNXSYSTM:00/LNXPWRBN:00/input/input7
E: DEVPATH=/devices/LNXSYSTM:00/LNXPWRBN:00/input/input7
E: EV=3
E: ID_FOR_SEAT=input-acpi-LNXPWRBN_00
E: ID_INPUT=1
E: ID_INPUT_KEY=1
E: ID_PATH=acpi-LNXPWRBN:00
E: ID_PATH_TAG=acpi-LNXPWRBN_00
E: KEY=10 0
E: MODALIAS=input:b0019vp0001e-e0,1,k74,ramlsfw
E: NAME="Power Button"
E: PHYS="LNXPWRBN/button/input0"
E: PRODUCT=19/0/1/0
E: PROP=0
E: SUBSYSTEM=input
E: TAGS=:seat:
E: USEC_INITIALIZED=12889923

P: /devices/LNXSYSTM:00/LNXPWRBN:00/input/input7/event4
N: input/event4
E: BACKSPACE=guess
E: DEVNAME=/dev/input/event4
E: DEVPATH=/devices/LNXSYSTM:00/LNXPWRBN:00/input/input7/event4
E: ID_INPUT=1
E: ID_INPUT_KEY=1
E: ID_PATH=acpi-LNXPWRBN:00
E: ID_PATH_TAG=acpi-LNXPWRBN_00
E: LIBINPUT_DEVICE_GROUP=19/0/1/0:LNXPWRBN/button
E: MAJOR=13
E: MINOR=68
E: SUBSYSTEM=input
E: TAGS=:power-switch:
E: USEC_INITIALIZED=13081371
E: XKBLAYOUT=us
E: XKBMODEL=pc104
E: XKBOPTIONS=compose:menu

P: /devices/LNXSYSTM:00/LNXSYBUS:00/PNP0C0C:00/input/input6
E: DEVPATH=/devices/LNXSYSTM:00/LNXSYBUS:00/PNP0C0C:00/input/input6
E: EV=3
E: ID_FOR_SEAT=input-acpi-PNP0C0C_00
E: ID_INPUT=1
E: ID_INPUT_KEY=1
E: ID_PATH=acpi-PNP0C0C:00
E: ID_PATH_TAG=acpi-PNP0C0C_00
E: KEY=10 0
E: MODALIAS=input:b0019vp0001e-e0,1,k74,ramlsfw
E: NAME="Power Button"
E: