Bug 3711

Hi, thanks for reporting this bug.
Assigned to the package maintainer.

(Please set the status to 'assigned' if you are working on it)

Keywords: (none) => Triaged
CC: (none) => thierry.vignaud
Assignee: bugsquad => anssi.hannula

Test button causing issues is bug #226.

As for not installing the driver package, did XFdrake ask if you want to use the proprietary driver?

Status: NEW => ASSIGNED

I used test button on previous development cycle with no problem (the same box).

Anyway, I just agreed to use proprietary driver, rebooted without testing and there was a simple black screen with LCD screen backlight on. 

Here is a part of xorg.cong

   Section "Device"
    Identifier "device1"
    VendorName "ATI Technologies Inc"
    BoardName "ATI Radeon HD 2000 and later (radeon/fglrx)"
    Driver "fglrx"
    BusID "PCI:2:0:0"
    Option "DPMS"
EndSection

Anssi, could you add a check in XFdrake for the case we know testing will fail (eg: KMS -> proprietary) and they disable the "test" button or add a warning like we do for some cards. See

See check_bad_card() in /usr/lib/libDrakX/Xconfig/cqrd.pm, which is called from /usr/lib/libDrakX/Xconfig/test.pm
:

sub test {
    my ($in, $raw_X, $card, $auto, $skip_badcard) = @_;

    my $bad_card = !Xconfig::card::check_bad_card($card);
    return 1 if $skip_badcard && $bad_card;

    if ($bad_card || !$auto) {
        $in->ask_yesorno(N("Test of the configuration"),
                         N("Do you want to test the configuration?") .
 ($bad_card ? "\n" . N("Warning: testing this graphic card may freeze your computer") : ''),
                         !$bad_card) or return 1;
    }

we could check if whether both current driver & driver to test are both not KMS or prevent testing.

Changing summary. Palm, please attach the full /etc/X11/xorg.conf, /var/log/Xorg.0.log and /var/log/Xorg.0.log.old, and attach the output of "lspcidrake -v".
Also, tell your card model if you know it, and the laptop model if this is a laptop.

Thierry, I'll look into it. We also have to prevent testing of any proprietary driver when using another proprietary driver.

Summary: x11-driver-video-fglrx installation from XFdrake doesn't install driver => black screen with fglrx

Created attachment 1229 [details]
suggestion

we should split the new tests I added in check_bad_card() into a new check_very_bad_card(), add those new tests there (meaning passing $raw_X to this new function too), and rejecting testing.
Sg like this patch (totally untested)

(get_xorg_driver) move it from service_harddrake

(is_proprietary_driver) split it out of pkgs_for_Driver2()

(check_very_bad_card) detect testing untestable cases

(test) use check_very_bad_card() in order to prevent testing untestable cases

If you agree to go this way, we'll commit in smaller pieces.

@Thierry:

- get_xorg_driver() is somewhat non-robust, as it doesn't handle multiple
  graphics cards nor cases where there is no Driver in xorg.conf
  (i.e. when X server autodetection is used, something we'll move into
   at some point as well, though not in the near future), or cases
  where we are in console, or cases where a /etc/X11/xorg.conf.d/ file is used.
  So better use modules::loaded_modules() instead (against 'nvidia', 'fglrx', 'i915', 'radeon', 'nouveau').

- 4th if clause in check_very_bad_card() is superfluous, it only triggers
  when either 1st or 2nd if clause had triggered as well

- in
  $in->ask_warn(N("Warning"),
  N("This configuration cannot be tested since:") .
  N("Testing this graphic card with this driver will break the display"));
  there are typos, and it is not actually a "warning" since the action is
  refused, and the use of ':' is strange here.
  I suggest simply:
  $in->ask_warn('', N("Cannot test configuration with this driver."));
  or alternatively get a more detailed description from check_very_bad_card().

Otherwise seems ok.

Created attachment 1230 [details]
lspcidrake -v as text

Created attachment 1231 [details]
xorg.conf accessed from outside installation

Created attachment 1232 [details]
copy of Xorg.0.log accessed from outside installation

Created attachment 1233 [details]
copy of Xorg.0.log.old accessed from outside installation

Slightly different story with Mageia2-alpha2 free DVD.

After fglrx installation, there is a black screen with suggestion
"You need to reboot because screen driver was change."
Pressing OK cycles to the question again. 
I've done reboot twice and then pressed Cancel. It gave me well known black screen (see #3).

Hi,

This bug was filed against cauldron, but we do not have cauldron at the moment.

Please report whether this bug is still valid for Mageia 2.

Thanks :)

Cheers,
marja

Keywords: (none) => NEEDINFO

Is this still valid? It concerns fglrx-8.911-2.mga2.nonfree.src.rpm
Now we are at 9.012

CC: (none) => micheelsen

I'm running Mageia 2 

# XFdrake installs 

x11-driver-video-fglrx-8.961-2.mga2.nonfree.x86_64

Pressing Test button makes screen black with no respond to Ctr-Alt-Backspase or Ctrl-Alt-Delete. Xard reboot returns to X with ATI driver


Section "Device"
    Identifier "device1"
    VendorName "ATI Technologies Inc"
    BoardName "ATI Radeon HD 2000 to HD 6300 (radeon/fglrx)"
    Driver "ati"
    BusID "PCI:2:0:0"
    Option "DPMS"
EndSection

Still valid now?

CC: (none) => nic

(In reply to Nic Baxter from comment #17)
> Still valid now?

5 weeks later, no reply

Closing as OLD

Feel free to reopen if needed

Status: ASSIGNED => RESOLVED
CC: (none) => marja11
Resolution: (none) => OLD