xorg-server-1.7.1 seems to be broken.
Andrzej Rosa
Black screen, nothing in logs, vesa doesn't work either. I have an
Intel GMA graphic card (driver was rebuilt during xorg-server
upgrade). After I followed this downgrade guide:
http://forums.gentoo.org/viewtopic-t-799568-highlight-xorgserver.html?sid=98c0c077fbe432ef67f585e7dab54a90
all is working again.
(As a sidenote - how VIDEO_CARDS are supposed to work nowadays?
During an upgrade I could add new cards, which showed that xorg-server
will be built with support for new hardware. Now I can add new card
into make.conf, but it doesn't seem to affect emerge (and it didn't do
anything when we tried to do it for real, on some other system). What
I'm saying, it worked fine during an upgrade, but doesn't seem to work
when you try just rebuilding an existing installation. A link to some
documentation would be great. Thanks.)
Andrzej Rosa.
Jeff Mitchell
> (As a sidenote - how VIDEO_CARDS are supposed to work nowadays?
> During an upgrade I could add new cards, which showed that xorg-server
> will be built with support for new hardware. Now I can add new card
> into make.conf, but it doesn't seem to affect emerge (and it didn't do
> anything when we tried to do it for real, on some other system). What
> I'm saying, it worked fine during an upgrade, but doesn't seem to work
> when you try just rebuilding an existing installation. A link to some
> documentation would be great. Thanks.)
Take a look at x11-base/xorg-drivers. The driver specs have moved from
xorg-server to there (thankfully). VIDEO_CARDS is still used.
--Jeff
Piotr Karbowski
rebuild all x11 drivers
emerge -1 $(qlist -IC |grep x11-drivers)
I almost sure that you didn't read elog info after merging new
xorg-server about *rebuild all drivers*.
Dominik Riva
the xorg-server by disabling kms for the Intel driver in my kernel.
I would like to have the kms fixed but I had not much time to
investigate yet.
Regards, Dominik
Piotr Karbowski
uvesafb, intelfb or vesafb.
Andrzej Rosa
crashes my box hard.
BTW - what's the matter with x11-libs/libXxf86vm? xorg-server-1.7.1
won't build without it, complaining of missing package at the
configure level. It looks like a dependency issue to me, not?
lspci output:
00:00.0 Host bridge: Intel Corporation 82852/82855 GM/GME/PM/GMV
Processor to I/O Controller (rev 02)
00:00.1 System peripheral: Intel Corporation 82852/82855 GM/GME/PM/GMV
Processor to I/O Controller (rev 02)
00:00.3 System peripheral: Intel Corporation 82852/82855 GM/GME/PM/GMV
Processor to I/O Controller (rev 02)
00:02.0 VGA compatible controller: Intel Corporation 82852/855GM
Integrated Graphics Device (rev 02)
00:02.1 Display controller: Intel Corporation 82852/855GM Integrated
Graphics Device (rev 02)
00:1d.0 USB Controller: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/
ICH4-M) USB UHCI Controller #1 (rev 01)
00:1d.1 USB Controller: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/
ICH4-M) USB UHCI Controller #2 (rev 01)
00:1d.2 USB Controller: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/
ICH4-M) USB UHCI Controller #3 (rev 01)
00:1d.7 USB Controller: Intel Corporation 82801DB/DBM (ICH4/ICH4-M)
USB2 EHCI Controller (rev 01)
00:1e.0 PCI bridge: Intel Corporation 82801 Mobile PCI Bridge (rev 81)
00:1f.0 ISA bridge: Intel Corporation 82801DBM (ICH4-M) LPC Interface
Bridge (rev 01)
00:1f.1 IDE interface: Intel Corporation 82801DBM (ICH4-M) IDE
Controller (rev 01)
00:1f.5 Multimedia audio controller: Intel Corporation 82801DB/DBL/DBM
(ICH4/ICH4-L/ICH4-M) AC'97 Audio Controller (rev 01)
01:00.0 Ethernet controller: Broadcom Corporation NetXtreme BCM5705M
Gigabit Ethernet (rev 01)
01:01.0 CardBus bridge: Texas Instruments PCI7510 PC card Cardbus
Controller (rev 01)
01:01.1 CardBus bridge: Texas Instruments PCI7510,7610 PC card Cardbus
Controller (rev 01)
01:01.2 FireWire (IEEE 1394): Texas Instruments PCI7410,7510,7610 OHCI-
Lynx Controller
01:01.3 System peripheral: Texas Instruments PCI7410,7510,7610 PCI
Firmware Loading Function
01:03.0 Network controller: Intel Corporation PRO/Wireless 2200BG
[Calexico2] Network Connection (rev 05)
Andrzej Rosa
On Nov 3, 3:56 pm, Dominik Riva <slalomsk...@gmail.com> wrote:
> I have more or less a similar problem on my hp laptop but I can "fix"
> the xorg-server by disabling kms for the Intel driver in my kernel.
/usr/src/linux $ grep -i kms .config
# CONFIG_DRM_I915_KMS is not set
Andrzej Rosa.
Daniel Robbins
>
> Upgrade of xorg-server from 1.6.5 to 1.7.1 hangs my system, hard.
> Black screen, nothing in logs, vesa doesn't work either. I have an
> Intel GMA graphic card (driver was rebuilt during xorg-server
> upgrade). After I followed this downgrade guide:
> http://forums.gentoo.org/viewtopic-t-799568-highlight-xorgserver.html?sid=98c0c077fbe432ef67f585e7dab54a90
> all is working again.
Maybe I should add these masks to our funtoo cautionary mask? Anyone
really want 1.7.1? Looks like it is causing a lot of problems.
-Daniel
Andrzej Rosa
> xorg-server to there (thankfully). VIDEO_CARDS is still used.
one machine to the other, and then tried to just build some missing
drivers. It seems that xorg-drivers is not scheduled for rebuild once
you update the VIDEO_CARDS line. I just tried it with emerge --newuse -
p world. Shouldn't it take a look into make.conf? mesa package
recognizes that something changed in the system.
Andrzej Rosa.
Piotr Karbowski
> BTW - what's the matter with x11-libs/libXxf86vm? xorg-server-1.7.1
> won't build without it, complaining of missing package at the
> configure level. It looks like a dependency issue to me, not?
Much easier will be help you if you will attach error + build.log.
Anyway, x11-libs/libXxf86vm-1.1.0 builded for me a few secunds ago
without any issue.
Kamil Muzylo
That is probably not a bad idea. I am running 1.7.1 without issues however it looks it causes a lot of headache for a lot of people.
Kamil.
jeanfrancis
Running 1.7.1 without any problem here also. I've got an nVidia. I had
to rebuild all the xorg drivers, but no other issue.
~JF
On Nov 3, 11:42 am, Kamil Muzylo <kamilmuz...@gmail.com> wrote:
> Daniel,
>
> That is probably not a bad idea. I am running 1.7.1 without issues however
> it looks it causes a lot of headache for a lot of people.
>
> Kamil.
>
>
>
> On Tue, Nov 3, 2009 at 11:21 AM, Daniel Robbins <drobb...@funtoo.org> wrote:
>
> > On Tue, Nov 3, 2009 at 5:31 AM, Andrzej Rosa <ar...@tlen.pl> wrote:
>
> > > Upgrade of xorg-server from 1.6.5 to 1.7.1 hangs my system, hard.
> > > Black screen, nothing in logs, vesa doesn't work either. I have an
> > > Intel GMA graphic card (driver was rebuilt during xorg-server
> > > upgrade). After I followed this downgrade guide:
>
Kamil Muzylo
I should have mentioned also that I am using NVIDIA card and I was required to rebuilt the drivers.
Kamil.
Piotr Karbowski
emerge says after emerging new xorg-server
> You must rebuild all drivers if upgrading from xorg-server 1.6
> or earlier, because the ABI changed. If you cannot start X because
> of module version mismatch errors, this is your problem.
> You can generate a list of all installed packages in the x11-drivers
> category using this command:
> emerge portage-utils; qlist -I -C x11-drivers/
So the problem is that most of users never doing it, never reading it...
Andrzej Rosa
> On Tue, Nov 3, 2009 at 5:14 PM, Andrzej Rosa <ar...@tlen.pl> wrote:
> > BTW - what's the matter with x11-libs/libXxf86vm? xorg-server-1.7.1
> > won't build without it, complaining of missing package at the
> > configure level. It looks like a dependency issue to me, not?
>
> Much easier will be help you if you will attach error + build.log.
prepared.
Anyway, I tried to upgrade xorg-server without libXxf86vm (for some
reason, probably a conflict...). Then as you try to merge xorg-
server-1.7.1 it fails at the configure stage complaining something
about xxf86vm package. Merging libXxf86vm solves the problem and xorg-
server builds properly.
> Anyway, x11-libs/libXxf86vm-1.1.0 builded for me a few secunds ago
> without any issue.
server-1.7.1 to try and reproduce my report? If not, I'll redo
everything tomorrow.
Andrzej Rosa.
Andrzej Rosa
> > BTW - what's the matter with x11-libs/libXxf86vm? xorg-server-1.7.1
> > won't build without it, complaining of missing package at the
> > configure level. It looks like a dependency issue to me, not?
>
> Much easier will be help you if you will attach error + build.log.
Sorry, my bad. It came to me a bit as an afterthought, so I wasn't prepared.
Anyway, I tried to upgrade xorg-server without libXxf86vm (for some reason,
probably a conflict...). Then as you try to merge xorg-server-1.7.1 it fails
at the configure stage complaining something about xxf86vm package. Merging
libXxf86vm solves the problem and xorg-server builds properly.
> Anyway, x11-libs/libXxf86vm-1.1.0 builded for me a few secunds ago
> without any issue.
Couldn't you maybe unmerge libXxf86vm and try to merge xorg-server-1.7.1 to
Dominik Riva
time. Do you have more things for me to check?
Jeff Mitchell
(sometimes from people not reading the warning about rebuilding all
drivers; sometimes for other reasons). I'm not sure that masking it now
will help, since they'll probably run into the same problems anyways
when the mask is removed. Often they're not issues that are simply
corrected by waiting for another micro release, since they are either
user error (like not rebuilding drivers) or some other new
incompatibility (like a new change that must be made in kernel config).
FWIW, the discussion on this list seems to be adhering to the above
statement :-)
--Jeff
Dominik Riva
Maybe it should be unmasked in ~x86?
Here some background http://bugs.gentoo.org/show_bug.cgi?id=239008
I haven't found any bugs yet that would hinder the unmasking
/usr/portage/profiles/package.mask/gentoo
# Markus Meier <mae...@gentoo.org> (13 Jun 2009)
# mask pre releases of media-gfx/inkscape-0.47
>=media-gfx/inkscape-0.47_pre0
Juan Fco. Giordana
> Maybe I should add these masks to our funtoo cautionary mask? Anyone
> really want 1.7.1? Looks like it is causing a lot of problems.
I am also running 1.7.1 without issues.
My question is, why aren't these drivers rebuilt after every xorg-server
upgrade using a similar method as mentioned by Piotr Karbowski above:
emerge -p $(qlist -IC x11-drivers/)
Andrzej Rosa
> Daniel Robbins wrote:
> > On Tue, Nov 3, 2009 at 5:31 AM, Andrzej Rosa <ar...@tlen.pl> wrote:
> >> Upgrade of xorg-server from 1.6.5 to 1.7.1 hangs my system, hard.
> >> Black screen, nothing in logs, vesa doesn't work either. I have an
> >> Intel GMA graphic card (driver was rebuilt during xorg-server
> >> upgrade). After I followed this downgrade guide:
> >> http://forums.gentoo.org/viewtopic-t-799568-highlight-xorgserver.html?si
> >
> > Maybe I should add these masks to our funtoo cautionary mask? Anyone
> > really want 1.7.1? Looks like it is causing a lot of problems.
>
> IME, xorg-server minor bumps *always* cause problems for some users
> (sometimes from people not reading the warning about rebuilding all
> drivers; sometimes for other reasons).
As far as I've seen, video drivers are rebuilt by default, so what other
driver has a "legitimate" reason to crash the system? Mouse or keyboard?
> I'm not sure that masking it now
> will help, since they'll probably run into the same problems anyways
> when the mask is removed. Often they're not issues that are simply
> corrected by waiting for another micro release, since they are either
> user error (like not rebuilding drivers) or some other new
> incompatibility (like a new change that must be made in kernel config).
Which here means turning off default setting. One would hope that at some
point in the future, drivers will get fixed to the point when turning off KMS is
no longer required. Especially if you assume that KMS used to work with the
older package.
> FWIW, the discussion on this list seems to be adhering to the above
> statement :-)
I wouldn't say so. I wrote it crashes my box and I made doubly sure to
rebuild all drivers. Others just wrote that they heard about issues, plus one
person reported KMS conflict. So we know this server crashes fairly popular
hardware, plus it requires turning off KMS on some other hardware, hence it is
broken there. Both cases strongly hint at driver issues, which could get fixed
in the future.
I'm not suggesting masking or unmasking anything, I don't have enough
knowledge to do so. I'm just trying to make sure you guys base your decision
on reported facts.
Andrzej Rosa.
Andrzej Rosa
> > Much easier will be help you if you will attach error + build.log.
>
> Sorry, my bad. It came to me a bit as an afterthought, so I wasn't
> prepared.
>
> Anyway, I tried to upgrade xorg-server without libXxf86vm (for some reason,
> probably a conflict...). Then as you try to merge xorg-server-1.7.1 it
> fails at the configure stage complaining something about xxf86vm package.
> Merging libXxf86vm solves the problem and xorg-server builds properly.
The relevant part of emerge output is probably this:
checking for GL... configure: error: Package requirements (glproto >= 1.4.9 gl
>= 7.1.0) were not met:
Package xxf86vm was not found in the pkg-config search path.
Perhaps you should add the directory containing `xxf86vm.pc'
to the PKG_CONFIG_PATH environment variable
Package 'xxf86vm', required by 'gl', not found
To reproduce unmerge x11-libs/libXxf86vm and try to build xorg-server-1.7.1.
Andrzej Rosa.
Jeff Mitchell
> On Tuesday 03 November 2009 22:19:36 Jeff Mitchell wrote:
>
>> IME, xorg-server minor bumps *always* cause problems for some users
>> (sometimes from people not reading the warning about rebuilding all
>> drivers; sometimes for other reasons).
>>
>
> As far as I've seen, video drivers are rebuilt by default, so what other
> driver has a "legitimate" reason to crash the system? Mouse or keyboard?
>
never noticed nvidia-drivers, for instance, being rebuilt by default
when xorg-server is bumped.
But yes -- input drivers. Anything you can find via "eix xf86-input"
>> I'm not sure that masking it now
>> will help, since they'll probably run into the same problems anyways
>> when the mask is removed. Often they're not issues that are simply
>> corrected by waiting for another micro release, since they are either
>> user error (like not rebuilding drivers) or some other new
>> incompatibility (like a new change that must be made in kernel config).
>>
>
> Which here means turning off default setting. One would hope that at some
> point in the future, drivers will get fixed to the point when turning off KMS is
> no longer required. Especially if you assume that KMS used to work with the
> older package.
>
KMS isn't on by default.
>
>
>> FWIW, the discussion on this list seems to be adhering to the above
>> statement :-)
>>
>
> I wouldn't say so. I wrote it crashes my box and I made doubly sure to
> rebuild all drivers. Others just wrote that they heard about issues
Most of the issues were resolved in 1.7.1, which is why 1.7.0 wasn't
unmasked.
> person reported KMS conflict. So we know this server crashes fairly popular
> hardware
But works very well for other people on the same hardware.
> plus it requires turning off KMS on some other hardware, hence it is
> broken there.
Again, KMS isn't on by default.
I'd like to point out that this is part of the pain that comes with
running ~. You assume some risk by pegging yourself to testing, which is
what the ~ keyword is for. This isn't a bad thing -- without people
testing and reporting bugs, things could never get fixed and work
properly in stable -- but I don't really see evidence here of a systemic
failure in xorg-server 1.7.1 that requires a full-on mask. Rather it
seems like people are running into bugs in testing software, which is
why it's marked testing, and that finding and reporting them will
benefit everyone. You guys are the front line (and we thank you for it).
--Jeff