[SOLVED] [Xorg-1.7.5 in Slackware64 -current] still no workie.. Xorg bug?
SlackwareThis Forum is for the discussion of Slackware Linux.
Notices
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
If you have any problems with the registration process or your account login, please contact us. If you need to reset your password, click here.
Having a problem logging in? Please visit this page to clear all LQ-related cookies.
Get a virtual cloud desktop with the Linux distro that you want in less than five minutes with Shells! With over 10 pre-installed distros to choose from, the worry-free installation life is here! Whether you are a digital nomad or just looking for flexibility, Shells can put your Linux machine on the device that you want to use.
Exclusive for LQ members, get up to 45% off per month. Click here for more info.
Hmm, yes, it does look VERY similar, and from what I gather from reading that convo, HAL is not directly part of the problem, but rather, it appears to be an Xorg bug. Thanks for that link. I have yet to follow the one or two off-shoot links from it.
This evening, I'll hopefully make time to try some other Xorg-1.7.5 packages such as linked above in this thread by another member, and see if it helps.
But with any luck, *IF* it is an Xorg bug and it gets "fixed" soon, then -current will get a new Xorg package soon after, so if I don't manage to try some of the patched Xorgs that are flying around, I'll definitely be looking forward to a new package in Slack-current.
Caveat emptor: I am running Slackware64-13, not current and didn't try 2.6.33 nor Xorg7.5 (yet) and don't use HAL (I have reconfigured xorg-server-1.6.3 with the '--disable-hal-config" option) BUT...
... I am very happy with the nouveau driver so ...
... Sasha, may I suggest you give it a try ?
- Follow this http://nouveau.freedesktop.org/wiki/InstallNouveau
- Skip only step 2 as the nouveau kernel driver is included in the 2.6.33 tree
- Blacklist nouveau
- But load it at end of startup (I suggest adding /sbin/modprobe/nouveau at the end of /etc/rc.d/rc.local)
Of course follow all installation steps at runlevel 3
Please let us now the outcome.
PS I'm gonna try -current to see what I get ASAP - but I need to make some room for it on my hard disk first...
yes, I have no problems about getting off of 'the blob' so nouveau is already on my radar. I don't play 3D games or design space-shuttles, or simulate earthquakes so nouveau may do all I need from a video driver-- provided it can use either Xinerama or XrandR to make both my displays work. Thank you for the link and supplemental instructions there, I am going to check it out soon.
At the risk of stating/asking the obvious though: using nouveau (or not) has no bearing on whether or not Xorg-1.7.5 works for me, does it? I mean, if the VESA driver doesn't work for me, chances are good that nouveau won't help.
However either way, to try nouveau, I'll be needing to rebuild my current (not "-current") kernel to enable the nouveau DRM, so I'll look into this shortly for the heck of it.
@ Matt -- which version of the log would you like? One from a failed X session using Xorg-1.7.5? Or the current log from my session I am running right now?
OK, well reports are becoming a little more easy to locate, about at least one bug in Xorg which is leading to a lot of folks using small WM's having very similar problems to what I am experiencing.
I think for the time being, I'm going to just not upgrade my Xorg-server package until there's convincing evidence that this *is* a bug and that it has been fixed.
# Didier -- I am just now back from a couple hours spent trying out nouveau. I see it has potential for sure, but...
To make the long story short: the results were about the same as I currently get using Xorg-1.7.5 with any other driver, except that it was harder to actually get the server started because it kept erroring out with something like (EE) Unable to open device once for each card. I tried at least 10 different xorg.conf files, ranging from none at all, to the minimal one given as an example on the nouveau wiki page you linked for me, and ranging up to the full xorg.conf I am currently usung. Most of the time, X would not start, giving the above error. After un-blacklisting nouveau and rebooting (it would not work if I modprobed nouveau after bootup) I managed to eventually get X to start ONCE, and ONCE only -- it activated ONE screen, with no mouse or keyboard support working, and then promptly locked up. I rebooted using Magic SysRq.
So... I've had about enough frigging with Xorg for a day or two and will take a break from this until I find some convincing evidence that there's a new package or something to try upgrading to. It *really* looks like an Xorg bug.
I'm very sure I have some other projects to fiddle with here!
Thank you to everyone who had some input here -- I am not done with all your suggestions, but again, need a little break.
No worries Didier it was worth a try; and, I'm happy that nouveau is working for some people, and continues to improve. Thanks for the idea; unfortunately (and I kinda suspected as much) it did not work for me.
I changed the title of this thread, and edited the first post, to better reflect what I've learned so far from everyone's feedback, i.e. that it probably is not my kernel, but it appears to be an Xorg bug (that's now my stance ).
Best of luck trying out your -current install! We just got a batch of new packages again a few hours ago -- do let us know how it goes -- if you need help, you don't hafta look far!
Would you mind please posting the link for this driver!
Quote:
Originally Posted by GazL
I'm just running the stock vmlinuz-generic-2.6.33 kernel + modules (note: not a monolithic/'huge' based one).
Blacklisting the nouveau module seems vital to getting anything working for me.
The 'nv' driver just doesn't work at all, no matter what I do.
The 'vesa' driver works fine.
The proprietary 'nvidia' 190.53 driver works fine once you put the nvidia-190.53-2.6.33.patch.txt patch on it
This is just a single card/single display system though, so I don't know how much value this will be to you.
I'd like to know where to find this. I've already searched the net and haven't found it. You can try this yourself to see that nothing comes up:
search terms (nvidia-190.53-2.6.33.patch index-of)
NADA!!
Here is what I did and got with slackware64-current as of yesterday:
1) Installed the huge kernel
I got a black screen shortly after booting, may be because nouveau was loaded at boot time (it's modularized even in config-huge-2.6.33) but I missed the nouveau firmware, which is requested at time module is loaded if using 2.6.33 (see chapter 3 here about that)
2) Installed the generic kernel + initrd + blacklist nouveau.
I could boot to runlevel 3 but after startx (under Fluxbox) I saw only the mouse cursor on a black screen. I could move it... that's it. Hopefully Ctrl+alt+del sent me back to runlevel 3.
3) Installed libdrm with support for nouveau + xf86-video-nouveau + the firmware for nouveau. "modprobe nouveau" worked but "startx" failed with this (quoted from Xorg.0.log)
Code:
drmOpenDevice: node name is /dev/dri/card0
drmOpenDevice: open result is 10, (OK)
drmOpenByBusid: Searching for BusID pci:0000:01:00.0
drmOpenDevice: node name is /dev/dri/card0
drmOpenDevice: open result is 10, (OK)
drmOpenByBusid: drmOpenMinor returns 10
drmOpenByBusid: drmGetBusid reports pci:0000:01:00.0
(EE) [drm] failed to open device
(EE) No devices detected.
Fatal server error:
no screens found
4) Just in case, re-compiled xorg-server with --disable-config-hal, killed the HAL daemon, re-used my /etc/X11/xorg.conf from slackware64-13 which works with nouveau. Alas, got the same result.
So I will stick to -stable and xorg-server 1.6.3 for now.
Meanwhile I will try the 2.6.33 kernel on it in order to help narrowing the origin of the problem.
PS In a former post I forgot to mention the necessity of firmware installation and also that libdrm should be installed in /usr/lib64 for Slackware64-current, that is to say using following command to configure libdrm :
And of course you can skip step 1 as well as step 2 in this document provided that xorg7.5 be already installed.
And in fact it's not necessary to load the nouveau kernel module at all, as if its name is included in xorg.conf, X will load it when it (hopefully) starts.
Sorry about missing information, which by the way I included in this other thread some time ago.
Best regards,
Last edited by Didier Spaier; 03-13-2010 at 10:41 AM.
Sasha, can you post the Xorg log from the failed 1.7.5 start?
I just double-checked my settings, and I'm running Xorg 1.7.5 with nvidia-190.53 (patched to install against kernel 2.6.33). I'm on a laptop, so I don't use xinerama.
1. Xorg fails to start with "(EE) [drm] failed to open device"
Did you miss the basic question "Are you using the latest code" above ? If you are using a 2.6.33 kernel from kernel.org , that nouveau code is not compatible with libdrm 2.4.18 or git. Either upgrade nouveau drm code (see InstallDRM) or use an old git version of libdrm and DDX, 2010-02-15 or older.
As a punishment, I'll upgrade nouveau drm code and report the outcome
LinuxQuestions.org is looking for people interested in writing
Editorials, Articles, Reviews, and more. If you'd like to contribute
content, let us know.