Summary: | x11-misc/slim-1.3.2-r3 hangs on start-up, leaving X running with blank screen | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Christopher Head <bugs> |
Component: | Current packages | Assignee: | Jeremy Olexa (darkside) (RETIRED) <darkside> |
Status: | RESOLVED WORKSFORME | ||
Severity: | normal | CC: | axs, desktop-misc, jer, johannes |
Priority: | Normal | ||
Version: | unspecified | ||
Hardware: | AMD64 | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: | emerge.info |
Description
Christopher Head
2011-05-22 06:56:26 UTC
I don't think this has to do with the version of xorg-server - that combination of versions runs fine here on several systems. Entirely possible. I have other machines where it works fine; it's only on a machine running nvidia-drivers that I have problems (and maybe only on one of two such machines; I'm not physically sitting in front of the other right now so I can't check). Oh, the machine in question also has $ emerge --info x11-drivers/xf86-input-synaptics … x11-drivers/xf86-input-synaptics-1.4.0 was built with the following: USE="(multilib)" LDFLAGS="-Wl,-O1 -Wl,--as-needed -Wl,-z,lazy" but that version hasn't changed for ages and the slim problem only appeared recently. I have the same problem. I use CoreI3 2310 built in graphics. Sometimes but not very often slim succesfully starts. It seemed to me rather random if it will start or not. gdm works fine for me, but I'd prefer to use slim. Created attachment 275797 [details]
emerge.info
I can't reproduce this issue on my available hardware (dual-core x86, intel 945GM; dual-core amd64, geforce 6150 via nvidia-drivers)... I'm also not seeing anything in the X11 docs that would say a call to XOpenDisplay could cause a race condition; it should either succeed or fail, it doesn't seem to block. Now; there are actually two calls to XOpenDisplay; the one inside App::WaitForServer , and the one inside the App::Run(), that's called right after App::StartServer finishes. From what I can tell, it's possible that App::StartServer could complete with a ServerPID of -1 (which I assume would be invalid) when that ServerTimeout() call above takes parameter 0 and ends up returning false, in which case the App::WaitForServer function call is actually skipped, but pretty well everything else goes on as normal. App::StartServer returns the ServerPID, but for some reason, this return value is not checked. It could be that this is, for whatever reason, what's happening on your systems? If possible could either of you try with the latest revision (-r6)? I don't know if it'll be any different or not. When you do, though, please post your /var/log/slim.log file so that I can trace what's going on. (I may need to provide you with a patch as well, to add more debug messages to the slim log) Since rebuilding everything including slim (for GCC4.5) and I think upgrading xorg-server in the mix, I can't seem to reproduce this any more after lots of boots, so I'm going to say it's probably gone. Please don't do that. Could you please point to where in the Bugzilla help pages it says I am not supposed to mark things I have verified as fixed as verified? I have found the documentation for exactly what different resolutions mean in the Gentoo context somewhat lacking, so I may be looking in the wrong place. |