After around six days uptime the size of my Xorg.0.log is already 240Mb and is filled with the following error messages: (EE) BUG: triggered 'if (priv->num_active_touches > priv->num_slots)' (EE) BUG: ../../src/synaptics.c:2991 in UpdateTouchState() (EE) (EE) Backtrace: (EE) 0: /usr/bin/X (xorg_backtrace+0x56) [0x7f78693cd556] (EE) 1: /usr/lib/xorg/modules/input/synaptics_drv.so (0x7f7860bb3000+0x527e) [0x7f7860bb827e] (EE) 2: /usr/lib/xorg/modules/input/synaptics_drv.so (0x7f7860bb3000+0x75a2) [0x7f7860bba5a2] (EE) 3: /usr/bin/X (0x7f786921a000+0x96708) [0x7f78692b0708] (EE) 4: /usr/bin/X (0x7f786921a000+0xbfa79) [0x7f78692d9a79] (EE) 5: /lib/x86_64-linux-gnu/libc.so.6 (0x7f7866ee0000+0x352f0) [0x7f7866f152f0] (EE) 6: /lib/x86_64-linux-gnu/libc.so.6 (__select+0x33) [0x7f7866fdd743] (EE) 7: /usr/bin/X (WaitForSomething+0x1ac) [0x7f78693ca82c] (EE) 8: /usr/bin/X (0x7f786921a000+0x57ec1) [0x7f7869271ec1] (EE) 9: /usr/bin/X (0x7f786921a000+0x5c29b) [0x7f786927629b] (EE) 10: /lib/x86_64-linux-gnu/libc.so.6 (__libc_start_main+0xf0) [0x7f7866f00a40] (EE) 11: /usr/bin/X (0x7f786921a000+0x4662e) [0x7f786926062e] (EE) (EE) [dix] SYN1B7F:01 06CB:2970 UNKNOWN: unable to find touch point 0 (EE) [dix] SYN1B7F:01 06CB:2970 UNKNOWN: unable to find touch point 0
Created attachment 115856 [details] Xorg.0.log.old (old smaller logfile)
Attach the evemu-record for your device please, for the following three sequences (separate recordings please): * single finger touch + move * two-finger tap * three-finger tap I'm pretty sure this is a kernel bug that upsets the synaptics driver with some sequence.
Created attachment 115871 [details] evemu-record for single finger touch + move
Created attachment 115872 [details] evemu-record for two-finger tap
Created attachment 115873 [details] evemu-record for three-finger tap
sorry about the long silence. I just tried replaying the events but I can't trigger it. When the finger count is off, this usually happened around suspend/resume or on startup and once the driver is in that state it just keeps going. But it's tricky to get it into that state. I recommend figuring out if there is a specific sequence that triggers that bug and then running evemu-record while reproducing it. That should tell us what's going on. Note that only the first time this bug is triggered in the log is reliable, anything subsequent may be triggered because of previously mangled state.
Created attachment 120608 [details] Xorg log after crash
I have this issue. X crashes daily at random. I can't reproduce it on demand. Observations: - Running Linux Mint 17.3 on Dell XPS 13. Had problem with 17.2 also - Touchpad driver synaptics-i2c-hid.3.13.0-32-backport.dkms v1.6.3 - Upgraded BIOS recently from A05 to A07. Before with A05: - Crashes occurred more often with Disable Touchpad While Typing (DTWT). Pretty much on demand I could turn on DTWT and start typing and it would crash in minutes. - With DTWT off, it crashes once-twice daily. - Many times, when I launched Chrome browser it would crash. This would happen once and I could immediately relaunch Chrome w/o crash. - After BIOS A07, I turned DTWT on and went a couple of days before a crash. So I guess it's better.
Note that the log message posted here isn't a crash, the macro to check the bug uses the backtrace to easier isolate it, but other than that it's just an error message. It's not a crash and shouldn't bring the server down. So if your server crashes, that is a different bug, please file it separately. I also recommend trying xf86-input-libinput, maybe that one will spit out a more useful error message when this happens.
There's no CANTFIX resolution here, sorry. But without a reproducible test case we've been unable to find the cause for this.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.