Bug 62541

Summary: Kernel oops/panic during system boot with systemd - "Unable to handle kernel NULL pointer dereference"
Product: Drivers Reporter: Adrian Byszuk (a.byszuk)
Component: Video(DRI - non Intel)Assignee: drivers_video-dri
Status: RESOLVED OBSOLETE    
Severity: high CC: a.byszuk, szg00000, xerofoify
Priority: P1    
Hardware: x86-64   
OS: Linux   
Kernel Version: 3.11 Subsystem:
Regression: No Bisected commit-id:
Attachments: dmesg, lswh, .config outputs

Description Adrian Byszuk 2013-10-05 06:50:50 UTC 
Created attachment 110251 [details]
dmesg, lswh, .config outputs

Hello,

I have a bug (100 % repeatable) that prevents booting Linux when systemd is used as an init system. However, booting with traditional init works fine.
I'm posting this bug to the DRI subsystem because I was (barely) able to capture to capture kernel message before the screen becomes blank.
I can verify that this problem happens in all 3.11.x kernels and in Debian's 3.10 kernel.
The hardware is Dell Vostro 3560 - Intel Ivy Bridge, hybrid graphics (Intel + Radeon), OCZ Vertex 4 SSD.
My suspicion is that maybe the system bootup is too fast for the kernel; I was able to boot kernel once even with systemd, but that was when I turned on all the debugging messages which flooded console a bit.

Photo of captured BUG message: http://s23.postimg.org/qgk843rzv/vga1.png
dmesg, lshw and .config outputs attached.

Kind regards,
Adrian
Comment 1 Adrian Byszuk 2013-10-05 08:14:45 UTC 
OK, I was able to track down the problem more precisely.

In my rc.local I have:
echo OFF > /sys/kernel/debug/vgaswitcheroo/switch

for switching discrete graphics after boot. With this line booting fails.
But when I change it to:
sleep 5
echo OFF > /sys/kernel/debug/vgaswitcheroo/switch

kernel is booting just fine.

So I think the problem is that systemd is, unlike the old init, executing rc.local just in the beginning which switches discrete graphics just in the beginning of system startup.
Comment 2 xerofoify 2014-06-20 04:11:50 UTC 
Is this bug closed? Seems the problem is with systemd and not the 
kernel. Can we please close this bug as it's fairly old as of me
writing this in June 2014.
Thanks A Fellow Developer,
Nick
Comment 3 Szőgyényi Gábor 2017-03-06 20:22:03 UTC 
Please try to reproduce this bug with latest kernel image & latest systemd.
Comment 4 Jani Nikula 2017-03-08 11:21:37 UTC 
(In reply to Szőgyényi Gábor from comment #3)
> Please try to reproduce this bug with latest kernel image & latest systemd.

Sorry, what's the point? If you're scrubbing the bugs here, please just close the old bugs, and ask for reopen if the bug persists with latests bits and pieces. It's been 3½ years since the reporter was last seen here.

Ditto for all the other bugs you've commented on.