Re: kern/49195 (drm2 panic in i915_gem.c)

[Alan Barrett <apb%cequrux.com@localhost>]

The following reply was made to PR kern/49195; it has been noted by GNATS.

From: Alan Barrett <apb%cequrux.com@localhost>
To: gnats-bugs%netbsd.org@localhost
Cc: 
Subject: Re: kern/49195 (drm2 panic in i915_gem.c)
Date: Mon, 17 Nov 2014 18:16:06 +0200

 On Mon, 17 Nov 2014, Alan Barrett wrote:
 > >This seems to be fixed with revision 1.18 of i915_gem.c.  Please test.
 >
 > I haven't yet tested that change.  However, I can report that the
 > problem appears with both AccelMethod "UXA" and AccelMethod "SNA"
 > but not with Accel off, in this part of the xorg.conf file:
 
 I have now tested with i915_gem.c revision 1.18.  I still get the
 write_domain panic with AccelMethod "SNA", although the back trace may
 not always be the same.
 
 With AccelMethod "UXA", I get such badly corrupted images that I didn't
 test for long enough to see whtether it panics.
 
 Here's a backtrace from the most recent write_domain panic.
 
 #11 0xc0b62a34 in kern_assert (
     fmt=fmt@entry=0xc0df5830 "kernel %sassertion \"%s\" failed: file \"%s\", line %d ") at src/sys/lib/libkern/kern_assert.c:51
 #12 0xc03e7ad7 in i915_gem_object_move_to_inactive (obj=0xc868ef00)
     at src/sys/external/bsd/drm2/dist/drm/i915/i915_gem.c:2854
 #13 0xc03e7b46 in i915_gem_retire_requests_ring (ring=0xc4b24344)
     at src/sys/external/bsd/drm2/dist/drm/i915/i915_gem.c:3249
 #14 0xc03e9a57 in i915_gem_object_flush_active (obj=0xc868ef00)
     at src/sys/external/bsd/drm2/dist/drm/i915/i915_gem.c:3347
 #15 i915_gem_busy_ioctl (dev=0xc504564c, data=0xdd8e1eb0, file=0xc659f5e8)
     at src/sys/external/bsd/drm2/dist/drm/i915/i915_gem.c:4836
 #16 0xc02ea1dc in drm_ioctl (fp=0xc65a9c00, cmd=3221775447, data=0xdd8e1eb0)
     at src/sys/external/bsd/drm2/drm/drm_drv.c:663
 #17 0xc09b8fd0 in sys_ioctl (l=0xc53fbd40, uap=0xdd8e1f68, retval=0xdd8e1f60)
     at src/sys/kern/sys_generic.c:681
 #18 0xc09c4693 in sy_call (rval=0xdd8e1f60, uap=0xdd8e1f68, l=0xc53fbd40,
     sy=<optimized out>) at src/sys/sys/syscallvar.h:61
 #19 sy_invoke (code=54, rval=0xdd8e1f60, uap=0xdd8e1f68, l=0xc53fbd40,
     sy=<optimized out>) at src/sys/sys/syscallvar.h:85
 #20 syscall (frame=0xdd8e1fa8)
     at src/sys/arch/x86/x86/syscall.c:156
 #21 0xc01005c6 in Xsyscall ()
 #22 0xdd8e1fa8 in ?? ()
 Backtrace stopped: previous frame inner to this frame (corrupt stack?)
 
 --apb (Alan Barrett)