Kernel-Crash when working with ubt0
Miranda Maria Sophie Van den Breukelingen
mms.vanbreukelingen at gmail.com
Tue Aug 27 23:39:29 UTC 2019
I recompiled the kernel with the patch enabled and now we get at 'service
bluetooth start ubt0 (ubt1...)' :
*/etc/rc.d/bluetooth: ERROR: Unable to setup Bluetooth stack for device
ubt0root at freeBSD13:/usr/home/miranda # service bluetooth start
ubt1/etc/rc.d/bluetooth: ERROR: Unable to setup Bluetooth stack for device
ubt1*
Should I recompile with the MTX_SPIN=y option now?
kldstat:
Id Refs Address Size Name
1 82 0xffffffff80200000 232ad30 kernel
2 1 0xffffffff8252b000 96f98 linux64.ko
3 3 0xffffffff825c2000 b760 linux_common.ko
4 1 0xffffffff825ce000 b4be8 linux.ko
5 1 0xffffffff82683000 3170 splash_bmp.ko
* 6 1 0xffffffff826a9000 18c28 ng_l2cap.ko*
7 4 0xffffffff826c2000 2dc0 ng_bluetooth.ko
8 6 0xffffffff826c5000 18d50 netgraph.ko
9 1 0xffffffff826df000 68840 if_em_updated.ko
10 1 0xffffffff82748000 a468 ng_ubt.ko
11 3 0xffffffff82753000 12d10 ng_hci.ko
12 1 0xffffffff82d18000 7b040 i915kms.ko
13 1 0xffffffff82d94000 3d9e8 drm2.ko
14 4 0xffffffff82dd2000 1f40 iicbus.ko
15 1 0xffffffff82dd4000 f70 iic.ko
16 1 0xffffffff82dd5000 1570 iicbb.ko
17 1 0xffffffff82dd7000 15720 if_iwm.ko
18 1 0xffffffff82ded000 e045f iwm3160fw.ko
19 1 0xffffffff82ece000 1840 uhid.ko
20 1 0xffffffff82ed0000 2928 ums.ko
21 1 0xffffffff82ed3000 4570 autofs.ko
22 1 0xffffffff82ed8000 19600 ng_btsocket.ko
23 1 0xffffffff82ef2000 acf mac_ntpd.ko
24 1 0xffffffff82ef3000 19728 ext2fs.ko
25 1 0xffffffff82f0d000 3a8c geom_linux_lvm.ko
6 has to be started via /boot/loader.conf - I can't insert a driver into
the kernel at all: kldload; pE: can't load wlan_xauth.ko: Operation not
permitted (as root, probably some subroutine in permanent use? Is it a
toggle at the startup services?)
but it was good compilation training, svn up had some changes in the main
files, so it was not a mistake upgrading to r351552M now. Fixed my Plasma5
and I added Enlightment as a DM. For my grafic card there's the
drm-legacy-kmod-driver the best and this works, too now - but compositor's
still on XRender as the most common problem is to loose the insights of the
windows after some minutes with OpenGL (2 and 3.1).
Any ideas except /dev/null the whole stuff from the beginnning on?
btw, 'pkg: gstreamer1-plugins-lame has a missing dependency: lame' ... does
this happen to everyone when using pkg install ? there's no dependency lame
and lame means something like 'like a horse that's gone lame.' but freeBSD
isn't, just read UPDATING!
Gonna build some old darwin kernel now, the influence of BSD is magnificant
in comparison to the mach-part - but that's another chapter.
tomorrow,
Miranda
On Tue, 27 Aug 2019 at 15:03, Maksim Yevmenkin <maksim.yevmenkin at gmail.com>
wrote:
> > > Hmm... interesting....
> > >
> > > I only took a brief look at it. I suppose I can ensure user space
> address is wired and then copyout() can be called with mutex held
> >
> > >No, you cannot do this, at least without making the kernel to panic.
> > User might unmap the wired mapping at any time still.
>
> Kostik,
>
> i was thinking along the lines of vslock/vsunlock and copyout_nofault.
> basically similar to the sysctl code. do you think this would not
> work?
>
> thanks,
> max
>
More information about the freebsd-current
mailing list