Re: netbsd-9 radeon GPU lockup, X stuck in "radfen/N"

["John D. Baker" <jdbaker%consolidated.net@localhost>]

On Mon, 26 Aug 2019, Thomas Klausner wrote:

> On Mon, Aug 26, 2019 at 10:44:37AM -0500, John D. Baker wrote:
> > [ 50615.096841] radeon0: autoconfiguration error: error: ring 0 stalled for more than 6308000msec
> > [ 50615.096841] radeon0: autoconfiguration error: warn: GPU lockup (current fence id 0x00000000002505b3 last fence id 0x00000000002507ad on ring 0)
> 
> I saw this back in January:
> 
> https://mail-index.netbsd.org/current-users/2019/01/09/msg034911.html

It did it again just a short while ago, but this time was followed by a
panic:

[...]
[ 59126.416743] radeon0: autoconfiguration error: error: ring 0 stalled for more than 117510msec
[ 59126.416743] radeon0: autoconfiguration error: warn: GPU lockup (current fence id 0x00000000001aa307 last fence id 0x00000000001aa37f on ring 0)
[ 59126.657265] radeon0: autoconfiguration error: error: failed to get a new IB (-11)
[ 59126.657265] kern error: [drm:(/x/netbsd-9/src/sys/external/bsd/drm2/dist/drm/radeon/radeon_cs.c:643)radeon_cs_ib_fill] *ERROR* Failed to get ib !
[ 59126.837383] kern warning: radeon: wait for empty RBBM fifo failed ! Bad things might happen.
[ 59126.917432] panic: kernel diagnostic assertion "ci->ci_mtx_count == -1" failed: file "/x/netbsd-9/src/sys/kern/kern_synch.c", line 666 mi_switch: cpu1: ci_mtx_count (-2) != -1 (block with spin-mutex held)
[ 59126.917432] cpu1: Begin traceback...
[ 59126.917432] vpanic() at netbsd:vpanic+0x160
[ 59126.917432] ugen_get_alt_index() at netbsd:ugen_get_alt_index
[ 59126.917432] mi_switch() at netbsd:mi_switch+0x7c0
[ 59126.917432] sleepq_block() at netbsd:sleepq_block+0x1b3
[ 59126.917432] cv_wait() at netbsd:cv_wait+0xfb
[ 59126.917432] linux_workqueue_thread() at netbsd:linux_workqueue_thread+0x47d
[ 59126.917432] cpu1: End traceback...

[ 59126.927444] dumping to dev 0,1 (offset=1574999, size=851823):
[ 59126.927444] dump kern warning: Failed to wait GUI idle while programming pipes. Bad things might happen.
[ 59127.037508] radeon0: info: Saved 3835 dwords of commands on ring 0.
[ 59127.037508] radeon0: info: (rs600_asic_reset:465) RBBM_STATUS=0xF4116100
[ 59128.118173] Skipping crash dump on recursive panic
[ 59128.118173] panic: atastart: channel 0 busy, xfer not possible
[ 59128.118173] cpu1: Begin traceback...
[ 59128.118173] vpanic() at netbsd:vpanic+0x160
[ 59128.118173] snprintf() at netbsd:snprintf
[ 59128.118173] ata_get_xfer() at netbsd:ata_get_xfer
[ 59128.118173] wdc_ata_bio() at netbsd:wdc_ata_bio+0x7b
[ 59128.118173] wd_dumpblocks() at netbsd:wd_dumpblocks+0x111
[ 59128.128184] dk_dump() at netbsd:dk_dump+0x172
[ 59128.128184] dump_header_flush() at netbsd:dump_header_flush+0x6d
[ 59128.128184] dump_header_addbytes() at t cpu0: ACPI CPU
[     1.018432] acpicpu0: C1: HLT, lat   0 us, pow     0 mW
[     1.018432] acpddseg+0x1e
[ 59128.128184] dump_seg_iter() at netbsd:dump_seg_iter+0x107
[ 59128.128184] cpu_dump() at netbsd:cpu_dump+0x6a
[ 59128.128184] dodumpsys() at netbsd:dodumpsys+0xfc
[ 59128.138193] dumpsys() at netbsd:dumpsys+0x1d
[ 59128.138193] vpanic() at netbsd:vpanic+0x169
[ 59128.138193] ugen_get_alt_index() at netbsd:ugen_get_alt_index
[ 59128.138193] mi_switch() at netbsd:mi_switch+0x7c0
[ 59128.138193] sleepq_block() at netbsd:sleepq_block+0x1b3
[ 59128.138193] cv_wait() at netbsd:cv_wait+0xfb
[ 59128.148197] linux_workqueue_thread() at    1.2197390] uhub0: 4 ports with 4 removable, self powered
[     1.219739] u
[reboot]

It seemed to happen just as 'xscreensaver' was activating.

-- 
|/"\ John D. Baker, KN5UKS               NetBSD     Darwin/MacOS X
|\ / jdbaker[snail]consolidated[flyspeck]net  OpenBSD            FreeBSD
| X  No HTML/proprietary data in email.   BSD just sits there and works!
|/ \ GPGkeyID:  D703 4A7E 479F 63F8 D3F4  BD99 9572 8F23 E4AD 1645