Re: port-amd64/56815: Lenovo ThinkCentre with i915drmkms graphics fails to boot -current

[Jarle Greipsland <jarle.greipsland%norid.no@localhost>]

The following reply was made to PR port-amd64/56815; it has been noted by GNATS.

From: Jarle Greipsland <jarle.greipsland%norid.no@localhost>
To: riastradh%NetBSD.org@localhost
Cc: gnats-bugs%NetBSD.org@localhost
Subject: Re: port-amd64/56815: Lenovo ThinkCentre with i915drmkms graphics
 fails to boot -current
Date: Wed, 12 Jul 2023 12:56:00 +0200 (CEST)

 Jarle Greipsland <jarle.greipsland%norid.no@localhost> writes:
 > Taylor R Campbell <riastradh%NetBSD.org@localhost> writes:
 [ ... ]
 >> Excellent!  Can you see if gdb does better about a stack trace?
 >> 
 >> $ gdb netbsd.gdb
 >> (gdb) target kvm netbsd.1.core
 >> (gdb) bt
 >> 
 >> Also, can you try with a newer crash?
 > Extracts using a -current gdb and -current crash below.  The
 > crash one seems relevant.
 I built a kernel with the LOCKDEBUG option set, and the panic
 messages now allude to there being an uninitialized lock:
 [ ... ]
 [   4.4428248] [drm] Supports vblank timestamp caching Rev 2 (21.10.2013).
 [   4.4460886] [drm] Driver supports precise vblank timestamp query.
 [   4.4460886] i915drmkms0: interrupting at msi4 vec 0 (i915drmkms0)
 [   4.6660696] {drm:netbsd:gen8_reset_engines+0x394} *ERROR* bcs'0 reset request
  timed out: {request: 00000001, RESET_CTL: 00000001}
 [   4.6660696] {drm:netbsd:gen8_reset_engines+0x394} *ERROR* bcs'0 reset request
  timed out: {request: 00000001, RESET_CTL: 00000001}
 [   4.6660696] {drm:netbsd:gen8_reset_engines+0x394} *ERROR* vcs'0 reset request
  timed out: {request: 00000001, RESET_CTL: 00000001}
 [   4.6760782] {drm:netbsd:gen8_reset_engines+0x394} *ERROR* vecs'0 reset reques
 t timed out: {request: 00000001, RESET_CTL: 00000001}
 [   4.6860750] [drm] Initialized i915 1.6.0 20200114 for i915drmkms0 on minor 0
 [   5.5760841] intelfb0 at i915drmkms0
 [   5.5760841] [drm] DRM_I915_DEBUG enabled
 [   5.5760841] [drm] DRM_I915_DEBUG_GEM enabled
 [   5.5760841] intelfb0: framebuffer at 0xb8020000, size 1920x1200, depth 32, st
 ride 7680
 [   5.5760841] panic: mutex_vector_enter,515: uninitialized lock (lock=0xffffc60
 010a6d460, from=ffffffff807eff7c)
 [   5.5760841] cpu1: Begin traceback...
 [   5.5760841] vpanic() at netbsd:vpanic+0x173
 [   5.5760841] panic() at netbsd:panic+0x3c
 [   5.5760841] lockdebug_wantlock() at netbsd:lockdebug_wantlock+0x180
 [   5.5860788] mutex_enter() at netbsd:mutex_enter+0x213
 [   5.5860788] intel_rps_mark_interactive() at netbsd:intel_rps_mark_interactive
 +0x1e
 [   5.5860788] intel_prepare_plane_fb() at netbsd:intel_prepare_plane_fb+0x230
 [   5.5860788] drm_atomic_helper_prepare_planes() at netbsd:drm_atomic_helper_pr
 epare_planes+0x8f
 [   5.5860788] intel_atomic_commit() at netbsd:intel_atomic_commit+0xb8
 [   5.5960744] drm_client_modeset_commit_atomic() at netbsd:drm_client_modeset_c
 ommit_atomic+0x19a
 [   5.5960744] drm_client_modeset_commit_force() at netbsd:drm_client_modeset_co
 mmit_force+0x4c
 [   5.5960744] drm_fb_helper_restore_fbdev_mode_unlocked() at netbsd:drm_fb_help
 er_restore_fbdev_mode_unlocked+0x55
 [   5.5960744] drmfb_genfb_setmode() at netbsd:drmfb_genfb_setmode+0x17
 [   5.5960744] genfb_attach() at netbsd:genfb_attach+0x1af
 [   5.6060732] drmfb_attach() at netbsd:drmfb_attach+0x23c
 [   5.6060732] intelfb_attach_task() at netbsd:intelfb_attach_task+0x4e
 [   5.6060732] i915drmkms_attach_real() at netbsd:i915drmkms_attach_real+0xb6
 [   5.6060732] config_mountroot_thread() at netbsd:config_mountroot_thread+0x60
 [   5.6060732] cpu1: End traceback...
 [   5.6060732] fatal breakpoint trap in supervisor mode
 [   5.6060732] trap type 1 code 0 rip 0xffffffff80235425 cs 0x8 rflags 0x202 cr2
  0 ilevel 0x8 rsp 0xffffc6023a6e6a20
 [   5.6060732] curlwp 0xffff90e0a73ea9c0 pid 0.219 lowest kstack 0xffffc6023a6e2
 2c0
 
 [   5.6060732] dumping to dev 0,1 (offset=179551, size=2074708):
 [   5.6060732] dump
 
 					-jarle
 -- 
 "Then I realized that I had spelled '-' wrong."
 				-- John Whitmore learns APL