On 6/5/25 01:22, Bjoern A. Zeeb wrote:
On Wed, 4 Jun 2025, Steve Kargl wrote:
On 6/4/25 13:11, Steve Kargl wrote:
On Tue, Jun 03, 2025 at 06:45:31PM -0700, Steve Kargl wrote:
I was able to rebuild world/kernel for a 2025-04-15
src/ (aka adc33d3288) and rebuilt drm-515-kmod and
gpu-firmware. The system boots and I can kldload
radeonkmd.ko. startx brings up the expected desktop.
Onward to src/ from 2025-05-01 (aka 8d136fb027).
The problem appears between adc33d3288 and
8d136fb027. There are a few commits in this
range that change the linuxkpi.
Sorting away wireless changes from sys/compat/.. here's what's left:
% git log --oneline adc33d3288..8d136fb027 sys/compat/ | grep -v 802.11
| grep -v skbuff | grep -v wsum | grep -v ASMEDIA
325aa4dbd10d linuxkpi: Introduce a properly typed jiffies <<
jiffies changed to proper type
8b51cd07f69e LinuxKPI: define time64_t << new typedef
28efbf9d2f67 LinuxKPI: add dummy header file linux/unaligned.h <<
empty header file
e29d72ac3ddd LinuxKPI: pci: add pci_info() << new macro
for logging
f94d7319540b LinuxKPI: sysfs: implement sysfs_match_string() <<
new macro/func
6841b9987e83 LinuxKPI: add container_of_const() << new macro
69880fede78f LinuxKPI: extend struct and enum for leds << LED
additions to struct/enum (unused)
059136a95aca LinuxKPI: add cleanup.h to mutex.h <<
#include added
15581af7c2d3 exec: Remove parameter 'segflg' from exec_copyin_args()
<< linuxolator
97f3a1565d88 linuxkpi: use iterator in zap_vma_ptes << VM
On a known good version, did you go twice through the dump_stack() as
well according to dmesg? Would give another hint if we now hit
that resume path for other reasons maybe?
I forgot to check for the dump_stack() output. As I iterate over the
above list of hash tags, I make a note. I suspect that one these
has uncovered a latent bug in drm.
--
steve
