Krunner GPU process launch failed: error_code=1002 --- FATAL:gpu_data_manager_impl_private.cc(423)] GPU process isn't usable. Goodbye

On Debian 12, 6.1.0-37-amd64, Plasma 5.27.5

happened two times that it closed many apps which has its configuration located on my system drive (which I have tried to snapshot using Timeshift and that always failed, destination is ZFS external drive and it always started using system drive disk space/exhausted it).

The journal:

Jul 15 10:25:07 * krunner[1948841]: [2025-07-15 10:25:07.944] [ ] [debug] kill_callback_handler 15
Jul 15 10:25:08 * systemd[3845]: plasma-krunner.service: Consumed 38.808s CPU time.
░░ Subject: Resources consumed by unit runtime
░░ Defined-By: systemd
░░ Support: Debian -- User Support
░░
░░ The unit UNIT completed and consumed the indicated resources.
Jul 15 10:25:09 * krunner[1952412]: [1952412:0715/102509.134085:ERROR:network_service_instance_impl.cc(613)] Network service crashed, restarting service.
Jul 15 10:25:09 * krunner[1952412]: [1952412:0715/102509.397310:ERROR:zygote_communication_linux.cc(297)] Failed to send GetTerminationStatus message to zygote
Jul 15 10:25:09 * krunner[1952411]: [1952411:0715/102509.430650:ERROR:network_service_instance_impl.cc(975)] Network service crashed, restarting service.
Jul 15 10:25:09 * krunner[1952412]: [1952412:0715/102509.497970:ERROR:gpu_process_host.cc(976)] GPU process launch failed: error_code=1002
Jul 15 10:25:09 * krunner[1952412]: [1952412:0715/102509.502529:ERROR:gpu_process_host.cc(976)] GPU process launch failed: error_code=1002
Jul 15 10:25:09 * krunner[1952412]: [1952412:0715/102509.503282:ERROR:gpu_process_host.cc(976)] GPU process launch failed: error_code=1002
Jul 15 10:25:09 * krunner[1952412]: [1952412:0715/102509.503716:ERROR:gpu_process_host.cc(976)] GPU process launch failed: error_code=1002
Jul 15 10:25:09 * krunner[1952412]: [1952412:0715/102509.504072:ERROR:gpu_process_host.cc(976)] GPU process launch failed: error_code=1002
Jul 15 10:25:09 * krunner[1952412]: [1952412:0715/102509.504512:ERROR:gpu_process_host.cc(976)] GPU process launch failed: error_code=1002
Jul 15 10:25:09 * kernel: traps: Proton Pass[1952412] trap int3 ip:55a81341b8de sp:7ffd9fbb0340 error:0 in Proton Pass[55a80f46f000+8997000]
Jul 15 10:25:09 * krunner[1952412]: [1952412:0715/102509.509321:ERROR:gpu_process_host.cc(976)] GPU process launch failed: error_code=1002
Jul 15 10:25:09 * krunner[1952412]: [1952412:0715/102509.509705:ERROR:gpu_process_host.cc(976)] GPU process launch failed: error_code=1002
Jul 15 10:25:09 * krunner[1952412]: [1952412:0715/102509.510087:ERROR:gpu_process_host.cc(976)] GPU process launch failed: error_code=1002
Jul 15 10:25:09 * krunner[1952412]: [1952412:0715/102509.510096:FATAL:gpu_data_manager_impl_private.cc(423)] GPU process isn’t usable. Goodbye.

Graphics details here.

This crash happens always when i cancel the Timeshift system backup task which is “stuck” at first stage “Estimating System Size”. A system snapshot location is a ZFS external drive.