On Thu, Jan 09, 2025 at 12:29:05PM +0100, Robert NEMKIN wrote:
> Hello everybody!
>
> I saw a possibly kernel bug.
> It's a debian trixie with kernel compiled by myself.
> With 6.12.6 everything seems to be ok.
> With 6.12.7 and 6.12.8 I saw crasges in dmesg.
> I attached dmesg of some tries and lsusb.
>
> Thanks for any hints.
>
> Robert
>
> dec 28 07:44:08 brick kernel: ------------[ cut here ]------------
> dec 28 07:44:08 brick kernel: WARNING: CPU: 0 PID: 296 at kernel/trace/trace_events.c:577 trace_event_raw_init+0x159/0x670
> dec 28 07:44:08 brick kernel: Modules linked in: xe(+) gpu_sched drm_suballoc_helper drm_gpuvm drm_exec drm_ttm_helper hid_generic usbhid hid i915 i2c_algo_bit drm_buddy drm_display_helper cec xhci_pci rc_core xhci_hcd ahci ttm libahci drm_kms_helper libata nvme intel_lpss_pci crc32_pclmul usbcore thunderbolt crc32c_intel drm scsi_mod igc intel_lpss i2c_i801 nvme_core idma64 scsi_common i2c_smbus video usb_common fan wmi
This isn't a USB issue, it's a tracing core issue with the xe driver.
It is fixed in commit afc6717628f9 ("tracing: Have process_string() also
allow arrays") which will be in the next 6.12.9 kernel release in a few
minutes.
It's a harmless warning for now, no need to worry.
thanks,
greg k-h
