[PATCH v6 00/22] Add generic memory shrinker to VirtIO-GPU and Panfrost DRM drivers

Dmitry Osipenko dmitry.osipenko at collabora.com
Tue Jun 28 12:51:11 UTC 2022


On 6/28/22 15:31, Robin Murphy wrote:
> [  100.511411]
> ==================================================================
> [  100.511419] BUG: KASAN: use-after-free in irq_work_single+0xa4/0x110
> [  100.511445] Write of size 4 at addr ffff0000107f5830 by task
> glmark2-es2-drm/280
> [  100.511458]
> [  100.511464] CPU: 1 PID: 280 Comm: glmark2-es2-drm Not tainted
> 5.19.0-rc3+ #400
> [  100.511479] Hardware name: ARM LTD ARM Juno Development Platform/ARM
> Juno Development Platform, BIOS EDK II Sep  3 2019
> [  100.511489] Call trace:
> [  100.511494]  dump_backtrace+0x1e4/0x1f0
> [  100.511512]  show_stack+0x20/0x70
> [  100.511523]  dump_stack_lvl+0x8c/0xb8
> [  100.511543]  print_report+0x16c/0x668
> [  100.511559]  kasan_report+0x80/0x208
> [  100.511574]  kasan_check_range+0x100/0x1b8
> [  100.511590]  __kasan_check_write+0x34/0x60
> [  100.511607]  irq_work_single+0xa4/0x110
> [  100.511619]  irq_work_run_list+0x6c/0x88
> [  100.511632]  irq_work_run+0x28/0x48
> [  100.511644]  ipi_handler+0x254/0x468
> [  100.511664]  handle_percpu_devid_irq+0x11c/0x518
> [  100.511681]  generic_handle_domain_irq+0x50/0x70
> [  100.511699]  gic_handle_irq+0xd4/0x118
> [  100.511711]  call_on_irq_stack+0x2c/0x58
> [  100.511725]  do_interrupt_handler+0xc0/0xc8
> [  100.511741]  el1_interrupt+0x40/0x68
> [  100.511754]  el1h_64_irq_handler+0x18/0x28
> [  100.511767]  el1h_64_irq+0x64/0x68
> [  100.511778]  irq_work_queue+0xc0/0xd8
> [  100.511790]  drm_sched_entity_fini+0x2c4/0x3b0
> [  100.511805]  drm_sched_entity_destroy+0x2c/0x40
> [  100.511818]  panfrost_job_close+0x44/0x1c0
> [  100.511833]  panfrost_postclose+0x38/0x60
> [  100.511845]  drm_file_free.part.0+0x33c/0x4b8
> [  100.511862]  drm_close_helper.isra.0+0xc0/0xd8
> [  100.511877]  drm_release+0xe4/0x1e0
> [  100.511891]  __fput+0xf8/0x390
> [  100.511904]  ____fput+0x18/0x28
> [  100.511917]  task_work_run+0xc4/0x1e0
> [  100.511929]  do_exit+0x554/0x1168
> [  100.511945]  do_group_exit+0x60/0x108
> [  100.511960]  __arm64_sys_exit_group+0x34/0x38
> [  100.511977]  invoke_syscall+0x64/0x180
> [  100.511993]  el0_svc_common.constprop.0+0x13c/0x170
> [  100.512012]  do_el0_svc+0x48/0xe8
> [  100.512028]  el0_svc+0x5c/0xe0
> [  100.512038]  el0t_64_sync_handler+0xb8/0xc0
> [  100.512051]  el0t_64_sync+0x18c/0x190
> [  100.512064]

This one shall be fixed by [1] that is not in the RC kernel yet, please
use linux-next.

[1]
https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/commit/?h=next-20220628&id=7d64c40a7d96190d9d06e240305389e025295916

-- 
Best regards,
Dmitry


More information about the amd-gfx mailing list