[PATCH 2/3] drm/meson: explicitly remove aggregate driver at module unload time
Neil Armstrong
neil.armstrong at linaro.org
Mon Sep 19 12:57:48 UTC 2022
On 19/09/2022 03:09, Adrián Larumbe wrote:
> Because component_master_del wasn't being called when unloading the
> meson_drm module, the aggregate device would linger forever in the global
> aggregate_devices list. That means when unloading and reloading the
> meson_dw_hdmi module, component_add would call into
> try_to_bring_up_aggregate_device and find the unbound meson_drm aggregate
> device.
>
> This would in turn dereference some of the aggregate_device's struct
> entries which point to memory automatically freed by the devres API when
> unbinding the aggregate device from meson_drv_unbind, and trigger an
> use-after-free bug:
>
> [ +0.000014] =============================================================
> [ +0.000007] BUG: KASAN: use-after-free in find_components+0x468/0x500
> [ +0.000017] Read of size 8 at addr ffff000006731688 by task modprobe/2536
> [ +0.000018] CPU: 4 PID: 2536 Comm: modprobe Tainted: G C O 5.19.0-rc6-lrmbkasan+ #1
> [ +0.000010] Hardware name: Hardkernel ODROID-N2Plus (DT)
> [ +0.000008] Call trace:
> [ +0.000005] dump_backtrace+0x1ec/0x280
> [ +0.000011] show_stack+0x24/0x80
> [ +0.000007] dump_stack_lvl+0x98/0xd4
> [ +0.000010] print_address_description.constprop.0+0x80/0x520
> [ +0.000011] print_report+0x128/0x260
> [ +0.000007] kasan_report+0xb8/0xfc
> [ +0.000007] __asan_report_load8_noabort+0x3c/0x50
> [ +0.000009] find_components+0x468/0x500
> [ +0.000008] try_to_bring_up_aggregate_device+0x64/0x390
> [ +0.000009] __component_add+0x1dc/0x49c
> [ +0.000009] component_add+0x20/0x30
> [ +0.000008] meson_dw_hdmi_probe+0x28/0x34 [meson_dw_hdmi]
> [ +0.000013] platform_probe+0xd0/0x220
> [ +0.000008] really_probe+0x3ac/0xa80
> [ +0.000008] __driver_probe_device+0x1f8/0x400
> [ +0.000008] driver_probe_device+0x68/0x1b0
> [ +0.000008] __driver_attach+0x20c/0x480
> [ +0.000009] bus_for_each_dev+0x114/0x1b0
> [ +0.000007] driver_attach+0x48/0x64
> [ +0.000009] bus_add_driver+0x390/0x564
> [ +0.000007] driver_register+0x1a8/0x3e4
> [ +0.000009] __platform_driver_register+0x6c/0x94
> [ +0.000007] meson_dw_hdmi_platform_driver_init+0x30/0x1000 [meson_dw_hdmi]
> [ +0.000014] do_one_initcall+0xc4/0x2b0
> [ +0.000008] do_init_module+0x154/0x570
> [ +0.000010] load_module+0x1a78/0x1ea4
> [ +0.000008] __do_sys_init_module+0x184/0x1cc
> [ +0.000008] __arm64_sys_init_module+0x78/0xb0
> [ +0.000008] invoke_syscall+0x74/0x260
> [ +0.000008] el0_svc_common.constprop.0+0xcc/0x260
> [ +0.000009] do_el0_svc+0x50/0x70
> [ +0.000008] el0_svc+0x68/0x1a0
> [ +0.000009] el0t_64_sync_handler+0x11c/0x150
> [ +0.000009] el0t_64_sync+0x18c/0x190
>
> [ +0.000014] Allocated by task 902:
> [ +0.000007] kasan_save_stack+0x2c/0x5c
> [ +0.000009] __kasan_kmalloc+0x90/0xd0
> [ +0.000007] __kmalloc_node+0x240/0x580
> [ +0.000010] memcg_alloc_slab_cgroups+0xa4/0x1ac
> [ +0.000010] memcg_slab_post_alloc_hook+0xbc/0x4c0
> [ +0.000008] kmem_cache_alloc_node+0x1d0/0x490
> [ +0.000009] __alloc_skb+0x1d4/0x310
> [ +0.000010] alloc_skb_with_frags+0x8c/0x620
> [ +0.000008] sock_alloc_send_pskb+0x5ac/0x6d0
> [ +0.000010] unix_dgram_sendmsg+0x2e0/0x12f0
> [ +0.000010] sock_sendmsg+0xcc/0x110
> [ +0.000007] sock_write_iter+0x1d0/0x304
> [ +0.000008] new_sync_write+0x364/0x460
> [ +0.000007] vfs_write+0x420/0x5ac
> [ +0.000008] ksys_write+0x19c/0x1f0
> [ +0.000008] __arm64_sys_write+0x78/0xb0
> [ +0.000007] invoke_syscall+0x74/0x260
> [ +0.000008] el0_svc_common.constprop.0+0x1a8/0x260
> [ +0.000009] do_el0_svc+0x50/0x70
> [ +0.000007] el0_svc+0x68/0x1a0
> [ +0.000008] el0t_64_sync_handler+0x11c/0x150
> [ +0.000008] el0t_64_sync+0x18c/0x190
>
> [ +0.000013] Freed by task 2509:
> [ +0.000008] kasan_save_stack+0x2c/0x5c
> [ +0.000007] kasan_set_track+0x2c/0x40
> [ +0.000008] kasan_set_free_info+0x28/0x50
> [ +0.000008] ____kasan_slab_free+0x128/0x1d4
> [ +0.000008] __kasan_slab_free+0x18/0x24
> [ +0.000007] slab_free_freelist_hook+0x108/0x230
> [ +0.000010] kfree+0x110/0x35c
> [ +0.000008] release_nodes+0xf0/0x16c
> [ +0.000008] devres_release_all+0xfc/0x180
> [ +0.000008] device_unbind_cleanup+0x24/0x164
> [ +0.000008] device_release_driver_internal+0x3e8/0x5b0
> [ +0.000010] driver_detach+0xac/0x1b0
> [ +0.000008] bus_remove_driver+0x158/0x29c
> [ +0.000008] driver_unregister+0x70/0xb0
> [ +0.000009] platform_driver_unregister+0x20/0x2c
> [ +0.000007] 0xffff800003722d98
> [ +0.000012] __do_sys_delete_module+0x288/0x400
> [ +0.000009] __arm64_sys_delete_module+0x5c/0x80
> [ +0.000008] invoke_syscall+0x74/0x260
> [ +0.000008] el0_svc_common.constprop.0+0xcc/0x260
> [ +0.000008] do_el0_svc+0x50/0x70
> [ +0.000007] el0_svc+0x68/0x1a0
> [ +0.000008] el0t_64_sync_handler+0x11c/0x150
> [ +0.000009] el0t_64_sync+0x18c/0x190
>
> [ +0.000013] Last potentially related work creation:
> [ +0.000007] kasan_save_stack+0x2c/0x5c
> [ +0.000007] __kasan_record_aux_stack+0xb8/0xf0
> [ +0.000009] kasan_record_aux_stack_noalloc+0x14/0x20
> [ +0.000008] insert_work+0x54/0x290
> [ +0.000009] __queue_work+0x48c/0xd24
> [ +0.000008] queue_work_on+0x90/0x11c
> [ +0.000008] call_usermodehelper_exec+0x188/0x404
> [ +0.000010] kobject_uevent_env+0x5a8/0x794
> [ +0.000010] kobject_uevent+0x14/0x20
> [ +0.000008] driver_register+0x230/0x3e4
> [ +0.000009] __platform_driver_register+0x6c/0x94
> [ +0.000007] gxbb_driver_init+0x28/0x34
> [ +0.000010] do_one_initcall+0xc4/0x2b0
> [ +0.000008] do_initcalls+0x20c/0x24c
> [ +0.000010] kernel_init_freeable+0x22c/0x278
> [ +0.000009] kernel_init+0x3c/0x170
> [ +0.000008] ret_from_fork+0x10/0x20
>
> [ +0.000013] The buggy address belongs to the object at ffff000006731600
> which belongs to the cache kmalloc-256 of size 256
> [ +0.000009] The buggy address is located 136 bytes inside of
> 256-byte region [ffff000006731600, ffff000006731700)
>
> [ +0.000015] The buggy address belongs to the physical page:
> [ +0.000008] page:fffffc000019cc00 refcount:1 mapcount:0 mapping:0000000000000000 index:0xffff000006730a00 pfn:0x6730
> [ +0.000011] head:fffffc000019cc00 order:2 compound_mapcount:0 compound_pincount:0
> [ +0.000008] flags: 0xffff00000010200(slab|head|node=0|zone=0|lastcpupid=0xffff)
> [ +0.000016] raw: 0ffff00000010200 fffffc00000c3d08 fffffc0000ef2b08 ffff000000002680
> [ +0.000009] raw: ffff000006730a00 0000000000150014 00000001ffffffff 0000000000000000
> [ +0.000006] page dumped because: kasan: bad access detected
>
> [ +0.000011] Memory state around the buggy address:
> [ +0.000007] ffff000006731580: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
> [ +0.000007] ffff000006731600: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> [ +0.000007] >ffff000006731680: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> [ +0.000007] ^
> [ +0.000006] ffff000006731700: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
> [ +0.000007] ffff000006731780: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
> [ +0.000006] ==================================================================
>
> Fix by adding 'remove' driver callback for meson-drm, and explicitly deleting the
> aggregate device.
>
> Signed-off-by: Adrián Larumbe <adrian.larumbe at collabora.com>
> ---
> drivers/gpu/drm/meson/meson_drv.c | 8 ++++++++
> 1 file changed, 8 insertions(+)
>
> diff --git a/drivers/gpu/drm/meson/meson_drv.c b/drivers/gpu/drm/meson/meson_drv.c
> index 8da454a17b77..f3da1c214a7c 100644
> --- a/drivers/gpu/drm/meson/meson_drv.c
> +++ b/drivers/gpu/drm/meson/meson_drv.c
> @@ -490,6 +490,13 @@ static int meson_drv_probe(struct platform_device *pdev)
> return 0;
> };
>
> +static int meson_drv_remove(struct platform_device *pdev)
> +{
> + component_master_del(&pdev->dev, &meson_drv_master_ops);
> +
> + return 0;
> +}
> +
> static struct meson_drm_match_data meson_drm_gxbb_data = {
> .compat = VPU_COMPATIBLE_GXBB,
> };
> @@ -527,6 +534,7 @@ static const struct dev_pm_ops meson_drv_pm_ops = {
>
> static struct platform_driver meson_drm_platform_driver = {
> .probe = meson_drv_probe,
> + .remove = meson_drv_remove,
> .shutdown = meson_drv_shutdown,
> .driver = {
> .name = "meson-drm",
Reviewed-by: Neil Armstrong <neil.armstrong at linaro.org>
More information about the dri-devel
mailing list