kernel.git/drivers/gpu/drm/ttm, branch linux-5.13.y

drm/ttm: add a check against null pointer dereference

2021-07-31T06:13:46Z

[ Upstream commit 9e5c772954406829e928dbe59891d08938ead04b ] When calling ttm_range_man_fini(), 'man' may be uninitialized, which may cause a null pointer dereference bug. Fix this by checking if it is a null pointer. This log reveals it: [ 7.902580 ] BUG: kernel NULL pointer dereference, address: 0000000000000058 [ 7.905721 ] RIP: 0010:ttm_range_man_fini+0x40/0x160 [ 7.911826 ] Call Trace: [ 7.911826 ] radeon_ttm_fini+0x167/0x210 [ 7.911826 ] radeon_bo_fini+0x15/0x40 [ 7.913767 ] rs400_fini+0x55/0x80 [ 7.914358 ] radeon_device_fini+0x3c/0x140 [ 7.914358 ] radeon_driver_unload_kms+0x5c/0xe0 [ 7.914358 ] radeon_driver_load_kms+0x13a/0x200 [ 7.914358 ] ? radeon_driver_unload_kms+0xe0/0xe0 [ 7.914358 ] drm_dev_register+0x1db/0x290 [ 7.914358 ] radeon_pci_probe+0x16a/0x230 [ 7.914358 ] local_pci_probe+0x4a/0xb0 Signed-off-by: Zheyu Ma Reviewed-by: Christian König Link: https://patchwork.freedesktop.org/patch/msgid/1626274459-8148-1-git-send-email-zheyuma97@gmail.com Signed-off-by: Christian König Signed-off-by: Sasha Levin

drm/ttm: Force re-init if ttm_global_init() fails

2021-07-28T12:37:32Z

[ Upstream commit 235c3610d5f02ee91244239b43cd9ae8b4859dff ] If we have a failure, decrement the reference count so that the next call to ttm_global_init() will actually do something instead of assume everything is all set up. Signed-off-by: Jason Ekstrand Fixes: 62b53b37e4b1 ("drm/ttm: use a static ttm_bo_global instance") Reviewed-by: Christian König Link: https://patchwork.freedesktop.org/patch/msgid/20210720181357.2760720-5-jason@jlekstrand.net Signed-off-by: Christian König Signed-off-by: Sasha Levin

drm/ttm: fix deref of bo->ttm without holding the lock v2

2021-06-08T11:24:15Z

We need to grab the resv lock first before doing that check. v2 (chk): simplify the change for -fixes Signed-off-by: Christian König Signed-off-by: Thomas Hellström Reviewed-by: Huang Rui Link: https://patchwork.freedesktop.org/patch/msgid/20210528130041.1683-1-christian.koenig@amd.com

drm/ttm: Skip swapout if ttm object is not populated

2021-05-26T13:02:47Z

Swapping a ttm object which has no backend pages makes no sense. Suggested-by: Christian König Signed-off-by: xinhui pan Reviewed-by: Christian König Link: https://patchwork.freedesktop.org/patch/msgid/20210521083112.33176-1-xinhui.pan@amd.com CC: stable@kernel.org Signed-off-by: Christian König

Merge tag 'drm-misc-next-fixes-2021-04-29' of git://anongit.freedesktop.org/drm/drm-misc into drm-next

2021-04-30T00:13:48Z

Two patches in drm-misc-next-fixes this week, one to fix the error handling in TTM when a BO can't be swapped out and one to prevent a wrong dereference in efifb. Signed-off-by: Dave Airlie From: Maxime Ripard Link: https://patchwork.freedesktop.org/patch/msgid/20210429090308.k3fuqvenf6vupfmg@gilmour

drm/ttm: fix error handling if no BO can be swapped out v4

2021-04-23T12:20:38Z

In case that all pre-allocated BOs are busy, just continue to populate BOs since likely half of system memory in total is still free. v4 (chk): fix code moved to VMWGFX as well Signed-off-by: Shiwu Zhang Reviewed-by: Christian König Signed-off-by: Christian König Link: https://patchwork.freedesktop.org/patch/msgid/20210422115757.3946-1-christian.koenig@amd.com

Merge tag 'drm-misc-next-fixes-2021-04-22' of git://anongit.freedesktop.org/drm/drm-misc into drm-next

2021-04-23T03:53:07Z

A few fixes for the next merge window, with some build fixes for anx7625 and lt8912b bridges, incorrect error handling for lt8912b and TTM, and one fix for TTM page limit accounting. Signed-off-by: Dave Airlie From: Maxime Ripard Link: https://patchwork.freedesktop.org/patch/msgid/20210422163329.dvbuwre3akwdmzjt@gilmour

drm/ttm: Don't count pages in SG BOs against pages_limit

2021-04-21T13:35:20Z

Pages in SG BOs were not allocated by TTM. So don't count them against TTM's pages limit. Signed-off-by: Felix Kuehling Reviewed-by: Christian König Signed-off-by: Christian König Link: https://patchwork.freedesktop.org/patch/msgid/20210414064804.29356-9-Felix.Kuehling@amd.com

drm/ttm: fix return value check

2021-04-21T13:35:11Z

The function returns the number of swapped pages here. Only abort when we get a negative error code. Signed-off-by: Christian König Reviewed-by: Matthew Auld Link: https://patchwork.freedesktop.org/patch/msgid/20210409130113.1459-1-christian.koenig@amd.com

drm/ttm: ioremap buffer according to TTM mem caching setting

2021-04-08T10:21:22Z

If tbo.mem.bus.caching is cached, buffer is intended to be mapped as cached from CPU. Map it with ioremap_cache. This wasn't necessary before as device memory was never mapped as cached from CPU side. It becomes necessary for aldebaran as device memory is mapped cached from CPU. Signed-off-by: Oak Zeng Reviewed-by: Christian Konig Signed-off-by: Alex Deucher Link: https://patchwork.freedesktop.org/patch/msgid/1614638628-10508-1-git-send-email-Oak.Zeng@amd.com Signed-off-by: Christian König