tjh.dev / kernel
Linux kernel mirror (for testing) git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel os linux
1
fork

Configure Feed

Select the types of activity you want to include in your feed.

drm/panthor: Flush shmem writes before mapping buffers CPU-uncached

The shmem layer zeroes out the new pages using cached mappings, and if
we don't CPU-flush we might leave dirty cachelines behind, leading to
potential data leaks and/or asynchronous buffer corruption when dirty
cachelines are evicted.

Fixes: 8a1cc07578bf ("drm/panthor: Add GEM logical block")
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Steven Price <steven.price@arm.com>
Reviewed-by: Liviu Dudau <liviu.dudau@arm.com>
Signed-off-by: Steven Price <steven.price@arm.com>
Link: https://patch.msgid.link/20251107171214.1186299-1-boris.brezillon@collabora.com

authored by

Boris Brezillon and committed by
Steven Price 576c930e eef295a8

+18
+18
drivers/gpu/drm/panthor/panthor_gem.c
··· 288 288 289 289 panthor_gem_debugfs_set_usage_flags(bo, 0); 290 290 291 + /* If this is a write-combine mapping, we query the sgt to force a CPU 292 + * cache flush (dma_map_sgtable() is called when the sgt is created). 293 + * This ensures the zero-ing is visible to any uncached mapping created 294 + * by vmap/mmap. 295 + * FIXME: Ideally this should be done when pages are allocated, not at 296 + * BO creation time. 297 + */ 298 + if (shmem->map_wc) { 299 + struct sg_table *sgt; 300 + 301 + sgt = drm_gem_shmem_get_pages_sgt(shmem); 302 + if (IS_ERR(sgt)) { 303 + ret = PTR_ERR(sgt); 304 + goto out_put_gem; 305 + } 306 + } 307 + 291 308 /* 292 309 * Allocate an id of idr table where the obj is registered 293 310 * and handle has the id what user can see. ··· 313 296 if (!ret) 314 297 *size = bo->base.base.size; 315 298 299 + out_put_gem: 316 300 /* drop reference from allocate - handle holds it now. */ 317 301 drm_gem_object_put(&shmem->base); 318 302