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

dma-buf: clarify dma_fence_ops->wait documentation

This callback is pretty much deprecated and should not be used by new implementations.

Clarify that in the documentation as well.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20210901120240.7339-2-christian.koenig@amd.com

+3 -7
+3 -7
include/linux/dma-fence.h
··· 214 214 * Custom wait implementation, defaults to dma_fence_default_wait() if 215 215 * not set. 216 216 * 217 - * The dma_fence_default_wait implementation should work for any fence, as long 218 - * as @enable_signaling works correctly. This hook allows drivers to 219 - * have an optimized version for the case where a process context is 220 - * already available, e.g. if @enable_signaling for the general case 221 - * needs to set up a worker thread. 217 + * Deprecated and should not be used by new implementations. Only used 218 + * by existing implementations which need special handling for their 219 + * hardware reset procedure. 222 220 * 223 221 * Must return -ERESTARTSYS if the wait is intr = true and the wait was 224 222 * interrupted, and remaining jiffies if fence has signaled, or 0 if wait 225 223 * timed out. Can also return other error values on custom implementations, 226 224 * which should be treated as if the fence is signaled. For example a hardware 227 225 * lockup could be reported like that. 228 - * 229 - * This callback is optional. 230 226 */ 231 227 signed long (*wait)(struct dma_fence *fence, 232 228 bool intr, signed long timeout);