On Mon, 6 Oct 2025 13:30:59 +0200 Alice Ryhl <[email protected]> wrote:
> On Wed, Oct 1, 2025 at 12:41 PM Alice Ryhl <[email protected]> wrote: > > > > When using GPUVM in immediate mode, it is necessary to call > > drm_gpuvm_unlink() from the fence signalling critical path. However, > > unlink may call drm_gpuvm_bo_put(), which causes some challenges: > > > > 1. drm_gpuvm_bo_put() often requires you to take resv locks, which you > > can't do from the fence signalling critical path. > > 2. drm_gpuvm_bo_put() calls drm_gem_object_put(), which is often going > > to be unsafe to call from the fence signalling critical path. > > > > To solve these issues, add a deferred version of drm_gpuvm_unlink() that > > adds the vm_bo to a deferred cleanup list, and then clean it up later. > > > > The new methods take the GEMs GPUVA lock internally rather than letting > > the caller do it because it also needs to perform an operation after > > releasing the mutex again. This is to prevent freeing the GEM while > > holding the mutex (more info as comments in the patch). This means that > > the new methods can only be used with DRM_GPUVM_IMMEDIATE_MODE. > > > > Reviewed-by: Boris Brezillon <[email protected]> > > Signed-off-by: Alice Ryhl <[email protected]> > > In this version, I got rid of the kref_put_mutex() usage, but I > realized that maybe we should bring it back. With the current code, > it's actually possible to observe a zombie vm_bo in the GEM's list > because we don't drop the refcount while holding the mutex. Alright, let's get back to the kref_put_mutex() approach then.
