On Wed, Jun 24, 2020 at 01:21:03PM +0100, Chris Wilson wrote: > Quoting Jason Gunthorpe (2020-06-24 13:10:53) > > On Wed, Jun 24, 2020 at 09:02:47AM +0100, Chris Wilson wrote: > > > When direct reclaim enters the shrinker and tries to reclaim pages, it > > > has to opportunitically unmap them [try_to_unmap_one]. For direct > > > reclaim, the calling context is unknown and may include attempts to > > > unmap one page of a dma object while attempting to allocate more pages > > > for that object. Pass the information along that we are inside an > > > opportunistic unmap that can allow that page to remain referenced and > > > mapped, and let the callback opt in to avoiding a recursive wait. > > > > i915 should already not be holding locks shared with the notifiers > > across allocations that can trigger reclaim. This is already required > > to use notifiers correctly anyhow - why do we need something in the > > notifiers? > > for (n = 0; n < num_pages; n++) > pin_user_page() > > may call try_to_unmap_page from the lru shrinker for [0, n-1]. Yes, of course you can't hold any locks that intersect with notifiers across pin_user_page()/get_user_page() It has always been that way. I consolidated all this tricky locking into interval notifiers, maybe updating i915 to use them will give it a solution. I looked at it once, it was straightforward enough until it got to all the #ifdefery > We're in the middle of allocating the object, how are we best to untangle > that? I don't know anything about i915, but this is clearly i915 not using notifiers properly, it needs proper fixing, not hacking up notifiers. Jason
