On 3/29/21 6:20 PM, Song Bao Hua (Barry Song) wrote: > > >> -----Original Message----- >> From: Mike Kravetz [mailto:[email protected]] >> Sent: Tuesday, March 30, 2021 12:24 PM >> To: [email protected]; [email protected] >> Cc: Roman Gushchin <[email protected]>; Michal Hocko <[email protected]>; Shakeel >> Butt >> <[email protected]>; Oscar Salvador <[email protected]>; David Hildenbrand >> <[email protected]>; Muchun Song <[email protected]>; David Rientjes >> <[email protected]>; linmiaohe <[email protected]>; Peter Zijlstra >> <[email protected]>; Matthew Wilcox <[email protected]>; HORIGUCHI NAOYA >> <[email protected]>; Aneesh Kumar K . V <[email protected]>; >> Waiman Long <[email protected]>; Peter Xu <[email protected]>; Mina Almasry >> <[email protected]>; Hillf Danton <[email protected]>; Joonsoo Kim >> <[email protected]>; Song Bao Hua (Barry Song) >> <[email protected]>; Will Deacon <[email protected]>; Andrew Morton >> <[email protected]>; Mike Kravetz <[email protected]> >> Subject: [PATCH v2 1/8] mm/cma: change cma mutex to irq safe spinlock >> >> Ideally, cma_release could be called from any context. However, that is >> not possible because a mutex is used to protect the per-area bitmap. >> Change the bitmap to an irq safe spinlock. >> >> Signed-off-by: Mike Kravetz <[email protected]> > > It seems mutex_lock is locking some areas with bitmap operations which > should be safe to atomic context. > > Reviewed-by: Barry Song <[email protected]>
Thanks Barry, Not sure if you saw questions from Michal in previous series? There was some concern from Joonsoo in the past about lock hold time due to bitmap scans. You may have some insight into the typical size of CMA areas on arm64. I believe the calls to set up the areas specify one bit per page. -- Mike Kravetz
