Re: [PATCH v4 0/4] mm/gup: page unpining improvements

2021-02-21 Thread Christoph Hellwig
On Thu, Feb 18, 2021 at 03:33:39PM +, Joao Martins wrote: > in a bvec at once? e.g. something like from this: > > bio_for_each_segment_all(bvec, bio, iter_all) { > if (mark_dirty && !PageCompound(bvec->bv_page)) > set_page_dirty_lock(bvec->bv_pag

Re: [PATCH v4 0/4] mm/gup: page unpining improvements

2021-02-18 Thread Joao Martins
On 2/18/21 3:33 PM, Joao Martins wrote: > On 2/18/21 7:24 AM, Christoph Hellwig wrote: >> On Fri, Feb 12, 2021 at 01:08:39PM +, Joao Martins wrote: >>> Hey, >>> >>> This series improves page unpinning, with an eye on improving MR >>> deregistration for big swaths of memory (which is bound by

Re: [PATCH v4 0/4] mm/gup: page unpining improvements

2021-02-18 Thread Joao Martins
On 2/18/21 7:24 AM, Christoph Hellwig wrote: > On Fri, Feb 12, 2021 at 01:08:39PM +, Joao Martins wrote: >> Hey, >> >> This series improves page unpinning, with an eye on improving MR >> deregistration for big swaths of memory (which is bound by the page >> unpining), particularly: > > Can you

Re: [PATCH v4 0/4] mm/gup: page unpining improvements

2021-02-18 Thread Christoph Hellwig
On Fri, Feb 12, 2021 at 01:08:39PM +, Joao Martins wrote: > Hey, > > This series improves page unpinning, with an eye on improving MR > deregistration for big swaths of memory (which is bound by the page > unpining), particularly: Can you also take a look at the (bdev and iomap) direct I/O co

[PATCH v4 0/4] mm/gup: page unpining improvements

2021-02-12 Thread Joao Martins
Hey, This series improves page unpinning, with an eye on improving MR deregistration for big swaths of memory (which is bound by the page unpining), particularly: 1) Decrement the head page by @ntails and thus reducing a lot the number of atomic operations per compound page. This is done by comp