From mboxrd@z Thu Jan 1 00:00:00 1970 From: John Hubbard Subject: Re: [PATCH v2 5/6] mm: track gup pages with page->dma_pinned_* fields Date: Mon, 2 Jul 2018 21:30:28 -0700 Message-ID: References: <20180702005654.20369-1-jhubbard@nvidia.com> <20180702005654.20369-6-jhubbard@nvidia.com> <20180702095331.n5zfz35d3invl5al@quack2.suse.cz> <010001645d77ee2c-de7fedbd-f52d-4b74-9388-e6435973792b-000000@email.amazonses.com> Mime-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <010001645d77ee2c-de7fedbd-f52d-4b74-9388-e6435973792b-000000@email.amazonses.com> Content-Language: en-US Sender: linux-kernel-owner@vger.kernel.org To: Christopher Lameter Cc: Jan Kara , john.hubbard@gmail.com, Matthew Wilcox , Michal Hocko , Jason Gunthorpe , Dan Williams , linux-mm@kvack.org, LKML , linux-rdma , linux-fsdevel@vger.kernel.org List-Id: linux-rdma@vger.kernel.org On 07/02/2018 05:08 PM, Christopher Lameter wrote: > On Mon, 2 Jul 2018, John Hubbard wrote: > >>> >>> These two are just wrong. You cannot make any page reference for >>> PageDmaPinned() account against a pin count. First, it is just conceptually >>> wrong as these references need not be long term pins, second, you can >>> easily race like: >>> >>> Pinner Random process >>> get_page(page) >>> pin_page_for_dma() >>> put_page(page) >>> -> oops, page gets unpinned too early >>> >> >> I'll drop this approach, without mentioning any of the locking that is hiding in >> there, since that was probably breaking other rules anyway. :) Thanks for your >> patience in reviewing this. > > Mayb the following would work: > > If you establish a reference to a page then increase the page count. If > the reference is a dma pin action also then increase the pinned count. > > That way you know how many of the references to the page are dma > pins and you can correctly manage the state of the page if the dma pins go > away. > I think this sounds like what this patch already does, right? See: __put_page_for_pinned_dma(), __get_page_for_pinned_dma(), and pin_page_for_dma(). The locking seems correct to me, but I suspect it's too heavyweight for such a hot path. But without adding a new put_user_page() call, that was the best I could come up with. What I'm hearing now from Jan and Michal is that the desired end result is a separate API call, put_user_pages(), so that we can explicitly manage these pinned pages. thanks, -- John Hubbard NVIDIA