From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andrew Morton Subject: [patch 046/155] mm/gup_benchmark: support pin_user_pages() and related calls Date: Wed, 01 Apr 2020 21:05:41 -0700 Message-ID: <20200402040541.c2n8b-a92%akpm@linux-foundation.org> References: <20200401210155.09e3b9742e1c6e732f5a7250@linux-foundation.org> Reply-To: linux-kernel@vger.kernel.org Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Return-path: Received: from mail.kernel.org ([198.145.29.99]:53646 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1725789AbgDBEFn (ORCPT ); Thu, 2 Apr 2020 00:05:43 -0400 In-Reply-To: <20200401210155.09e3b9742e1c6e732f5a7250@linux-foundation.org> Sender: mm-commits-owner@vger.kernel.org List-Id: mm-commits@vger.kernel.org To: akpm@linux-foundation.org, corbet@lwn.net, dan.j.williams@intel.com, david@fromorbit.com, hch@infradead.org, ira.weiny@intel.com, jack@suse.cz, jgg@ziepe.ca, jglisse@redhat.com, jhubbard@nvidia.com, kirill.shutemov@linux.intel.com, linux-mm@kvack.org, mhocko@suse.com, mike.kravetz@oracle.com, mm-commits@vger.kernel.org, shuah@kernel.org, torvalds@linux-foundation.org, vbabka@suse.cz, viro@zeniv.linux.org.uk, willy@infradead.org =46rom: John Hubbard Subject: mm/gup_benchmark: support pin_user_pages() and related calls Up until now, gup_benchmark supported testing of the following kernel functions: * get_user_pages(): via the '-U' command line option * get_user_pages_longterm(): via the '-L' command line option * get_user_pages_fast(): as the default (no options required) Add test coverage for the new corresponding pin_*() functions: * pin_user_pages_fast(): via the '-a' command line option * pin_user_pages(): via the '-b' command line option Also, add an option for clarity: '-u' for what is now (still) the default choice: get_user_pages_fast(). Also, for the commands that set FOLL_PIN, verify that the pages really are dma-pinned, via the new is_dma_pinned() routine. Those commands are: PIN_FAST_BENCHMARK : calls pin_user_pages_fast() PIN_BENCHMARK : calls pin_user_pages() In between the calls to pin_*() and unpin_user_pages(), check each page: if page_maybe_dma_pinned() returns false, then WARN and return. Do this outside of the benchmark timestamps, so that it doesn't affect reported times. Link: http://lkml.kernel.org/r/20200211001536.1027652-10-jhubbard@nvidia.com Signed-off-by: John Hubbard Reviewed-by: Ira Weiny Acked-by: Kirill A. Shutemov Cc: Jan Kara Cc: J=C3=A9r=C3=B4me Glisse Cc: "Matthew Wilcox (Oracle)" Cc: Al Viro Cc: Christoph Hellwig Cc: Dan Williams Cc: Dave Chinner Cc: Jason Gunthorpe Cc: Jonathan Corbet Cc: Michal Hocko Cc: Mike Kravetz Cc: Shuah Khan Cc: Vlastimil Babka Signed-off-by: Andrew Morton --- mm/gup_benchmark.c | 71 +++++++++++++++++-- tools/testing/selftests/vm/gup_benchmark.c | 15 +++- 2 files changed, 80 insertions(+), 6 deletions(-) --- a/mm/gup_benchmark.c~mm-gup_benchmark-support-pin_user_pages-and-relate= d-calls +++ a/mm/gup_benchmark.c @@ -8,6 +8,8 @@ #define GUP_FAST_BENCHMARK _IOWR('g', 1, struct gup_benchmark) #define GUP_LONGTERM_BENCHMARK _IOWR('g', 2, struct gup_benchmark) #define GUP_BENCHMARK _IOWR('g', 3, struct gup_benchmark) +#define PIN_FAST_BENCHMARK _IOWR('g', 4, struct gup_benchmark) +#define PIN_BENCHMARK _IOWR('g', 5, struct gup_benchmark) =20 struct gup_benchmark { __u64 get_delta_usec; @@ -19,6 +21,48 @@ struct gup_benchmark { __u64 expansion[10]; /* For future use */ }; =20 +static void put_back_pages(unsigned int cmd, struct page **pages, + unsigned long nr_pages) +{ + unsigned long i; + + switch (cmd) { + case GUP_FAST_BENCHMARK: + case GUP_LONGTERM_BENCHMARK: + case GUP_BENCHMARK: + for (i =3D 0; i < nr_pages; i++) + put_page(pages[i]); + break; + + case PIN_FAST_BENCHMARK: + case PIN_BENCHMARK: + unpin_user_pages(pages, nr_pages); + break; + } +} + +static void verify_dma_pinned(unsigned int cmd, struct page **pages, + unsigned long nr_pages) +{ + unsigned long i; + struct page *page; + + switch (cmd) { + case PIN_FAST_BENCHMARK: + case PIN_BENCHMARK: + for (i =3D 0; i < nr_pages; i++) { + page =3D pages[i]; + if (WARN(!page_maybe_dma_pinned(page), + "pages[%lu] is NOT dma-pinned\n", i)) { + + dump_page(page, "gup_benchmark failure"); + break; + } + } + break; + } +} + static int __gup_benchmark_ioctl(unsigned int cmd, struct gup_benchmark *gup) { @@ -66,6 +110,14 @@ static int __gup_benchmark_ioctl(unsigne nr =3D get_user_pages(addr, nr, gup->flags, pages + i, NULL); break; + case PIN_FAST_BENCHMARK: + nr =3D pin_user_pages_fast(addr, nr, gup->flags, + pages + i); + break; + case PIN_BENCHMARK: + nr =3D pin_user_pages(addr, nr, gup->flags, pages + i, + NULL); + break; default: kvfree(pages); ret =3D -EINVAL; @@ -78,15 +130,22 @@ static int __gup_benchmark_ioctl(unsigne } end_time =3D ktime_get(); =20 + /* Shifting the meaning of nr_pages: now it is actual number pinned: */ + nr_pages =3D i; + gup->get_delta_usec =3D ktime_us_delta(end_time, start_time); gup->size =3D addr - gup->addr; =20 + /* + * Take an un-benchmark-timed moment to verify DMA pinned + * state: print a warning if any non-dma-pinned pages are found: + */ + verify_dma_pinned(cmd, pages, nr_pages); + start_time =3D ktime_get(); - for (i =3D 0; i < nr_pages; i++) { - if (!pages[i]) - break; - put_page(pages[i]); - } + + put_back_pages(cmd, pages, nr_pages); + end_time =3D ktime_get(); gup->put_delta_usec =3D ktime_us_delta(end_time, start_time); =20 @@ -105,6 +164,8 @@ static long gup_benchmark_ioctl(struct f case GUP_FAST_BENCHMARK: case GUP_LONGTERM_BENCHMARK: case GUP_BENCHMARK: + case PIN_FAST_BENCHMARK: + case PIN_BENCHMARK: break; default: return -EINVAL; --- a/tools/testing/selftests/vm/gup_benchmark.c~mm-gup_benchmark-support-p= in_user_pages-and-related-calls +++ a/tools/testing/selftests/vm/gup_benchmark.c @@ -18,6 +18,10 @@ #define GUP_LONGTERM_BENCHMARK _IOWR('g', 2, struct gup_benchmark) #define GUP_BENCHMARK _IOWR('g', 3, struct gup_benchmark) =20 +/* Similar to above, but use FOLL_PIN instead of FOLL_GET. */ +#define PIN_FAST_BENCHMARK _IOWR('g', 4, struct gup_benchmark) +#define PIN_BENCHMARK _IOWR('g', 5, struct gup_benchmark) + /* Just the flags we need, copied from mm.h: */ #define FOLL_WRITE 0x01 /* check pte is writable */ =20 @@ -40,8 +44,14 @@ int main(int argc, char **argv) char *file =3D "/dev/zero"; char *p; =20 - while ((opt =3D getopt(argc, argv, "m:r:n:f:tTLUwSH")) !=3D -1) { + while ((opt =3D getopt(argc, argv, "m:r:n:f:abtTLUuwSH")) !=3D -1) { switch (opt) { + case 'a': + cmd =3D PIN_FAST_BENCHMARK; + break; + case 'b': + cmd =3D PIN_BENCHMARK; + break; case 'm': size =3D atoi(optarg) * MB; break; @@ -63,6 +73,9 @@ int main(int argc, char **argv) case 'U': cmd =3D GUP_BENCHMARK; break; + case 'u': + cmd =3D GUP_FAST_BENCHMARK; + break; case 'w': write =3D 1; break; _