From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-9.6 required=3.0 tests=DKIM_INVALID,DKIM_SIGNED, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_PATCH,MAILING_LIST_MULTI,SIGNED_OFF_BY, SPF_HELO_NONE,SPF_PASS,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 5358DC43603 for ; Tue, 10 Dec 2019 16:25:07 +0000 (UTC) Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by mail.kernel.org (Postfix) with ESMTP id 0B68A2073B for ; Tue, 10 Dec 2019 16:25:06 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=fail reason="signature verification failed" (2048-bit key) header.d=kernel-dk.20150623.gappssmtp.com header.i=@kernel-dk.20150623.gappssmtp.com header.b="Zpixa69l" DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 0B68A2073B Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=kernel.dk Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=owner-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix) id EBEF26B2D3A; Tue, 10 Dec 2019 11:25:03 -0500 (EST) Received: by kanga.kvack.org (Postfix, from userid 40) id DDD686B2D3D; Tue, 10 Dec 2019 11:25:03 -0500 (EST) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id C72096B2D3D; Tue, 10 Dec 2019 11:25:03 -0500 (EST) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0018.hostedemail.com [216.40.44.18]) by kanga.kvack.org (Postfix) with ESMTP id B26846B2D3A for ; Tue, 10 Dec 2019 11:25:03 -0500 (EST) Received: from smtpin15.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay01.hostedemail.com (Postfix) with SMTP id 54761180AD81F for ; Tue, 10 Dec 2019 16:25:03 +0000 (UTC) X-FDA: 76249756086.15.beam97_ef2745811d38 X-HE-Tag: beam97_ef2745811d38 X-Filterd-Recvd-Size: 8271 Received: from mail-io1-f65.google.com (mail-io1-f65.google.com [209.85.166.65]) by imf16.hostedemail.com (Postfix) with ESMTP for ; Tue, 10 Dec 2019 16:25:02 +0000 (UTC) Received: by mail-io1-f65.google.com with SMTP id t26so7991166ioi.13 for ; Tue, 10 Dec 2019 08:25:02 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel-dk.20150623.gappssmtp.com; s=20150623; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=o/x6NjL9O5h2Ux7aJEFNccuj0M9CNUr2pjvWBBusJd4=; b=Zpixa69lHW83t2ZbBBOHojQxENx5w2X0sKlji5l+ZpBo6tA/3fj9yn2HfnjrlJQ6Ob U8FY3MOCg7q0XHIPvND9EpFk53qsIhRbPB/9HbD1ilihLRNcBtOuwWUXkIqNVoGnuzdD 5XOtCxdmCCmO2Arsi/KYjMVTyNg6egphSQmKqqQfmQkq1VyVM8Gj0Xtm9JJX4Led8WKe zmbD/SLsl+MyHYWuaQHEb1/hAAYeSbscwPIu0vVxe909vlonZb4pS3Y0AjqklnvTPNnZ 6/Cw7617YWeJqBJwlZJAG0ivRv7q+hmVouVZ8mhUmu2zs+i9PsKCc/s41pHKmnnSHHi9 5YdQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=o/x6NjL9O5h2Ux7aJEFNccuj0M9CNUr2pjvWBBusJd4=; b=ojuJOvsAR4Q0+gTqGUBL5akDSzoaQjB3An0aSW/fMS21iWDMDtBiOvXKvw6YU7FsfN ZRZ5UOs6XMBpjrl2Psw3VfPHCGi9tskZvC7Ay/zBSoYo1qgfZtdMkcmbun7vmJ5ejMl2 0WjlhWlK+tH2DDzfmLFIm6z2/KQ2+m6fJ3mIvIAXzx6hQP9udLa9yMl0dgPuMiMga8Fy +WRg3jMyc6ZAiwhp8clxhNzSizce0k/4ePwGWZn7SOJjdcyEjQuNU+F8zJKgLjNR+a69 xaL7KCifYKL/kBWan22R2gGxndFuG33rhp5+trE6Y5dXq91XsqK2UAbwXZLhGWJI5mnc PLYg== X-Gm-Message-State: APjAAAX20+kVBzMv6D7r+eN8v+uKdGRSsC7z0YARpGyF7CycXKplr5th kr0G4FYqZBO8a0NCl72fAKRe8zqwECL9HA== X-Google-Smtp-Source: APXvYqxxtEmqFA3byjJGnmrQmrrH3TRgfkvrkxj17L5ofRuCP+Z2MTkeccUYnzTdzqEGwVD/DbWwww== X-Received: by 2002:a02:c6d5:: with SMTP id r21mr32752498jan.129.1575995101208; Tue, 10 Dec 2019 08:25:01 -0800 (PST) Received: from x1.thefacebook.com ([65.144.74.34]) by smtp.gmail.com with ESMTPSA id y11sm791174iol.23.2019.12.10.08.24.59 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Tue, 10 Dec 2019 08:25:00 -0800 (PST) From: Jens Axboe To: linux-mm@kvack.org, linux-fsdevel@vger.kernel.org, linux-block@vger.kernel.org Cc: Jens Axboe Subject: [PATCH 3/5] mm: make buffered writes work with RWF_UNCACHED Date: Tue, 10 Dec 2019 09:24:52 -0700 Message-Id: <20191210162454.8608-4-axboe@kernel.dk> X-Mailer: git-send-email 2.24.0 In-Reply-To: <20191210162454.8608-1-axboe@kernel.dk> References: <20191210162454.8608-1-axboe@kernel.dk> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=1.2.4 Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: If RWF_UNCACHED is set for io_uring (or pwritev2(2)), we'll drop the cache instantiated for buffered writes. If new pages aren't instantiated, we leave them alone. This provides similar semantics to reads with RWF_UNCACHED set. Signed-off-by: Jens Axboe --- include/linux/fs.h | 3 ++ mm/filemap.c | 78 +++++++++++++++++++++++++++++++++++++++++++--- 2 files changed, 76 insertions(+), 5 deletions(-) diff --git a/include/linux/fs.h b/include/linux/fs.h index bf58db1bc032..bcf486c132a8 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -285,6 +285,7 @@ enum positive_aop_returns { #define AOP_FLAG_NOFS 0x0002 /* used by filesystem to direct * helper code (eg buffer layer) * to clear GFP_FS from alloc */ +#define AOP_FLAG_UNCACHED 0x0004 =20 /* * oh the beauties of C type declarations. @@ -3105,6 +3106,8 @@ extern ssize_t generic_file_write_iter(struct kiocb= *, struct iov_iter *); extern ssize_t generic_file_direct_write(struct kiocb *, struct iov_iter= *); extern ssize_t generic_perform_write(struct file *, struct iov_iter *, struct kiocb *); +extern void write_drop_cached_pages(struct page **, + struct address_space *mapping, unsigned *); =20 ssize_t vfs_iter_read(struct file *file, struct iov_iter *iter, loff_t *= ppos, rwf_t flags); diff --git a/mm/filemap.c b/mm/filemap.c index fe37bd2b2630..d6171bf705f9 100644 --- a/mm/filemap.c +++ b/mm/filemap.c @@ -3287,10 +3287,12 @@ struct page *grab_cache_page_write_begin(struct a= ddress_space *mapping, pgoff_t index, unsigned flags) { struct page *page; - int fgp_flags =3D FGP_LOCK|FGP_WRITE|FGP_CREAT; + int fgp_flags =3D FGP_LOCK|FGP_WRITE; =20 if (flags & AOP_FLAG_NOFS) fgp_flags |=3D FGP_NOFS; + if (!(flags & AOP_FLAG_UNCACHED)) + fgp_flags |=3D FGP_CREAT; =20 page =3D pagecache_get_page(mapping, index, fgp_flags, mapping_gfp_mask(mapping)); @@ -3301,21 +3303,67 @@ struct page *grab_cache_page_write_begin(struct a= ddress_space *mapping, } EXPORT_SYMBOL(grab_cache_page_write_begin); =20 +/* + * Start writeback on the pages in pgs[], and then try and remove those = pages + * from the page cached. Used with RWF_UNCACHED. + */ +void write_drop_cached_pages(struct page **pgs, struct address_space *ma= pping, + unsigned *nr) +{ + loff_t start, end; + int i; + + end =3D 0; + start =3D LLONG_MAX; + for (i =3D 0; i < *nr; i++) { + struct page *page =3D pgs[i]; + loff_t off; + + off =3D (loff_t) page_to_index(page) << PAGE_SHIFT; + if (off < start) + start =3D off; + if (off > end) + end =3D off; + get_page(page); + } + + __filemap_fdatawrite_range(mapping, start, end, WB_SYNC_NONE); + + for (i =3D 0; i < *nr; i++) { + struct page *page =3D pgs[i]; + + lock_page(page); + if (page->mapping =3D=3D mapping) { + wait_on_page_writeback(page); + if (!page_has_private(page) || + try_to_release_page(page, 0)) + remove_mapping(mapping, page); + } + unlock_page(page); + } + *nr =3D 0; +} +EXPORT_SYMBOL_GPL(write_drop_cached_pages); + +#define GPW_PAGE_BATCH 16 + ssize_t generic_perform_write(struct file *file, struct iov_iter *i, struct kiocb *iocb) { struct address_space *mapping =3D file->f_mapping; const struct address_space_operations *a_ops =3D mapping->a_ops; + struct page *drop_pages[GPW_PAGE_BATCH]; loff_t pos =3D iocb->ki_pos; long status =3D 0; ssize_t written =3D 0; - unsigned int flags =3D 0; + unsigned int flags =3D 0, drop_nr =3D 0; =20 do { struct page *page; unsigned long offset; /* Offset into pagecache page */ unsigned long bytes; /* Bytes to write to page */ size_t copied; /* Bytes copied from user */ + bool drop_page =3D false; /* drop page after IO */ void *fsdata; =20 offset =3D (pos & (PAGE_SIZE - 1)); @@ -3323,6 +3371,9 @@ ssize_t generic_perform_write(struct file *file, iov_iter_count(i)); =20 again: + if (iocb->ki_flags & IOCB_UNCACHED) + flags |=3D AOP_FLAG_UNCACHED; + /* * Bring in the user page that we will copy from _first_. * Otherwise there's a nasty deadlock on copying from the @@ -3343,10 +3394,17 @@ ssize_t generic_perform_write(struct file *file, break; } =20 +retry: status =3D a_ops->write_begin(file, mapping, pos, bytes, flags, &page, &fsdata); - if (unlikely(status < 0)) + if (unlikely(status < 0)) { + if (status =3D=3D -ENOMEM && (flags & AOP_FLAG_UNCACHED)) { + drop_page =3D true; + flags &=3D ~AOP_FLAG_UNCACHED; + goto retry; + } break; + } =20 if (mapping_writably_mapped(mapping)) flush_dcache_page(page); @@ -3376,12 +3434,22 @@ ssize_t generic_perform_write(struct file *file, iov_iter_single_seg_count(i)); goto again; } + if (drop_page && + ((pos >> PAGE_SHIFT) !=3D ((pos + copied) >> PAGE_SHIFT))) { + drop_pages[drop_nr] =3D page; + if (++drop_nr =3D=3D GPW_PAGE_BATCH) + write_drop_cached_pages(drop_pages, mapping, + &drop_nr); + } else + balance_dirty_pages_ratelimited(mapping); + pos +=3D copied; written +=3D copied; - - balance_dirty_pages_ratelimited(mapping); } while (iov_iter_count(i)); =20 + if (drop_nr) + write_drop_cached_pages(drop_pages, mapping, &drop_nr); + return written ? written : status; } EXPORT_SYMBOL(generic_perform_write); --=20 2.24.0