Linux-Fsdevel Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Kent Overstreet <kent.overstreet@gmail.com>
To: linux-kernel@vger.kernel.org, akpm@linux-foundation.org,
	viro@zeniv.linux.org.uk, linux-mm@kvack.org,
	linux-fsdevel@vger.kernel.org
Cc: Kent Overstreet <kent.overstreet@gmail.com>
Subject: [PATCH 2/2] fs: generic_file_buffered_read() now uses find_get_pages_contig
Date: Tue,  9 Jun 2020 20:10:36 -0400	[thread overview]
Message-ID: <20200610001036.3904844-3-kent.overstreet@gmail.com> (raw)
In-Reply-To: <20200610001036.3904844-1-kent.overstreet@gmail.com>

Convert generic_file_buffered_read() to get pages to read from in
batches, and then copy data to userspace from many pages at once - in
particular, we now don't touch any cachelines that might be contended
while we're in the loop to copy data to userspace.

This is is a performance improvement on workloads that do buffered reads
with large blocksizes, and a very large performance improvement if that
file is also being accessed concurrently by different threads.

On smaller reads (512 bytes), there's a very small performance
improvement (1%, within the margin of error).

Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
---
 mm/filemap.c | 266 ++++++++++++++++++++++++++++-----------------------
 1 file changed, 144 insertions(+), 122 deletions(-)

diff --git a/mm/filemap.c b/mm/filemap.c
index 206d51a1c9..0d1836081c 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -2051,67 +2051,6 @@ static void shrink_readahead_size_eio(struct file_ra_state *ra)
 	ra->ra_pages /= 4;
 }
 
-static int generic_file_buffered_read_page_ok(struct kiocb *iocb,
-			struct iov_iter *iter,
-			struct page *page)
-{
-	struct address_space *mapping = iocb->ki_filp->f_mapping;
-	struct inode *inode = mapping->host;
-	struct file_ra_state *ra = &iocb->ki_filp->f_ra;
-	unsigned offset = iocb->ki_pos & ~PAGE_MASK;
-	unsigned bytes, copied;
-	loff_t isize, end_offset;
-
-	BUG_ON(iocb->ki_pos >> PAGE_SHIFT != page->index);
-
-	/*
-	 * i_size must be checked after we know the page is Uptodate.
-	 *
-	 * Checking i_size after the check allows us to calculate
-	 * the correct value for "bytes", which means the zero-filled
-	 * part of the page is not copied back to userspace (unless
-	 * another truncate extends the file - this is desired though).
-	 */
-
-	isize = i_size_read(inode);
-	if (unlikely(iocb->ki_pos >= isize))
-		return 1;
-
-	end_offset = min_t(loff_t, isize, iocb->ki_pos + iter->count);
-
-	bytes = min_t(loff_t, end_offset - iocb->ki_pos, PAGE_SIZE - offset);
-
-	/* If users can be writing to this page using arbitrary
-	 * virtual addresses, take care about potential aliasing
-	 * before reading the page on the kernel side.
-	 */
-	if (mapping_writably_mapped(mapping))
-		flush_dcache_page(page);
-
-	/*
-	 * Ok, we have the page, and it's up-to-date, so
-	 * now we can copy it to user space...
-	 */
-
-	copied = copy_page_to_iter(page, offset, bytes, iter);
-
-	iocb->ki_pos += copied;
-
-	/*
-	 * When a sequential read accesses a page several times,
-	 * only mark it as accessed the first time.
-	 */
-	if (iocb->ki_pos >> PAGE_SHIFT != ra->prev_pos >> PAGE_SHIFT)
-		mark_page_accessed(page);
-
-	ra->prev_pos = iocb->ki_pos;
-
-	if (copied < bytes)
-		return -EFAULT;
-
-	return !iov_iter_count(iter) || iocb->ki_pos == isize;
-}
-
 static struct page *
 generic_file_buffered_read_readpage(struct file *filp,
 				    struct address_space *mapping,
@@ -2255,6 +2194,79 @@ generic_file_buffered_read_no_cached_page(struct kiocb *iocb,
 	return generic_file_buffered_read_readpage(filp, mapping, page);
 }
 
+static int generic_file_buffered_read_get_pages(struct kiocb *iocb,
+						struct iov_iter *iter,
+						struct page **pages,
+						unsigned nr)
+{
+	struct file *filp = iocb->ki_filp;
+	struct address_space *mapping = filp->f_mapping;
+	struct file_ra_state *ra = &filp->f_ra;
+	pgoff_t index = iocb->ki_pos >> PAGE_SHIFT;
+	pgoff_t last_index = (iocb->ki_pos + iter->count + PAGE_SIZE-1) >> PAGE_SHIFT;
+	int i, j, ret, err = 0;
+
+	nr = min_t(unsigned long, last_index - index, nr);
+find_page:
+	if (fatal_signal_pending(current))
+		return -EINTR;
+
+	ret = find_get_pages_contig(mapping, index, nr, pages);
+	if (ret)
+		goto got_pages;
+
+	if (iocb->ki_flags & IOCB_NOWAIT)
+		return -EAGAIN;
+
+	page_cache_sync_readahead(mapping, ra, filp, index, last_index - index);
+
+	ret = find_get_pages_contig(mapping, index, nr, pages);
+	if (ret)
+		goto got_pages;
+
+	pages[0] = generic_file_buffered_read_no_cached_page(iocb, iter);
+	err = PTR_ERR_OR_ZERO(pages[0]);
+	ret = !IS_ERR_OR_NULL(pages[0]);
+got_pages:
+	for (i = 0; i < ret; i++) {
+		struct page *page = pages[i];
+		pgoff_t pg_index = index +i;
+		loff_t pg_pos = max(iocb->ki_pos,
+				    (loff_t) pg_index << PAGE_SHIFT);
+		loff_t pg_count = iocb->ki_pos + iter->count - pg_pos;
+
+		if (PageReadahead(page))
+			page_cache_async_readahead(mapping, ra, filp, page,
+					pg_index, last_index - pg_index);
+
+		if (!PageUptodate(page)) {
+			if (iocb->ki_flags & IOCB_NOWAIT) {
+				for (j = i; j < ret; j++)
+					put_page(pages[j]);
+				ret = i;
+				err = -EAGAIN;
+				break;
+			}
+
+			page = generic_file_buffered_read_pagenotuptodate(filp,
+						iter, page, pg_pos, pg_count);
+			if (IS_ERR_OR_NULL(page)) {
+				for (j = i + 1; j < ret; j++)
+					put_page(pages[j]);
+				ret = i;
+				err = PTR_ERR_OR_ZERO(page);
+				break;
+			}
+		}
+	}
+
+	if (likely(ret))
+		return ret;
+	if (err)
+		return err;
+	goto find_page;
+}
+
 /**
  * generic_file_buffered_read - generic file read routine
  * @iocb:	the iocb to read
@@ -2275,83 +2287,93 @@ static ssize_t generic_file_buffered_read(struct kiocb *iocb,
 		struct iov_iter *iter, ssize_t written)
 {
 	struct file *filp = iocb->ki_filp;
+	struct file_ra_state *ra = &filp->f_ra;
 	struct address_space *mapping = filp->f_mapping;
 	struct inode *inode = mapping->host;
-	struct file_ra_state *ra = &filp->f_ra;
 	size_t orig_count = iov_iter_count(iter);
-	pgoff_t last_index;
-	int error = 0;
+	struct page *pages[64];
+	int i, pg_nr, error = 0;
+	bool writably_mapped;
+	loff_t isize, end_offset;
 
 	if (unlikely(iocb->ki_pos >= inode->i_sb->s_maxbytes))
 		return 0;
 	iov_iter_truncate(iter, inode->i_sb->s_maxbytes);
 
-	last_index = (iocb->ki_pos + iter->count + PAGE_SIZE-1) >> PAGE_SHIFT;
-
-	for (;;) {
-		pgoff_t index = iocb->ki_pos >> PAGE_SHIFT;
-		struct page *page;
-
+	do {
 		cond_resched();
-find_page:
-		if (fatal_signal_pending(current)) {
-			error = -EINTR;
-			goto out;
-		}
 
-		page = find_get_page(mapping, index);
-		if (!page) {
-			if (iocb->ki_flags & IOCB_NOWAIT)
-				goto would_block;
-			page_cache_sync_readahead(mapping,
-					ra, filp,
-					index, last_index - index);
-			page = find_get_page(mapping, index);
-			if (unlikely(page == NULL)) {
-				page = generic_file_buffered_read_no_cached_page(iocb, iter);
-				if (!page)
-					goto find_page;
-				if (IS_ERR(page)) {
-					error = PTR_ERR(page);
-					goto out;
-				}
-			}
-		}
-		if (PageReadahead(page)) {
-			page_cache_async_readahead(mapping,
-					ra, filp, page,
-					index, last_index - index);
+		i = 0;
+		pg_nr = generic_file_buffered_read_get_pages(iocb, iter, pages,
+							     ARRAY_SIZE(pages));
+		if (pg_nr < 0) {
+			error = pg_nr;
+			break;
 		}
-		if (!PageUptodate(page)) {
-			if (iocb->ki_flags & IOCB_NOWAIT) {
-				put_page(page);
-				error = -EAGAIN;
-				goto out;
-			}
 
-			page = generic_file_buffered_read_pagenotuptodate(filp,
-					iter, page, iocb->ki_pos, iter->count);
-			if (!page)
-				goto find_page;
-			if (IS_ERR(page)) {
-				error = PTR_ERR(page);
-				goto out;
-			}
-		}
+		/*
+		 * i_size must be checked after we know the pages are Uptodate.
+		 *
+		 * Checking i_size after the check allows us to calculate
+		 * the correct value for "nr", which means the zero-filled
+		 * part of the page is not copied back to userspace (unless
+		 * another truncate extends the file - this is desired though).
+		 */
+		isize = i_size_read(inode);
+		if (unlikely(iocb->ki_pos >= isize))
+			goto put_pages;
 
-		error = generic_file_buffered_read_page_ok(iocb, iter, page);
-		put_page(page);
+		end_offset = min_t(loff_t, isize, iocb->ki_pos + iter->count);
 
-		if (error) {
-			if (error > 0)
-				error = 0;
-			goto out;
+		while ((iocb->ki_pos >> PAGE_SHIFT) + pg_nr >
+		       (end_offset + PAGE_SIZE - 1) >> PAGE_SHIFT)
+			put_page(pages[--pg_nr]);
+
+		/*
+		 * Once we start copying data, we don't want to be touching any
+		 * cachelines that might be contended:
+		 */
+		writably_mapped = mapping_writably_mapped(mapping);
+
+		/*
+		 * When a sequential read accesses a page several times, only
+		 * mark it as accessed the first time.
+		 */
+		if (iocb->ki_pos >> PAGE_SHIFT !=
+		    ra->prev_pos >> PAGE_SHIFT)
+			mark_page_accessed(pages[0]);
+		for (i = 1; i < pg_nr; i++)
+			mark_page_accessed(pages[i]);
+
+		for (i = 0; i < pg_nr; i++) {
+			unsigned offset = iocb->ki_pos & ~PAGE_MASK;
+			unsigned bytes = min_t(loff_t, end_offset - iocb->ki_pos,
+					       PAGE_SIZE - offset);
+			unsigned copied;
+
+			/*
+			 * If users can be writing to this page using arbitrary
+			 * virtual addresses, take care about potential aliasing
+			 * before reading the page on the kernel side.
+			 */
+			if (writably_mapped)
+				flush_dcache_page(pages[i]);
+
+			copied = copy_page_to_iter(pages[i], offset, bytes, iter);
+
+			iocb->ki_pos += copied;
+			ra->prev_pos = iocb->ki_pos;
+
+			if (copied < bytes) {
+				error = -EFAULT;
+				break;
+			}
 		}
-	}
+put_pages:
+		for (i = 0; i < pg_nr; i++)
+			put_page(pages[i]);
+	} while (iov_iter_count(iter) && iocb->ki_pos < isize && !error);
 
-would_block:
-	error = -EAGAIN;
-out:
 	file_accessed(filp);
 	written += orig_count - iov_iter_count(iter);
 
-- 
2.27.0


  parent reply	other threads:[~2020-06-10  0:10 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-06-10  0:10 [PATCH 0/2] generic_file_buffered_read() refactoring & optimization Kent Overstreet
2020-06-10  0:10 ` [PATCH 1/2] fs: Break generic_file_buffered_read up into multiple functions Kent Overstreet
2020-06-10  0:10 ` Kent Overstreet [this message]
2020-06-10  0:47   ` [PATCH 2/2] fs: generic_file_buffered_read() now uses find_get_pages_contig Matthew Wilcox
2020-06-10  1:08     ` Kent Overstreet
2020-06-10  1:38   ` Matthew Wilcox
2020-06-10  1:46     ` Kent Overstreet
2020-06-10  1:36 ` [PATCH v2 1/2] fs: Break generic_file_buffered_read up into multiple functions Kent Overstreet
2020-06-10  1:36 ` [PATCH v2 2/2] fs: generic_file_buffered_read() now uses find_get_pages_contig Kent Overstreet
2020-06-18  1:05   ` Andrew Morton
2020-06-19  3:20     ` [PATCH v3 0/2] generic_file_buffered_read() refactoring & optimization Kent Overstreet
2020-06-19 12:59       ` Christoph Hellwig
2020-06-19 18:44         ` Kent Overstreet
2020-06-19  3:20     ` [PATCH v3 1/2] fs: Break generic_file_buffered_read up into multiple functions Kent Overstreet
2020-06-19  3:20     ` [PATCH v3 2/2] fs: generic_file_buffered_read() now uses find_get_pages_contig Kent Overstreet
2020-06-30  0:12 ` Fixup patch for [PATCH 0/2] generic_file_buffered_read() refactoring & optimization Kent Overstreet
  -- strict thread matches above, loose matches on Subject: below --
2020-10-17 20:10 [PATCH 0/2] generic_file_buffered_read() refactoring, perf improvements Kent Overstreet
2020-10-17 20:10 ` [PATCH 2/2] fs: generic_file_buffered_read() now uses find_get_pages_contig Kent Overstreet
2020-10-20 14:47   ` Jens Axboe
2018-08-15 23:26 [PATCH 0/2] generic_file_buffered_read improvements Kent Overstreet
2018-08-15 23:26 ` [PATCH 2/2] fs: generic_file_buffered_read() now uses find_get_pages_contig Kent Overstreet
2018-08-16 14:56   ` kbuild test robot

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200610001036.3904844-3-kent.overstreet@gmail.com \
    --to=kent.overstreet@gmail.com \
    --cc=akpm@linux-foundation.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=viro@zeniv.linux.org.uk \
    --subject='Re: [PATCH 2/2] fs: generic_file_buffered_read() now uses find_get_pages_contig' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).