From: Dan Williams <dan.j.williams@intel.com>
To: akpm@linux-foundation.org
Cc: linux-mm@kvack.org, Jeff Moyer <jmoyer@redhat.com>,
linux-nvdimm@lists.01.org, Jan Kara <jack@suse.com>
Subject: [-mm PATCH v2 02/25] dax: increase granularity of dax_clear_blocks() operations
Date: Wed, 09 Dec 2015 18:37:20 -0800 [thread overview]
Message-ID: <20151210023720.30368.14658.stgit@dwillia2-desk3.jf.intel.com> (raw)
In-Reply-To: <20151210023708.30368.92962.stgit@dwillia2-desk3.jf.intel.com>
dax_clear_blocks is currently performing a cond_resched() after every
PAGE_SIZE memset. We need not check so frequently, for example md-raid
only calls cond_resched() at stripe granularity. Also, in preparation
for introducing a dax_map_atomic() operation that temporarily pins a dax
mapping move the call to cond_resched() to the outer loop.
The worst case latency between calls to cond_resched() after this change
is 500us the average latency is 133us. This is up from a 10us max and
4us average.
Reviewed-by: Jan Kara <jack@suse.com>
Reviewed-by: Jeff Moyer <jmoyer@redhat.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
---
fs/dax.c | 22 ++++++++--------------
1 file changed, 8 insertions(+), 14 deletions(-)
diff --git a/fs/dax.c b/fs/dax.c
index 19492cc65a30..11721c0fc127 100644
--- a/fs/dax.c
+++ b/fs/dax.c
@@ -28,6 +28,7 @@
#include <linux/sched.h>
#include <linux/uio.h>
#include <linux/vmstat.h>
+#include <linux/sizes.h>
/*
* dax_clear_blocks() is called from within transaction context from XFS,
@@ -43,24 +44,17 @@ int dax_clear_blocks(struct inode *inode, sector_t block, long size)
do {
void __pmem *addr;
unsigned long pfn;
- long count;
+ long count, sz;
count = bdev_direct_access(bdev, sector, &addr, &pfn, size);
if (count < 0)
return count;
- BUG_ON(size < count);
- while (count > 0) {
- unsigned pgsz = PAGE_SIZE - offset_in_page(addr);
- if (pgsz > count)
- pgsz = count;
- clear_pmem(addr, pgsz);
- addr += pgsz;
- size -= pgsz;
- count -= pgsz;
- BUG_ON(pgsz & 511);
- sector += pgsz / 512;
- cond_resched();
- }
+ sz = min_t(long, count, SZ_128K);
+ clear_pmem(addr, sz);
+ size -= sz;
+ BUG_ON(sz & 511);
+ sector += sz / 512;
+ cond_resched();
} while (size);
wmb_pmem();
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2015-12-10 2:37 UTC|newest]
Thread overview: 48+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-12-10 2:37 [-mm PATCH v2 00/25] get_user_pages() for dax pte and pmd mappings Dan Williams
2015-12-10 2:37 ` [-mm PATCH v2 01/25] pmem, dax: clean up clear_pmem() Dan Williams
2015-12-10 2:37 ` Dan Williams [this message]
2015-12-10 2:37 ` [-mm PATCH v2 03/25] dax: guarantee page aligned results from bdev_direct_access() Dan Williams
2015-12-10 2:37 ` [-mm PATCH v2 04/25] dax: fix lifetime of in-kernel dax mappings with dax_map_atomic() Dan Williams
2015-12-11 18:11 ` [-mm PATCH v3 " Dan Williams
2015-12-17 22:00 ` Ross Zwisler
2015-12-17 22:16 ` Dan Williams
2015-12-10 2:37 ` [-mm PATCH v2 05/25] mm, dax: fix livelock, allow dax pmd mappings to become writeable Dan Williams
2015-12-10 2:37 ` [-mm PATCH v2 06/25] dax: Split pmd map when fallback on COW Dan Williams
2015-12-10 2:37 ` [-mm PATCH v2 07/25] um: kill pfn_t Dan Williams
2015-12-10 2:37 ` [-mm PATCH v2 08/25] kvm: rename pfn_t to kvm_pfn_t Dan Williams
2015-12-10 2:37 ` [-mm PATCH v2 09/25] mm, dax, pmem: introduce pfn_t Dan Williams
2015-12-11 18:22 ` [-mm PATCH v3 " Dan Williams
2015-12-10 2:38 ` [-mm PATCH v2 10/25] mm: introduce find_dev_pagemap() Dan Williams
2015-12-11 18:27 ` [-mm PATCH v3 " Dan Williams
2015-12-10 2:38 ` [-mm PATCH v2 11/25] x86, mm: introduce vmem_altmap to augment vmemmap_populate() Dan Williams
2015-12-15 16:50 ` Dan Williams
2015-12-15 23:28 ` Andrew Morton
2015-12-15 23:37 ` Dan Williams
2015-12-10 2:38 ` [-mm PATCH v2 12/25] libnvdimm, pfn, pmem: allocate memmap array in persistent memory Dan Williams
2015-12-10 2:38 ` [-mm PATCH v2 13/25] avr32: convert to asm-generic/memory_model.h Dan Williams
2015-12-10 2:38 ` [-mm PATCH v2 14/25] hugetlb: fix compile error on tile Dan Williams
2015-12-10 2:38 ` [-mm PATCH v2 15/25] frv: fix compiler warning from definition of __pmd() Dan Williams
2015-12-10 2:38 ` [-mm PATCH v2 16/25] x86, mm: introduce _PAGE_DEVMAP Dan Williams
2015-12-10 2:38 ` [-mm PATCH v2 17/25] mm, dax, gpu: convert vm_insert_mixed to pfn_t Dan Williams
2015-12-10 2:38 ` [-mm PATCH v2 18/25] mm, dax: convert vmf_insert_pfn_pmd() " Dan Williams
2015-12-10 2:38 ` [-mm PATCH v2 19/25] list: introduce list_del_poison() Dan Williams
2015-12-15 23:41 ` Andrew Morton
2015-12-16 0:17 ` Dan Williams
2015-12-10 2:39 ` [-mm PATCH v2 20/25] libnvdimm, pmem: move request_queue allocation earlier in probe Dan Williams
2015-12-10 2:39 ` [-mm PATCH v2 21/25] mm, dax, pmem: introduce {get|put}_dev_pagemap() for dax-gup Dan Williams
2015-12-15 23:46 ` Andrew Morton
2015-12-10 2:39 ` [-mm PATCH v2 22/25] mm, dax: dax-pmd vs thp-pmd vs hugetlbfs-pmd Dan Williams
2015-12-10 2:39 ` [-mm PATCH v2 23/25] mm, x86: get_user_pages() for dax mappings Dan Williams
2015-12-16 0:14 ` Andrew Morton
2015-12-16 2:18 ` Dan Williams
2015-12-18 0:09 ` Dan Williams
2015-12-10 2:39 ` [-mm PATCH v2 24/25] dax: provide diagnostics for pmd mapping failures Dan Williams
2015-12-10 2:39 ` [-mm PATCH v2 25/25] dax: re-enable dax pmd mappings Dan Williams
2015-12-10 18:08 ` [-mm PATCH v2 00/25] get_user_pages() for dax pte and " Jeff Moyer
2015-12-10 18:56 ` Dan Williams
2015-12-10 19:20 ` Jeff Moyer
2015-12-11 2:03 ` Dan Williams
2015-12-14 14:52 ` Jeff Moyer
2015-12-14 16:44 ` Dan Williams
2015-12-11 18:44 ` Dan Williams
2015-12-15 1:59 ` Dan Williams
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=20151210023720.30368.14658.stgit@dwillia2-desk3.jf.intel.com \
--to=dan.j.williams@intel.com \
--cc=akpm@linux-foundation.org \
--cc=jack@suse.com \
--cc=jmoyer@redhat.com \
--cc=linux-mm@kvack.org \
--cc=linux-nvdimm@lists.01.org \
/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
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).