From: Yin Fengwei <fengwei.yin@intel.com>
To: willy@infradead.org, david@redhat.com, linux-mm@kvack.org
Cc: dave.hansen@intel.com, tim.c.chen@intel.com,
ying.huang@intel.com, fengwei.yin@intel.com
Subject: [RFC PATCH 5/5] filemap: batched update mm counter,rmap when map file folio
Date: Mon, 30 Jan 2023 20:55:04 +0800 [thread overview]
Message-ID: <20230130125504.2509710-6-fengwei.yin@intel.com> (raw)
In-Reply-To: <20230130125504.2509710-1-fengwei.yin@intel.com>
Instead of update mm counter, rmap one by one, batched update
brings some level performance gain.
Signed-off-by: Yin Fengwei <fengwei.yin@intel.com>
---
mm/filemap.c | 32 ++++++++++++++++++++++++++++----
1 file changed, 28 insertions(+), 4 deletions(-)
diff --git a/mm/filemap.c b/mm/filemap.c
index fe0c226c8b1e..6d9438490025 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -3360,28 +3360,52 @@ static vm_fault_t filemap_map_folio_range(struct vm_fault *vmf,
struct vm_area_struct *vma = vmf->vma;
struct file *file = vma->vm_file;
unsigned int mmap_miss = READ_ONCE(file->f_ra.mmap_miss);
- int ref_count = 0, count = 0;
+ int ref_count = 0, count = 0, maplen = 0;
+ struct page *pg = page;
do {
- if (PageHWPoison(page))
+ if (PageHWPoison(page)) {
+ if (maplen) {
+ page_add_file_rmap_range(folio, pg, maplen,
+ vma, false);
+ add_mm_counter(vma->vm_mm,
+ mm_counter_file(pg), maplen);
+ }
+ pg = page + 1;
+ maplen = 0;
continue;
+ }
if (mmap_miss > 0)
mmap_miss--;
- if (!pte_none(*vmf->pte))
+ if (!pte_none(*vmf->pte)) {
+ if (maplen) {
+ page_add_file_rmap_range(folio, pg, maplen,
+ vma, false);
+ add_mm_counter(vma->vm_mm,
+ mm_counter_file(pg), maplen);
+ }
+ pg = page + 1;
+ maplen = 0;
continue;
+ }
if (vmf->address == addr)
ret = VM_FAULT_NOPAGE;
ref_count++;
+ maplen++;
- do_set_pte(vmf, page, addr);
+ do_set_pte_entry(vmf, page, addr);
update_mmu_cache(vma, addr, vmf->pte);
} while (vmf->pte++, page++, addr += PAGE_SIZE, ++count < len);
+ if (maplen) {
+ page_add_file_rmap_range(folio, pg, maplen, vma, false);
+ add_mm_counter(vma->vm_mm, mm_counter_file(pg), maplen);
+ }
folio_ref_add(folio, ref_count);
WRITE_ONCE(file->f_ra.mmap_miss, mmap_miss);
--
2.30.2
next prev parent reply other threads:[~2023-01-30 12:53 UTC|newest]
Thread overview: 20+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-01-30 12:54 [RFC PATCH 0/5] folio based filemap_map_pages() Yin Fengwei
2023-01-30 12:55 ` [RFC PATCH 1/5] mm: Enable fault around for shared file page fault Yin Fengwei
2023-01-30 13:21 ` Matthew Wilcox
2023-01-31 0:59 ` Yin, Fengwei
2023-01-30 12:55 ` [RFC PATCH 2/5] filemap: do file page mapping with folio granularity Yin Fengwei
2023-01-30 13:35 ` Matthew Wilcox
2023-01-31 1:03 ` Yin, Fengwei
2023-01-31 3:34 ` Huang, Ying
2023-01-31 6:32 ` Yin, Fengwei
2023-01-30 12:55 ` [RFC PATCH 3/5] rmap: add page_add_file_rmap_range() Yin Fengwei
2023-01-30 13:50 ` Matthew Wilcox
2023-01-31 7:24 ` Huang, Ying
2023-01-31 7:48 ` Yin, Fengwei
2023-01-30 12:55 ` [RFC PATCH 4/5] mm: add do_set_pte_entry() Yin Fengwei
2023-01-30 13:53 ` Matthew Wilcox
2023-01-31 1:06 ` Yin, Fengwei
2023-01-30 12:55 ` Yin Fengwei [this message]
2023-01-30 14:14 ` [RFC PATCH 5/5] filemap: batched update mm counter,rmap when map file folio Matthew Wilcox
2023-01-31 1:11 ` Yin, Fengwei
2023-01-30 12:58 ` [RFC PATCH 0/5] folio based filemap_map_pages() Yin, Fengwei
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=20230130125504.2509710-6-fengwei.yin@intel.com \
--to=fengwei.yin@intel.com \
--cc=dave.hansen@intel.com \
--cc=david@redhat.com \
--cc=linux-mm@kvack.org \
--cc=tim.c.chen@intel.com \
--cc=willy@infradead.org \
--cc=ying.huang@intel.com \
/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