Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Smtp-Origin-Hostprefix: devbig
From:   Song Liu <songliubraving@fb.com>
Smtp-Origin-Hostname: devbig006.ftw2.facebook.com
To:     <linux-mm@kvack.org>, <linux-fsdevel@vger.kernel.org>,
        <linux-kernel@vger.kernel.org>
CC:     <matthew.wilcox@oracle.com>, <kirill.shutemov@linux.intel.com>,
        <kernel-team@fb.com>, <william.kucharski@oracle.com>,
        <akpm@linux-foundation.org>, <hdanton@sina.com>,
        Song Liu <songliubraving@fb.com>
Smtp-Origin-Cluster: ftw2c04
Subject: [PATCH v9 0/6] Enable THP for text section of non-shmem files
Date:   Mon, 24 Jun 2019 17:12:40 -0700
Message-ID: <20190625001246.685563-1-songliubraving@fb.com>
MIME-Version: 1.0
Content-Type: text/plain
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Changes v8 => v9:
1. Fix bad use of IS_ENABLED (kbuild test robot)

Changes v7 => v8:
1. Use IS_ENABLED wherever possible (Kirill A. Shutemov);
2. Improve handling of !PageUptodate case (Kirill A. Shutemov);
3. Add comment for calling lru_add_drain (Kirill A. Shutemov);
4. Add more information about DENYWRITE dynamic (Johannes Weiner).

Changes v6 => v7:
1. Avoid accessing vma without holding mmap_sem (Hillf Dayton)
2. In collapse_file() use readahead API instead of gup API. This matches
   better with existing logic for shmem.
3. Add inline documentation for @nr_thps (kbuild test robot)

Changes v5 => v6:
1. Improve THP stats in 3/6, (Kirill).

Changes v4 => v5:
1. Move the logic to drop THP from pagecache to open() path (Rik).
2. Revise description of CONFIG_READ_ONLY_THP_FOR_FS.

Changes v3 => v4:
1. Put the logic to drop THP from pagecache in a separate function (Rik).
2. Move the function to drop THP from pagecache to exit_mmap().
3. Revise confusing commit log 6/6.

Changes v2 => v3:
1. Removed the limitation (cannot write to file with THP) by truncating
   whole file during sys_open (see 6/6);
2. Fixed a VM_BUG_ON_PAGE() in filemap_fault() (see 2/6);
3. Split function rename to a separate patch (Rik);
4. Updated condition in hugepage_vma_check() (Rik).

Changes v1 => v2:
1. Fixed a missing mem_cgroup_commit_charge() for non-shmem case.

This set follows up discussion at LSF/MM 2019. The motivation is to put
text section of an application in THP, and thus reduces iTLB miss rate and
improves performance. Both Facebook and Oracle showed strong interests to
this feature.

To make reviews easier, this set aims a mininal valid product. Current
version of the work does not have any changes to file system specific
code. This comes with some limitations (discussed later).

This set enables an application to "hugify" its text section by simply
running something like:

          madvise(0x600000, 0x80000, MADV_HUGEPAGE);

Before this call, the /proc/<pid>/maps looks like:

    00400000-074d0000 r-xp 00000000 00:27 2006927     app

After this call, part of the text section is split out and mapped to
THP:

    00400000-00425000 r-xp 00000000 00:27 2006927     app
    00600000-00e00000 r-xp 00200000 00:27 2006927     app   <<< on THP
    00e00000-074d0000 r-xp 00a00000 00:27 2006927     app

Limitations:

1. This only works for text section (vma with VM_DENYWRITE).
2. Original limitation #2 is removed in v3.

We gated this feature with an experimental config, READ_ONLY_THP_FOR_FS.
Once we get better support on the write path, we can remove the config and
enable it by default.

Tested cases:
1. Tested with btrfs and ext4.
2. Tested with real work application (memcache like caching service).
3. Tested with "THP aware uprobe":
   https://patchwork.kernel.org/project/linux-mm/list/?series=131339

This set (plus a few uprobe patches) is also available at

   https://github.com/liu-song-6/linux/tree/uprobe-thp

Please share your comments and suggestions on this.

Thanks!

Song Liu (6):
  filemap: check compound_head(page)->mapping in filemap_fault()
  filemap: update offset check in filemap_fault()
  mm,thp: stats for file backed THP
  khugepaged: rename collapse_shmem() and khugepaged_scan_shmem()
  mm,thp: add read-only THP support for (non-shmem) FS
  mm,thp: avoid writes to file with THP in pagecache

 drivers/base/node.c    |   6 +++
 fs/inode.c             |   3 ++
 fs/namei.c             |  23 +++++++-
 fs/proc/meminfo.c      |   4 ++
 fs/proc/task_mmu.c     |   4 +-
 include/linux/fs.h     |  32 +++++++++++
 include/linux/mmzone.h |   2 +
 mm/Kconfig             |  11 ++++
 mm/filemap.c           |   9 ++--
 mm/khugepaged.c        | 117 ++++++++++++++++++++++++++++++++---------
 mm/rmap.c              |  12 +++--
 mm/vmstat.c            |   2 +
 12 files changed, 189 insertions(+), 36 deletions(-)

--
2.17.1