linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: "Zach O'Keefe" <zokeefe@google.com>
To: Alex Shi <alex.shi@linux.alibaba.com>,
	David Hildenbrand <david@redhat.com>,
	 David Rientjes <rientjes@google.com>,
	Michal Hocko <mhocko@suse.com>,
	 Pasha Tatashin <pasha.tatashin@soleen.com>,
	SeongJae Park <sj@kernel.org>,  Song Liu <songliubraving@fb.com>,
	Vlastimil Babka <vbabka@suse.cz>, Zi Yan <ziy@nvidia.com>,
	 linux-mm@kvack.org
Cc: Andrea Arcangeli <aarcange@redhat.com>,
	Andrew Morton <akpm@linux-foundation.org>,
	 Arnd Bergmann <arnd@arndb.de>,
	Axel Rasmussen <axelrasmussen@google.com>,
	 Chris Kennelly <ckennelly@google.com>,
	Chris Zankel <chris@zankel.net>, Helge Deller <deller@gmx.de>,
	 Hugh Dickins <hughd@google.com>,
	Ivan Kokshaysky <ink@jurassic.park.msu.ru>,
	 "James E.J. Bottomley" <James.Bottomley@HansenPartnership.com>,
	Jens Axboe <axboe@kernel.dk>,
	 "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>,
	Matthew Wilcox <willy@infradead.org>,
	 Matt Turner <mattst88@gmail.com>,
	Max Filippov <jcmvbkbc@gmail.com>,
	 Miaohe Lin <linmiaohe@huawei.com>,
	Minchan Kim <minchan@kernel.org>,
	 Patrick Xia <patrickx@google.com>,
	Pavel Begunkov <asml.silence@gmail.com>,
	 Peter Xu <peterx@redhat.com>,
	Richard Henderson <rth@twiddle.net>,
	 Thomas Bogendoerfer <tsbogend@alpha.franken.de>,
	Yang Shi <shy828301@gmail.com>,
	 "Zach O'Keefe" <zokeefe@google.com>
Subject: [RFC PATCH 11/14] mm/madvise: introduce MADV_COLLAPSE sync hugepage collapse
Date: Tue,  8 Mar 2022 13:34:14 -0800	[thread overview]
Message-ID: <20220308213417.1407042-12-zokeefe@google.com> (raw)
In-Reply-To: <20220308213417.1407042-1-zokeefe@google.com>

The idea of hugepage collapse in process context was previously
introduced by David Rientjes to linux-mm[1].

The idea is to introduce a new madvise mode, MADV_COLLAPSE, that allows
users to request a synchronous collapse of memory.

The benefits of this approach are:

* cpu is charged to the process that wants to spend the cycles for the
  THP
* avoid unpredictable timing of khugepaged collapse
* flexible separation of sync userspace and async khugepaged THP collapse
  policies

Immediate users of this new functionality include:

* malloc implementations that manage memory in hugepage-sized chunks,
  but sometimes subrelease memory back to the system in native-sized
  chunks via MADV_DONTNEED; zapping the pmd.  Later, when the memory
  is hot, the implementation could madvise(MADV_COLLAPSE) to re-back the
  memory by THP to regain TLB performance.
* immediately back executable text by hugepages.  Current support
  provided by CONFIG_READ_ONLY_THP_FOR_FS may take too long on a large
  system.

To keep patches digestible, introduce MADV_COLLAPSE in a few stages.

Add plumbing to existing madvise infrastructure, as well as populate
uapi header files, leaving the actual madvise(MADV_COLLAPSE) handler
stubbed out.  Only privately-mapped anon memory is supported for now.

[1] https://lore.kernel.org/linux-mm/d098c392-273a-36a4-1a29-59731cdf5d3d@google.com/

Signed-off-by: Zach O'Keefe <zokeefe@google.com>
---
 include/linux/huge_mm.h                | 12 +++++++
 include/uapi/asm-generic/mman-common.h |  2 ++
 mm/khugepaged.c                        | 46 ++++++++++++++++++++++++++
 mm/madvise.c                           |  5 +++
 4 files changed, 65 insertions(+)

diff --git a/include/linux/huge_mm.h b/include/linux/huge_mm.h
index fd905b0b2c71..407b63ab4185 100644
--- a/include/linux/huge_mm.h
+++ b/include/linux/huge_mm.h
@@ -226,6 +226,9 @@ void __split_huge_pud(struct vm_area_struct *vma, pud_t *pud,
 
 int hugepage_madvise(struct vm_area_struct *vma, unsigned long *vm_flags,
 		     int advice);
+int madvise_collapse(struct vm_area_struct *vma,
+		     struct vm_area_struct **prev,
+		     unsigned long start, unsigned long end);
 void vma_adjust_trans_huge(struct vm_area_struct *vma, unsigned long start,
 			   unsigned long end, long adjust_next);
 spinlock_t *__pmd_trans_huge_lock(pmd_t *pmd, struct vm_area_struct *vma);
@@ -383,6 +386,15 @@ static inline int hugepage_madvise(struct vm_area_struct *vma,
 	BUG();
 	return 0;
 }
+
+static inline int madvise_collapse(struct vm_area_struct *vma,
+				   struct vm_area_struct **prev,
+				   unsigned long start, unsigned long end)
+{
+	BUG();
+	return 0;
+}
+
 static inline void vma_adjust_trans_huge(struct vm_area_struct *vma,
 					 unsigned long start,
 					 unsigned long end,
diff --git a/include/uapi/asm-generic/mman-common.h b/include/uapi/asm-generic/mman-common.h
index 6c1aa92a92e4..6ce1f1ceb432 100644
--- a/include/uapi/asm-generic/mman-common.h
+++ b/include/uapi/asm-generic/mman-common.h
@@ -77,6 +77,8 @@
 
 #define MADV_DONTNEED_LOCKED	24	/* like DONTNEED, but drop locked pages too */
 
+#define MADV_COLLAPSE	25		/* Synchronous hugepage collapse */
+
 /* compatibility flags */
 #define MAP_FILE	0
 
diff --git a/mm/khugepaged.c b/mm/khugepaged.c
index 12ae765c5c32..ca1e523086ed 100644
--- a/mm/khugepaged.c
+++ b/mm/khugepaged.c
@@ -2519,3 +2519,49 @@ void khugepaged_min_free_kbytes_update(void)
 		set_recommended_min_free_kbytes();
 	mutex_unlock(&khugepaged_mutex);
 }
+
+/*
+ * Returns 0 if successfully able to collapse range into THPs (or range already
+ * backed by THPs). Due to implementation detail, THPs collapsed here may be
+ * split again before this function returns.
+ */
+static int _madvise_collapse(struct mm_struct *mm,
+			     struct vm_area_struct *vma,
+			     struct vm_area_struct **prev,
+			     unsigned long start,
+			     unsigned long end, gfp_t gfp,
+			     struct collapse_control *cc)
+{
+	/* Implemented in later patch */
+	return -ENOSYS;
+}
+
+int madvise_collapse(struct vm_area_struct *vma,
+		     struct vm_area_struct **prev, unsigned long start,
+		     unsigned long end)
+{
+	struct collapse_control cc;
+	gfp_t gfp;
+	int error;
+	struct mm_struct *mm = vma->vm_mm;
+
+	/* Requested to hold mmap_lock in read */
+	mmap_assert_locked(mm);
+
+	mmgrab(mm);
+	collapse_control_init(&cc, /* enforce_pte_scan_limits= */ false);
+	gfp = vma_thp_gfp_mask(vma);
+	lru_add_drain(); /* lru_add_drain_all() too heavy here */
+	error = _madvise_collapse(mm, vma, prev, start, end, gfp, &cc);
+	mmap_assert_locked(mm);
+	mmdrop(mm);
+
+	/*
+	 * madvise() returns EAGAIN if kernel resources are temporarily
+	 * unavailable.
+	 */
+	if (error == -ENOMEM)
+		error = -EAGAIN;
+
+	return error;
+}
diff --git a/mm/madvise.c b/mm/madvise.c
index 5b6d796e55de..292aa017c150 100644
--- a/mm/madvise.c
+++ b/mm/madvise.c
@@ -58,6 +58,7 @@ static int madvise_need_mmap_write(int behavior)
 	case MADV_FREE:
 	case MADV_POPULATE_READ:
 	case MADV_POPULATE_WRITE:
+	case MADV_COLLAPSE:
 		return 0;
 	default:
 		/* be safe, default to 1. list exceptions explicitly */
@@ -1046,6 +1047,8 @@ static int madvise_vma_behavior(struct vm_area_struct *vma,
 		if (error)
 			goto out;
 		break;
+	case MADV_COLLAPSE:
+		return madvise_collapse(vma, prev, start, end);
 	}
 
 	anon_name = anon_vma_name(vma);
@@ -1139,6 +1142,7 @@ madvise_behavior_valid(int behavior)
 #ifdef CONFIG_TRANSPARENT_HUGEPAGE
 	case MADV_HUGEPAGE:
 	case MADV_NOHUGEPAGE:
+	case MADV_COLLAPSE:
 #endif
 	case MADV_DONTDUMP:
 	case MADV_DODUMP:
@@ -1328,6 +1332,7 @@ int madvise_set_anon_name(struct mm_struct *mm, unsigned long start,
  *  MADV_NOHUGEPAGE - mark the given range as not worth being backed by
  *		transparent huge pages so the existing pages will not be
  *		coalesced into THP and new pages will not be allocated as THP.
+ *  MADV_COLLAPSE - synchronously coalesce pages into new THP.
  *  MADV_DONTDUMP - the application wants to prevent pages in the given range
  *		from being included in its core dump.
  *  MADV_DODUMP - cancel MADV_DONTDUMP: no longer exclude from core dump.
-- 
2.35.1.616.g0bdcbb4464-goog



  parent reply	other threads:[~2022-03-08 21:35 UTC|newest]

Thread overview: 57+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-03-08 21:34 [RFC PATCH 00/14] mm: userspace hugepage collapse Zach O'Keefe
2022-03-08 21:34 ` [RFC PATCH 01/14] mm/rmap: add mm_find_pmd_raw helper Zach O'Keefe
2022-03-09 22:48   ` Yang Shi
2022-03-08 21:34 ` [RFC PATCH 02/14] mm/khugepaged: add struct collapse_control Zach O'Keefe
2022-03-09 22:53   ` Yang Shi
2022-03-08 21:34 ` [RFC PATCH 03/14] mm/khugepaged: add __do_collapse_huge_page() helper Zach O'Keefe
2022-03-08 21:34 ` [RFC PATCH 04/14] mm/khugepaged: separate khugepaged_scan_pmd() scan and collapse Zach O'Keefe
2022-03-08 21:34 ` [RFC PATCH 05/14] mm/khugepaged: add mmap_assert_locked() checks to scan_pmd() Zach O'Keefe
2022-03-08 21:34 ` [RFC PATCH 06/14] mm/khugepaged: add hugepage_vma_revalidate_pmd_count() Zach O'Keefe
2022-03-09 23:15   ` Yang Shi
2022-03-08 21:34 ` [RFC PATCH 07/14] mm/khugepaged: add vm_flags_ignore to hugepage_vma_revalidate_pmd_count() Zach O'Keefe
2022-03-09 23:17   ` Yang Shi
2022-03-10  0:00     ` Zach O'Keefe
2022-03-10  0:41       ` Yang Shi
2022-03-10  1:09         ` Zach O'Keefe
2022-03-10  2:16           ` Yang Shi
2022-03-10 15:50             ` Zach O'Keefe
2022-03-10 18:17               ` Yang Shi
2022-03-10 18:46                 ` David Rientjes
2022-03-10 18:58                   ` Zach O'Keefe
2022-03-10 19:54                   ` Yang Shi
2022-03-10 20:24                     ` Zach O'Keefe
2022-03-10 18:53                 ` Zach O'Keefe
2022-03-10 15:56   ` David Hildenbrand
2022-03-10 18:39     ` Zach O'Keefe
2022-03-10 18:54     ` David Rientjes
2022-03-21 14:27       ` Michal Hocko
2022-03-08 21:34 ` [RFC PATCH 08/14] mm/thp: add madv_thp_vm_flags to __transparent_hugepage_enabled() Zach O'Keefe
2022-03-08 21:34 ` [RFC PATCH 09/14] mm/khugepaged: record SCAN_PAGE_COMPOUND when scan_pmd() finds THP Zach O'Keefe
2022-03-09 23:40   ` Yang Shi
2022-03-10  0:46     ` Zach O'Keefe
2022-03-10  2:05       ` Yang Shi
2022-03-10  8:37         ` Zach O'Keefe
2022-03-08 21:34 ` [RFC PATCH 10/14] mm/khugepaged: rename khugepaged-specific/not functions Zach O'Keefe
2022-03-08 21:34 ` Zach O'Keefe [this message]
2022-03-09 23:43   ` [RFC PATCH 11/14] mm/madvise: introduce MADV_COLLAPSE sync hugepage collapse Yang Shi
2022-03-10  1:11     ` Zach O'Keefe
2022-03-08 21:34 ` [RFC PATCH 12/14] mm/madvise: introduce batched madvise(MADV_COLLPASE) collapse Zach O'Keefe
2022-03-10  0:06   ` Yang Shi
2022-03-10 19:26     ` David Rientjes
2022-03-10 20:16       ` Matthew Wilcox
2022-03-11  0:06         ` Zach O'Keefe
2022-03-25 16:51           ` Zach O'Keefe
2022-03-25 19:54             ` Yang Shi
2022-03-08 21:34 ` [RFC PATCH 13/14] mm/madvise: add __madvise_collapse_*_batch() actions Zach O'Keefe
2022-03-08 21:34 ` [RFC PATCH 14/14] mm/madvise: add process_madvise(MADV_COLLAPSE) Zach O'Keefe
2022-03-21 14:32 ` [RFC PATCH 00/14] mm: userspace hugepage collapse Zi Yan
2022-03-21 14:51   ` Zach O'Keefe
2022-03-21 14:37 ` Michal Hocko
2022-03-21 15:46   ` Zach O'Keefe
2022-03-22 12:11     ` Michal Hocko
2022-03-22 15:53       ` Zach O'Keefe
2022-03-29 12:24         ` Michal Hocko
2022-03-30  0:36           ` Zach O'Keefe
2022-03-22  6:40 ` Zach O'Keefe
2022-03-22 12:05   ` Michal Hocko
2022-03-23 13:30     ` Zach O'Keefe

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=20220308213417.1407042-12-zokeefe@google.com \
    --to=zokeefe@google.com \
    --cc=James.Bottomley@HansenPartnership.com \
    --cc=aarcange@redhat.com \
    --cc=akpm@linux-foundation.org \
    --cc=alex.shi@linux.alibaba.com \
    --cc=arnd@arndb.de \
    --cc=asml.silence@gmail.com \
    --cc=axboe@kernel.dk \
    --cc=axelrasmussen@google.com \
    --cc=chris@zankel.net \
    --cc=ckennelly@google.com \
    --cc=david@redhat.com \
    --cc=deller@gmx.de \
    --cc=hughd@google.com \
    --cc=ink@jurassic.park.msu.ru \
    --cc=jcmvbkbc@gmail.com \
    --cc=kirill.shutemov@linux.intel.com \
    --cc=linmiaohe@huawei.com \
    --cc=linux-mm@kvack.org \
    --cc=mattst88@gmail.com \
    --cc=mhocko@suse.com \
    --cc=minchan@kernel.org \
    --cc=pasha.tatashin@soleen.com \
    --cc=patrickx@google.com \
    --cc=peterx@redhat.com \
    --cc=rientjes@google.com \
    --cc=rth@twiddle.net \
    --cc=shy828301@gmail.com \
    --cc=sj@kernel.org \
    --cc=songliubraving@fb.com \
    --cc=tsbogend@alpha.franken.de \
    --cc=vbabka@suse.cz \
    --cc=willy@infradead.org \
    --cc=ziy@nvidia.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;
as well as URLs for NNTP newsgroup(s).