Skip to contentSkip to conversation

avoid xfs transaction reservation recursion

PATCH Superseded Yafang Shao 2 participants 5 messages Aug 19, 2020

Subsystems (6)

v5 of 15 Superseded Latest is v15 →

lore
Share & tools
lore
#0 lore raw

v5 cover avoid xfs transaction reservation recursion

This patchset avoids transaction reservation recursion by reintroducing the discarded PF_FSTRANS in a new way, suggested by Dave. In this new implementation, four new helpers are introduced, which are xfs_trans_context_{set, clear, update} and fstrans_context_active, suggested by Dave. And re-using the task->journal_info to indicates whehter the task is in fstrans or not, suggested by Willy

Patch #1 is picked from Willy's patchset "Overhaul memalloc_no*"[1]

[1] https://lore.kernel.org/linux-mm/[email protected]/

v5:
- pick one of Willy's patch
- introduce four new helpers, per Dave

v4:
- retitle from "xfs: introduce task->in_fstrans for transaction reservation recursion protection"
- reuse current->journal_info, per Willy

Matthew Wilcox (Oracle) (1):
mm: Add become_kswapd and restore_kswapd

Yafang Shao (1):
xfs: avoid transaction reservation recursion

 fs/iomap/buffered-io.c    |  4 ++--
 fs/xfs/libxfs/xfs_btree.c | 14 ++++++++------
 fs/xfs/xfs_aops.c         |  5 +++--
 fs/xfs/xfs_linux.h        |  4 ----
 fs/xfs/xfs_trans.c        | 19 +++++++++----------
 fs/xfs/xfs_trans.h        | 23 +++++++++++++++++++++++
 include/linux/iomap.h     |  7 +++++++
 include/linux/sched/mm.h  | 28 ++++++++++++++++++++++++++++
 mm/vmscan.c               | 16 +---------------
 9 files changed, 81 insertions(+), 39 deletions(-)
Signature · 1 line
2.18.1
#1 lore raw

PATCH v5 1/2 mm: Add become_kswapd and restore_kswapd

From: "Matthew Wilcox (Oracle)" <[email protected]>

Since XFS needs to pretend to be kswapd in some of its worker threads, create methods to save & restore kswapd state. Don't bother restoring kswapd state in kswapd -- the only time we reach this code is when we're exiting and the task_struct is about to be destroyed anyway.

Cc Dave Chinner <[email protected]>
Cc Christoph Hellwig <[email protected]>
Cc Michal Hocko <[email protected]>
Cc Darrick J. Wong <[email protected]>
Cc Matthew Wilcox <[email protected]>
---
 fs/xfs/libxfs/xfs_btree.c | 14 ++++++++------
 include/linux/sched/mm.h  | 28 ++++++++++++++++++++++++++++
 mm/vmscan.c               | 16 +---------------
 3 files changed, 37 insertions(+), 21 deletions(-)
fs/xfs/libxfs/xfs_btree.c +8 -6
index 2d25bab68764..a04a44238aab 100644
@@ -2813,8 +2813,9 @@ xfs_btree_split_worker(
 {
 	struct xfs_btree_split_args	*args = container_of(work,
 						struct xfs_btree_split_args, work);
+	bool			is_kswapd = args->kswapd;
 	unsigned long		pflags;
-	unsigned long		new_pflags = PF_MEMALLOC_NOFS;
+	int			memalloc_nofs;
 
 	/*
 	 * we are in a transaction context here, but may also be doing work
@@ -2822,16 +2823,17 @@ xfs_btree_split_worker(
 	 * temporarily to ensure that we don't block waiting for memory reclaim
 	 * in any way.
 	 */
-	if (args->kswapd)
-		new_pflags |= PF_MEMALLOC | PF_SWAPWRITE | PF_KSWAPD;
-
-	current_set_flags_nested(&pflags, new_pflags);
+	if (is_kswapd)
+		pflags = become_kswapd();
+	memalloc_nofs = memalloc_nofs_save();
 
 	args->result = __xfs_btree_split(args->cur, args->level, args->ptrp,
 					 args->key, args->curp, args->stat);
 	complete(args->done);
 
-	current_restore_flags_nested(&pflags, new_pflags);
+	memalloc_nofs_restore(memalloc_nofs);
+	if (is_kswapd)
+		restore_kswapd(pflags);
 }
 
 /*
include/linux/sched/mm.h +28 -0
index f889e332912f..80cc132f13a1 100644
@@ -303,6 +303,34 @@ static inline void memalloc_nocma_restore(unsigned int flags)
 }
 #endif
 
+/*
+ * Tell the memory management that we're a "memory allocator",
+ * and that if we need more memory we should get access to it
+ * regardless (see "__alloc_pages()"). "kswapd" should
+ * never get caught in the normal page freeing logic.
+ *
+ * (Kswapd normally doesn't need memory anyway, but sometimes
+ * you need a small amount of memory in order to be able to
+ * page out something else, and this flag essentially protects
+ * us from recursively trying to free more memory as we're
+ * trying to free the first piece of memory in the first place).
+ */
+#define KSWAPD_PF_FLAGS		(PF_MEMALLOC | PF_SWAPWRITE | PF_KSWAPD)
+
+static inline unsigned long become_kswapd(void)
+{
+	unsigned long flags = current->flags & KSWAPD_PF_FLAGS;
+
+	current->flags |= KSWAPD_PF_FLAGS;
+
+	return flags;
+}
+
+static inline void restore_kswapd(unsigned long flags)
+{
+	current->flags &= ~(flags ^ KSWAPD_PF_FLAGS);
+}
+
 #ifdef CONFIG_MEMCG
 /**
  * memalloc_use_memcg - Starts the remote memcg charging scope.
mm/vmscan.c +1 -15
index 99e1796eb833..3a2615bfde35 100644
@@ -3859,19 +3859,7 @@ static int kswapd(void *p)
 	if (!cpumask_empty(cpumask))
 		set_cpus_allowed_ptr(tsk, cpumask);
 
-	/*
-	 * Tell the memory management that we're a "memory allocator",
-	 * and that if we need more memory we should get access to it
-	 * regardless (see "__alloc_pages()"). "kswapd" should
-	 * never get caught in the normal page freeing logic.
-	 *
-	 * (Kswapd normally doesn't need memory anyway, but sometimes
-	 * you need a small amount of memory in order to be able to
-	 * page out something else, and this flag essentially protects
-	 * us from recursively trying to free more memory as we're
-	 * trying to free the first piece of memory in the first place).
-	 */
-	tsk->flags |= PF_MEMALLOC | PF_SWAPWRITE | PF_KSWAPD;
+	become_kswapd();
 	set_freezable();
 
 	WRITE_ONCE(pgdat->kswapd_order, 0);
@@ -3921,8 +3909,6 @@ static int kswapd(void *p)
 			goto kswapd_try_sleep;
 	}
 
-	tsk->flags &= ~(PF_MEMALLOC | PF_SWAPWRITE | PF_KSWAPD);
-
 	return 0;
 }
 
Signature · 1 line
2.18.1
Acked-by Michal Hocko on v5 1/2
#2 lore raw

Re: [PATCH v5 1/2] mm: Add become_kswapd and restore_kswapd

Quoted 7 lines · Yafang Shao

From: "Matthew Wilcox (Oracle)" <[email protected]>

Since XFS needs to pretend to be kswapd in some of its worker threads, create methods to save & restore kswapd state. Don't bother restoring kswapd state in kswapd -- the only time we reach this code is when we're exiting and the task_struct is about to be destroyed anyway.

Quoted 7 lines
Cc Dave Chinner <[email protected]>
Cc Christoph Hellwig <[email protected]>
Cc Michal Hocko <[email protected]>
Cc Darrick J. Wong <[email protected]>
Cc Matthew Wilcox <[email protected]>

See https://lore.kernel.org/linux-mm/[email protected]/

Please add:

Quoted 4 lines

+/*
+ * Tell the memory management that we're a "memory allocator",
+ * and that if we need more memory we should get access to it
+ * regardless (see "__alloc_pages()"). "kswapd" should

Quoted 8 lines

+ * never get caught in the normal page freeing logic.
+ *
+ * (Kswapd normally doesn't need memory anyway, but sometimes
+ * you need a small amount of memory in order to be able to
+ * page out something else, and this flag essentially protects
+ * us from recursively trying to free more memory as we're
+ * trying to free the first piece of memory in the first place).
+ */

And let's change that comment as suggested by Michal (slightly edited by me):

/*
* Tell the memory management code that this thread is working on behalf
* of background memory reclaim (like kswapd). That means that it will
* get access to memory reserves should it need to allocate memory in
* order to make forward progress. With this great power comes great
* responsibility to not exhaust those reserves.
*/

Quoted 56 lines
include/linux/sched/mm.h +16 -0
+#define KSWAPD_PF_FLAGS		(PF_MEMALLOC | PF_SWAPWRITE | PF_KSWAPD)
+
+static inline unsigned long become_kswapd(void)
+{
+	unsigned long flags = current->flags & KSWAPD_PF_FLAGS;
+
+	current->flags |= KSWAPD_PF_FLAGS;
+
+	return flags;
+}
+
+static inline void restore_kswapd(unsigned long flags)
+{
+	current->flags &= ~(flags ^ KSWAPD_PF_FLAGS);
+}
+
 #ifdef CONFIG_MEMCG
 /**
  * memalloc_use_memcg - Starts the remote memcg charging scope.
mm/vmscan.c +1 -15
index 99e1796eb833..3a2615bfde35 100644
@@ -3859,19 +3859,7 @@ static int kswapd(void *p)
 	if (!cpumask_empty(cpumask))
 		set_cpus_allowed_ptr(tsk, cpumask);
 
-	/*
-	 * Tell the memory management that we're a "memory allocator",
-	 * and that if we need more memory we should get access to it
-	 * regardless (see "__alloc_pages()"). "kswapd" should
-	 * never get caught in the normal page freeing logic.
-	 *
-	 * (Kswapd normally doesn't need memory anyway, but sometimes
-	 * you need a small amount of memory in order to be able to
-	 * page out something else, and this flag essentially protects
-	 * us from recursively trying to free more memory as we're
-	 * trying to free the first piece of memory in the first place).
-	 */
-	tsk->flags |= PF_MEMALLOC | PF_SWAPWRITE | PF_KSWAPD;
+	become_kswapd();
 	set_freezable();
 
 	WRITE_ONCE(pgdat->kswapd_order, 0);
@@ -3921,8 +3909,6 @@ static int kswapd(void *p)
 			goto kswapd_try_sleep;
 	}
 
-	tsk->flags &= ~(PF_MEMALLOC | PF_SWAPWRITE | PF_KSWAPD);
-
 	return 0;
 }
 
Signature · 1 line
2.18.1
#3 lore raw

Re: [PATCH v5 1/2] mm: Add become_kswapd and restore_kswapd

Quoted 17 lines · Matthew Wilcox
Quoted 7 lines · Yafang Shao

From: "Matthew Wilcox (Oracle)" <[email protected]>

Since XFS needs to pretend to be kswapd in some of its worker threads, create methods to save & restore kswapd state. Don't bother restoring kswapd state in kswapd -- the only time we reach this code is when we're exiting and the task_struct is about to be destroyed anyway.

Quoted 7 lines
Cc Dave Chinner <[email protected]>
Cc Christoph Hellwig <[email protected]>
Cc Michal Hocko <[email protected]>
Cc Darrick J. Wong <[email protected]>
Cc Matthew Wilcox <[email protected]>
Quoted 6 lines

See https://lore.kernel.org/linux-mm/[email protected]/

Please add:

Sure.
I missed that discussion.

Quoted 16 lines
Quoted 4 lines

+/*
+ * Tell the memory management that we're a "memory allocator",
+ * and that if we need more memory we should get access to it
+ * regardless (see "__alloc_pages()"). "kswapd" should

Quoted 8 lines

+ * never get caught in the normal page freeing logic.
+ *
+ * (Kswapd normally doesn't need memory anyway, but sometimes
+ * you need a small amount of memory in order to be able to
+ * page out something else, and this flag essentially protects
+ * us from recursively trying to free more memory as we're
+ * trying to free the first piece of memory in the first place).
+ */

And let's change that comment as suggested by Michal (slightly edited by me):

Quoted 8 lines

/*
* Tell the memory management code that this thread is working on behalf
* of background memory reclaim (like kswapd). That means that it will
* get access to memory reserves should it need to allocate memory in
* order to make forward progress. With this great power comes great
* responsibility to not exhaust those reserves.
*/

I will update it with that comment.

Quoted 56 lines
Quoted 56 lines
include/linux/sched/mm.h +16 -0
+#define KSWAPD_PF_FLAGS              (PF_MEMALLOC | PF_SWAPWRITE | PF_KSWAPD)
+
+static inline unsigned long become_kswapd(void)
+{
+     unsigned long flags = current->flags & KSWAPD_PF_FLAGS;
+
+     current->flags |= KSWAPD_PF_FLAGS;
+
+     return flags;
+}
+
+static inline void restore_kswapd(unsigned long flags)
+{
+     current->flags &= ~(flags ^ KSWAPD_PF_FLAGS);
+}
+
 #ifdef CONFIG_MEMCG
 /**
  * memalloc_use_memcg - Starts the remote memcg charging scope.
mm/vmscan.c +1 -16
index 99e1796eb833..3a2615bfde35 100644
@@ -3859,19 +3859,7 @@ static int kswapd(void *p)
      if (!cpumask_empty(cpumask))
              set_cpus_allowed_ptr(tsk, cpumask);
 
-     /*
-      * Tell the memory management that we're a "memory allocator",
-      * and that if we need more memory we should get access to it
-      * regardless (see "__alloc_pages()"). "kswapd" should
-      * never get caught in the normal page freeing logic.
-      *
-      * (Kswapd normally doesn't need memory anyway, but sometimes
-      * you need a small amount of memory in order to be able to
-      * page out something else, and this flag essentially protects
-      * us from recursively trying to free more memory as we're
-      * trying to free the first piece of memory in the first place).
-      */
-     tsk->flags |= PF_MEMALLOC | PF_SWAPWRITE | PF_KSWAPD;
+     become_kswapd();
      set_freezable();
 
      WRITE_ONCE(pgdat->kswapd_order, 0);
@@ -3921,8 +3909,6 @@ static int kswapd(void *p)
                      goto kswapd_try_sleep;
      }
 
-     tsk->flags &= ~(PF_MEMALLOC | PF_SWAPWRITE | PF_KSWAPD);
-
      return 0;
 }
 
--

2.18.1

Signature · 2 lines
Thanks
Yafang
#4 lore raw

PATCH v5 2/2 xfs: avoid transaction reservation recursion

In reply to Yafang Shao

PF_FSTRANS which is used to avoid transaction reservation recursion, is dropped since commit 9070733b4efa ("xfs: abstract PF_FSTRANS to PF_MEMALLOC_NOFS") and commit 7dea19f9ee63 ("mm: introduce memalloc_nofs_{save,restore} API") and replaced by PF_MEMALLOC_NOFS which means to avoid filesystem reclaim recursion. That change is subtle. Let's take the exmple of the check of WARN_ON_ONCE(current->flags & PF_MEMALLOC_NOFS)) to explain why this abstraction from PF_FSTRANS to PF_MEMALLOC_NOFS is not proper.
Below comment is quoted from Dave,

Quoted 5 lines

It wasn't for memory allocation recursion protection in XFS - it was for transaction reservation recursion protection by something trying to flush data pages while holding a transaction reservation. Doing this could deadlock the journal because the existing reservation could prevent the nested reservation for being able to reserve space

Quoted 8 lines

in the journal and that is a self-deadlock vector.
IOWs, this check is not protecting against memory reclaim recursion bugs at all (that's the previous check [1]). This check is protecting against the filesystem calling writepages directly from a context where it can self-deadlock.
So what we are seeing here is that the PF_FSTRANS -> PF_MEMALLOC_NOFS abstraction lost all the actual useful information about what type of error this check was protecting against.

As a result, we should reintroduce PF_FSTRANS. As current->journal_info isn't used in XFS, we can reuse it to indicate whehter the task is in fstrans or not, Per Willy. To achieve that, four new helpers are introduce in this patch, per Dave:
- xfs_trans_context_set()
Used in xfs_trans_alloc()
- xfs_trans_context_clear()
Used in xfs_trans_commit() and xfs_trans_cancel()
- xfs_trans_context_update()
Used in xfs_trans_roll()
- fstrans_context_active()
To check whehter current is in fs transcation or not

[1]. Below check is to avoid memory reclaim recursion. if (WARN_ON_ONCE((current->flags & (PF_MEMALLOC|PF_KSWAPD)) ==

        PF_MEMALLOC))
        goto redirty;
Cc Dave Chinner <[email protected]>
Cc Christoph Hellwig <[email protected]>
Cc Michal Hocko <[email protected]>
Cc Darrick J. Wong <[email protected]>
Cc Matthew Wilcox <[email protected]>
---
 fs/iomap/buffered-io.c |  4 ++--
 fs/xfs/xfs_aops.c      |  5 +++--
 fs/xfs/xfs_linux.h     |  4 ----
 fs/xfs/xfs_trans.c     | 19 +++++++++----------
 fs/xfs/xfs_trans.h     | 23 +++++++++++++++++++++++
 include/linux/iomap.h  |  7 +++++++
 6 files changed, 44 insertions(+), 18 deletions(-)
fs/iomap/buffered-io.c +2 -2
index bcfc288dba3f..8043224ec079 100644
@@ -1500,9 +1500,9 @@ iomap_do_writepage(struct page *page, struct writeback_control *wbc, void *data)
 
 	/*
 	 * Given that we do not allow direct reclaim to call us, we should
-	 * never be called in a recursive filesystem reclaim context.
+	 * never be called while in a filesystem transaction.
 	 */
-	if (WARN_ON_ONCE(current->flags & PF_MEMALLOC_NOFS))
+	if (WARN_ON_ONCE(fstrans_context_active()))
 		goto redirty;
 
 	/*
fs/xfs/xfs_aops.c +3 -2
index b35611882ff9..83e0a1840221 100644
@@ -62,7 +62,8 @@ xfs_setfilesize_trans_alloc(
 	 * We hand off the transaction to the completion thread now, so
 	 * clear the flag here.
 	 */
-	current_restore_flags_nested(&tp->t_pflags, PF_MEMALLOC_NOFS);
+	xfs_trans_context_clear(tp);
+
 	return 0;
 }
 
@@ -125,7 +126,7 @@ xfs_setfilesize_ioend(
 	 * thus we need to mark ourselves as being in a transaction manually.
 	 * Similarly for freeze protection.
 	 */
-	current_set_flags_nested(&tp->t_pflags, PF_MEMALLOC_NOFS);
+	xfs_trans_context_set(tp);
 	__sb_writers_acquired(VFS_I(ip)->i_sb, SB_FREEZE_FS);
 
 	/* we abort the update if there was an IO error */
fs/xfs/xfs_linux.h +0 -4
index ab737fed7b12..8a4f6db77e33 100644
@@ -102,10 +102,6 @@ typedef __u32			xfs_nlink_t;
 #define xfs_cowb_secs		xfs_params.cowb_timer.val
 
 #define current_cpu()		(raw_smp_processor_id())
-#define current_set_flags_nested(sp, f)		\
-		(*(sp) = current->flags, current->flags |= (f))
-#define current_restore_flags_nested(sp, f)	\
-		(current->flags = ((current->flags & ~(f)) | (*(sp) & (f))))
 
 #define NBBY		8		/* number of bits per byte */
 
fs/xfs/xfs_trans.c +9 -10
index ed72867b1a19..5f3a4ff51b3c 100644
@@ -153,8 +153,6 @@ xfs_trans_reserve(
 	int			error = 0;
 	bool			rsvd = (tp->t_flags & XFS_TRANS_RESERVE) != 0;
 
-	/* Mark this thread as being in a transaction */
-	current_set_flags_nested(&tp->t_pflags, PF_MEMALLOC_NOFS);
 
 	/*
 	 * Attempt to reserve the needed disk blocks by decrementing
@@ -163,10 +161,8 @@ xfs_trans_reserve(
 	 */
 	if (blocks > 0) {
 		error = xfs_mod_fdblocks(mp, -((int64_t)blocks), rsvd);
-		if (error != 0) {
-			current_restore_flags_nested(&tp->t_pflags, PF_MEMALLOC_NOFS);
+		if (error != 0)
 			return -ENOSPC;
-		}
 		tp->t_blk_res += blocks;
 	}
 
@@ -241,8 +237,6 @@ xfs_trans_reserve(
 		tp->t_blk_res = 0;
 	}
 
-	current_restore_flags_nested(&tp->t_pflags, PF_MEMALLOC_NOFS);
-
 	return error;
 }
 
@@ -284,6 +278,8 @@ xfs_trans_alloc(
 	INIT_LIST_HEAD(&tp->t_dfops);
 	tp->t_firstblock = NULLFSBLOCK;
 
+	/* Mark this thread as being in a transaction */
+	xfs_trans_context_set(tp);
 	error = xfs_trans_reserve(tp, resp, blocks, rtextents);
 	if (error) {
 		xfs_trans_cancel(tp);
@@ -878,7 +874,8 @@ __xfs_trans_commit(
 
 	xfs_log_commit_cil(mp, tp, &commit_lsn, regrant);
 
-	current_restore_flags_nested(&tp->t_pflags, PF_MEMALLOC_NOFS);
+	if (!regrant)
+		xfs_trans_context_clear(tp);
 	xfs_trans_free(tp);
 
 	/*
@@ -910,7 +907,8 @@ __xfs_trans_commit(
 			xfs_log_ticket_ungrant(mp->m_log, tp->t_ticket);
 		tp->t_ticket = NULL;
 	}
-	current_restore_flags_nested(&tp->t_pflags, PF_MEMALLOC_NOFS);
+
+	xfs_trans_context_clear(tp);
 	xfs_trans_free_items(tp, !!error);
 	xfs_trans_free(tp);
 
@@ -971,7 +969,7 @@ xfs_trans_cancel(
 	}
 
 	/* mark this thread as no longer being in a transaction */
-	current_restore_flags_nested(&tp->t_pflags, PF_MEMALLOC_NOFS);
+	xfs_trans_context_clear(tp);
 
 	xfs_trans_free_items(tp, dirty);
 	xfs_trans_free(tp);
@@ -1013,6 +1011,7 @@ xfs_trans_roll(
 	if (error)
 		return error;
 
+	xfs_trans_context_update(trans, *tpp);
 	/*
 	 * Reserve space in the log for the next transaction.
 	 * This also pushes items in the "AIL", the list of logged items,
fs/xfs/xfs_trans.h +23 -0
index b752501818d2..895f560229d6 100644
@@ -243,4 +243,27 @@ void		xfs_trans_buf_copy_type(struct xfs_buf *dst_bp,
 
 extern kmem_zone_t	*xfs_trans_zone;
 
+static inline void
+xfs_trans_context_set(struct xfs_trans *tp)
+{
+	ASSERT(!current->journal_info);
+	current->journal_info = tp;
+	tp->t_pflags = memalloc_nofs_save();
+}
+
+static inline void
+xfs_trans_context_update(struct xfs_trans *old, struct xfs_trans *new)
+{
+	ASSERT(current->journal_info == old);
+	current->journal_info = new;
+}
+
+static inline void
+xfs_trans_context_clear(struct xfs_trans *tp)
+{
+	ASSERT(current->journal_info == tp);
+	current->journal_info = NULL;
+	memalloc_nofs_restore(tp->t_pflags);
+}
+
 #endif	/* __XFS_TRANS_H__ */
include/linux/iomap.h +7 -0
index 4d1d3c3469e9..54194dd6009d 100644
@@ -271,4 +271,11 @@ int iomap_swapfile_activate(struct swap_info_struct *sis,
 # define iomap_swapfile_activate(sis, swapfile, pagespan, ops)	(-EIO)
 #endif /* CONFIG_SWAP */
 
+/* Use the journal_info to indicate current is in a transaction */
+static inline bool
+fstrans_context_active(void)
+{
+	return current->journal_info != NULL;
+}
+
 #endif /* LINUX_IOMAP_H */
Signature · 1 line
2.18.1