Switch KM_SLEEP to KM_PUSHPAGE

Differences between how paging is done on Solaris and Linux can cause
deadlocks if KM_SLEEP is used in any the following contexts.

  * The txg_sync thread
  * The zvol write/discard threads
  * The zpl_putpage() VFS callback

This is because KM_SLEEP will allow for direct reclaim which may result
in the VM calling back in to the filesystem or block layer to write out
pages.  If a lock is held over this operation the potential exists to
deadlock the system.  To ensure forward progress all memory allocations
in these contexts must us KM_PUSHPAGE which disables performing any I/O
to accomplish the memory allocation.

Previously, this behavior was acheived by setting PF_MEMALLOC on the
thread.  However, that resulted in unexpected side effects such as the
exhaustion of pages in ZONE_DMA.  This approach touchs more of the zfs
code, but it is more consistent with the right way to handle these cases
under Linux.

This is patch lays the ground work for being able to safely revert the
following commits which used PF_MEMALLOC:

  21ade34 Disable direct reclaim for z_wr_* threads
  cfc9a5c Fix zpl_writepage() deadlock
  eec8164 Fix ASSERTION(!dsl_pool_sync_context(tx->tx_pool))

Signed-off-by: Richard Yao <[email protected]>
Signed-off-by: Brian Behlendorf <[email protected]>
Issue #726

4 files changed, 15 insertions, 15 deletions

diff --git a/include/sys/dbuf.h b/include/sys/dbuf.h
index f3e14aad3..ef93eb54c 100644
--- a/include/sys/dbuf.h
+++ b/include/sys/dbuf.h
@@ -345,13 +345,13 @@ boolean_t dbuf_is_metadata(dmu_buf_impl_t *db);
 	} \
 _NOTE(CONSTCOND) } while (0)
 
-#define	dprintf_dbuf_bp(db, bp, fmt, ...) do {			\
-	if (zfs_flags & ZFS_DEBUG_DPRINTF) {			\
-	char *__blkbuf = kmem_alloc(BP_SPRINTF_LEN, KM_SLEEP);	\
-	sprintf_blkptr(__blkbuf, bp);				\
-	dprintf_dbuf(db, fmt " %s\n", __VA_ARGS__, __blkbuf);	\
-	kmem_free(__blkbuf, BP_SPRINTF_LEN);			\
-	}							\
+#define	dprintf_dbuf_bp(db, bp, fmt, ...) do {				\
+	if (zfs_flags & ZFS_DEBUG_DPRINTF) {				\
+	char *__blkbuf = kmem_alloc(BP_SPRINTF_LEN, KM_PUSHPAGE);	\
+	sprintf_blkptr(__blkbuf, bp);					\
+	dprintf_dbuf(db, fmt " %s\n", __VA_ARGS__, __blkbuf);		\
+	kmem_free(__blkbuf, BP_SPRINTF_LEN);				\
+	}								\
 _NOTE(CONSTCOND) } while (0)
 
 #define	DBUF_VERIFY(db)	dbuf_verify(db)
diff --git a/include/sys/dsl_dataset.h b/include/sys/dsl_dataset.h
index c4530a8f0..948abb020 100644
--- a/include/sys/dsl_dataset.h
+++ b/include/sys/dsl_dataset.h
@@ -271,7 +271,7 @@ int dsl_destroy_inconsistent(const char *dsname, void *arg);
 #ifdef ZFS_DEBUG
 #define	dprintf_ds(ds, fmt, ...) do { \
 	if (zfs_flags & ZFS_DEBUG_DPRINTF) { \
-	char *__ds_name = kmem_alloc(MAXNAMELEN, KM_SLEEP); \
+	char *__ds_name = kmem_alloc(MAXNAMELEN, KM_PUSHPAGE); \
 	dsl_dataset_name(ds, __ds_name); \
 	dprintf("ds=%s " fmt, __ds_name, __VA_ARGS__); \
 	kmem_free(__ds_name, MAXNAMELEN); \
diff --git a/include/sys/dsl_dir.h b/include/sys/dsl_dir.h
index 2191635dd..65ad202bb 100644
--- a/include/sys/dsl_dir.h
+++ b/include/sys/dsl_dir.h
@@ -150,7 +150,7 @@ timestruc_t dsl_dir_snap_cmtime(dsl_dir_t *dd);
 #define	dprintf_dd(dd, fmt, ...) do { \
 	if (zfs_flags & ZFS_DEBUG_DPRINTF) { \
 	char *__ds_name = kmem_alloc(MAXNAMELEN + strlen(MOS_DIR_NAME) + 1, \
-	    KM_SLEEP); \
+	    KM_PUSHPAGE); \
 	dsl_dir_name(dd, __ds_name); \
 	dprintf("dd=%s " fmt, __ds_name, __VA_ARGS__); \
 	kmem_free(__ds_name, MAXNAMELEN + strlen(MOS_DIR_NAME) + 1); \
diff --git a/include/sys/spa.h b/include/sys/spa.h
index 9de5736ed..28bb4e1de 100644
--- a/include/sys/spa.h
+++ b/include/sys/spa.h
@@ -690,12 +690,12 @@ extern void spa_configfile_set(spa_t *, nvlist_t *, boolean_t);
 extern void spa_event_notify(spa_t *spa, vdev_t *vdev, const char *name);
 
 #ifdef ZFS_DEBUG
-#define	dprintf_bp(bp, fmt, ...) do {				\
-	if (zfs_flags & ZFS_DEBUG_DPRINTF) { 			\
-	char *__blkbuf = kmem_alloc(BP_SPRINTF_LEN, KM_SLEEP);	\
-	sprintf_blkptr(__blkbuf, (bp));				\
-	dprintf(fmt " %s\n", __VA_ARGS__, __blkbuf);		\
-	kmem_free(__blkbuf, BP_SPRINTF_LEN);			\
+#define	dprintf_bp(bp, fmt, ...) do {					\
+	if (zfs_flags & ZFS_DEBUG_DPRINTF) {				\
+	char *__blkbuf = kmem_alloc(BP_SPRINTF_LEN, KM_PUSHPAGE);	\
+	sprintf_blkptr(__blkbuf, (bp));					\
+	dprintf(fmt " %s\n", __VA_ARGS__, __blkbuf);			\
+	kmem_free(__blkbuf, BP_SPRINTF_LEN);				\
 	} \
 _NOTE(CONSTCOND) } while (0)
 #else