Allow for lock-free reading zfsdev_state_list.

Restructure the zfsdev_state_list to allow for lock-free reading by
converting to a simple singly-linked list from which items are never
deleted and over which only forward iterations are performed.  It depends
on, among other things, the atomicity of accessing the zs_minor integer
and zs_next pointer.

This fixes a lock inversion in which the zfsdev_state_lock is used by
both the sync task (txg_sync) and indirectly by any user program which
uses /dev/zfs; the zfsdev_release method uses the same lock and then
blocks on the sync task.

The most typical failure scenerio occurs when the sync task is cleaning
up a user hold while various concurrent "zfs" commands are in progress.

Neither Illumos nor Solaris are affected by this issue because they use
DDI interface which provides lock-free reading of device state via the
ddi_get_soft_state() function.

Signed-off-by: Tim Chase <tim@chase2k.com>
Signed-off-by: Chunwei Chen <tuxoko@gmail.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #2301
This commit is contained in:
Tim Chase
2014-05-08 09:51:01 -05:00
committed by Brian Behlendorf
parent 1cbae971c5
commit 3937ab20f3
2 changed files with 52 additions and 18 deletions
+8 -1
View File
@@ -371,8 +371,15 @@ enum zfsdev_state_type {
ZST_ALL,
};
/*
* The zfsdev_state_t structure is managed as a singly-linked list
* from which items are never deleted. This allows for lock-free
* reading of the list so long as assignments to the zs_next and
* reads from zs_minor are performed atomically. Empty items are
* indicated by storing -1 into zs_minor.
*/
typedef struct zfsdev_state {
list_node_t zs_next; /* next zfsdev_state_t link */
struct zfsdev_state *zs_next; /* next zfsdev_state_t link */
struct file *zs_file; /* associated file struct */
minor_t zs_minor; /* made up minor number */
void *zs_onexit; /* onexit data */