1
0
mirror of git://sourceware.org/git/lvm2.git synced 2024-12-29 15:22:30 +03:00

Always use blocking lock for VGs and orphan locks.

Because we have now strong rule for lock ordering:
 - VG locks must be taken in alphabetical order
 - ORPHAN locks must be the last
vgs_locked() is now not needed.

This fixes problem with orphan locking, e.g.
vgremove VG1    |    vgremove VG2
lock(VG1)       |    lock(VG2)
lock(ORPHAN)    |    lock(ORPHAN) -> fail, non-blocking

https://bugzilla.redhat.com/show_bug.cgi?id=578413

(More similar places in code.)
This commit is contained in:
Milan Broz 2010-03-31 17:23:56 +00:00
parent 6733116a19
commit d7cbaae1fd
2 changed files with 3 additions and 4 deletions

View File

@ -1,5 +1,6 @@
Version 2.02.63 -
================================
Always use blocking lock for VGs and orphan locks.
Allocate all segments memory from vg private mempool.
Return newly allocated PV segment after segment split.
Optimise PV segments search for the most last segment search case.

View File

@ -420,11 +420,9 @@ int lock_vol(struct cmd_context *cmd, const char *vol, uint32_t flags)
switch (flags & LCK_SCOPE_MASK) {
case LCK_VG:
/*
* Automatically set LCK_NONBLOCK if one or more VGs locked.
* This will enforce correctness and prevent deadlocks rather
* than relying on the caller to set the flag properly.
* VG locks alphabetical, ORPHAN lock last
*/
if (!_blocking_supported || vgs_locked())
if (!_blocking_supported)
flags |= LCK_NONBLOCK;
if (vol[0] != '#' &&