Drop v_interlock for zfs_range_lock(), another thread might hold part
of this range and wait for v_interlock.
This commit is contained in:
parent
398893d51c
commit
2595f06922
|
@ -5985,7 +5985,9 @@ zfs_netbsd_putpages(void *v)
|
|||
len = UINT64_MAX;
|
||||
else
|
||||
len = offhi - offlo;
|
||||
mutex_exit(vp->v_interlock);
|
||||
rl = zfs_range_lock(zp, offlo, len, RL_WRITER);
|
||||
mutex_enter(vp->v_interlock);
|
||||
tsd_set(zfs_putpage_key, &cleaned);
|
||||
}
|
||||
error = genfs_putpages(v);
|
||||
|
|
Loading…
Reference in New Issue