gfs2: clean_journal improperly set sd_log_flush_head
authorBob Peterson <rpeterso@redhat.com>
Mon, 25 Mar 2019 15:34:19 +0000 (09:34 -0600)
committerAndreas Gruenbacher <agruenba@redhat.com>
Tue, 7 May 2019 21:39:04 +0000 (23:39 +0200)
commit7c70b896951c84d63e6d71b82668f9c8b8bbd440
tree87b69c01020de67faf1fe3536413146436cd1636
parent7881ef3f33bb80f459ea6020d1e021fc524a6348
gfs2: clean_journal improperly set sd_log_flush_head

This patch fixes regressions in 588bff95c94efc05f9e1a0b19015c9408ed7c0ef.
Due to that patch, function clean_journal was setting the value of
sd_log_flush_head, but that's only valid if it is replaying the node's
own journal. If it's replaying another node's journal, that's completely
wrong and will lead to multiple problems. This patch tries to clean up
the mess by passing the value of the logical journal block number into
gfs2_write_log_header so the function can treat non-owned journals
generically. For the local journal, the journal extent map is used for
best performance. For other nodes from other journals, new function
gfs2_lblk_to_dblk is called to figure it out using gfs2_iomap_get.

This patch also tries to establish more consistency when passing journal
block parameters by changing several unsigned int types to a consistent
u32.

Fixes: 588bff95c94e ("GFS2: Reduce code redundancy writing log headers")
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
fs/gfs2/bmap.c
fs/gfs2/bmap.h
fs/gfs2/incore.h
fs/gfs2/log.c
fs/gfs2/log.h
fs/gfs2/lops.c
fs/gfs2/lops.h
fs/gfs2/recovery.c
fs/gfs2/recovery.h