[PATCH 51/51] mkfs: hack around not having enough log blocks

[Darrick J. Wong]

Under some circumstances I don't yet understand, mkfs underestimates
the minimum log size and the kernel refuses to mount.  Put in a crude
hack so that this doesn't happen.

DO NOT APPLY!!!

Singed-off-by: Darrick J. Wong <darrick.wong at oracle.com>
---
 mkfs/xfs_mkfs.c |    1 +
 1 file changed, 1 insertion(+)


diff --git a/mkfs/xfs_mkfs.c b/mkfs/xfs_mkfs.c
index 556b691..ab1e8f9 100644
--- a/mkfs/xfs_mkfs.c
+++ b/mkfs/xfs_mkfs.c
@@ -2401,6 +2401,7 @@ an AG size that is one stripe unit smaller, for example %llu.\n"),
 				   logversion, lsunit);
 	ASSERT(min_logblocks);
 	min_logblocks = MAX(XFS_MIN_LOG_BLOCKS, min_logblocks);
+if (min_logblocks < 860) min_logblocks = 860;
 	if (!logsize && dblocks >= (1024*1024*1024) >> blocklog)
 		min_logblocks = MAX(min_logblocks, XFS_MIN_LOG_BYTES>>blocklog);
 	if (logsize && xi.logBBsize > 0 && logblocks > DTOBT(xi.logBBsize)) {