Revise and improve ntfs_subr.c 1.30: read only a single cluster at a time

in ntfs_writentvattr_plain and ntfs_readntvattr_plain, and purge the boot
block from the buffer cache if isn't exactly one cluster long. These two
changes work around the same buffer cache bug that ntfs_subr.c 1.30 tried
to, but in a different way. This may decrease throughput by reading smaller
amounts of data from the disk at a time, but may increase it by avoiding
bogus writes of clean buffers.
Problem (re)reported by Karel J. Bosschaart on -current.

1 files changed, 8 insertions, 0 deletions

diff --git a/sys/fs/ntfs/ntfs_vfsops.c b/sys/fs/ntfs/ntfs_vfsops.c
index 0f3e60e..f01c6ff 100644
--- a/sys/fs/ntfs/ntfs_vfsops.c
+++ b/sys/fs/ntfs/ntfs_vfsops.c
@@ -311,6 +311,14 @@ ntfs_mountfs(devvp, mp, argsp, td)
 		goto out;
 	ntmp = malloc( sizeof *ntmp, M_NTFSMNT, M_WAITOK | M_ZERO);
 	bcopy( bp->b_data, &ntmp->ntm_bootfile, sizeof(struct bootfile) );
+	/*
+	 * We must not cache the boot block if its size is not exactly
+	 * one cluster in order to avoid confusing the buffer cache when
+	 * the boot file is read later by ntfs_readntvattr_plain(), which
+	 * reads a cluster at a time.
+	 */
+	if (ntfs_cntob(1) != BBSIZE)
+		bp->b_flags |= B_NOCACHE;
 	brelse( bp );
 	bp = NULL;