3.16.55-rc1 review patch. If anyone has any objections, please let me know.
------------------
From: Omar Sandoval osandov@fb.com
commit 1b9e619c5bc8235cfba3dc4ced2fb0e3554a05d4 upstream.
I was seeing disk flushes still happening when I mounted a Btrfs filesystem with nobarrier for testing. This is because we use FUA to write out the first super block, and on devices without FUA support, the block layer translates FUA to a flush. Even on devices supporting true FUA, using FUA when we asked for no barriers is surprising.
Fixes: 387125fc722a8ed ("Btrfs: fix barrier flushes") Signed-off-by: Omar Sandoval osandov@fb.com Reviewed-by: Qu Wenruo wqu@suse.com Reviewed-by: David Sterba dsterba@suse.com Signed-off-by: David Sterba dsterba@suse.com [bwh: Backported to 3.16: - I/O flag names are different, and are combined with the operation type - Use the do_barrier parameter instead of checking the NOBARRIER option again] Signed-off-by: Ben Hutchings ben@decadent.org.uk --- --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -3140,6 +3140,7 @@ static int write_dev_supers(struct btrfs int errors = 0; u32 crc; u64 bytenr; + int op_flags;
if (max_mirrors == 0) max_mirrors = BTRFS_SUPER_MIRROR_MAX; @@ -3204,10 +3205,10 @@ static int write_dev_supers(struct btrfs * we fua the first super. The others we allow * to go down lazy. */ - if (i == 0) - ret = btrfsic_submit_bh(WRITE_FUA, bh); - else - ret = btrfsic_submit_bh(WRITE_SYNC, bh); + op_flags = REQ_SYNC | REQ_NOIDLE; + if (i == 0 && do_barriers) + op_flags |= REQ_FUA; + ret = btrfsic_submit_bh(WRITE | op_flags, bh); if (ret) errors++; }