diff options
author | Mike Snitzer <snitzer@redhat.com> | 2017-02-16 23:57:17 -0500 |
---|---|---|
committer | Mike Snitzer <snitzer@redhat.com> | 2017-02-17 00:54:09 -0500 |
commit | 37a098e9d10db6e2efc05fe61e3a6ff2e9802c53 (patch) | |
tree | 3800aa0a250b90b0325476a7b5ab1680dda5aac7 /fs/nfsd/state.h | |
parent | 6085831883c25860264721df15f05bbded45e2a2 (diff) | |
download | linux-37a098e9d10db6e2efc05fe61e3a6ff2e9802c53.tar.bz2 |
dm round robin: revert "use percpu 'repeat_count' and 'current_path'"
The sloppy nature of lockless access to percpu pointers
(s->current_path) in rr_select_path(), from multiple threads, is
causing some paths to used more than others -- which results in less
IO performance being observed.
Revert these upstream commits to restore truly symmetric round-robin
IO submission in DM multipath:
b0b477c dm round robin: use percpu 'repeat_count' and 'current_path'
802934b dm round robin: do not use this_cpu_ptr() without having preemption disabled
There is no benefit to all this complexity if repeat_count = 1 (which is
the recommended default).
Cc: stable@vger.kernel.org # 4.6+
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Diffstat (limited to 'fs/nfsd/state.h')
0 files changed, 0 insertions, 0 deletions