Added in stricter no merge semantics for block I/O
Subsystems (2)
Added in stricter no merge semantics for block I/O
Updated 'nomerges' tunable to accept a value of '2' - indicating that _no_ merges at all are to be attempted (not even the simple one-hit cache).
The following table illustrates the additional benefit - 5 minute runs of a random I/O load were applied to a dozen devices on a 16-way x86_64 system.
nomerges Throughput %System Improvement (tput / %sys) -------- ------------ ----------- ------------------------- 0 12.45 MB/sec 0.669365609 1 12.50 MB/sec 0.641519199 0.40% / 2.71% 2 12.52 MB/sec 0.639849750 0.56% / 2.96%
--- Documentation/ABI/testing/sysfs-block | 14 ++++++++++++++ Documentation/block/queue-sysfs.txt | 10 +++++----- block/blk-sysfs.c | 11 +++++++---- block/elevator.c | 11 ++++++++++- include/linux/blkdev.h | 3 +++ 5 files changed, 39 insertions(+), 10 deletions(-)
Documentation/ABI/testing/sysfs-block +14 -0
@@ -128,3 +128,17 @@ Description: preferred request size for workloads where sustained throughput is desired. If no optimal I/O size is reported this file contains 0. + +What: /sys/block/<disk>/queue/nomerges +Date: January 2010 +Contact: +Description: + Standard I/O elevator operations include attempts to + merge contiguous I/Os. For known random I/O loads these + attempts will always fail and result in extra cycles + being spent in the kernel. This allows one to turn off + this behavior on one of two ways: When set to 1, complex + merge checks are disabled, but the simple one-shot merges + with the previous I/O request are enabled. When set to 2, + all merge tries are disabled. The default value is 0 - + which enables all types of merge tries.
Documentation/block/queue-sysfs.txt +5 -5
@@ -25,11 +25,11 @@ size allowed by the hardware. nomerges (RW) ------------- -This enables the user to disable the lookup logic involved with IO merging -requests in the block layer. Merging may still occur through a direct -1-hit cache, since that comes for (almost) free. The IO scheduler will not -waste cycles doing tree/hash lookups for merges if nomerges is 1. Defaults -to 0, enabling all merges. +This enables the user to disable the lookup logic involved with IO +merging requests in the block layer. By default (0) all merges are +enabled. When set to 1 only simple one-hit merges will be tried. When +set to 2 no merge algorithms will be tried (including one-hit or more +complex tree/hash lookups). nr_requests (RW) ----------------
block/blk-sysfs.c +7 -4
@@ -189,7 +189,8 @@ static ssize_t queue_nonrot_store(struct request_queue *q, const char *page,static ssize_t queue_nomerges_show (struct request_queue *q,char *page) { -return queue_var_show (blk_queue_nomerges (q), page); +return queue_var_show ((blk_queue_nomerges (q) <<1 ) | +blk_queue_noxmerges (q), page); } static ssize_t queue_nomerges_store(struct request_queue *q, const char *page,
@@ -199,10 +200,12 @@ static ssize_t queue_nomerges_store(struct request_queue *q, const char *page,ssize_t ret =queue_var_store (&nm, page, count);spin_lock_irq (q->queue_lock); -if (nm) +queue_flag_clear (QUEUE_FLAG_NOMERGES , q); +queue_flag_clear (QUEUE_FLAG_NOXMERGES , q); +if (nm ==2 )queue_flag_set (QUEUE_FLAG_NOMERGES , q); -else -queue_flag_clear (QUEUE_FLAG_NOMERGES , q); +else if (nm) +queue_flag_set (QUEUE_FLAG_NOXMERGES , q);spin_unlock_irq (q->queue_lock);return ret;
block/elevator.c +10 -1
@@ -474,6 +474,15 @@ int elv_merge(struct request_queue *q, struct request **req, struct bio *bio)int ret;/* +* Levels of merges: +* nomerges: No merges at all attempted +* noxmerges: Only simple one-hit cache try +* merges: All merge tries attempted +*/ +if (blk_queue_nomerges (q)) +return ELEVATOR_NO_MERGE ; + +/* * First try one-hit cache. */ if (q->last_merge) {
@@ -484,7 +493,7 @@ int elv_merge(struct request_queue *q, struct request **req, struct bio *bio) } } -if (blk_queue_nomerges (q)) +if (blk_queue_noxmerges (q))return ELEVATOR_NO_MERGE ;/*
include/linux/blkdev.h +3 -0
@@ -463,6 +463,7 @@ struct request_queue#define QUEUE_FLAG_IO_STAT 15/* do IO stats */ #define QUEUE_FLAG_CQ 16/* hardware does queuing */ #define QUEUE_FLAG_DISCARD 17/* supports DISCARD */ +#define QUEUE_FLAG_NOXMERGES 18/* No extended merges */ #define QUEUE_FLAG_DEFAULT ((1 << QUEUE_FLAG_IO_STAT) | \ (1 << QUEUE_FLAG_CLUSTER) | \
@@ -589,6 +590,8 @@ enum {#define blk_queue_queuing (q) test_bit(QUEUE_FLAG_CQ, &(q)->queue_flags)#define blk_queue_stopped (q) test_bit(QUEUE_FLAG_STOPPED, &(q)->queue_flags)#define blk_queue_nomerges (q) test_bit(QUEUE_FLAG_NOMERGES, &(q)->queue_flags) +#define blk_queue_noxmerges (q) \ + test_bit(QUEUE_FLAG_NOXMERGES, &(q)->queue_flags)#define blk_queue_nonrot (q) test_bit(QUEUE_FLAG_NONROT, &(q)->queue_flags)#define blk_queue_io_stat (q) test_bit(QUEUE_FLAG_IO_STAT, &(q)->queue_flags)#define blk_queue_flushing (q) ((q)->ordseq)
Signature · 1 line
1.6.5
Quoted 16 lines · Alan D. Brunelle
Added in stricter no merge semantics for block I/O
Updated 'nomerges' tunable to accept a value of '2' - indicating that _no_ merges at all are to be attempted (not even the simple one-hit cache).
The following table illustrates the additional benefit - 5 minute runs of a random I/O load were applied to a dozen devices on a 16-way
x86_64system.nomerges Throughput %System Improvement (tput / %sys) -------- ------------ ----------- ------------------------- 0 12.45 MB/sec 0.669365609 1 12.50 MB/sec 0.641519199 0.40% / 2.71% 2 12.52 MB/sec 0.639849750 0.56% / 2.96%Signed-off-by Alan D. Brunelle <[email protected]>
Quoted 7 lines
--- Documentation/ABI/testing/sysfs-block | 14 ++++++++++++++ Documentation/block/queue-sysfs.txt | 10 +++++----- block/blk-sysfs.c | 11 +++++++---- block/elevator.c | 11 ++++++++++- include/linux/blkdev.h | 3 +++ 5 files changed, 39 insertions(+), 10 deletions(-)
Thanks Alan, I think this is a good addition, I have often before hacked up the "no merges at all" logic for testing purposes as well. I have queued it up for 2.6.34.
Signature · 1 line
Jens Axboe