Skip to contentSkip to conversation

Recent changes (master)

PATCH Jens Axboe 1 participant 1 message Sep 6, 2019

lore
Share & tools
lore

Jens Axboe

author
#0 lore raw

The following changes since commit 4a479420d50eada0a7b9a972c529d75e2884732d:

smalloc: use SMALLOC_BPI instead of SMALLOC_BPB in add_pool() (2019-09-03 12:32:01 -0600)

are available in the Git repository at:

git://git.kernel.dk/fio.git master

for you to fetch changes up to 27f436d9f72a9d2d3da3adfdf712757152eab29e:

engines/io_uring: use its own option group (2019-09-05 09:15:41 -0600)

----------------------------------------------------------------
Jens Axboe (2):
engines/io_uring: add support for registered files engines/io_uring: use its own option group

 HOWTO              |   6 +++
 engines/io_uring.c | 117 +++++++++++++++++++++++++++++++++++++++++++++++++----
 fio.1              |   6 +++
 optgroup.h         |   2 +
 4 files changed, 123 insertions(+), 8 deletions(-)

---

Diff of recent changes:

HOWTO +6 -0
index 4201e2e9..6b449e97 100644
@@ -2033,6 +2033,12 @@ with the caveat that when used on the command line, they must come after the
 	map and release for each IO. This is more efficient, and reduces the
 	IO latency as well.
 
+.. option:: registerfiles : [io_uring]
+	With this option, fio registers the set of files being used with the
+	kernel. This avoids the overhead of managing file counts in the kernel,
+	making the submission and completion part more lightweight. Required
+	for the below :option:`sqthread_poll` option.
+
 .. option:: sqthread_poll : [io_uring]
 
 	Normally fio will submit IO by issuing a system call to notify the
engines/io_uring.c +109 -8

200 lines not on this page · raw message

fio.1 +6 -0
index 3e872bce..e0283f7f 100644
@@ -1791,6 +1791,12 @@ release them when IO is done. If this option is set, the pages are pre-mapped
 before IO is started. This eliminates the need to map and release for each IO.
 This is more efficient, and reduces the IO latency as well.
 .TP
+.BI (io_uring)registerfiles
+With this option, fio registers the set of files being used with the kernel.
+This avoids the overhead of managing file counts in the kernel, making the
+submission and completion part more lightweight. Required for the below
+sqthread_poll option.
+.TP
 .BI (io_uring)sqthread_poll
 Normally fio will submit IO by issuing a system call to notify the kernel of
 available items in the SQ ring. If this option is set, the act of submitting IO
optgroup.h +2 -0
index 8009bf25..55ef5934 100644
@@ -64,6 +64,7 @@ enum opt_category_group {
 	__FIO_OPT_G_MMAP,
 	__FIO_OPT_G_ISCSI,
 	__FIO_OPT_G_NBD,
+	__FIO_OPT_G_IOURING,
 	__FIO_OPT_G_NR,
 
 	FIO_OPT_G_RATE		= (1ULL << __FIO_OPT_G_RATE),
@@ -104,6 +105,7 @@ enum opt_category_group {
 	FIO_OPT_G_INVALID	= (1ULL << __FIO_OPT_G_NR),
 	FIO_OPT_G_ISCSI         = (1ULL << __FIO_OPT_G_ISCSI),
 	FIO_OPT_G_NBD		= (1ULL << __FIO_OPT_G_NBD),
+	FIO_OPT_G_IOURING	= (1ULL << __FIO_OPT_G_IOURING),
 };
 
 extern const struct opt_group *opt_group_from_mask(uint64_t *mask);