Improve IEC binary and SI decimal prefix handling

[fio.git] / fio.1

diff --git a/fio.1 b/fio.1
index 270798a47bc0ddaffb7c3a2c4d4497ce2f20456d..7363a375432f351c031fbeb9e4e243808def8c71 100644 (file)
--- a/fio.1
+++ b/fio.1
@@ -287,7 +287,7 @@ Sequential reads.
 Sequential writes.
 .TP
 .B trim
 Sequential writes.
 .TP
 .B trim

-Sequential trim (Linux block devices only).
+Sequential trims (Linux block devices only).

 .TP
 .B randread
 Random reads.
 .TP
 .B randread
 Random reads.


@@ -296,7 +296,7 @@ Random reads.
 Random writes.
 .TP
 .B randtrim
 Random writes.
 .TP
 .B randtrim

-Random trim (Linux block devices only).
+Random trims (Linux block devices only).

 .TP
 .B rw, readwrite
 Mixed sequential reads and writes.
 .TP
 .B rw, readwrite
 Mixed sequential reads and writes.


@@ -305,8 +305,8 @@ Mixed sequential reads and writes.
 Mixed random reads and writes.
 .TP
 .B trimwrite
 Mixed random reads and writes.
 .TP
 .B trimwrite

-Trim and write mixed workload. Blocks will be trimmed first, then the same
-blocks will be written to.
+Sequential trim and write mixed workload. Blocks will be trimmed first, then
+the same blocks will be written to.

 .RE
 .P
 Fio defaults to read if the option is not specified.
 .RE
 .P
 Fio defaults to read if the option is not specified.


@@ -353,7 +353,7 @@ reasons. Allowed values are 1024 or 1000, with 1024 being the default.
 .TP
 .BI unified_rw_reporting \fR=\fPbool
 Fio normally reports statistics on a per data direction basis, meaning that
 .TP
 .BI unified_rw_reporting \fR=\fPbool
 Fio normally reports statistics on a per data direction basis, meaning that

-read, write, and trim are accounted and reported separately. If this option is
+reads, writes, and trims are accounted and reported separately. If this option is

 set fio sums the results and reports them as "mixed" instead.
 .TP
 .BI randrepeat \fR=\fPbool
 set fio sums the results and reports them as "mixed" instead.
 .TP
 .BI randrepeat \fR=\fPbool


@@ -396,9 +396,27 @@ available on Linux. If using ZFS on Solaris this must be set to 'none'
 because ZFS doesn't support it. Default: 'posix'.
 .RE
 .TP
 because ZFS doesn't support it. Default: 'posix'.
 .RE
 .TP

-.BI fadvise_hint \fR=\fPbool
+.BI fadvise_hint \fR=\fPstr

 Use \fBposix_fadvise\fR\|(2) to advise the kernel what I/O patterns
 Use \fBposix_fadvise\fR\|(2) to advise the kernel what I/O patterns

-are likely to be issued. Default: true.
+are likely to be issued. Accepted values are:
+.RS
+.RS
+.TP
+.B 0
+Backwards compatible hint for "no hint".
+.TP
+.B 1
+Backwards compatible hint for "advise with fio workload type". This
+uses \fBFADV_RANDOM\fR for a random workload, and \fBFADV_SEQUENTIAL\fR
+for a sequential workload.
+.TP
+.B sequential
+Advise using \fBFADV_SEQUENTIAL\fR
+.TP
+.B random
+Advise using \fBFADV_RANDOM\fR
+.RE
+.RE

 .TP
 .BI fadvise_stream \fR=\fPint
 Use \fBposix_fadvise\fR\|(2) to advise the kernel what stream ID the
 .TP
 .BI fadvise_stream \fR=\fPint
 Use \fBposix_fadvise\fR\|(2) to advise the kernel what stream ID the


@@ -717,11 +735,14 @@ properly.
 Read, write and erase an MTD character device (e.g., /dev/mtd0). Discards are
 treated as erases. Depending on the underlying device type, the I/O may have
 to go in a certain pattern, e.g., on NAND, writing sequentially to erase blocks
 Read, write and erase an MTD character device (e.g., /dev/mtd0). Discards are
 treated as erases. Depending on the underlying device type, the I/O may have
 to go in a certain pattern, e.g., on NAND, writing sequentially to erase blocks

-and discarding before overwriting. The writetrim mode works well for this
+and discarding before overwriting. The trimwrite mode works well for this

 constraint.
 .TP
 .B pmemblk
 Read and write through the NVML libpmemblk interface.
 constraint.
 .TP
 .B pmemblk
 Read and write through the NVML libpmemblk interface.

+.TP
+.B dev-dax
+Read and write through a DAX device exposed from persistent memory.

 .RE
 .P
 .RE
 .RE
 .P
 .RE