freebsd-dev/share/man/man3/Q_INI.3
Edward Tomasz Napierala 5a38af1fb8 Introduce <sys/qmath.h>, a fixed-point math library from Netflix.
This makes it possible to perform mathematical operations
on
fractional values without using floating point. It operates on Q
numbers, which are integer-sized, opaque structures initialized
to hold a chosen number of integer and fractional
bits.


For a general description of the Q number system, see the "Fixed Point
Representation & Fractional Math" whitepaper[1]; for the actual
API see the qmath(3) man page.

This is one of dependencies for the upcoming stats(3) framework[2]
that will be applied to the TCP stack in a later commit.

1. https://www.superkits.net/whitepapers/Fixed%20Point%20Representation%20&%20Fractional%20Math.pdf
2. https://reviews.freebsd.org/D20477

Reviewed by:	bcr (man pages, earlier version), sef (earlier version)
Discussed with:	cem, dteske, imp, lstewart
Sponsored By:	Klara Inc, Netflix
Obtained from:	Netflix
Differential Revision:	https://reviews.freebsd.org/D20116
2019-08-27 11:46:22 +00:00

262 lines
6.0 KiB
Groff

.\"
.\" Copyright (c) 2018 Netflix, Inc.
.\" All rights reserved.
.\"
.\" Redistribution and use in source and binary forms, with or without
.\" modification, are permitted provided that the following conditions
.\" are met:
.\" 1. Redistributions of source code must retain the above copyright
.\" notice, this list of conditions, and the following disclaimer,
.\" without modification, immediately at the beginning of the file.
.\" 2. The name of the author may not be used to endorse or promote products
.\" derived from this software without specific prior written permission.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
.\" ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR
.\" ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
.\" SUCH DAMAGE.
.\"
.\" $FreeBSD$
.\"
.Dd July 8, 2018
.Dt Q_INI 3
.Os
.Sh NAME
.Nm Q_INI ,
.Nm Q_NCBITS ,
.Nm Q_BT ,
.Nm Q_TC ,
.Nm Q_NTBITS ,
.Nm Q_NFCBITS ,
.Nm Q_MAXNFBITS ,
.Nm Q_NFBITS ,
.Nm Q_NIBITS ,
.Nm Q_RPSHFT ,
.Nm Q_ABS ,
.Nm Q_MAXSTRLEN ,
.Nm Q_TOSTR ,
.Nm Q_SHL ,
.Nm Q_SHR ,
.Nm Q_DEBUG
.Nd fixed-point math miscellaneous functions/variables
.Sh SYNOPSIS
.In sys/qmath.h
.Ft QTYPE
.Fn Q_INI "QTYPE *q" "ITYPE iv" "ITYPE dfv" "int rpshft"
.Fd Q_NCBITS
.Ft __typeof(q)
.Fn Q_BT "QTYPE q"
.Ft ITYPE
.Fn Q_TC "QTYPE q" "ITYPE v"
.Ft uint32_t
.Fn Q_NTBITS "QTYPE q"
.Ft uint32_t
.Fn Q_NFCBITS "QTYPE q"
.Ft uint32_t
.Fn Q_MAXNFBITS "QTYPE q"
.Ft uint32_t
.Fn Q_NFBITS "QTYPE q"
.Ft uint32_t
.Fn Q_NIBITS "QTYPE q"
.Ft uint32_t
.Fn Q_RPSHFT "QTYPE q"
.Ft NTYPE
.Fn Q_ABS "NTYPE n"
.Ft uint32_t
.Fn Q_MAXSTRLEN "QTYPE q" "int base"
.Ft char *
.Fn Q_TOSTR "QTYPE q" "int prec" "int base" "char *s" "int slen"
.Ft ITYPE
.Fn Q_SHL "QTYPE q" "ITYPE iv"
.Ft ITYPE
.Fn Q_SHR "QTYPE q" "ITYPE iv"
.Ft char *, ...
.Fn Q_DEBUG "QTYPE q" "char *prefmt" "char *postfmt" "incfmt"
.Ft ITYPE
.Fn Q_DFV2BFV "ITYPE dfv" "int nfbits"
.Sh DESCRIPTION
.Fn Q_INI
initialises a Q number with the supplied integral value
.Fa iv
and decimal fractional value
.Fa dfv ,
with appropriate control bits based on the requested radix shift point
.Fa rpshft .
.Fa dfv
must be passed as a preprocessor literal to preserve leading zeroes.
.Pp
The
.Dv Q_NCBITS
defined constant specifies the number of reserved control bits, currently 3.
.Pp
.Fn Q_NTBITS ,
.Fn Q_NFCBITS ,
.Fn Q_MAXNFBITS ,
.Fn Q_NFBITS
and
.Fn Q_NIBITS
return the
.Fa q Ns -specific
count of total, control-encoded fractional, maximum fractional, effective
fractional, and integer bits applicable to
.Fa q
respectively.
.Pp
.Fn Q_BT
returns the C data type of
.Fa q ,
while
.Fn Q_TC
returns
.Fa v
type casted to the C data type of
.Fa q .
.Pp
.Fn Q_RPSHFT
returns the bit position of
.Fa q Ap s
binary radix point relative to bit zero.
.Pp
.Fn Q_ABS
returns the absolute value of any standard numeric type
.Pq that uses the MSB as a sign bit, but not Q numbers
passed in as
.Fa n .
The function is signed/unsigned type safe.
.Pp
.Fn Q_SHL
and
.Fn Q_SHR
return the integral value
.Fa v
left or right shifted by the appropriate amount for
.Fa q .
.Pp
.Fn Q_MAXSTRLEN
calculates the maximum number of characters that may be required to render the
C-string representation of
.Fa q
with numeric base
.Fa base .
.Pp
.Fn Q_TOSTR
renders the C-string representation of
.Fa q
with numeric base
.Fa base
and fractional precision
.Fa prec
into
.Fa s
which has an available capacity of
.Fa slen
characters.
.Fa base
must be in range
.Bq 2,16 .
Specifying
.Fa prec
as -1 renders the number's fractional component with maximum precision.
If
.Fa slen
is greater than zero but insufficient to hold the complete C-string, the '\\0'
C-string terminator will be written to
.Fa *s ,
thereby returning a zero length C-string.
.Pp
.Fn Q_DEBUG
returns a format string and associated data suitable for printf-like rendering
of debugging information pertaining to
.Fa q .
If either
.Fa prefmt
and/or
.Fa postfmt
are specified, they are prepended and appended to the resulting format string
respectively.
The
.Fa incfmt
boolean specifies whether to include
.Pq Vt true
or exclude
.Pq Vt false
the raw format string itself in the debugging output.
.Pp
.Fn Q_DFV2BFV
converts decimal fractional value
.Fa dfv
to its binary-encoded representation with
.Fa nfbits
of binary precision.
.Fa dfv
must be passed as a preprocessor literal to preserve leading zeroes.
The returned value can be used to set a Q number's fractional bits, for example
using
.Fn Q_SFVAL .
.Pp
All of those functions operate on
the following data types:
.Vt s8q_t ,
.Vt u8q_t ,
.Vt s16q_t ,
.Vt u16q_t ,
.Vt s32q_t ,
.Vt u32q_t ,
.Vt s64q_t ,
and
.Vt u64q_t ,
which are referred to generically as
.Fa QTYPE .
The
.Fa ITYPE
refers to the
.Xr stdint 7
integer types.
.Fa NTYPE
is used to refer to any numeric type and is therefore a superset of
.Fa QTYPE
and
.Fa ITYPE .
.Pp
For more details, see
.Xr qmath 3 .
.Sh RETURN VALUES
.Fn Q_INI
returns the initialised Q number which can be used to chain initialise
additional Q numbers.
.Pp
.Fn Q_TOSTR
returns a pointer to the '\\0' C-string terminator appended to
.Fa s
after the rendered numeric data, or NULL on buffer overflow.
.Pp
.Fn Q_DFV2BFV
returns the binary-encoded representation of decimal fractional value
.Fa dfv
with
.Fa nfbits
of binary precision.
.Sh SEE ALSO
.Xr errno 2 ,
.Xr qmath 3 ,
.Xr stdint 7
.Sh HISTORY
The
.Xr qmath 3
functions first appeared in
.Fx 13.0 .
.Sh AUTHORS
.An -nosplit
The
.Xr qmath 3
functions and this manual page were written by
.An Lawrence Stewart Aq Mt lstewart@FreeBSD.org
and sponsored by Netflix, Inc.