> -----Original Message-----
> From: rtems-devel-boun...@rtems.org [mailto:rtems-devel-
> boun...@rtems.org] On Behalf Of Sebastian Huber
> Sent: Friday, October 25, 2013 2:15 AM
> To: rtems-devel@rtems.org
> Subject: Re: cpuset macro /vs/ inline implementation
>
> On 2013-10-24 22:26, Jennifer Averett wrote:
> > /*
> > * Copyright (c) 2013 On-Line Applications Research Corporation.
> > * All rights reserved.
> > *
> > * On-Line Applications Research Corporation
> > * 7047 Old Madison Pike Suite 320
> > * Huntsville Alabama 35806
> > *<i...@oarcorp.com>
> > *
> > * 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.
> > * 2. Redistributions in binary form must reproduce the above copyright
> > * notice, this list of conditions and the following disclaimer in the
> > * documentation and/or other materials provided with the distribution.
> > *
> > * 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.
> > */
> >
> > /*
> > * This file implements an API compatible with static portion of
> > * the GNU/Linux cpu_set_t macros but is independently implemented.
> > * The GNU/Linux manual page and the FreeBSD cpuset_t implementation
> > * were used as reference material.
> > *
> > * Not implemented:
> > * + Linux CPU_XXX_S
> > * + FreeBSD CPU_SUBSET
> > * + FreeBSD CPU_OVERLAP
> > */
> >
> > #ifndef_SYS_CPUSET_H_
> > #define_SYS_CPUSET_H_
> >
> > #include <sys/types.h>
> >
> > /* RTEMS supports a maximum of 32 CPU cores */ #define CPU_SETSIZE 32
>
> Why not use
>
> #ifndef CPU_SETSIZE
> #define CPU_SETSIZE 32
> #endif
> like in FreeBSD?
Fine
>
> >
> > /* Number of bits per cpu_set_t element */ #define _NCPUBITS
> > (sizeof(long) * NBBY) /* bits per mask */
>
> NBBY is defined in <sys/param.h>.
>
> > /* Number of words in the cpu_set_t array */ #define _NCPUWORDS
> > howmany(CPU_SETSIZE, _NCPUBITS)
>
> howmany() is defined in <sys/param.h>.
Will include in the defines at top.
> >
> > /* Define the cpu set structure */
> > typedef struct _cpuset {
> > long __bits[_NCPUWORDS];
> > } cpu_set_t;
>
> We need statically initialized CPU sets for RTEMS (e.g. for the
> configuration).
> So we should use this
>
> typedef uint32_t cpu_set_word_t;
>
> typedef struct {
> cpu_set_word_t bits[_NCPUWORDS];
> } cpu_set_t;
Will change
> >
> > /* determine the mask for a particular cpu within the element */
> > #define __cpuset_mask(_cpu) ((long)1 << ((_cpu) % _NCPUBITS))
> > /* determine the index for this cpu within the cpu set array */
> > #define __cpuset_index(_cpu) ((_cpu)/_NCPUBITS)
>
> Why not use an inline function?
Will change...
> > /* zero out set */
> > static inline void CPU_ZERO(cpu_set_t *set) {
> > size_t i;
> > for (i = 0; i < _NCPUWORDS; i++)
> > set->__bits[i] = 0;
> > }
> >
> > /* fill set */
> > static inline void CPU_FILL(cpu_set_t *set) {
> > size_t i;
> > for (i = 0; i < _NCPUWORDS; i++)
> > set->__bits[i] = -1;
> > }
>
> Why not use a const qualifier if possible?
I'll add const.
Do we also want to add restrict where appropriate?
> > /* set cpu within set */
> > static inline void CPU_SET(size_t cpu, cpu_set_t *set) {
> > set->__bits[__cpuset_index(cpu)] |= __cpuset_mask(cpu); }
> >
> > /* clear cpu within set */
> > static inline void CPU_CLR(size_t cpu, cpu_set_t *set) {
> > set->__bits[__cpuset_index(cpu)] &= ~__cpuset_mask(cpu); }
> >
> > /* Return 1 is cpu is set in set, 0 otherwise */ static inline int
> > CPU_ISSET(size_t cpu, cpu_set_t *set) {
> > return ((set->__bits[__cpuset_index(cpu)] & __cpuset_mask(cpu)) !=
> > 0); }
> >
> > /* copy src set to dest set */
> > static inline void CPU_COPY(cpu_set_t *dest, cpu_set_t *src) {
> > *dest = *src;
> > }
> >
> > /* logical and: dest set = src1 set and src2 set */ static inline void
> > CPU_AND(cpu_set_t *dest,cpu_set_t *src1, cpu_set_t *src2) {
> > size_t i;
> > for (i = 0; i < _NCPUWORDS; i++)
> > dest->__bits[i] = src1->__bits[i] & src2->__bits[i]; }
> >
> > /* logical or: dest set = src1 set or src2 set */ static inline void
> > CPU_OR(cpu_set_t *dest,cpu_set_t *src1, cpu_set_t *src2) {
> > size_t i;
> > for (i = 0; i < _NCPUWORDS; i++)
> > dest->__bits[i] = src1->__bits[i] | src2->__bits[i]; }
> >
> > /* logical xor: dest set = src1 set xor src2 set */ static inline void
> > CPU_XOR(cpu_set_t *dest,cpu_set_t *src1, cpu_set_t *src2) {
> > size_t i;
> > for (i = 0; i < _NCPUWORDS; i++)
> > dest->__bits[i] = src1->__bits[i] ^ src2->__bits[i]; }
> >
> > /* logical nand: dest set = src1 set nand src2 set */ static inline
> > void CPU_NAND(cpu_set_t *dest,cpu_set_t *src1, cpu_set_t *src2) {
> > size_t i;
> > for (i = 0; i < _NCPUWORDS; i++)
> > dest->__bits[i] = ~(src1->__bits[i] & src2->__bits[i]); }
> >
> > /* return the number of set cpus in set */ static inline int
> > CPU_COUNT(cpu_set_t *set) {
> > size_t i;
> > int count = 0;
> >
> > for (i=0; i < _NCPUWORDS; i++)
> > if (CPU_ISSET(i, set) != 0)
> > count++;
> > return count;
> > }
> >
> > /* return 1 if the sets set1 and set2 are equal, otherwise return 0 */
> > static inline int CPU_EQUAL(cpu_set_t *set1, cpu_set_t *set2) {
> > size_t i;
> >
> > for (i=0; i < _NCPUWORDS; i++)
> > if (set1->__bits[i] != set2->__bits[i] )
> > return 0;
> > return 1;
> > }
>
> Why not use bool as return type?
BSD and linux return 0 or 1 do we want to deviate that much?
> >
> > /* return 1 if the sets set1 and set2 are equal, otherwise return 0 */
> > static inline int CPU_CMP(cpu_set_t *set1, cpu_set_t *set2) {
> > return CPU_EQUAL(set1, set2);
> > }
> >
> > /* return 1 if the set is empty, otherwise return 0 */ static inline
> > int CPU_EMPTY(cpu_set_t *set) {
> > size_t i;
> >
> > for (i=0; i < _NCPUWORDS; i++)
> > if (set->__bits[i] != 0 )
> > return 0;
> > return 1;
> > }
> >
> > #endif
>
>
> --
> Sebastian Huber, embedded brains GmbH
>
> Address : Dornierstr. 4, D-82178 Puchheim, Germany
> Phone : +49 89 189 47 41-16
> Fax : +49 89 189 47 41-09
> E-Mail : sebastian.hu...@embedded-brains.de
> PGP : Public key available on request.
>
> Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG.
> _______________________________________________
> rtems-devel mailing list
> rtems-devel@rtems.org
> http://www.rtems.org/mailman/listinfo/rtems-devel
Jennifer Averett
On-Line Applications Research
256-319-2752
_______________________________________________
rtems-devel mailing list
rtems-devel@rtems.org
http://www.rtems.org/mailman/listinfo/rtems-devel
