From mboxrd@z Thu Jan  1 00:00:00 1970
X-Msuck: nntp://news.gmane.org/gmane.linux.lib.musl.general/5164
Path: news.gmane.org!not-for-mail
From: Szabolcs Nagy <nsz@port70.net>
Newsgroups: gmane.linux.lib.musl.general
Subject: Re: [UGLY PATCH v2] Support for no-legacy-syscalls archs
Date: Mon, 26 May 2014 23:13:49 +0200
Message-ID: <20140526211349.GJ12324@port70.net>
References: <20140525054237.GA18085@brightrain.aerifal.cx>
 <20140526184036.GZ507@brightrain.aerifal.cx>
Reply-To: musl@lists.openwall.com
NNTP-Posting-Host: plane.gmane.org
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
X-Trace: ger.gmane.org 1401138850 26961 80.91.229.3 (26 May 2014 21:14:10 GMT)
X-Complaints-To: usenet@ger.gmane.org
NNTP-Posting-Date: Mon, 26 May 2014 21:14:10 +0000 (UTC)
To: musl@lists.openwall.com
Original-X-From: musl-return-5169-gllmg-musl=m.gmane.org@lists.openwall.com Mon May 26 23:14:04 2014
Return-path: <musl-return-5169-gllmg-musl=m.gmane.org@lists.openwall.com>
Envelope-to: gllmg-musl@plane.gmane.org
Original-Received: from mother.openwall.net ([195.42.179.200])
	by plane.gmane.org with smtp (Exim 4.69)
	(envelope-from <musl-return-5169-gllmg-musl=m.gmane.org@lists.openwall.com>)
	id 1Wp2Da-0002xp-QB
	for gllmg-musl@plane.gmane.org; Mon, 26 May 2014 23:14:02 +0200
Original-Received: (qmail 5428 invoked by uid 550); 26 May 2014 21:14:02 -0000
Mailing-List: contact musl-help@lists.openwall.com; run by ezmlm
Precedence: bulk
List-Post: <mailto:musl@lists.openwall.com>
List-Help: <mailto:musl-help@lists.openwall.com>
List-Unsubscribe: <mailto:musl-unsubscribe@lists.openwall.com>
List-Subscribe: <mailto:musl-subscribe@lists.openwall.com>
Original-Received: (qmail 5420 invoked from network); 26 May 2014 21:14:01 -0000
Content-Disposition: inline
In-Reply-To: <20140526184036.GZ507@brightrain.aerifal.cx>
User-Agent: Mutt/1.5.21 (2010-09-15)
Xref: news.gmane.org gmane.linux.lib.musl.general:5164
Archived-At: <http://permalink.gmane.org/gmane.linux.lib.musl.general/5164>

* Rich Felker <dalias@libc.org> [2014-05-26 14:40:36 -0400]:
> +#ifdef SYS_fork
>  	ret = syscall(SYS_fork);
> +#else
> +	ret = syscall(SYS_clone, SIGCHLD);
> +#endif

clone has more args

>  int select(int n, fd_set *restrict rfds, fd_set *restrict wfds, fd_set *restrict efds, struct timeval *restrict tv)
>  {
> +#ifdef SYS_select
>  	return syscall_cp(SYS_select, n, rfds, wfds, efds, tv);
> +#else
> +	long data[2] = { 0, _NSIG/8 };
> +	struct timespec ts;
> +	if (tv) {
> +		ts.tv_sec = tv->tv_sec;
> +		ts.tv_nsec = tv->tv_usec > 999999 ?
> +			999999999 : tv->tv_usec * 1000;
> +	}
> +	return syscall_cp(SYS_pselect6, n, rfds, wfds, efds, tv ? &ts : 0, data);
> +#endif

tv_usec may be negative

isnt it better to adjust tv_sec if usec is large?
or fail with EINVAL like in futimensat:

> +	if (times) {
> +		int i;
> +		for (i=0; i<2; i++) {
> +			if (times[i].tv_usec >= 1000000U)
> +				return __syscall_ret(-EINVAL);
> +			ts[i].tv_sec = times[i].tv_sec;
> +			ts[i].tv_nsec = times[i].tv_usec * 1000;
> +		}
> +	}


> +	if (times) {
> +		struct timeval tmp[2];
> +		int i;
> +		tv = tmp;
> +		for (i=0; i<2; i++) {
> +			if (times[i].tv_nsec >= 1000000000U) {
> +				if (times[i].tv_nsec == UTIME_NOW &&
> +				    times[1-i].tv_nsec == UTIME_NOW) {
> +					tv = 0;
> +					break;
> +				}
> +				if (times[i].tv_nsec == UTIME_OMIT)
> +					return __syscall_ret(-ENOSYS);
> +				return __syscall_ret(-EINVAL);
> +			}
> +			tmp[i].tv_sec = times[i].tv_sec;
> +			tmp[i].tv_usec = times[i].tv_nsec / 1000;
> +		}
> +	}

(tv_nsec+500)/1000 would be better rounding

but i guess 500 ns does not matter much

>  pid_t getpgrp(void)
>  {
> +#ifdef SYS_getpgrp
>  	return __syscall(SYS_getpgrp);
> +#else
> +	return __syscall(SYS_getpgid, 0);
> +#endif

you said this can be just the new call