Date: Fri, 21 Jun 2002 12:42:35 -0400 From: Mikhail Teterin <mi+mx@aldan.algebra.com> To: Garance A Drosihn <drosih@rpi.edu>, Juli Mallett <jmallett@FreeBSD.ORG> Cc: arch@FreeBSD.ORG Subject: Re: feature request for xargs Message-ID: <200206211242.35819.mi%2Bmx@aldan.algebra.com> In-Reply-To: <p0511178cb938f5a581d7@[128.113.24.47]> References: <200206200706.g5K76M514469@freefall.freebsd.org> <200206211033.03948.mi%2Bmx@aldan.algebra.com> <p0511178cb938f5a581d7@[128.113.24.47]>
next in thread | previous in thread | raw e-mail | index | archive | help
On Friday 21 June 2002 11:45 am, Garance A Drosihn wrote: = > = >All operations, involving many light tasks can benefit, i.e. = > = > find /some/dir \! -type d -print0 | xargs -0 -j0 rm -f = = It's not clear to me that this specific case would benefit. If you = have a large number of files, then this will spawn off a large number = of processes, only to see those processes Not _that_ large -- each rm will (sequentionally) be removing up to MAX_ARGS (sp?) files. = In thinking about that example, I wonder if we should not allow '-j 0' = to spawn infinite processes. Maybe allow a -j value from 1 to 64 (or = some other arbitrary number), just to reduce the chances of a one-line = "fork bomb". Well, not infinite, but up to one's maxproc. gmake's -l <L> option may be a good idea too -- unlimited so long the system load remains at or below <L> (but at least one). -mi To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-arch" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200206211242.35819.mi%2Bmx>