From mboxrd@z Thu Jan  1 00:00:00 1970
Date: Mon, 14 Aug 2000 00:50:44 -0400
From: Alexander Viro <viro@math.psu.edu>
To: 9fans@cse.psu.edu
Subject: Re: [9fans] Potential race in syspipe()
In-Reply-To: <200008140430.AAA01256@cse.psu.edu>
Message-ID: <Pine.GSO.4.10.10008140035390.194-100000@weyl.math.psu.edu>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Topicbox-Message-UUID: fae012b4-eac8-11e9-9e20-41e7f4b1d025



On Mon, 14 Aug 2000 presotto@plan9.bell-labs.com wrote:

> magic in one place.  I noticed your snippet didn't check the
> need to growfd after getting i, perhaps a good reason for
> findfreefd().

	Didn't have to, actually - if the first for() would reach the end
of array we would have i==f->nfd and the second for() would immediately
terminate with j==f->nfd+1. After that we would get the right call of
growfd() and correct values at once.

	But yes, findfreefd() makes the things more obviously correct.

[snip]

> As for create(), I'ld rather change the man page.  Otherwise we'ld
> have to reserve an fd in some way before the open but not have that
> fd visible to other processes until it was assinged a channel
> or let it be step on-able by dup2...

<nods>

							Al