[3166] in Release_Engineering
NetProb ticket #1519, /usr/athena/etc/track, update.c
daemon@ATHENA.MIT.EDU (pjb@MIT.EDU)
Fri Mar 11 10:35:06 1994
From: pjb@MIT.EDU
Date: Fri, 11 Mar 94 10:35:00 -0500
To: rel-eng@MIT.EDU
Cc:
>From haynes@cats.ucsc.edu Thu Mar 10 14:50:50 1994
>Received: by gregor.MIT.EDU (5.57/4.7) id AA14765; Thu, 10 Mar 94 14:50:43 -05>00
>Received: from cats-po-1.UCSC.EDU by MIT.EDU with SMTP
> id AA27535; Thu, 10 Mar 94 14:50:36 EST
>Received: from hobbes.ucsc.edu by cats.ucsc.edu with SMTP
> id LAA16836; Thu, 10 Mar 1994 11:47:42 -0800
>Received: from localhost by hobbes.ucsc.edu (8.6.4/4.7) id LAA05907; Thu, 10 M>ar 1994 11:47:38 -0800
>Date: Thu, 10 Mar 1994 11:47:35 -0800 (PST)
>From: "James H. Haynes" <haynes@cats.ucsc.edu>
>Subject: /usr/athena/etc/track, update.c
>To: bugs@MIT.EDU
>Message-Id: <Pine.3.89.9403101138.A5903-0100000@hobbes.ucsc.edu>
>Mime-Version: 1.0
>Content-Type: TEXT/PLAIN; charset=US-ASCII
>
>In track, the file update.c contains this passage
> sprintf( temp,"%s_trk.tmp",to);
>
> if ( 0 <= ( fdt = open( temp, O_WRONLY | O_CREAT)));
>
> else if ( errno == ENOSPC) {
>
>Some systems, e.g. Linux, require that the open() call have a third argument
>for the file mode when O_CREAT is used; and for those that don't require it
>I wonder what determines what it will be? Anyway Linux without an
>explicit mode argument happens to create the file read-only, so the
>subsequent attempt to write it fails. I don't know what the correct mode
>might be, but I'm guessing 0755.
>
