Message-ID:	<B0000029172@stargate.astr.lu.lv>
From:	"Andris Pavenis" <pavenis AT lanet DOT lv>
To:	Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
Date:	Mon, 4 May 1998 11:45:02 +0300
MIME-Version:	1.0
Subject:	Re: mkstemp.c
CC:	djgpp-workers AT delorie DOT com
References:	<B0000028948 AT stargate DOT astr DOT lu DOT lv>
In-reply-to:	<Pine.SUN.3.91.980503171653.19648A@is>

Date sent:      	Sun, 3 May 1998 17:17:28 +0300 (IDT)
From:           	Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
Subject:        	Re: mkstemp.c

> 
> On Wed, 29 Apr 1998, Andris Pavenis wrote:
> 
> > For gcc I think we should avoid putting temporary directory on
> > network drive if possible.
> 
> You cannot rely on people to endure such a requirement.  Somebody,
> someday will do the unthinkable, and it is not nice for GCC to fail
> then.

I checked this with shared disk from WinNT 4.0 and all worked Ok. 
I don't have possibility to test this with other redirectors, but I think
the situation is not so terrible even if some of them is returning something 
weird as the current return value of biostime is used to generate the name
problems fore ms may occur only if first attempt to create temporary file 
fails that should be very rare event. If there will be some problems we'll
be able to return to this question later when next version of gcc will be released...

> 
> But I think your suggestion is safe enough, so I don't think we should
> worry too much about that.
> 
> > In suggested version (errno != ENOENT) program can go in unnecessary
> > loop on some I/O problem. Perhaps it would be better to post some test
> > example for checking this in different systems and ask to send the results.
> 
> It never hurts to ask people to test and report results.  So please go
> ahead and ask them.
> 
The only problem here is to write some test suite for checking results of
file operation in different situations. Currently I don't have one and I don't know 
when (or whether) I'll have

Andris

- Raw text -