NetNews Usenet Archive 1992 #19/NN_1992_19.iso/spool/comp/lang/c/12851

Text File | 1992-08-27 | 1.9 KB | 53 lines

Newsgroups: comp.lang.c
Path: sparky!uunet!mdisea!mitchell
From: [e-mail] (Bill Mitchell)
Subject: errno  (was Re: Too many open files)
Message-ID: <[e-mail]>
Sender: [e-mail]
Bcc:  mitchell
Reply-To: [e-mail] (Bill Mitchell)
Organization: Motorola, Mobile Data Division - Seattle, WA
References: <[e-mail]> <[e-mail]>
Date: Thu, 27 Aug 1992 14:22:24 GMT
Lines: 39

in comp.lang.c, [e-mail] (Lars Wirzenius) said:


>[e-mail] (Michel Trudeau) writes:
>> [* fopen fails, some compilers set errno, some don't *]
>
>Yes.  There is no guarantee that fopen set errno to anything at all.
>
>--
>[e-mail]

That's the bottom line, but a bit terse.  Let me expand a bit, reading
from PJP's book The Standard C Library:

Page 50, Using <errno.h> starts out:

"The C Standard leaves much unsaid about the errors that can be reported.
It says even less about the values of any error codes [.....] Even different
versions of UNIX define dirrerent sets of error codes.  If you are writing
code for a specific system you may have to learn its peculiar set of error
codes.  List the header <errno.h> if you can. [...]"

In section 7.1.14, the standard says, "The value of errno is zero at program
startup, but it is never set to zero by any library function.  The value of
errno may be set to nonzero by a library function call whether or not there
is an error, provided the use of errno is not documented in the description
of the function in this International Standard."

Section 7.9.5.3, on the fopen function, does not document the use of errno
by fopen.

So, there is no guarantee that fopen set errno to anything at all, but
it is allowed to set errno to any nonzero value whether or not there
was an error.  See implementation-specific documentation for implementation-
specific details.

--
[e-mail] (Bill Mitchell)