Date: Sun, 4 Apr 1999 14:06:37 +0200 (IST) From: Eli Zaretskii X-Sender: eliz AT is To: "Mark E." , djgpp-workers AT delorie DOT com Subject: Re: Bash 2.03 updated March 22nd In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Reply-To: djgpp-workers AT delorie DOT com X-Mailing-List: djgpp-workers AT delorie DOT com X-Unsubscribes-To: listserv AT delorie DOT com Precedence: bulk On Tue, 30 Mar 1999, I wrote: > > and anything else you can think of that might help me find the problem. > > I used a new version of `cat', but there are no real changes (I think) > except that it was built with v2.02. I will try with the old `cat' as > well. I tried this with the old `cat' from txt122b.zip (as distributed on SimTel.NET), and got the same results. To recap: the problem is that configure script, at least the one from texinfo-3.12g pretest (now tested with 3.12h as well, with the same result), starts failing halfway through the script with messages like this: c:/djgpp/bin/cat.exe: -: Access denied (EACCES) After the first time this message is printed, several additional tests fail silently (e.g., the scripts says that the library doesn't have getcwd and other functions). Other tests cause the above message be printed again. I looked into generated config.status and config.log, and both of them include several runs of binary nulls. This is certainly reason grave enough for the test programs to fail, but it isn't clear to me whether the nulls are the cause of `cat's failures or they are generated *because* `cat' fails. In any case, the problem seems to happen in handling ``here documents'', like in "cat > conftest.c <