Thanks for the report, Bob.
[Bob reported (included below) that the largefile test from autoconf-2.49c
doesn't work on HP-UX 10.20. ]
This test (note the ULL suffixes) is compiled without error by /bin/cc
on an HPUX10.20:
#define _FILE_OFFSET_BITS 64
#include
int a[(off_t) 92233720368547
| On Jan 31, 2001, Akim Demaille <[EMAIL PROTECTED]> wrote:
| >> "Markus" == Markus F X J Oberhumer <[EMAIL PROTECTED]>
|writes:
| Markus> After a look through the sources it seems that AS_EXECUTABLE_P
| Markus> should use both "test -f" and "test -x" on systems that
| Markus> support it, an
> "Tim" == Tim Van Holder <[EMAIL PROTECTED]> writes:
Tim> Yes and no. The reason we try to use 'test -x' is so that 'test
Tim> -x foo' will pick up foo.exe. I had not thought of this finding
Tim> directories (then again, I don't have . in my path). I agree this
Tim> is a serious problem; eve
> "Tom" == Tom Tromey <[EMAIL PROTECTED]> writes:
> "Akim" == Akim Demaille <[EMAIL PROTECTED]> writes:
Akim> Can there be any non source files which we want to ship?
Tom> Yes, sometimes. For instance with ansi2knr support we ship the
Tom> man page. I don't know whether this impacts au
Hello.
Akim Demaille wrote:
> Still, `-f' is definitely better where we can use it. The question
> is, `is test -x my-dir positive on DOS environments'? If so, then on
> this arg the second `test' can be just `:'.
test -x path returns 0 (true) for executables and directories. I tested
with DJG
Earnie:
Thanks for the excellent suggestion. However, it still would be best
if the aclocal Perl program were updated to use "$0" to calculate where
the "share" directory is located ("$0" should contain the path to the
bin directory and the share directory is just ../share from there).
The rea
Brian Cameron <[EMAIL PROTECTED]> wrote:
| Thanks for the excellent suggestion. However, it still would be best
| if the aclocal Perl program were updated to use "$0" to calculate where
| the "share" directory is located ("$0" should contain the path to the
| bin directory and the share directory
Jim:
> No. IMHO it should not do that.
> If it does that, then the binary will work differently depending
> on where it resides when you run it. That will cause grief for some
> installations. For example, if I currently configure each new version with
> --prefix=/whatever/autoconf-$(date +%Y
Akim Demaille wrote:
>
> > "Tim" == Tim Van Holder <[EMAIL PROTECTED]> writes:
>
> Tim> Yes and no. The reason we try to use 'test -x' is so that 'test
> Tim> -x foo' will pick up foo.exe. I had not thought of this finding
> Tim> directories (then again, I don't have . in my path). I agree t
Brian Cameron <[EMAIL PROTECTED]> wrote:
...
| > Why not just use aclocal's `-I include_dir' option?
|
| Since calls to automake are often buried, it would be time consuming to
| find & fix all places where this issue is a problem. We are using automake
| to build the 40 or so packages that make
The subject says it all. I consider it a bug.
Is this easy to fix?
Lars J
Jim:
> | > Why not just use aclocal's `-I include_dir' option?
> |
> | Since calls to automake are often buried, it would be time consuming to
> | find & fix all places where this issue is a problem. We are using automake
> | to build the 40 or so packages that make up Gnome, so finding each pl
"Lars J. Aas" <[EMAIL PROTECTED]> writes:
> The subject says it all. I consider it a bug.
> Is this easy to fix?
Nope :( Quite hard. The easiest would be backup/restore :(
"Lars J. Aas" <[EMAIL PROTECTED]> writes:
> The subject says it all. I consider it a bug.
> Is this easy to fix?
Hm, on a second thought, now that AS_MESSAGE tests whether the LOG fd
is set to output the message or not, it might be just as easy as
moving the set up of the LOG fd :)
# AS_MESSA
Earnie Boyd <[EMAIL PROTECTED]> writes:
> There is a caveat to "would work". If I have executable foo.exe and
> directory foo then both tests returns false. Given this, I don't see
> anyway out of `((test -x foo.exe && test -f foo.exe) || (test -x foo &&
> test -f foo))'.
Hm, I have some hesit
On Fri, Feb 02, 2001 at 07:07:57PM +0100, Akim Demaille wrote:
: "Lars J. Aas" <[EMAIL PROTECTED]> writes:
: > The subject says it all. I consider it a bug.
: > Is this easy to fix?
:
: Hm, on a second thought, now that AS_MESSAGE tests whether the LOG fd
: is set to output the message or not, i
> -Original Message-
> From: Akim Demaille [mailto:[EMAIL PROTECTED]]
> Sent: Friday, February 02, 2001 7:09 PM
> To: Earnie Boyd
> Cc: Tim Van Holder; [EMAIL PROTECTED];
> Alexandre Oliva
> Subject: Re: autoconf 2.49c fails if '.' is in PATH
>
>
> Earnie Boyd <[EMAIL PROTECTED]> writes
> > This would change the default behavior, but since the output of
> > autoreconf is intended mostly for humans, it shouldn't be an issue.
>
> Sounds pretty much the same :) That an OK in advance.
Only after fixing the testsuite. I didn't fix the infinite loop in
_AC_COMPUTE_INT_COMPILE on HP-U
18 matches
Mail list logo
