Bug#468938: AC_TRY_RUN generates invalid shell code
"Artur R. Czechowski" <[EMAIL PROTECTED]> writes: > I kindly ask to fix the bug ASAP, it makes me impossible (well, hard) to fix > http://bugs.debian.org/456686, which is serious. I have now read through this bug report a couple of times and I still don't see how it relates to Autoconf. Can you explain? -- "Let others praise ancient times; I am glad I was born in these." --Ovid (43 BC-18 AD) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#468938: AC_TRY_RUN generates invalid shell code
* Artur R. Czechowski wrote on Sun, Mar 02, 2008 at 02:47:13PM CET:
>
> AC_TRY_RUN([#include
> int main()
> {
> static struct random_data rand_data;
> static char rand_state[256];
> initstate_r(0, rand_state, sizeof(rand_state), &rand_data);
> exit(0);
> }], initstate_r_works=yes,, ])
Please drop the two consecutive commas and the trailing space, then
retry.
> checking for usability of initstate_r... ./configure: line 2573: syntax error
> near unexpected token `else'
> ./configure: line 2573: `else'
>
> Around the line 2573 there is code:
>
> 2571 if test "$cross_compiling" = yes; then
> 2572
> 2573 else
Cheers,
Ralf
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#468938: AC_TRY_RUN generates invalid shell code
* Ralf Wildenhues wrote on Sun, Mar 02, 2008 at 03:06:14PM CET:
> * Artur R. Czechowski wrote on Sun, Mar 02, 2008 at 02:47:13PM CET:
> >
> > AC_TRY_RUN([#include
> > int main()
> > {
> > static struct random_data rand_data;
> > static char rand_state[256];
> > initstate_r(0, rand_state, sizeof(rand_state), &rand_data);
> > exit(0);
> > }], initstate_r_works=yes,, ])
>
> Please drop the two consecutive commas and the trailing space, then
> retry.
Ah, now I understand why they are there. You are trying to fool
autoconf so that the configure script does not cause an error when cross
compiling. Well, don't do that. Do something like
AC_TRY_RUN([#include
int main()
{
static struct random_data rand_data;
static char rand_state[256];
initstate_r(0, rand_state, sizeof(rand_state), &rand_data);
exit(0);
}], [initstate_r_works=yes], [initstate_r_works=no], [initstate_r_works=no])
in order to be on the safe side. In the cross-compiling arg, you could
also test for $host or $host_os bits (in which case you have to ensure
that AC_CANONICAL_HOST has been invoked earlier) and set things for
systems you know about for sure.
BTW, AC_TRY_RUN has been obsoleted, autoupdate can typically update your
configure.ac script to use AC_RUN_IFELSE.
Cheers,
Ralf
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#468938: AC_TRY_RUN generates invalid shell code
Package: autoconf
Version: 2.61-6
Severity: important
Hello,
Please consider a minimal configure.ac containing:
8<-[cut here]-8<---8<-[cut here]-8<
AC_INIT([IMMS],[3.0.2],[EMAIL PROTECTED])
AC_PREREQ([2.52g])
AC_REVISION($Id: configure.ac,v 1.30 2003/10/23 23:53:20 mag Exp $)
AC_PROG_CC
AC_MSG_CHECKING([for usability of initstate_r])
AC_TRY_RUN([#include
int main()
{
static struct random_data rand_data;
static char rand_state[256];
initstate_r(0, rand_state, sizeof(rand_state), &rand_data);
exit(0);
}], initstate_r_works=yes,, ])
8<-[cut here]-8<---8<-[cut here]-8<
then run autoconf on it and then run ./configure. There result is:
[EMAIL PROTECTED]:/tmp/c$ ./configure
checking for gcc... gcc
checking for C compiler default output file name... a.out
checking whether the C compiler works... yes
checking whether we are cross compiling... no
checking for suffix of executables...
checking for suffix of object files... o
checking whether we are using the GNU C compiler... yes
checking whether gcc accepts -g... yes
checking for gcc option to accept ISO C89... none needed
checking for usability of initstate_r... ./configure: line 2573: syntax error
near unexpected token `else'
./configure: line 2573: `else'
Around the line 2573 there is code:
2571 if test "$cross_compiling" = yes; then
2572
2573 else
2574 { { echo "$as_me:$LINENO: error: cannot run test program while cross
comp iling
2575 See \`config.log' for more details." >&5
2576 echo "$as_me: error: cannot run test program while cross compiling
2577 See \`config.log' for more details." >&2;}
2578{ (exit 1); exit 1; }; }
2579 fi
2580 ]])
which is invalid. Please consider following shell script:
8<-[cut here]-8<---8<-[cut here]-8<
#!/bin/sh
if test "$1" = "foo"; then
else
echo There is no foo
fi
8<-[cut here]-8<---8<-[cut here]-8<
It always run with an error:
./foo: line 4: syntax error near unexpected token `else'
./foo: line 4: `else'
/bin/sh is a symlink to /bin/bash on my box.
Replacing the shebang line with #!/bin/dash doesn't help too:
./foo: 4: Syntax error: "else" unexpected
Valid script is:
8<-[cut here]-8<---8<-[cut here]-8<
#!/bin/sh
if test "$1" = "foo"; then
:;
else
echo There is no foo
fi
8<-[cut here]-8<---8<-[cut here]-8<
It runs correctly with bash and with dash.
My bash and dash versions:
||/ Name VersionDescription
+++-==-==-
ii bash 3.1dfsg-9 The GNU Bourne Again SHell
ii dash 0.5.4-7POSIX-compliant shell
I kindly ask to fix the bug ASAP, it makes me impossible (well, hard) to fix
http://bugs.debian.org/456686, which is serious.
Best regards
Artur
-- System Information:
Debian Release: lenny/sid
APT prefers unstable
APT policy: (500, 'unstable')
Architecture: amd64 (x86_64)
Kernel: Linux 2.6.24-1-amd64 (SMP w/1 CPU core)
Locale: LANG=C, LC_CTYPE=pl_PL (charmap=ISO-8859-2)
Shell: /bin/sh linked to /bin/bash
Versions of packages autoconf depends on:
ii debianutils 2.28.4 Miscellaneous utilities specific t
ii m41.4.10-1 a macro processing language
ii perl 5.8.8-12 Larry Wall's Practical Extraction
Versions of packages autoconf recommends:
ii automake [automaken] 1:1.10.1-3 A tool for generating GNU Standard
-- no debconf information
--
windows jest jak Odie - głupi jak but, cały czas się uśmiecha, a linux jak
Garfield - może i by coś zrobił, ale trzeba go najpierw do tego zmusić.
/yacoob/
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
