Your message dated Mon, 02 Aug 2010 13:15:39 -0700
with message-id <[email protected]>
and subject line Bug #539761: undefined ac_fn_cxx_try_run
has caused the Debian Bug report #539761,
regarding undefined ac_fn_cxx_try_run
to be marked as done.
This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.
(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact [email protected]
immediately.)
--
539761: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=539761
Debian Bug Tracking System
Contact [email protected] with problems
--- Begin Message ---
Package: autoconf
Version: 2.64-1
After upgrading from autoconf 2.63 to 2.64 our configure.in script is
starting to fail due to the ac_fn_cxx_try_run helper function being
undefined on first use.
I tracked it down to a works or fails case depending on whether the
first instance of AC_RUN_IFELSE is wrapped in AC_CACHE_CHECK
WORKING:
AC_RUN_IFELSE([AC_LANG_SOURCE([[int main(){return 0;}]])],[],[],[])
AC_CACHE_CHECK(if working, foo_works,
AC_RUN_IFELSE([AC_LANG_SOURCE([[int main(){return
1;}]])],[foo_works=yes],[foo_works=no],[])
)
FAILING:
AC_CACHE_CHECK(if working, foo_works,
AC_RUN_IFELSE([AC_LANG_SOURCE([[int main(){return
1;}]])],[foo_works=yes],[foo_works=no],[])
)
AC_RUN_IFELSE([AC_LANG_SOURCE([[int main(){return 0;}]])],[],[],[])
In the working case the first instance of AC_RUN_IFELSE seems to trigger
a global very early setup definition to be added.
The second failing case, seems to only add the definition right inside
the if-else-fi statement created by AC_CACHE_CHECK immediately before
that usage. This definition is not available to other code outside that
scope which needs it later.
AYJ
--- End Message ---
--- Begin Message ---
I'm looking through old bugs before uploading Autoconf 2.67.
Amos Jeffries <[email protected]> reported that a configure
script failed due to the ac_fn_cxx_try_run helper function being
undefined on first use.
In a reply, Yavor Doganov <[email protected]> pointed out that using
proper m4 quoting fixes the problem.
This seems to me a proper and complete response (unless Amos or
someone else can point out why the quoting should not be
necessary), so I am closing this bug (with this email).
--
Ben Pfaff
http://benpfaff.org
--- End Message ---
