On Wed, 14 Dec 2016 17:19:27 -0800
Paul Eggert wrote:
> I was referring to code with his proposed patch installed. 'delete' is
> O(N); 'replace' calls 'delete' in a loop and is therefore O(N**2).
> 'epsclosure' calls 'replace' in a loop and so I suppose it is O(N**3).
> I haven't looked into how
Pedro Alves wrote:
> m4/inet_pton.m4 has this:
>
> if test $ac_cv_have_decl_inet_pton = yes; then
> dnl It needs to be overridden, because the stdcall calling convention
> dnl is not compliant with POSIX.
> REPLACE_INET_PTON=1
> INET_PTON_LIB="-lws2_32"
> else
>
Gisle Vanem wrote:
> > Can't see how that can run fine? The compiler will set up the call
> > assuming cdecl convention, while the called function has stdcall
> > convention.
>
> I would expect the 'reinterpret_cast(::getaddrinfo)' to fix that.
But this will cause undefined behaviour. reinterpre
With MSVC, the implementation of the 'getlogin' and 'getlogin_r' functions
produces a link error:
/home/bruno/msvc/compile cl -nologo -DHAVE_CONFIG_H -DEXEEXT=\".exe\" -I.
-I../../gltests -I.. -DGNULIB_STRICT_CHECKING=1 -DIN_GNULIB_TESTS=1 -I.
-I../../gltests -I.. -I../../gltests/.. -I../gllib
Followup to this discussion from October 2016:
Jim:
> > I did not mean to imply by that message that we should eliminate every
> > use of the program_name module. My desire is more to avoid accidental
> > use of it when the getprogname module would be more appropriate.
Bruno:
> Fully agree on thi
Compiling a gnulib testdir with 32-bit MSVC 14, I get these errors:
C:\cygwin64\home\bruno\testdir-posix\gllib\acosf.c(24): error C2084: function
'float acosf(float)' already has a body
C:\cygwin64\home\bruno\testdir-posix\gllib\asinf.c(24): error C2084: function
'float asinf(float)' already has
Compiling a gnulib testdir with MSVC 14, I get these errors:
C:\cygwin64\home\bruno\testdir-posix\gllib\snprintf.c(38): error C2084:
function 'int snprintf(char *const ,const std::size_t,const char *const ,...)'
already has a body
C:\cygwin64\home\bruno\testdir-posix\gllib\vsnprintf.c(40): error
Compiling a gnulib testdir with MSVC 14, I get these errors:
/home/bruno/msvc/compile cl -nologo -DHAVE_CONFIG_H -DEXEEXT=\".exe\" -I.
-I../../gltests -I.. -DGNULIB_STRICT_CHECKING=1 -DIN_GNULIB_TESTS=1 -I.
-I../../gltests -I.. -I../../gltests/.. -I../gllib -I../../gltests/../gllib
-D_WIN32_WI
In an MSVC 14 build, test-stdint fails to compile:
test-stdint.c(312): error C2034: '_gl_verify_error_if_negative': type of bit
field too small for number of bits
The reason is that gnulib's (or it could also be is still in effect. This fixes it.
2016-12-17 Bruno Haible
stdint: Fi
On Fri, 2016 Dec 16 18:24-0800, Paul Eggert wrote:
>
> Thanks. We should address this underyling problem. I installed the
> attached to try to work around the problem. I don't want to have to
> modify lots of programs to include seemingly-irrelevant headers like
> stdint.h.
Good call, and I'm happ
$ ./gnulib-tool --create-testdir --with-tests --single-configure hypotl
...
configure.ac:278: warning: AC_REQUIRE: `gl_FUNC_FREXPL' was expanded before it
was required
configure.ac:278:
http://www.gnu.org/software/autoconf/manual/autoconf.html#Expanded-Before-Required
glm4/sqrtl.m4:7: gl_FUNC_SQR
On 17/12/16 11:08, Bruno Haible wrote:
> I'm seeing this compilation failure on a recent mingw:
looks good to me.
thanks
Eli Zaretskii wrote:
> You can now remove the workaround from stdio-impl.h.
Done.
Thanks for the rapid debugging. This is the kind of bug that you typically
cannot reproduce any more after two weeks, if you don't handle it quickly.
Bruno
> Date: Sat, 17 Dec 2016 09:51:15 +0200
> From: Eli Zaretskii
> Cc: bug-gnulib@gnu.org, emacs-de...@gnu.org
>
> I will try to look into this some more today, and see if I can unlock
> the mystery. It bothers me, too.
Bug squashed, see commit 0757b4f. It was always there, we were just
lucky unt
I'm seeing this compilation failure on a recent mingw:
../../gltests/test-getlogin_r.c:31:17: fatal error: pwd.h: No such file or
directory
compilation terminated.
make[4]: *** [Makefile:7987: test-getlogin_r.o] Error 1
This fixes it:
2016-12-17 Bruno Haible
getlogin_r tests: Port
I'm seeing this compilation failure on a recent mingw:
In file included from ../../gltests/test-getlogin.c:23:0:
../../gltests/test-getlogin.c:24:18: error: ‘getlogin’ undeclared here (not in
a function)
SIGNATURE_CHECK (getlogin, char *, (void));
^
../../gltests/signature.h:46
Pedro Alves wrote:
> Can't see how that can run fine? The compiler will set up the call
> assuming cdecl convention, while the called function has stdcall
> convention.
I would expect the 'reinterpret_cast(::getaddrinfo)' to fix that.
With this little example:
#include
#include
int main
17 matches
Mail list logo
