NetBSD-Users archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: gcc issues when compiling ArcticFox browser



Hi,

Leonardo Taccari wrote:
I think that's unrelated.  USE_PKGSRC_GCC_RUNTIME is used as a kludge
to depends on corresponding gcc-libs package (USE_GCC_RUNTIME should be
used).

I am still stuck with this and looked at it a bit further.

The error is:
76:42.39 In file included from /usr/pkg/gcc6/include/c++/math.h:36:0,
76:42.39                  from /home/multix/code/Arctic-Fox/intl/icu/source/common/putil.cpp:63: 76:42.39 /usr/pkg/gcc6/include/c++/cmath:1101:11: error: '::double_t' has not been declared
76:42.39    using ::double_t;
76:42.39            ^~~~~~~~


and the code in putil.cpp around that line is:

// Must be before any other #includes.
#include "uposixdefs.h"

/* include ICU headers */
#include "unicode/utypes.h"
#include "unicode/putil.h"
#include "unicode/ustring.h"
#include "putilimp.h"
#include "uassert.h"
#include "umutex.h"
#include "cmemory.h"
#include "cstring.h"
#include "locmap.h"
#include "ucln_cmn.h"
#include "charstr.h"

/* Include standard headers. */
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <math.h>

There is no code, just include and the compiler barfs at math.h - so it looks to me that somehow GCC's 6.5 headers are inconsistent and I think this is a pkg issue, but can't really prove that. This is an older version of ICU, so nto even real FF code. If gcc 5.5 could compile it and gcc 7 too, I am even more convinced that it is a system vs. pkg compiler issue. WHy, I don't know.

Are any tricks needed when using the pkg compiler?

I just used:
export CC="/usr/pkg/gcc6/bin/gcc"
export CXX="/usr/pkg/gcc6/bin/g++"


no other tricks or settings.

Riccardo

Riccardo


Home | Main Index | Thread Index | Old Index