[test][cw] problems with __declspec() in floating_point_comparison.hpp
CW 9.4 fails to compile floating_point_comparison.hpp (see http://tinyurl.com/ybzaze ) because of line 36: enum BOOST_TEST_DECL floating_point_comparison_type { where BOOST_TEST_DECL resolves to __declspec(dllimport) or __declspec(dllexport) depending on the context. Seemingly, this compiler does not support __declspec applied to enum types --usage with classes and functions is supported, though. The obvious workaround is to replace the offending line with #if BOOST_WORKAROUND(__MWERKS__,BOOST_TESTED_AT(0x3205)) enum floating_point_comparison_type { #else enum BOOST_TEST_DECL floating_point_comparison_type { #endif Is it OK for me to commit this change to the HEAD and RC_1_34_0? Thank you, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo
Hi Jaoquin ! An'n Dunnersdag 30 November 2006, 14:44 hett Joaquín Mª López Muñoz schreven:
CW 9.4 fails to compile floating_point_comparison.hpp (see http://tinyurl.com/ybzaze ) because of line 36:
enum BOOST_TEST_DECL floating_point_comparison_type {
where BOOST_TEST_DECL resolves to __declspec(dllimport) or __declspec(dllexport) depending on the context. Seemingly, this compiler does not support __declspec applied to enum types --usage with classes and functions is supported, though.
The obvious workaround is to replace the offending line with
#if BOOST_WORKAROUND(__MWERKS__,BOOST_TESTED_AT(0x3205)) enum floating_point_comparison_type { #else enum BOOST_TEST_DECL floating_point_comparison_type { #endif
Is it OK for me to commit this change to the HEAD and RC_1_34_0?
I've just checked with msvc-8.0 and simply removed the BOOST_TEST_DECL. So please just remove it _completely_ as it is not needed. And gcc will complain about it when visibility-support is enabled, too. Yours, Jürgen -- * Dipl.-Math. Jürgen Hunold ! Ingenieurgesellschaft für * voice: ++49 511 262926 57 ! Verkehrs- und Eisenbahnwesen mbH * fax : ++49 511 262926 99 ! Lister Straße 15 * juergen.hunold@ivembh.de ! www.ivembh.de
"Jürgen Hunold" <hunold@ivembh.de> wrote in message news:200611301515.29992.hunold@ivembh.de... Hi Jaoquin ! An'n Dunnersdag 30 November 2006, 14:44 hett Joaquín Mª López Muñoz schreven:
CW 9.4 fails to compile floating_point_comparison.hpp (see ...
#if BOOST_WORKAROUND(__MWERKS__,BOOST_TESTED_AT(0x3205)) enum floating_point_comparison_type { #else enum BOOST_TEST_DECL floating_point_comparison_type { #endif
Is it OK for me to commit this change to the HEAD and RC_1_34_0?
I've just checked with msvc-8.0 and simply removed the BOOST_TEST_DECL. So please just remove it _completely_ as it is not needed. And gcc will complain about it when visibility-support is enabled, too.
Go ahead and commit. Gennadiy
Gennadiy Rozental ha escrito:
"Jürgen Hunold" <hunold@ivembh.de> wrote in message news:200611301515.29992.hunold@ivembh.de... Hi Jaoquin !
An'n Dunnersdag 30 November 2006, 14:44 hett Joaquín Mª López Muñoz schreven:
CW 9.4 fails to compile floating_point_comparison.hpp (see ...
#if BOOST_WORKAROUND(__MWERKS__,BOOST_TESTED_AT(0x3205)) enum floating_point_comparison_type { #else enum BOOST_TEST_DECL floating_point_comparison_type { #endif
Is it OK for me to commit this change to the HEAD and RC_1_34_0?
I've just checked with msvc-8.0 and simply removed the BOOST_TEST_DECL. So please just remove it _completely_ as it is not needed. And gcc will complain about it when visibility-support is enabled, too.
Go ahead and commit.
Gennadiy
Hello Gennadiy, Commit what? The workaround on __MWERKS__ or total elmination of BOOST_TEST_DECL in this declaration as Jürgen proposes? Joaquín M López Muñoz Telefónica, Investigación y Desarrollo
Hi Joaquin ! An'n Dunnersdag 30 November 2006, 15:37 hett Joaquín Mª López Muñoz schreven:
Gennadiy Rozental ha escrito:
#if BOOST_WORKAROUND(__MWERKS__,BOOST_TESTED_AT(0x3205)) enum floating_point_comparison_type { #else enum BOOST_TEST_DECL floating_point_comparison_type { #endif
Is it OK for me to commit this change to the HEAD and RC_1_34_0?
I've just checked with msvc-8.0 and simply removed the BOOST_TEST_DECL. So please just remove it _completely_ as it is not needed. And gcc will complain about it when visibility-support is enabled, too.
Go ahead and commit.
Gennadiy
Commit what? The workaround on __MWERKS__ or total elmination of BOOST_TEST_DECL in this declaration as Jürgen proposes?
Well, I've just blown the macro away :-)) My msvc just does not need it (and we never export enums in our own sources, too...) Let's wait for the regression runners to pick it up... Yours, Jürgen -- * Dipl.-Math. Jürgen Hunold ! Ingenieurgesellschaft für * voice: ++49 511 262926 57 ! Verkehrs- und Eisenbahnwesen mbH * fax : ++49 511 262926 99 ! Lister Straße 15 * juergen.hunold@ivembh.de ! www.ivembh.de
participants (3)
-
Gennadiy Rozental -
Joaquín Mª López Muñoz -
Jürgen Hunold