[Bimap][1.35] Regressions problems in Sun-5.9 and HP-UX
Hi, There are a some regression problems for Boost.Bimap running on SunOS and HP-UX. Detailed report: http://boost.org/regression/trunk/developer/issues.html [ sun-5.9 ] ------------------------------------------------------------------------------- assign, foreach, lambda, property_map, range, serialization test_bimap_assign, test_bimap_convenience_header, test_bimap_info test_bimap_lambda, test_bimap_modify, test_bimap_mutable test_bimap_operator_bracket, test_bimap_ordered, test_bimap_project test_bimap_property_map, test_bimap_range, test_bimap_sequenced test_bimap_unconstrained, test_bimap_unordered, typeof, xpressive [ hp_cxx-71_006_tru64 ] ------------------------------------------------------------------------------- foreach, test_bimap_info, test_bimap_ordered, test_bimap_unordered I am working blind here because I do not have access to these platforms. I wish Boost.Bimap could support them for Boost 1.35. Is there somebody willing to help me cleaning this regressions? Best regards Matias
Matias Capeletto wrote:
Hi,
There are a some regression problems for Boost.Bimap running on SunOS and HP-UX.
Detailed report: http://boost.org/regression/trunk/developer/issues.html
[...]
[ hp_cxx-71_006_tru64 ] ------------------------------------------------------------------------------- foreach, test_bimap_info, test_bimap_ordered, test_bimap_unordered
I am working blind here because I do not have access to these platforms. I wish Boost.Bimap could support them for Boost 1.35. Is there somebody willing to help me cleaning this regressions?
I can give you a hand for Tru64/CXX, what do you need? (BTW, this is not HP-UX but Tru64, formerly known as OSF1.) Markus
Matias Capeletto wrote:
There are a some regression problems for Boost.Bimap running on SunOS and HP-UX.
On HP-UX, it is only on the machine using gcc 3.4.2. From the website, it seems to me that gcc 3.4 series has a problem with bimap library. I'll see if I can spot something obvious. In the meantime, if you have access to a machine with gcc 3.4.x, any platform, you can investigate yourself. If gcc 3.4 is indeed the culprit, the tests will fail to compile. Thanks, Boris ----- Original Message ----- From: "Matias Capeletto" <matias.capeletto@gmail.com> To: <boost@lists.boost.org> Sent: Sunday, November 18, 2007 12:14 AM Subject: [boost] [Bimap][1.35] Regressions problems in Sun-5.9 and HP-UX
Hi,
There are a some regression problems for Boost.Bimap running on SunOS and HP-UX.
Detailed report: http://boost.org/regression/trunk/developer/issues.html
[ sun-5.9 ] ------------------------------------------------------------------------------- assign, foreach, lambda, property_map, range, serialization test_bimap_assign, test_bimap_convenience_header, test_bimap_info test_bimap_lambda, test_bimap_modify, test_bimap_mutable test_bimap_operator_bracket, test_bimap_ordered, test_bimap_project test_bimap_property_map, test_bimap_range, test_bimap_sequenced test_bimap_unconstrained, test_bimap_unordered, typeof, xpressive
[ hp_cxx-71_006_tru64 ] ------------------------------------------------------------------------------- foreach, test_bimap_info, test_bimap_ordered, test_bimap_unordered
I am working blind here because I do not have access to these platforms. I wish Boost.Bimap could support them for Boost 1.35. Is there somebody willing to help me cleaning this regressions?
Best regards Matias _______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
Boris Gubenko ha escrito:
Matias Capeletto wrote:
There are a some regression problems for Boost.Bimap running on SunOS and HP-UX.
On HP-UX, it is only on the machine using gcc 3.4.2. From the website, it seems to me that gcc 3.4 series has a problem with bimap library.
I'll see if I can spot something obvious. In the meantime, if you have access to a machine with gcc 3.4.x, any platform, you can investigate yourself. If gcc 3.4 is indeed the culprit, the tests will fail to compile.
Matías, Line 97 of boost/bimap/relation/support/member_with_tag.hpp, which is the one where GCC 3.4.x stumbles upon, reads: template< class Relation > struct member_with_tag < BOOST_DEDUCED_TYPENAME member_at::info, Relation, void // line 97
{ ... I think the problem is that BOOST_DEDUCED_TYPENAME shouldn't be there as member_at::info does not depend on Relation. In fact, you've got other similar specializations at the same file all of which do not have BOOST_DEDUCED_TYPENAME. HTH, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo
On Nov 19, 2007 11:06 AM, Joaquín Mª López Muñoz <joaquin@tid.es> wrote:
Matías,
Line 97 of boost/bimap/relation/support/member_with_tag.hpp, which is the one where GCC 3.4.x stumbles upon, reads:
template< class Relation > struct member_with_tag < BOOST_DEDUCED_TYPENAME member_at::info, Relation, void // line 97
{ ...
I think the problem is that BOOST_DEDUCED_TYPENAME shouldn't be there as member_at::info does not depend on Relation. In fact, you've got other similar specializations at the same file all of which do not have BOOST_DEDUCED_TYPENAME.
Thanks for spotting this Joaquin. I have removed it, lets see if this affects the regressions. Best regards Matias
I thought it may be some bad conditionalization in the Boost sources for gcc 3.4, but preprocessed file generated by gcc 3.4.2 compiles just fine with gcc 4.2.1. So, it really looks like a bug/feature of gcc 3.4 series. FWIW, Boris ----- Original Message ----- From: "Boris Gubenko" <Boris.Gubenko@hp.com> To: <boost@lists.boost.org> Cc: "Boris Gubenko" <Boris.Gubenko@hp.com> Sent: Monday, November 19, 2007 8:56 AM Subject: Re: [boost] [Bimap][1.35] Regressions problems in Sun-5.9 and HP-UX
Matias Capeletto wrote:
There are a some regression problems for Boost.Bimap running on SunOS and HP-UX.
On HP-UX, it is only on the machine using gcc 3.4.2. From the website, it seems to me that gcc 3.4 series has a problem with bimap library.
I'll see if I can spot something obvious. In the meantime, if you have access to a machine with gcc 3.4.x, any platform, you can investigate yourself. If gcc 3.4 is indeed the culprit, the tests will fail to compile.
Thanks, Boris
----- Original Message ----- From: "Matias Capeletto" <matias.capeletto@gmail.com> To: <boost@lists.boost.org> Sent: Sunday, November 18, 2007 12:14 AM Subject: [boost] [Bimap][1.35] Regressions problems in Sun-5.9 and HP-UX
Hi,
There are a some regression problems for Boost.Bimap running on SunOS and HP-UX.
Detailed report: http://boost.org/regression/trunk/developer/issues.html
[ sun-5.9 ] ------------------------------------------------------------------------------- assign, foreach, lambda, property_map, range, serialization test_bimap_assign, test_bimap_convenience_header, test_bimap_info test_bimap_lambda, test_bimap_modify, test_bimap_mutable test_bimap_operator_bracket, test_bimap_ordered, test_bimap_project test_bimap_property_map, test_bimap_range, test_bimap_sequenced test_bimap_unconstrained, test_bimap_unordered, typeof, xpressive
[ hp_cxx-71_006_tru64 ] ------------------------------------------------------------------------------- foreach, test_bimap_info, test_bimap_ordered, test_bimap_unordered
I am working blind here because I do not have access to these platforms. I wish Boost.Bimap could support them for Boost 1.35. Is there somebody willing to help me cleaning this regressions?
Best regards Matias _______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
On Nov 19, 2007 5:33 PM, Boris Gubenko <Boris.Gubenko@hp.com> wrote:
I thought it may be some bad conditionalization in the Boost sources for gcc 3.4, but preprocessed file generated by gcc 3.4.2 compiles just fine with gcc 4.2.1.
So, it really looks like a bug/feature of gcc 3.4 series.
Thanks for this test Boris. I have marked it as explicit failure. This leaves Boost.Bimap with SunOS and Tru64/CXX regressions only. Regards Matias
participants (4)
-
Boris Gubenko -
Joaquín Mª López Muñoz -
Markus Schöpflin -
Matias Capeletto