lexical_cast issues with older gcc's, vc 7.1
I think this patch RCS file: /cvsroot/boost/boost/boost/lexical_cast.hpp,v ... revision 1.30 date: 2006/10/25 06:43:44; author: alnsn; state: Exp; lines: +276 -59 Optimization for floating types (1.25 faster on gcc / FreeBSD 6.2). VC6 test is still broken. Causes all these new failures (look for "error" in each of the build logs): http://cci.lbl.gov/misc_build/results/2006_10_27_1935/mac_os_10_3_build_log http://cci.lbl.gov/misc_build/results/2006_10_27_1935/redhat80_py25_build_lo... http://cci.lbl.gov/misc_build/results/2006_10_27_1935/redhat90_py25_build_lo... http://cci.lbl.gov/misc_build/results/2006_10_27_1935/redhat_ws3_py25_build_... http://cci.lbl.gov/misc_build/results/2006_10_27_1935/win_xp_vc71_py25_build... If this is not enough information to fix the bug, please let me know how I can help. BTW: It works on a number of other platforms: http://cci.lbl.gov/misc_build/show_error_summary.cgi?build_tag=2006_10_27_19... Thanks! Cheers, Ralf
Ralf W. Grosse-Kunstleve wrote:
All errors except vc71 are boil down to this error: std::basic_streambuf<char, std::char_traits<char> > is a inaccessible base of boost::detail::lexical_streambuf<char>" I fixed this in revision 1.31. Let me know if it fixes your tests. -- Alexander Nasonov Project Manager at Akmosoft ( http://www.akmosoft.com ) Blog: http://nasonov.blogspot.com Email: $(FirstName) dot $(LastName) at gmail dot com
Ralf W. Grosse-Kunstleve wrote:
http://cci.lbl.gov/misc_build/results/2006_10_27_1935/win_xp_vc71_py25_build...
It seems that you run the test with "-sBUILD=<native-wchar_t>off". I didn't know about this option. It should be fixed now. At least lexical_cast test runs fine with <native-wchar_t>off. -- Alexander Nasonov Project Manager at Akmosoft ( http://www.akmosoft.com ) Blog: http://nasonov.blogspot.com Email: $(FirstName) dot $(LastName) at gmail dot com
participants (2)
-
Alexander Nasonov -
Ralf W. Grosse-Kunstleve