Re: diff from 2019-10-04 13:33 to 2019-10-05 13:54

[Roland Illig <roland.illig%gmx.de@localhost>]

Am 06.10.2019 um 13:21 schrieb Makoto Fujiwara:
> reporting URL:
>     http://www.ki.nu/pkgsrc/reports/current/NetBSD-9.0_BETA/20191005.1354/meta/report.html

The protobuf failure looks funny in the preprocessed source:

  generator->PrintString(!std::
# 1623 "google/protobuf/text_format.cc" 3 4
                              ((sizeof (
# 1623 "google/protobuf/text_format.cc"
                              val
# 1623 "google/protobuf/text_format.cc" 3 4
                              ) == sizeof (float)) ? __isnanf (
# 1623 "google/protobuf/text_format.cc"
                              val
# 1623 "google/protobuf/text_format.cc" 3 4
                              ) : (sizeof (
# 1623 "google/protobuf/text_format.cc"
                              val
# 1623 "google/protobuf/text_format.cc" 3 4
                              ) == sizeof (double)) ? __isnand (
# 1623 "google/protobuf/text_format.cc"
                              val
# 1623 "google/protobuf/text_format.cc" 3 4
                              ) : __isnanl (
# 1623 "google/protobuf/text_format.cc"
                              val
# 1623 "google/protobuf/text_format.cc" 3 4
                              )).
# 1623 "google/protobuf/text_format.cc"
                                         ? SimpleFtoa(val) : "nan");


Any idea why isnan is a macro in C++?