I thought the problem would be related to the timestamps of the .o and .d files for the source file in question in some way that shows up when using clang. (With gcc I don't see this problem.)
Although I don't understand how. Even if I add a ls -lT to show the full timestamp of the .o and .d files after the compilation into gb_CxxObject__command, I still see that both with gcc and clang the files have the same timestamp to the second... So I don't understand. (I even wrote a quick program to show the timestamp with sub-second resolution, but that shows just zeros for the nanoseconds...) --tml _______________________________________________ LibreOffice mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/libreoffice
