Hi Tony, hallo Marc I've attached a patch for the interpreter that should fix this issue for gcc-compiled interpreter on all platforms. It works for me on mips-irix and i386-linux interpreters.
* It removes the *_MODULO_BROKEN and *_DIVISION_BROKEN test from the configure script, since it's pointless to test for undefined behaviour. * It changes long_rem and int_rem definitions to explicitely test for -1 as the dividend, and return 0 in that case, like the spec demands. * It also fixes int_div and long_div to check for MIN_VALUE / -1 and return MIN_VALUE in that case, according to the spec. * It adds JINT_MIN and JLONG_MIN to jtypes.h.in. could you give it a try on m68k and powerpc? cheers, dalibor topic __________________________________ Do you Yahoo!? SBC Yahoo! DSL - Now only $29.95 per month! http://sbc.yahoo.com
division.diff
Description: division.diff
