cwitty@PORTIA.STANFORD.EDU (Carl Witty) (05/24/89)
GCC version 1.35, from Labrea on May 21 Compiled with "config.gcc next"; with no changes to the source. NeXT, release 0.9. GCC generates an invalid assembler pseudo-op when attempting to output a float constant. NeXT> cat foo.c float x = 1.0; NeXT> gcc -v -c foo.c gcc version 1.35 /Net/odysseus/Users/students/cwitty/OpticalDisk/lib/gcc-cpp -v -undef -D__GNUC__ -Dmc68000 -DNeXT -Dunix -D__MACH__ -D__mc68000__ -D__NeXT__ -D__unix__ -D____MACH____ -D__HAVE_68881__ -Dmc68020 foo.c /tmp/cc002761.cpp GNU CPP version 1.35 /Net/odysseus/Users/students/cwitty/OpticalDisk/lib/gcc-cc1 /tmp/cc002761.cpp -quiet -dumpbase foo.c -version -o /tmp/cc002761.s GNU C version 1.35 (68k, MIT syntax) compiled by GNU C version 1.35. as -mc68020 -o foo.o /tmp/cc002761.s /tmp/cc002761.s:7:Unknown pseudo-op /tmp/cc002761.s:7:Rest of line ignored. 1st junk character valued 48. NeXT> gcc -S foo.c NeXT> cat foo.s #NO_APP gcc_compiled.: .globl _x .data .even _x: .single 0r1.00000000000000000000e+00 NeXT> One question: In gen*.c, why isn't the first argument s of the function fatal declared as pointer to character? The NeXT include files declare fprintf as extern int fprintf(FILE *stream, const char *format, ...); so there is a warning generated for each use of fatal with an int as the second argument. Carl Witty cwitty@portia.Stanford.EDU