On Sun, Oct 8, 2017 at 10:34 PM, Ralf Friedl <ralf.fri...@online.de> wrote: > > Because the way the line is now is one line with 8 values. Changing that to > use ifdefs for each value would change that one line to 3*8+2=26 lines, that > makes the code much harder to read and maintain. And there may be other > places where it is used. It's perfectly legitimate to define the values as > 0, like you would define O_BINARY as 0 on platforms that don't need/supply > it.
Okay. I'm partially convinced about this argument. Consider that there are other places in busybox code that temporarily defines IUCLC to 0 for similar purpose: libbb/xfuncs.c libbb/bb_askpass.c coreutils/stty.c loginutils/getty.c > There is nothing wrong with that. It is possible to use "#if SIGUSR3" to > test for SIGUSR being defined and nonzero. It has the additional benefit > that you can use the value in a C statement (as opposed to a C preprocessor > test) and let the compiler optimize it out. Busybox does that a lot. My concern is that there're some macros that are defined to 0 for a purpose, e.g. #define O_RDONLY 0x0000 so it's not a good idea in general to arbitrarily define a macro to 0 without knowing what you are doing. But for IUCLC and such, it might be safe. James (jrtc27), I apologize for my misunderstanding of the issue. _______________________________________________ busybox mailing list busybox@busybox.net http://lists.busybox.net/mailman/listinfo/busybox