- Add check in roundx() functions for infinite or NaN cases
- Add block to avoid style warnings
- Define long double constants and macros for infinity and nan
- Correct return syntax to match NuttX style
- Make c89 compliant
- Fix definitions of INFINITY_L/NAN_L
* include/nuttx/lib/math.h - match standard naming conventions
- Rename isinf_l to isinfl
- Rename isinf_f to isinff
- Add finite()
- Add finitel()
- Add finitef()
- Define isnanl and isnanf
- Define isfinite() so that it uses the appropriate macro
* libs/libc/math/lib_asinf.c
libs/libc/math/lib_asinl.c
libs/libc/math/lib_roundf.c
libs/libc/math/lib_roundl.c
libs/libc/math/lib_sqrtf.c
libs/libc/math/lib_sqrtl.c
- Use renamed macros
- Use correct NAN_x or INFINIT_x macro on returns
Probably this is a bug of a GCC, but on AMRv7 the code "if (relax_factor > 1)"
generates "bne.n" instruction if "relax_factor" is "int". Few lines above
"relax_factor *= LOG_RELAX_MULTIPLIER;" is done without overflow check hence
at some moment overflow occurs and "relax_factor" becomes a zero and condition
"if (relax_factor > 1)" becomes always evaluated to "true" hence "epsilon"
becomes zero always.
Probably this is not the best way to fix the bug (The best way is to report it
to GCC), but this change allows to get correct behavior of "log" and "logf" for
ARMv7 based MCUs
Signed-off-by: Petro Karashchenko <petro.karashchenko@gmail.com>
Fix a problem where the log function would loop infinitely by
calculation error when a specific value was passed. For example,
in the following case, the log function loops infinitely and
never returns. So, this commit fixes to return the right value.
double a = 7883961.5;
double b = log(a);
Squashed commit of the following:
libs/libxx: Fix some confusing in naming. If the directory is called libxx, then the library must be libxx.a (unless perhaps LIBCXX is selected).
libs/: Fix paths in moved library directories.
libs: Brute force move of libc, libnx, and libxx to libs. Cannot yet build it in that configuration.
