C compilers may fold const values at compile time, so expressions
which try to elicit underflow/overflow by performing simple
arithemetic on suitable values will not generate the required
exceptions.
Work around this by replacing code which does these arithmetic
operations with calls to the existing __math_xflow functions that are
designed to do this correctly.
Signed-off-by: Keith Packard <keithp@keithp.com>
----
v2:
libm/math: Pass sign to __math_xflow instead of muliplying result
* libc/stdlib/mprec.c (ulp, b2d, d2b): changed a few expressions
like x << y-z to the equivalent x << (y-z).
(d2b): changed if statements with assignment to perform the
assignment prior to the if check.
* libc/reent/reent.c: included stdlib.h for "_free_r" prototype.
* libc/unix/getpass.c (getpass): moved "echo" assignment out of if.
* libc/unix/ttyname.c: included string.h for "strcpy" prototype.
* libc/unix/getcwd.c (ISDOT): added parentheses to clarify && and ||
precedence.
* libc/include/sys/unistd.h: added "vfork" prototype (for popen.c).
Added "_execve" prototype (for execl.c, execle.c, execv.c, and
execve.c).
* libc/posix/popen.c (popen): added parentheses to clarify && and ||
precedence.
* libm/math/e_cosh.c (__ieee754_cosh): changed parentheses to
clarify && and || precendence (and to remove pascalism).
* libm/math/e_sinh.c (__ieee754_sinh): Ditto.
* libm/math/s_infconst.c: added another pair of braces to all
initializers for __infinity (need three: for __infinity[1] array,
for union __dmath, and for i[2]).
Keith Packard
Jeff Johnston
Christopher Faylor