C applications

Note: You need to rebuild your binaries. You can't run binaries from an earlier version of the OS on QNX Neutrino 7.

We've discontinued some QNX 4 functions:

Instead of using:	Use:
atoh()	strtoul()
chsize()	ftruncate()
_cmdfd()	No direct replacement
eof()	Check the result of read()
fgetchar()	fgetc(stdin)
flushall()	fflush(NULL)
fputchar()	fputc(c, stdout)
getprio()	pthread_getschedparam() or SchedGet()
input_line()	No direct replacement
setprio()	pthread_setschedparam() or SchedSet()
strcmpi(), stricmp()	strcasecmp()
strnicmp()	strncasecmp()
strnset(), strset()	memset()
tell(), tell64()	lseek(fd, 0, SEEK_CUR)

We've replaced the Dinkumware math library; the new version of libm is based on FreeBSD's, with a few functions from NetBSD's. The .so number has changed from 2 to 3. The differences include:

We've discontinued the following functions:

Instead of using:	Use:
clog10(), clog10f(), clog10l()	No direct replacement
fegettrapenable(), fesettrapenable(), fp_exception_mask()	fedisableexcept(), feenableexcept()
fp_exception_value()	fetestexcept()
fp_precision()	fegetprec(), fesetprec()
fp_rounding()	fegetround(), fesetround()
norm(), normf(), norml()	No direct replacement

The old math library returned errors via errno and fp exceptions (math_errhandling == MATH_ERRNO | MATH_ERREXCEPT). The new one returns errors only via fp exceptions (math_errhandling == MATH_ERREXCEPT).
The headers <ieeefp.h> and <fpstatus.h> have been removed. You can use <fenv.h> as a more portable replacement. This header file declares the following functions, most of which are defined by C99 and POSIX, except as noted below:

feclearexcept()

Clear the specified floating-point status flags

fedisableexcept(), feenableexcept()

Mask and unmask the specified floating-point exceptions (BSD, QNX Neutrino)

fegetenv(), fesetenv()

Get and set the current floating-point environment

fegetexceptflag(), fesetexceptflag()

Get and set the states of the specified floating-point status flags

fegetprec(), fesetprec()

Get and set the current floating-point precision (QNX Neutrino)

fegetround(), fesetround()

Get and set the current floating-point rounding direction

feholdexcept()

Save the floating-point environment, clear all flags, and ignore future errors

feraiseexcept()

Raise the specified floating-point exceptions

fetestexcept()

Determine which of the specified floating-point status flags are set

feupdateenv()

Restore the floating-point environment and raise any previously raised exceptions
The j0(), j1(), jn(), y0(), y1(), and yn() functions are now in libm instead of libbessel.

For more information about the changes to the libraries, see the release notes.