az_platform_clock_msec POSIX implementation

[CIPop]

az_posix.c implements the az_platform_clock_msec by measuring CPU time, not wall time. This API isn't used today.
In ANSI C/POSIX, clock_t clock() is to approximate the CPU time spent in a process and the current naming creates confusion.

A few ways to fix this:

1. clock_gettime() clock_getres(). The problem is that POSIX doesn't guarantee availability of (all) timing functions. We'll need runtime code to discover enabled functionality.
2. Use the c time() epoch timer with 1 second accuracy.

Also, please consider the API breaking change to rename the API name to AZ_NODISCARD az_result az_platform_time_msec(int64_t* out_clock_msec) to further reduce confusion on the purpose of the API.

[CIPop]

The decision is to:

1. Require POSIX clock_gettime / clock_getres and CLOCK_MONOTONIC and provide a POSIX compliant implementation using these APIs.
2. Maintain the name of the function as-is to avoid breaking changes.

Based on man-pages:
CLOCK_MONOTONIC: Clock that cannot be set and represents monotonic time since—as described by POSIX—"some unspecified point in the past". On Linux, that point corresponds to the number of seconds that the system has been running since it was booted.
The CLOCK_MONOTONIC clock is not affected by discontinuous jumps in the system time (e.g., if the system administrator manually changes the clock), but is affected by the incremental adjustments performed by adjtime(3) and NTP. This clock does not count time that the system is suspended.

I have checked that RTOSes implementing POSIX layers do have support for above APIs:

• Zephyr - Support for CLOCK_MONOTONIC is available.
• FreeRTOS+POSIX - Clock type is ignored.