Fix #1412, Wrong memory alignment calculation

[frank-kue]

Checklist (Please check before submitting)

• I reviewed the Contributing Guide.
• I signed and emailed the appropriate Contributor License Agreement to GSFC-SoftwareRelease@mail.nasa.gov and copied cfs-program@lists.nasa.gov.

Describe the contribution
A clear and concise description of what the contribution is.

• Fix Wrong memory alignment calculation in libc-stdlib-stubs.c function OCS_malloc() #1412
• This corrects the memory alignment calculation as already described in the issue.

Testing performed
Steps taken to test the contribution:

• I rebuild and tested this fix with exactly the cFS/OSAL code, configuration and environment in which bug caused a crash of the test coverage-vxworks-tasks-testrunner.exe. The test executed successfully (with out crash) when build with -O2 as well as -O0. This was RTEMS 6, ARM architecture and BSP xilinx_zynq_a9_qemu and tests run on QEMU.
• Moreover, I build and run the current cFS (with OSAL) git head with and without this patch and did run all tests. cFS: 2402f753b6e45341437bb352fad59a97485f27ce and OSAL 7e1ffbb61053e1805a0f7dd401b17f9d90e25fbd. These where the standard "ouf-of-box" cFS builds for SIMULATION=native (Linux) and for SIMULATION=i686-rtems6.

Expected behavior changes
Test coverage-vxworks-tasks-testrunner.exe should not crash.

System(s) tested on
See above.

Additional context
None.

Third party code
None.

Contributor Info - All information REQUIRED for consideration of pull request
Frank Kuehndel, embedded brains GmbH & Co. KG

[jphickey]

Thanks for submitting this, I don't know how this had been in there for years and never noticed.

[jphickey]

I just reverted the change from cpuaddr back to size_t -- as the NextSize local variable is a size_t here. Pedantic as they are almost certainly the same underlying type, but correct nonetheless. Just need the "-1".