Fix warnings in unit test BSP

[pcolberg]

commit caa98ebd373d48e5ec063d604347248d9df82ba6
Author: Peter Colberg <peter.colberg@intel.com>
Date:   Thu Nov 17 18:40:39 2022

    fake_bsp: fix out-of-bounds read of OFFSET_CONFIGURATION_ROM
    
    The check failed to account for the trailing null byte, and
    memcpy() read beyond the end of the copied string literal.
    
    Signed-off-by: Peter Colberg <peter.colberg@intel.com>

commit 7f737cdc93826a36087417004b99c34e7296e9f8
Author: Peter Colberg <peter.colberg@intel.com>
Date:   Thu Nov 17 18:04:25 2022

    fake_bsp: fix stringop-overread warnings with string literals
    
    GCC 11 correctly points out that there is no point in limiting strnlen()
    to a size that is greater than the length of the string literal argument.
    
    test/fake_bsp/fakegoodbsp.cpp:47:26: warning: 'size_t strnlen(const char*, size_t)' specified bound 1204 exceeds source size 16 [-Wstringop-overread]
       47 |     size_t Xlen = strnlen(X, MAX_NAME_SIZE) + 1;                               \
          |                   ~~~~~~~^~~~~~~~~~~~~~~~~~
    
    strnlen() is commonly used with an input buffer that is possibly not
    null-terminated, i.e., contains a truncated string. strnlen() is also
    useful to set an upper bound on the length to consume, e.g., to avoid
    a denial of service. Neither of these use cases apply here.
    
    MAX_NAME_SIZE was previously introduced in acl.h to address Klocwork
    issues relating to missing null termination of strings. The chosen
    1204 is a typo of 1024, but even then, it should be 1023 such that a
    buffer with the trailing null byte occupies 1023 + 1 = 1024 bytes.
    
    Signed-off-by: Peter Colberg <peter.colberg@intel.com>

[zibaiwan]

Thanks @pcolberg !