test_nvs_host: fix asan reported bugs (IDFGH-9860)

[codysch]

• Use of sizeof(uint32_t *) instead of sizeof(uint32_t) in arguments
• Use of memcmp without checking that length agreed (add memeq)

These were observed when running these tests built with clang 14 on a MacOS 12.6.3 system (Xcode 13.1). I suspect some of them will appear on any system where uint32_t * and uint32_t are different sizes, and where reading some data should cause the test to fail (due to lack of certain python tooling)

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[igrr]

sha=630343e4b9a706b25cf06c8c407463c774ad6db7

[igrr]

@codysch I've tried to run our test pipeline but it failed on memeq comparisons, here is the log: https://gist.github.com/igrr/abff87b6b6f920203badc1f3f2543bde
I haven't had time to dig deeper into this yet; if you have an idea why this happens, please let me know!

Edit: I have now fixed the failing tests, will merge the PR soon.