Fix compile errors when MBEDTLS_HAVE_TIME is not defined

[raoulstrackx]

When the library was being compiled without MBEDTLS_HAVE_TIME defined, it resulted in compile errors. This commit fixes those problems

[gilles-peskine-arm]

• Please add a changelog entry file under Changelog.d.
• Apparently there are also build errors in X.509 code. This pull request is complementary with Fix ssl_context_info.c to correctly check for EOF - type-limit bug #3449. It would be great if you could synchronize with @naynajain to fix the problem everywhere.
• To avoid the problem coming up again, there should be a build in tests/scripts/all.sh with MBEDTLS_HAVE_TIME disabled but other options enabled. Currently, I think all the builds with MBEDTLS_HAVE_TIME disabled are variants of the baremetal configuration. Would you mind investigating why this build doesn't catch the problems here, and adding a configuration that does catch the problem, so that we have a non-regression test?

[gilles-peskine-arm]

The Mbed OS CI is failing for unrelated reasons at the moment, so you can ignore failures of the Mbed OS CI jobs.

[naynajain]

* Please add a changelog entry file under `Changelog.d`.

* Apparently there are also build errors in X.509 code. This pull request is complementary with #3449. It would be great if you could synchronize with @naynajain to fix the problem everywhere.

Hi,

Gilles referred to your PR as it seems one of my commit in my PR 3449 is fixing same issue as submitted by your PR ? How would you want to go about it ?

Thanks & Regards,
- Nayna

[daxtens]

Hi, I've been working with @naynajain and I'd like to help land this.

I looked into the non-regression test. It's tricky. To pick up the bug we found, we need a build with !defined(MBEDTLS_HAVE_TIME) && defined(MBEDTLS_X509_CRT_PARSE_C), and we need including time.h to fail. scripts/config.py baremetal generates a suitable config, but because we're building on a real system that does have time.h, everything still builds. I'm working on a better test for this that set up a fake include directory.

[raoulstrackx]

Hi @daxtens, @naynajain, sorry for the late response, I've just come back from holiday. I don't have many cycles to burn, but I'd like to help/learn.

[tom-cosgrove-arm]

Closing this PR as these changes are being does as part of #3624