[feature] ITM functionality for ST-Link V2 and STM32Fxx chipsets

[simplerobot]

This change implements feature #136.

The main changes:

• Adds enable/disable/read trace methods to stlink library.
• Adds flags to stlink and chipid for trace support.
• Adds an additional command line tool st-trace.

I did bench testing while developing this, including testing with incorrect clock speeds and with overruns. However, this testing has been done with an extremely limited number of only STM32F4 processors. I could use some help running this on a wider variety of hardware. I am currently using st-trace in a production environment that automatically runs test suites on multiple hardware devices.

This code review is larger than I would hope for my first to this project. I've tried to make the code fit in with the project, but please don't hold back if I've missed the mark anywhere.

Thanks,
John

[simplerobot]

@Nightwalker-87 I will squash all commits into a single commit after all CR feedback has been implemented.

[simplerobot]

I had added the max_trace_freq field to the stlink structure in the middle. I've corrected that by moving it to the end.

[slyshykO]

@Nightwalker-87, this PR looks good to me.

[Nightwalker-87]

@slyshykO It still fails: https://travis-ci.org/github/stlink-org/stlink/builds/748613319
Unfortunately I can only test when triggering a manual build after merging.
This Travis CI integration has become such a mess! You can't use it before merging any more. 😠

[slyshykO]

@Nightwalker-87 think it is time for Github Actions

[simplerobot]

@Nightwalker-87 Sorry I was waiting for additional code review feedback and had not squashed all commits into a single commit yet.

It also looks like I broke the build in clang by mixing a 64 bit size_t type and a 32 bit int type. I will work on fixing this now and get clang installed to verify the results.

[slyshykO]

There is #1076, so maybe we beg @Ant-ON to fix one more clang's warning?

[Nightwalker-87]

... he seems to collect everything up, involuntarily 🤣
@simplerobot You should be fine, it's ok to keep it in mind for future PRs.

[Nightwalker-87]

@Nightwalker-87 think it is time for Github Actions

At least the recent custom triggered build of the develop branch now showed up, so there still is some hope... ⏳

[slyshykO]

@Nightwalker-87 Travis CI has changed its pricing model.
There is a blog post - https://blog.travis-ci.com/2020-11-02-travis-ci-new-billing .
So somebody who has access to Travis account should check credits.

[Nightwalker-87]

I know, but it is still free for OSS-Projects, so we don't pay anything for it.

[slyshykO]

I know, but it is still free for OSS-Projects, so we don't pay anything for it.

Free, but more limited than before. Looks like we encounter those limits.

[Nightwalker-87]

I'll have an eye on it. Should we run into further problems, one may think about appropriate alternatives.
Thanks for the input. 👍