Update Subprojects to newest stable Version.

[MtoLStoN]

Closes #840.

[bjoe2k4]

I've tested this on our hpc cluster (ubuntu 20.04/intel 2023.1) and it works well.

Only downside are flaky toml-f/tftest-* tests when tests are not limited to xtb only (run tests without --suite xtb). These test fails (usually just one) occur randomly when tests are run multiple times, more likely when test concurrency is increased. If i recall correctly i had this problem already with v6.6.0 but I didn't take it serious.

[awvwgk]

Only downside are flaky toml-f/tftest-* tests when tests are not limited to xtb only (run tests without --suite xtb). These test fails (usually just one) occur randomly when tests are run multiple times, more likely when test concurrency is increased. If i recall correctly i had this problem already with v6.6.0 but I didn't take it serious.

Could you report the flaky tests in TOML Fortran? https://github.com/toml-f/toml-f/issues/new/choose

[bjoe2k4]

Unfortunately i cannot reproduce these flaky tests with standalone toml-f, otherwise i would have reported it upstream. Not sure how to continue from here.

[MtoLStoN]

While these flaky tests are definitely something we should look at, I can also partly reproduce them with the "old" toml-f version. However, as of now, I could not see any downside except these flaky tests. Interestingly, it seems, that these hiccups only happen with the Intel Fortran Compiler and not with gfortran.

Therefore, I believe it is safe to disregard these failing tests for the time being.