rtcvb32: Workarounds are
NEVER done/supported by the original distributor/author.
Not entirely true. Sometimes, when a piece of software (or hardware!) has a bug, and that bug hasn't been fixed, the original distributor does suggest a workaround.
For example, when there is a security flaw that hasn't been patched, sometimes there is a way to mitigate or eliminate the danger. For example, when Heartbleed was a threat, the risk could be eliminated by recompiling OpenSSL without TLS heartbeat support (which happens to be easy for the end user to do on Gentoo GNU/Linux, but not so easy on other distributions).
Another example: I mentioned hardware. Well, sure enough, there are plenty of errata issued for modern CPUs; quite often, a kernel will need to add extra code to work around such issues. One famous example was the Pentium FDIV flaw; certain division operations would give incorrect results.