[llvm] [Bolt] fix a wrong relocation update issue with weak references (PR #69136)

[Vladislav Khmelevsky via llvm-commits]

yota9 wrote:

> . But I wonder which symbols would have no address, and still be strong definitions (non-weak)

Any PLT symbol located in another binary (DEFAULT UND)

> Maksim pointed out that we should be skipping processing static relocations in places that also have dynamic relocations

As for the further discussions - we need to consider it carefully. Maybe you're right and we need to add "white" list of dynamic relocations that would also have static relocations. To be honest I can't say what way would be the best here right away, it needs deep consideration and long list of tests I think :) I would try to ponder about this at leisure cause it doesn't look as simple question to me..

https://github.com/llvm/llvm-project/pull/69136