hvdijk wrote:
Given that https://llvm.org/docs/RemoveDIsDebugInfo.html specifically documents
these functions as "no plans to deprecate", even though I personally feel like
they are a bad idea, I just updated them so they compile with no deprecation
warnings. Does this look like a good approac
hvdijk wrote:
Thanks again, it ends up looking much better (IMO) with `InsertPosition`, and
it manages to avoid the change on the users' side that I was hoping to avoid.
https://github.com/hvdijk/llvm-project/commit/dibuilder-insertposition Since
this allows us to have a single method rather t
hvdijk wrote:
Thanks, I knew it was possible (it had to be: it would be possible to create a
dummy instruction, insert it at the specified point, and inspect its parent)
but had not yet found the best way of doing it, I was trying to get it working
with `IP.getNodePtr()->getParent()`. I'll upd
hvdijk wrote:
I'm taking a look, but I'm having some doubts about the change itself,
`DIBuilder::insertDbgValueIntrinsic`'s `BasicBlock *BB =
InsertBefore->getParent();` (where `InsertBefore` is an iterator) doesn't seem
safe, it seems like it should be valid to pass an at-the-end iterator the
hvdijk wrote:
Yes, thank you, something like that looks great! I should be able to do some
testing with that tomorrow to check to make sure I am not missing anything.
https://github.com/llvm/llvm-project/pull/124287
___
lldb-commits mailing list
lldb-
hvdijk wrote:
Hi, in principle this change makes sense but it causes a bit of an issue with
[`DIBuilder`](https://llvm.org/doxygen/classllvm_1_1DIBuilder.html) not yet
having been updated to take iterators. The workaround for that at the moment is
trivial (change e.g. `DIB.insertDeclare(..., B
