jimingham wrote:
This seems like a generic module loading observer. I don't see anything JIT
specific about it. Not saying a generic module loading observer is not a good
idea. But calling it a JITLoader seems pretty confusing to me.
https://github.com/llvm/llvm-project/pull/142514
https://github.com/medismailben approved this pull request.
Hey @clayborg, this is pretty cool. I'm glad you were able to use and extend
the ScriptedPythonInterface to implement this, hopefully it wasn't too
complicated. LGTM!
https://github.com/llvm/llvm-project/pull/142514
__
jimingham wrote:
One thing that's nicer about this approach than independent callbacks for each
of the hooks is that it allows you to group the three callbacks in the same
class, so that you can more easily keep shared state.
But that's a general problem with these affordances. For instance,
jimingham wrote:
But if you don't care so much about shared state, then I think a better way of
doing what you want is just to add `target module-hook add`...
https://github.com/llvm/llvm-project/pull/142514
___
lldb-commits mailing list
lldb-commits@
jimingham wrote:
I changed the stop-hooks recently so they optionally fire when lldb first gets
control of the process, so you can already write python code that intervenes
when your "did_attach" and "did_launch" callbacks fire.
https://github.com/llvm/llvm-project/pull/142514
jimingham wrote:
This also seems like an awkward way to do what we've wanted for a while, which
is the equivalent of stop hooks for "launch", "attach" and "module loaded",
since you have to do all three, even if you only wanted to do one of the set.
https://github.com/llvm/llvm-project/pull/14
