I agree it's nice to have somewhere, but git history should be fine. I should elaborate on motivations, 1. ensure everyone is using the script so that any issues are surfaced, 2. prevent missteps in the manual process, which may also be bit rotting if everyone is using the script.

sgtm

Nice improvement!

Should lldb avoid calling asan (or any other sanitizer's) mmap? If so, maybe this function can be sanitizer aware and ignore those symbols.

for context this caused standalone swift-lldb builds to have warnings disabled via -w

@compnerd Saleem, what do you think? (see also my reply to you)

@teemperor I'm not sure what these build errors mean, for example https://reviews.llvm.org/B70998. Also there's the linux build failure, but that's not relevant for debugserver, right?

Add LLVM_ENABLE_WARNINGS to LLVMConfig.cmake.in

If another project defines LLVM_ENABLE_WARNINGS before loading HandleLLVMOptions, it seems correct to me that the first one is used. This change ensures the default value of ON is setup at the last possible opportunity, before LLVM_ENABLE_WARNINGS is read and acted on.

clang-format

Follow up to 8aeb212887024a615ca02437cd12fa055bd54b6f

If an LLVM install disabled LLVM_ENABLE_WARNINGS, should other builds inherit that? I would think no, but is there a precedent for that that to be the case?

The LLVM_ENABLE_WARNINGS variable is read only within HandleLLVMOptions.cmake. Outside declarations/defaults have effect only when HandleLLVMOptions is loaded, one way or another.

@lebedev.ri clang/CMakeLists.txt contains include(HandleLLVMOptions) for its standalone build, I think this covers the issue you're pointing out

Given the expr status quo, I'd stick with that.

In D86493#2236454, @aprantl wrote:

Sometimes these odd leaf APIs are used by custom downstream variants of LLDB

I wonder if it is time to take the gdb commands out and just make this a task -> lldb commands table?

fwiw I have a python command called abbrev that takes a command and shows you the alias you can use instead.

Thanks. I thought this was reasonable since the documentation shows many uses of lldb aliases already.

Sounds good @aprantl

Remove unintended changes.

Thanks for reviewing @clayborg.

I'm getting build failures that appear to be caused by this patch. I'm compiling on macOS with Xcode 10.1. The failures are uses of LLDB_RECORD_DUMMY(), passing a callback function pointer as one of the args. This results in errors like:

@clayborg this has been updated the approach you suggested, adding an interactive parameter.

Add interactive parameter

Thanks @compnerd!

@pcc I have this fix related to -no-llvm-bc (D4371). I'm looking for a reviewer: could you review, or could you suggest someone? Thanks.

@compnerd would you be willing to accept?

@jingham If you're still willing to review this change to command regex, it is updated per our previous discussion, thanks.

@keith I screwed up and committed without amending the commit message to attribute you as the author :( – I'm sorry

There are other commands that print instructions in the same way that command regex does, should they be updated in this change too?

Added IOHandlerActivatedInteractively

@smeenai or @compnerd – much late follow up, but either of you willing to accept?