[Lldb-commits] [lldb] DynamicLoaderDarwin load images in parallel (PR #110439)

[Jason Molenda via lldb-commits]

jasonmolenda wrote:

It is a bit interesting that the user and system time are so much higher -- looks like around 40% higher? -- than the single-threaded approach, I wonder if that's thread creation/teardown overhead, or if it's costlier to acquire locks and it's adding up.  Wallclock is the right thing to optimize for, but it was a little surprising to see that.  I'm not curious enough to look into it more, or ask anyone else to.

Was the setting intended for testing purposes only, or did you intend to include that in a final PR?  Sometimes when we land a new feature that may have unintended consequences, we'll have a flag enabling the feature by default as a way for people impacted to have a way to quickly disable it.  After the feature/change has had some more widespread use, we drop the setting.

https://github.com/llvm/llvm-project/pull/110439