Replace all uses of log::log_enabled with Debug printers
#74876
Conversation
|
r? @eddyb (rust_highfive has picked a reviewer for you, use r? to override) |
rust-highfive
added
the
S-waiting-on-review
|
r? @RalfJung |
|
cc @hawkw |
src/librustc_middle/ty/context.rs
Outdated
| } | ||
| } | ||
|
|
||
| struct DebugStats<'tcx>(TyCtxt<'tcx>); |
There was a problem hiding this comment.
Same here for encapsulating the type inside debug_stats.
What does this mean, do you have an example? |
|
Basically instead of we now get in cases where we had lots of |
|
@bors r+ |
|
📌 Commit b81d164 has been approved by |
bors
added
S-waiting-on-bors
|
☀️ Test successful - checks-actions, checks-azure |
| #[doc(hidden)] | ||
| pub struct RenderAllocation<'a, 'tcx, Tag, Extra> { |
There was a problem hiding this comment.
Why is the doc hidden?
Also this might be pub(crate) if it suit.
There was a problem hiding this comment.
I think this may get called from miri. I hid it from docs because the function is supposed to be the entry point for this and the only reason this is public is because impl Debug didn't work. It can work though, and I think we should just figure out how to make it work, but it'll require some lifetime experiments with the InterpCx
| @@ -56,7 +56,7 @@ impl<'mir, 'tcx> InterpCx<'mir, 'tcx, CompileTimeInterpreter<'mir, 'tcx>> { | |||
| self.copy_op(place.into(), dest)?; | |||
|
|
|||
| self.return_to_block(ret.map(|r| r.1))?; | |||
| self.dump_place(*dest); | |||
| trace!("{:?}", self.dump_place(*dest)); | |||
There was a problem hiding this comment.
Btw, why are these Debug impls? They are definitely made for human consumption, so sounds more like a case of Display to me TBH... (but changing it now is probably not worth it because Miri also needs adjusting).
There was a problem hiding this comment.
idk, the choice was fairly arbitrary
|
This was a slight performance regression on the ctfe stress test. I suspect we're optimizing more poorly or something like that? Edit: it's also possible that this was just noise, unclear -- but seems unlikely. |
|
Given the changes in this PR it's possible |
|
hmm... maybe some targetted placement of |
cc @RalfJung this touches a bunch of logging in the miri engine. There are some visual changes, mainly that in several cases we stop prepending lines with the module path and just have a newline.