Conversation
|
Tagging subscribers to this area: @JulieLeeMSFT, @jakobbotsch Issue DetailsThis PR addresses Jakob's feedback in #95249 (last comment) So we now validate that
|
|
You can also remove the |
src/coreclr/jit/morph.cpp
Outdated
| return tree; | ||
| } | ||
|
|
||
| if (tree->IsNothingNode()) |
There was a problem hiding this comment.
Is this actually reachable? Presumably, NOPs should go through fgMorphLeaf now.
There was a problem hiding this comment.
Yes, gtFoldExpr called above may return a NOP (to be precise, gtFoldExprConst does that for:
* BOUNDS_CHECK_Rng void <l:$285, c:$284>
+--* CNS_INT int 0 $40
\--* CNS_INT int 1 $41
Co-authored-by: SingleAccretion <62474226+SingleAccretion@users.noreply.github.com>
|
/azp list |
This comment was marked as resolved.
This comment was marked as resolved.
|
/azp run runtime-coreclr outerloop, runtime-coreclr jitstress, Fuzzlyn |
|
Azure Pipelines successfully started running 3 pipeline(s). |
|
@jakobbotsch @dotnet/jit-contrib PTAL, this is ready for review, diffs |
jakobbotsch
left a comment
There was a problem hiding this comment.
LGTM, nice cleanup. Why is there a diff?
It seems to only 1 method slightly regressed, but that method is quite huge (5k bytes of codegen).
|
This PR addresses SingleAccretion and Jakob's feedback in #95249 (last comment)
So we now validate that
GT_NOPhas no child nodes and is always ofTYP_VOIDtype