Explicitly represent BBJ_EHFINALLYRET successors#93377
Explicitly represent BBJ_EHFINALLYRET successors#93377BruceForstall merged 2 commits intodotnet:mainfrom
Conversation
Currently, BBJ_EHFINALLYRET blocks have no explicit successors in the IR. To implement successor iteration, a very expensive process is followed to (1) find the region of blocks where a BBJ_CALLFINALLY block calling the `finally` might be found, (2) search the region for such blocks, and (3) return as a successor all the BBJ_ALWAYS blocks in the corresponding BBJ_CALLFINALLY/BBJ_ALWAYS pair. Change the IR to explicitly represent and maintain this list of successors for BBJ_EHFINALLYRET blocks. The representation is a simple array of `BasicBlock*`, similar to how BBJ_SWITCH block targets are represented. Fixes dotnet#84278 Notes: 1. The BBJ_EHFINALLYRET successors are computed in `impFixPredLists()`. There are various dumpers that run before this, so we need to tolerate incomplete successor information in some places. 2. `ehGetCallFinallyBlockRange()` is still used by some code. I changed the semantics to return a `[first..last]` range inclusive of `last` instead of the previous `[beginning..end)` range exclusive of `end`. This makes it easier to use with our BasicBlock iterators.
ghost
added
the
area-CodeGen-coreclr
|
Tagging subscribers to this area: @JulieLeeMSFT, @jakobbotsch Issue DetailsCurrently, BBJ_EHFINALLYRET blocks have no explicit successors in the IR. To implement successor iteration, a very expensive process is followed to (1) find the region of blocks where a BBJ_CALLFINALLY block calling the Change the IR to explicitly represent and maintain this list of successors for BBJ_EHFINALLYRET blocks. The representation is a simple array of Fixes #84278 Notes:
|
|
No diffs, generally a slight TP improvement, in some cases significant. |
|
@AndyAyersMS PTAL |
AndyAyersMS
left a comment
AndyAyersMS
left a comment
There was a problem hiding this comment.
Looks good overall. Left a few suggestions for possibly simplifying the code.
|
@AndyAyersMS I addressed your suggestions. |
|
The superpmi-diffs failure is out of disk space. |
|
Looks like @AndyAyersMS had a PR with the same win-arm64 failures: #93371 |
Maybe the build machines picked up a new MSVC arm64 compiler that has a bug? It does look like the compiler changed recently. |
2 participants
Currently, BBJ_EHFINALLYRET blocks have no explicit successors in the IR. To implement successor iteration, a very expensive process is followed to (1) find the region of blocks where a BBJ_CALLFINALLY block calling the
finallymight be found, (2) search the region for such blocks, and (3) return as a successor all the BBJ_ALWAYS blocks in the corresponding BBJ_CALLFINALLY/BBJ_ALWAYS pair.Change the IR to explicitly represent and maintain this list of successors for BBJ_EHFINALLYRET blocks. The representation is a simple array of
BasicBlock*, similar to how BBJ_SWITCH block targets are represented.Fixes #84278
Notes:
impFixPredLists(). There are various dumpers that run before this, so we need to tolerate incomplete successor information in some places.ehGetCallFinallyBlockRange()is still used by some code. I changed the semantics to return a[first..last]range inclusive oflastinstead of the previous[beginning..end)range exclusive ofend. This makes it easier to use with our BasicBlock iterators.