diff options
author | Joseph Tremoulet <jotrem@microsoft.com> | 2017-08-18 12:13:41 -0400 |
---|---|---|
committer | Joseph Tremoulet <jotrem@microsoft.com> | 2017-08-18 13:42:25 -0400 |
commit | 73304d84c008380b1fed2d28bf48a16fdf3ec7b4 (patch) | |
tree | 75e3cb0b87e5ce9e41c91b0c3827666714bed15f /Documentation | |
parent | 7145b1e1ccf2d5d77138be9589e221b3e4e472ef (diff) | |
download | coreclr-73304d84c008380b1fed2d28bf48a16fdf3ec7b4.tar.gz coreclr-73304d84c008380b1fed2d28bf48a16fdf3ec7b4.tar.bz2 coreclr-73304d84c008380b1fed2d28bf48a16fdf3ec7b4.zip |
Mention Byref Store/Load VN and Helper Call Kills
These are two more areas that have received recent scrutiny.
Diffstat (limited to 'Documentation')
-rw-r--r-- | Documentation/performance/JitOptimizerTodoAssessment.md | 17 |
1 files changed, 17 insertions, 0 deletions
diff --git a/Documentation/performance/JitOptimizerTodoAssessment.md b/Documentation/performance/JitOptimizerTodoAssessment.md index 24b9a4fa53..5a34535800 100644 --- a/Documentation/performance/JitOptimizerTodoAssessment.md +++ b/Documentation/performance/JitOptimizerTodoAssessment.md @@ -122,6 +122,15 @@ benchmarking story set up to measure the effect of such changes when they go in, we should do that. +### Helper Call Register Kill Set Improvements + +We have some facility to allocate caller-save registers across calls to runtime +helpers that are known not to trash them, but the information about which +helpers trash which registers is spread across a few places in the codebase, +and has some puzzling quirks like separate "GC" and "NoGC" kill sets for the +same helper. Unifying the information sources and then refining the recorded +kill sets would help avoid more stack traffic. See [#12940](https://github.com/dotnet/coreclr/issues/12940). + Low-Hanging Fruit ----------------- @@ -144,6 +153,14 @@ patterns and emit specialized write barriers that avoid various overheads -- see [#13006](https://github.com/dotnet/coreclr/issues/13006) and [#12812](https://github.com/dotnet/coreclr/issues/12812). +### Byref-Exposed Store/Load Value Propagation + +There are a few tweaks to our value-numbering for byref-exposed loads and stores +to share some of the machinery we use for heap loads and stores that would +allow better propagation through byref-exposed locals and out parameters -- +see [#13457](https://github.com/dotnet/coreclr/issues/13457) and +[#13458](https://github.com/dotnet/coreclr/issues/13458). + Miscellaneous ------------- |