Reviewers: Søren Gjesse, Mads Ager, Description: Generate inline code for contextual loads.
Contextual load requires only a map check followed by a cell hole check so we can generate pretty compact code for that. The fact that we have inlined code is marked by mov ecx, offset instruction after the IC call. Inlining is only enabled inside loops and in non-builtin functions. The generated code size increase is about 3%. This descreased the pc-to-code cache hit rate in some of the benchmarks that trigger GC. To compensate we now have 4 times as much entries in the cache. Please review this at http://codereview.chromium.org/3402014/show Affected files: M src/arm/ic-arm.cc M src/arm/stub-cache-arm.cc M src/frames.h M src/ia32/codegen-ia32.cc M src/ia32/ic-ia32.cc M src/ia32/stub-cache-ia32.cc M src/ic.h M src/ic.cc M src/v8-counters.h M src/x64/ic-x64.cc M src/x64/stub-cache-x64.cc -- v8-dev mailing list [email protected] http://groups.google.com/group/v8-dev
