Flex Java Compiler Infrastructure

Settings | Report Duplicate

1

I Use This!

Activity Not Available

Commits : Listings

Analyzed 12 months ago. based on code collected 12 months ago.

Commit Message	Contributor	Files Modified	Lines Added	Lines Removed	Code Location	Date
Jan 16, 2023 — Jan 16, 2024 Showing page 1 of 108 Search / Filter on:
Support uniqueRW statistics to find average virgin write object size.	C. Scott Ananian	More...				over 17 years ago
Keep track of average object size for virgin reads/writes.	C. Scott Ananian	More...				over 17 years ago
Add computed cost per read/write.	C. Scott Ananian	More...				over 17 years ago
Fix PowerPC failure when invoking a method with a long or double parameter from native code.	C. Scott Ananian	More...				over 17 years ago
Count the number of method invocations inside transactions.	C. Scott Ananian	More...				over 17 years ago
Don't clone harpoon.Runtime.Transactions.ImplHelper, or we won't be able to substitute the appropriate native code.	C. Scott Ananian	More...				over 17 years ago
Fix Alex's class constant hack.	C. Scott Ananian	More...				over 17 years ago
Remove unnecessary assert which might be distorting measurements.	C. Scott Ananian	More...				over 17 years ago
Oops: really fix transaction statistics bug.	C. Scott Ananian	More...				over 17 years ago
Fix statistics when compiling with transactions.	C. Scott Ananian	More...				over 17 years ago
Allowing using LL/SC stubs for benchmarking transactions.	C. Scott Ananian	More...				over 17 years ago
Fix for 'no array' transactional builds: make sure JNI also doesn't do transformations on arrays.	C. Scott Ananian	More...				over 17 years ago
Add configuration-checking bits to make sure runtime and FLEX transaction variant options match up.	C. Scott Ananian	More...				over 17 years ago
Add configuration-checking bits for NT and NO_ARRAY transactional builds.	C. Scott Ananian	More...				over 17 years ago
Compute appropriate numbers for graph.	C. Scott Ananian	More...				over 17 years ago
Results from re-running the benchmarks on PowerPC.	C. Scott Ananian	More...				over 17 years ago
Add some code to do multi-threaded operations (sorta).	C. Scott Ananian	More...				over 17 years ago
Final results on x86.	C. Scott Ananian	More...				over 17 years ago
Lock-free functional arrays.	C. Scott Ananian	More...				over 17 years ago
First numbers from array microbenchmark.	C. Scott Ananian	More...				over 17 years ago
Better makefile rules for benchmarking arrays.	C. Scott Ananian	More...				over 17 years ago
Tweak implementation to match thesis better.	C. Scott Ananian	More...				over 17 years ago
Rename old benchmark results to results-bench.txt, to make room for a new results-array.txt file.	C. Scott Ananian	More...				over 17 years ago
Use semi-precise collection; workaround bug in x86 optimized code, where top-level obj is (conservatively) held, which makes non-randomized SHALLOW really slow because we can't GC any of it.	C. Scott Ananian	More...				over 17 years ago
Add randomized re-rooting.	C. Scott Ananian	More...				over 17 years ago
Add makefile/.cvsignore support for the array benchmarks.	C. Scott Ananian	More...				over 17 years ago
Add a #ifdef so that PPC assembly isn't built on x86.	C. Scott Ananian	More...				over 17 years ago
First draft of array benchmark. We've got imperative arrays, naive functional arrays, and shallow-binding functional arrays.	C. Scott Ananian	More...				over 17 years ago
Normalized times.	C. Scott Ananian	More...				over 17 years ago
The weird compare to FFFFCACB was actually there for a reason...	C. Scott Ananian	More...				over 17 years ago

←
1
2
3
4
5
6
7
8
9
…
107
108
→