aboutsummaryrefslogtreecommitdiff
path: root/src
AgeCommit message (Collapse)Author
2011-12-02more strict mode fixesAlon Zakai
2011-12-02various fixes from js strict mode testingAlon Zakai
2011-12-01support for running emscripten in the browserAlon Zakai
2011-12-01code cleanup, do not do heavy operations in parameters to print or assertAlon Zakai
2011-12-01misc minor but important memory fixesAlon Zakai
2011-11-30minor optimization to main loop in frameworkAlon Zakai
2011-11-30disable unneeded (and slow) JSifying of typesAlon Zakai
2011-11-30refactor compiler to allow future batching of types and globals. move ↵Alon Zakai
postsets from run() to the toplevel to boost performance and simplify compilation. fix various bugs that were noticed during this
2011-11-29refactor parseFunctions ==> sidePassAlon Zakai
2011-11-29refactor unparsedFunctions a tiny bitAlon Zakai
2011-11-29more memory debugging stuff, and a disabled option to throttle the framework ↵Alon Zakai
inbox
2011-11-29fix makeReintegrator bugAlon Zakai
2011-11-29discard almost all .tokens when intertyper ends, to save memory (all except ↵Alon Zakai
for inside Globals)
2011-11-29more memory debugging tools and minor optimizationsAlon Zakai
2011-11-28support for nested objects in library objects, including functionsAlon Zakai
2011-11-28fix test_lua by supporting lines beginning with ; inside functionsAlon Zakai
2011-11-28allow more GCing inside our big compiler loopsAlon Zakai
2011-11-28partial fix for breakinthemiddle in test_casesAlon Zakai
2011-11-28fix for test_closure_compiler with spidermonkey as compiler engineAlon Zakai
2011-11-28memory debugging code, and some minor optimizations from itAlon Zakai
2011-11-27When I64_MODE==1, split each global constant int64 at compile time.Benjamin M. Schwartz
This change allows the generated code to be processed by the closure compiler. Without this change, commit 81f82a6a0dd932344027a408ff7bf905de1d2e52 prevents the closure compiler from running (see http://code.google.com/closure/compiler/docs/error-ref.html#error). This change should also slightly improve startup speed and reduce code size.
2011-11-27move printing of type info to right positionAlon Zakai
2011-11-27tiny fix for needAnalysis being a dictionaryAlon Zakai
2011-11-27do not save function JS in memory, print it out and forget it to save memory ↵Alon Zakai
during compilation
2011-11-27fix for i64 mode 1 conversion of double to i64Alon Zakai
2011-11-27refactor finalCombinerAlon Zakai
2011-11-27fix test_emscripten_apiAlon Zakai
2011-11-26fixes for 64-bit values in globals in i64 mode 1Alon Zakai
2011-11-26optimize overflow correction for 32 bitsAlon Zakai
2011-11-26inline calculations for sign/unsigning of >32 bitsAlon Zakai
2011-11-26optimize resigning of <32 bit intsAlon Zakai
2011-11-26clean up signing codeAlon Zakai
2011-11-26comment on byvalAlon Zakai
2011-11-25fix for passing structs by value, and warning for unfixable case of passing ↵Alon Zakai
structs by value between C and C++ calling conventions
2011-11-25fix varargs calls in function pointersAlon Zakai
2011-11-24optimize tempInt and tempBigInt's with makeInlineCalculationAlon Zakai
2011-11-24rewrite makeRounding to fix a bugAlon Zakai
2011-11-23replace compiler tricks for one-time loop removal and label removal with a ↵Alon Zakai
proper pass in the js optimizer
2011-11-23optimize relooper optimizerAlon Zakai
2011-11-22remove buggy condition stealing optimization, that is no longer useful anyhowAlon Zakai
2011-11-22fix bugs in (unused) setSubAlon Zakai
2011-11-22experimental label removal diffAlon Zakai
2011-11-21some unused code possibilities regarding switchAlon Zakai
2011-11-21fix i64 comparison bug, and various test fixesAlon Zakai
2011-11-21llvm_uadd_with_overflow_i32Alon Zakai
2011-11-21properly handle branches to phi from invokes in assignsAlon Zakai
2011-11-20js optimizer fix, and remove assigns to undefined in libraryAlon Zakai
2011-11-19partial work to remove assigns of undefined in js optimizerAlon Zakai
2011-11-19fix for phis from invokeAlon Zakai
2011-11-18support reaching phi from invoke, and a workaround with vars in phi sets for ↵Alon Zakai
closure compiler