Age | Commit message (Collapse) | Author | |
---|---|---|---|
2014-05-07 | Split alloca liveness propagation into separate forward and backward phases. | Dan Gohman | |
This makes the code simpler, and avoids problems with the two directions interfering with each other. This fixes asm2.test_sqlite and other failures. | |||
2014-05-07 | Make bottom-up and top-down liveness propagation independent. This fixes Bug ↵ | Dan Gohman | |
1006301. | |||
2014-05-06 | remove unneeded |0 also in memcpy | Alon Zakai | |
2014-05-06 | avoid unneeded |0 now that we >>0 HEAP8 accesses | Alon Zakai | |
2014-05-06 | emit >>0 in HEAP8, for consistency with HEAP16|32 | Alon Zakai | |
2014-04-29 | do not emit a branch to the relooper if the target is the default ↵ | Alon Zakai | |
destination anyhow; fixes #32 | |||
2014-04-29 | relooper update | Alon Zakai | |
2014-04-29 | relooper update | Alon Zakai | |
2014-04-25 | emscripten-no-aliasing-function-pointers option | Alon Zakai | |
2014-04-24 | comment | Alon Zakai | |
2014-04-16 | relooper update | Alon Zakai | |
2014-04-08 | break phi var cycles only between actual phi vars | Alon Zakai | |
2014-03-31 | suppress args warnings when the number is forced anyhow | Alon Zakai | |
2014-03-31 | ensure a cast on everything in a return, even a nan or an inf | Alon Zakai | |
2014-03-31 | truncate arguments that are not in the actual function being called | Alon Zakai | |
2014-03-28 | expand constantExprs for all illegal values, not just >=64bit, so that ↵ | Alon Zakai | |
things like nested i24 CEs are fixed up by promoteIntegers | |||
2014-03-26 | optimize x < 0 and generalize it to non-i64 values too | Alon Zakai | |
2014-03-25 | emit output variable for calls to absolute values, even though they ↵ | Alon Zakai | |
segfault, to not break validation | |||
2014-03-24 | apply switch-related phi changes after illegal phis were erased | Alon Zakai | |
2014-03-23 | handle switches that reach phis appearing before them; fixes emscripen issue ↵ | Alon Zakai | |
2244 | |||
2014-03-15 | Handle conflicting lifetime markers conservatively. | Dan Gohman | |
Avoid infinite looping in AllocaManager::computeInterBlockLiveness, without using new temporary variables. This also makes it clear that the algorithm is conservative in the face of conflicting liveness intrinsics. | |||
2014-03-15 | update relooper | Alon Zakai | |
2014-03-14 | warn if given the wrong triple | Alon Zakai | |
2014-03-14 | prettyWarning utility | Alon Zakai | |
2014-03-14 | update relooper; 1.13.21.13.2 | Alon Zakai | |
2014-03-11 | tiny indentation fix | Alon Zakai | |
2014-03-11 | mention the function in the asm() error | Alon Zakai | |
2014-03-10 | update float ffi rules | Alon Zakai | |
2014-03-08 | resolve constants fully, through both multiple aliases and bitcasts | Alon Zakai | |
2014-03-08 | avoid infinite loops in AllocaManager::computeInterBlockLiveness | Alon Zakai | |
2014-03-07 | fix the case where exceptions are lowered and we already had ↵ | Alon Zakai | |
getHigh/preInvoke/postInvoke defined in the module | |||
2014-03-06 | Implement integer promotion for urem, udiv, srem, and sdiv | Dan Gohman | |
The optimizer sometimes thinks it's beneficial to truncate all manner of i64 operators to narrower types, even when still wider than the platform's widest legal type. | |||
2014-03-06 | handle aliases in getConstAsOffset | Alon Zakai | |
2014-03-06 | Fix a use-after-free error in GlobalOpt CleanupConstantGlobalUsers | Dan Gohman | |
This is a backport of r197178 from LLVM trunk. | |||
2014-03-06 | do not warn on invalid argument numbers or types in varargs calls | Alon Zakai | |
2014-03-06 | Improve linker messages when modules differ in triple or data layout. | Dan Gohman | |
This is a backport of r203009 from LLVM trunk. Since we recently changed triples and data layouts, this is particularly apropos. | |||
2014-03-05 | Disable the ResolveAliases and GlobalCleanup passes. | Dan Gohman | |
Also add a test for handling of global aliases. | |||
2014-03-05 | Minor CallHandler cleanups. | Dan Gohman | |
2014-03-05 | Disable the LowerExpectIntrinsic pass. | Dan Gohman | |
2014-03-05 | Don't emit redirects for llvm.memcpy.p0i8.p0i8.i32 and friends. | Dan Gohman | |
We never reference these intrinsics by name, so we don't need redirects for them. | |||
2014-03-05 | Don't emit JS declarations for LLVM intrinsics expanded inline. | Dan Gohman | |
2014-03-05 | Tidy up the output for memset and memcpy, and add a test. | Dan Gohman | |
2014-03-05 | Don't run a pass to strip dead prototypes; just ignore them. | Dan Gohman | |
Also, fix the code for supressing declarations for no-op intrinsics. | |||
2014-03-05 | emscripten_debugger() intrinsic | Alon Zakai | |
2014-03-05 | Clean up the handling of inline asm. | Dan Gohman | |
Make inline asm a report_fatal_error instead of an assertion failure so that it's a little friendlier, and add a test to make sure llc -march=js rejects inline asm. Also disable the PNaCl inline asm("":::"memory") lowering pass. If people are using this, it's best that we diagnose it as it likely isn't portable. There are usually better alternatives. | |||
2014-03-05 | Teach phi translation to also look through bitcasts. | Dan Gohman | |
This fixes a regression introduced in d95cd364f0c049d6c1b8d78746d44c00ed2f69f2; when regular expression translation looks through bitcasts but phi translation doesn't, phi translation may fail to properly detect dependencies. | |||
2014-03-05 | Fix missing LLVMBuild.txt and CMake dependencies. | Dan Gohman | |
2014-03-05 | Trim unneeded #includes. | Dan Gohman | |
This also helps clarify that we're not using the MC framework. | |||
2014-03-05 | Eliminate a few more temporary std::strings. | Dan Gohman | |
2014-03-05 | Fix a broken CMake build file. | Dan Gohman | |