summaryrefslogtreecommitdiff
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* Fix Relooper leaking Branches (#3097)Daniel Wirtz2020-09-084-67/+100
| | | Fixes the `Relooper` leaking `Branch`es in `Optimizer::SkipEmptyBlocks`, by refactoring the API so a `std::unique_ptr` is ensured for each `Block`, `Branch` and `Shape` upon adding to the relooper.
* Stack utils (#3083)Thomas Lively2020-09-074-3/+516
| | | | | | Implement and test utilities for manipulating and analyzing a new stacky form of Binaryen IR that is able to express arbitrary stack machine code. This new Poppy IR will eventually replace Stack IR, and new optimization passes will be built with these utilities. See #3059.
* Improve inlining "heavyweight" (#3085)Max Graey2020-09-046-29/+35
| | | | | | | | | Split that mode into an option to check for loops (which indicate a function is "heavy") and a constant check for having calls. The case of calls is different as we would need more logic to avoid infinite recursion if we are willing to inling functions with calls. Practically, this renames allowHeavyweight to allowFunctionsWithLoops.
* Optimize MergeBlocks by caching branch results (#3102)Alon Zakai2020-09-033-16/+115
| | | | | | | | | | | BranchSeekerCache caches the set of branches in a node + its children, and helps compute new results by looking in the cache and using data for the children. This avoids quadratic time in the common case of a post-walk on a tower of nested blocks which is common in a switch. Fixes #3090 . On the testcase there this pass goes from over a minute to less than a second.
* Remove old stack function from StackCheck (#3100)Alon Zakai2020-09-031-12/+1
|
* wasm-emscripten-finalize: Don't rename the imported table (#3101)Alon Zakai2020-09-031-9/+4
| | | | | | | | | When minimizing wasm changes, leave it as __indirect_function_table which is what LLVM emits. This also removes the renaming of the memory. That was never needed as LLVM already emits "memory" there. See #3043
* Fix Wasm2JSBuilder leaking temporary Function (#3098)Daniel Wirtz2020-09-031-7/+8
| | | Fixes `Wasm2JSBuilder` leaking a temporary `Function` (`WASM_FETCH_HIGH_BITS`) in `Wasm2JSBuilder::processWasm`. The function is created to be converted to JS, but is not actually part of the module, so it either needs to be cleaned up separately or be added to the module. This PR does the latter in case it is useful.
* Fix LegalizeJSInterface leaking duplicate stub Functions (#3095)Daniel Wirtz2020-09-021-6/+9
| | | Fixes `LegalizeJSInterface::makeLegalStub` forgetting to `delete` stub Functions that turned out to be not needed. Now checks whether a stub is needed and otherwise skips creating the redundant stub right away.
* MinifyImportsAndExports: Minify the memory and table as well. (#3089)Alon Zakai2020-09-022-16/+12
| | | | | | | | | | | | We were careful not to minify those, as well as the stack pointer, which makes sense in dynamic linking. But we don't run this pass in dynamic linking anyhow - we need the proper names of symbols in that case. So this was not helping us, and was just a leftover from an early state. This both a useful optimization and also important for #3043, as the wasm backend exports the table as __indirect_function_table - a much longer name than emscripten's table. So just changing to that would regress code size on small projects. Once we land this, the name won't matter as it will be minified anyhow.
* StackCheck: Check both under and overflow (#3091)Alon Zakai2020-09-021-24/+53
| | | | | | | | | | | | | | | | | | | See emscripten-core/emscripten#9039 (comment) The valid stack area is a region [A, B] in memory. Previously we just checked that new stack positions S were S >= A, which prevented us from growing too much (the stack grows down). But that only worked if the growth was small enough to not overflow and become a big unsigned value. This PR makes us check the other way too, which requires us to know where the stack starts out at. This still supports the old way of just passing in the growth limit. We can remove it after the roll. In principle this can all be done on the LLVM side too after emscripten-core/emscripten#12057 but I'm not sure of the details there, and this is easy to fix here and get testing up (which can help with later LLVM work). This helps emscripten-core/emscripten#11860 by allowing us to clean up some fastcomp-specific stuff in tests.
* Fix DataFlowOpts leaking temporary Functions (#3093)Daniel Wirtz2020-09-021-4/+4
| | | Fixes `DataFlowOpts` leaking allocated temporary functions created to precompute an expression as their body. Reusing the `body` afterwards is fine since expressions are arena allocated separately.
* Harden exnref literals (#3092)Daniel Wirtz2020-09-025-18/+19
| | | | | * Make `Literal::type` immutable to guarantee that we do not lose track of `Literal::exn` by changing the literal's type * Add an assert to guarantee that we don't create `exnref` literals without an `ExceptionPackage` (for now) * Enforce rvalue reference when creating an `exnref` Literal from a `std::unique_ptr<ExceptionPackage>`, avoiding a redundant copy by means of requiring `std::move`
* Fix ExceptionPackage memory errors (#3088)Thomas Lively2020-09-013-9/+16
| | | | | | | | | | | First, adds an explicit destructor call to fix a memory leak in `Literal::operator=` in which existing `ExceptionPackage`s would be silently dropped. Next, changes `Literal::getExceptionPackage` to return the `ExceptionPackage` by value to avoid a use-after-free bug in the interpreter that was surfaced by the new destructor call. A future improvement would be to switch to using `std::variant`. Fixes #3087.
* Add Binaryen(Get|Set)AllowHeavyweight to binaryen-c.h (#3082)Max Graey2020-08-283-3/+11
| | | These declarations were previously missing causing the respective APIs to be not exposed. Also makes sure that a Boolean is returned by the JS API and adds a test to verify that it is working now.
* Add allowHeavyweight inlining option (#3032)Max Graey2020-08-265-1/+29
| | | | | As discussed in #2921, this allows inlining of functions not identified as "lightweight" (that include a loop, for example).
* Add new compound Rtt type (#3076)Daniel Wirtz2020-08-262-137/+311
| | | Extends compound types introduced in #3012 with a representation of `Rtt`s as described in the GC proposal, by also introducing the concept of a `HeapType` shared between `TypeInfo` and `Rtt`. Again, this should be a non-functional change since `Rtt`s are not used anywhere yet. Subtyping rules and updating the `xref` aliases is left for future work.
* wasm-emscripten-finalize: Add flags to limit dynCall creation (#3070)Sam Clegg2020-08-266-15/+63
| | | | | | Two new flags here, one to completely removes dynCalls, and another to limit them to only signatures that contains i64. See #3043
* SAFE_HEAP: remove fastcomp, prepare for new emscripten approach (#3078)Alon Zakai2020-08-251-15/+19
| | | | | | | | | | | | | | | In fastcomp we implemented emscripten_get_sbrk_ptr in wasm, and exported _emscripten_get_sbrk_ptr. We don't need that anymore and can remove it. However I want to switch us to implementing emscripten_get_sbrk_ptr in wasm in upstream too, as part of removing DYNAMICTOP_PTR and other silliness that we have around link (#3043). This makes us support an export of emscripten_get_sbrk_ptr (no prefix), and also it makes sure not to instrument that function, which may contain some memory operations itself, but if we SAFE_HEAP-ify them we'd get infinite recursion, as the SAFE_HEAP methods need to call that.
* Avoid wasted work when there are no locals (#3060)Max Graey2020-08-243-4/+11
| | | Adds early returns to local optimizations passes in cases where there are no locals to optimize.
* also drop size for memory.copy(x, x, y) (#3075)Max Graey2020-08-241-2/+5
| | | This fixes a bug in which a side effect in the calculation of the size could be lost.
* Add new compound Signature, Struct and Array types (#3012)Daniel Wirtz2020-08-2414-157/+587
| | | | | Extends the `Type` hash-consing infrastructure to handle type-parameterized and constructed types introduced in the typed function references and GC proposals. This should be a non-functional change since the new types are not used anywhere yet. Recursive type construction and canonicalization is also left as future work. Co-authored-by: Thomas Lively <tlively@google.com>
* memory.copy: use nop reductions only for ignoreImplicitTraps (#3074)Max Graey2020-08-241-3/+15
| | | | | | | | | According to changes in spec: WebAssembly/bulk-memory-operations#124 WebAssembly/bulk-memory-operations#145 we unfortunately can't fold to nop even for memory.copy(x, y, 0). So this PR revert all reductions to nop but do this only under ignoreImplicitTraps flag
* Remove optimization for memory.copy(x, x, C) (#3073)Max Graey2020-08-231-11/+1
| | | | | That can trap, so we can only remove it if traps are ignored, which was not handled properly. Revert it as we consider the options.
* OptimizeInstructions on memory.copy: check size for side effect as well (#3072)Max Graey2020-08-231-0/+2
| | | Fix issue found by fuzzer: #3038 (comment)
* Optimize bulk memory.copy (#3038)Max Graey2020-08-221-0/+68
| | | Replace it with a load and a store when the size is a small constant and remove it entirely when it would be a nop.
* Remove old EM_ASM handling method (#3069)Alon Zakai2020-08-212-63/+7
| | | | | | | The minimizeWasmChanges flag now does nothing (but new changes are coming, so keep it around) - this moves us to always doing the new way of things. With that we can update the tests. See #3043
* Use const modifier when dealing with types (#3064)Daniel Wirtz2020-08-2022-50/+50
| | | Since they make the code clearer and more self-documenting.
* Replace Type::expand() with an iterator-based approach (#3061)Daniel Wirtz2020-08-1926-183/+200
| | | This leads to simpler code and is a prerequisite for #3012, which makes it so that not all `Type`s are backed by vectors that `expand` could return.
* wasm-emscripten-finalize: Make EM_ASM modifications optional (#3044)Alon Zakai2020-08-193-13/+42
| | | | | | | | | | | | | | | | | | | | | | wasm-emscripten-finalize renames EM_ASM calls to have the signature in the name. This isn't actually useful - emscripten doesn't benefit from that. I think it was optimized in fastcomp, and in upstream we copied the general form but not the optimizations, and then EM_JS came along which is easier to optimize anyhow. This PR makes those changes optional: when not doing them, it just leaves the calls as they are. Emscripten will need some changes to handle that, but those are simple. For convenience this adds a flag to "minimize wasm changes". The idea is that this flag avoids needing a double-roll or other inconvenience as the changes need to happen in tandem on the emscripten side. The same flag can be reused for later changes similar to this one. When they are all done we can remove the flag. (Note how the code ifdefed by the flag can be removed once we no longer need the old way of doing things - that is, the new approach is simpler on the binaryen side). See #3043
* Remove asm2wasm (#3042)Alon Zakai2020-08-172-3589/+0
| | | | | | | Now that fastcomp has been removed from Emscripten, there is no need for the asm2wasm tool which it used to compile fastcomp's asm.js output to wasm. See emscripten-core/emscripten#11860
* Make wasm-emscripten-finalize's output optional (#3055)Alon Zakai2020-08-171-18/+22
| | | | | | | | | | This helps towards the goal of allowing emscripten to not always modify the wasm during link. Until now wasm-emscripten-finalize always wrote an output, while with this PR it only does so if it was asked to, either by giving it an output filename, or asking for text output. The only noticeable change from this should be to make what was an error before (not specify an output or ask for text) into a non-error (run and print metadata, but do not write the wasm).
* Prepare for compound types that are single but not basic (#3046)Daniel Wirtz2020-08-1731-188/+256
| | | | | | | | | | | | | | As a follow-up to https://github.com/WebAssembly/binaryen/pull/3012#pullrequestreview-459686171 this PR prepares for the new compound Signature, Struct and Array types that are single but not basic. This includes: * Renames `Type::getSingle` to `Type::getBasic` (NFC). Previously, its name was not representing its implementation (`isSingle` excluded `none` and `unreachable` while `getSingle` didn't, i.e. `getSingle` really was `getBasic`). Note that a hypothetical `Type::getSingle` cannot return `ValueType` anyway (new compound types are single but don't map to `ValueType`), so I figured it's best to skip implementing it until we actually need it. * Marks locations where we are (still) assuming that all single types are basic types, as suggested in https://github.com/WebAssembly/binaryen/pull/3012#discussion_r465356708, but using a macro, so we get useful errors once we start implementing the new types and can quickly traverse the affected locations. The macro is added where * there used to be a `switch (type.getSingle())` or similar that handled any basic type (NFC), but in the future will also have to handle single types that are not basic types. * we are not dealing with `Unary`, `Binary`, `Load`, `Store` or `AtomicXY` instructions, since these don't deal with compound types anyway.
* Add 64-bit hash_combine (#3041)Daniel Wirtz2020-08-161-1/+9
| | | Currently only the low 32-bits of a hash are guaranteed to be shuffled before combining with the other hash, so this PR also adds a 64-bit variant of hash_combine, including a comment on where the constants are coming from.
* Refactor hashing (#3023)Daniel Wirtz2020-08-1211-92/+86
| | | | | | | | | * Unifies internal hashing helpers to naturally integrate with std::hash * Removes the previous custom implementation * Computed hashes are now always size_t * Introduces a hash_combine helper * Fixes an overwritten partial hash in Relooper.cpp
* Added headers to CMake files (#3037)Wouter van Oortmerssen2020-08-107-0/+14
| | | This is needed for headers to show up in IDE projects, and has no other effect on the build.
* Fix typo in Asyncify comment (#3031)Nikita Baksalyar2020-08-101-1/+1
|
* StubUnsupportedJSOps: Remove CallIndirects (#3027)Alon Zakai2020-08-061-6/+23
| | | | | | wasm2js does not have full call_indirect support as we don't trap if the type is incorrect, which wasm does. Therefore the StubUnsupportedJSOps pass needs to remove those operations so that the fuzzer doesn't find spurious issues.
* Fix CountLeadingZeroes on MSVC (#3028)Alon Zakai2020-08-061-2/+5
| | | | | | | We just had the logic there wrong - MSVC's intrinsic returns the bit index, not the number of leading zeros. That's identical when scanning forward but not in reverse... Fixes #2942
* Asyncify verbose option (#3022)Alon Zakai2020-08-063-11/+58
| | | | | | | | | | | | | | | | This logs out the decisions made about instrumenting functions, which can help figure out why a function is instrumented, or to get a list of what might need to be. As the test shows, it can print things like this: [asyncify] import is an import that can change the state [asyncify] calls-import can change the state due to import [asyncify] calls-calls-import can change the state due to calls-import [asyncify] calls-calls-calls-import can change the state due to calls-calls-import (the test has calls-calls-calls-import => calls-calls-import => calls-import -> import).
* Refactor wasm-emscripten-finalize to use a single pass runner (#2987)Sam Clegg2020-08-051-47/+41
|
* Add StubUnsupportedJSOps to remove operations that JS does not support (#3024)Alon Zakai2020-08-054-1/+82
| | | | | | | | This doesn't lower them - it just replaces the unsupported operation with a drop. This will be useful for fuzzing, where to compare JS to the correct semantics we must avoid operations where JS is not always accurate. Also fully document the i64 -> f32 conversion issue in JS.
* Move generateDynCallThunks into its own pass. NFC. (#3000)Sam Clegg2020-08-047-18/+62
| | | | | | The core logic is still living in EmscriptenGlueGenerator because its used also by fixInvokeFunctionNames. As a followup we can figure out how to make these more independent.
* Refactor getMaxBits() out of OptimizeInstructions and add beginnings of unit ↵Alon Zakai2020-08-042-218/+236
| | | | | | | | | testing for it (#3019) getMaxBits just moves around, no logic is changed. Aside from adding getMaxBits, the change in bits.h is 99% whitespace. helps #2879
* Modernize binaryen.js glue code (#3005)Max Graey2020-08-041-1142/+1032
|
* Implement prototype v128.load{32,64}_zero instructions (#3011)Thomas Lively2020-08-0312-7/+76
| | | | | | | | Specified in https://github.com/WebAssembly/simd/pull/237. Since these are just prototypes necessary for benchmarking, this PR does not add support for these instructions to the fuzzer or the C or JS APIs. This PR also renumbers the QFMA instructions that previously used the opcodes for these new instructions. The renumbering matches the renumbering in V8 and LLVM.
* AlignmentLowering: Handle all possible cases for i64, f32, f64 (#3008)Alon Zakai2020-07-311-18/+160
| | | | | | | | | | Previously we only handled i32. That was enough for all real-world code people have run through wasm2js apparently (which is the only place the pass is needed - it lowers unaligned loads to individual loads etc., as unaligned operations fail in JS). Apparently it's pretty rare to have unaligned f32 loads for example. This will be useful in fuzzing wasm2js, as without this we can't compare results to the interpreter (which does alignment properly).
* New Dealign pass: reduce load/store alignment to 1 (#3010)Alon Zakai2020-07-314-0/+46
| | | | | Pretty trivial, but will be useful in wasm2js testing, where we can't assume an incorrectly-aligned load/store will still work, so we'll need to be pessimistic about alignment there.
* Better const fuzzing (#2972)Alon Zakai2020-07-301-14/+35
| | | | | | | | Tweak floating-point numbers with not just a +-1 integer, but also a float in [-1, 1]. Apply a tweak to powers of 2 as well. This found bugs in various codebases, see WebAssembly/spec#1224
* wasm2js: Add an "Export" scope for name resolution (#2998)Alon Zakai2020-07-301-33/+44
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Previously we used "Top" for both exports and the top level (which has functions and globals). The warning about name collisions there was meant only for exports (where if a name collides and so it must be renamed, means that there will be an externally-visible oddness for the user). But it applied to functions too, which could be annoying, and was not dangerous (at worst, it might be confusing when reading the emitted JS and seeing NAME_1, NAME_2, but there is no effect on execution or on exports). To fix this, add a new Export name scope. This separates function names from export names. However, it runs into another issue which is that when checking for a name conflict we had a big set of all the names in all the scopes. That is, FOO would only ever be used in one scope, period, and other appearances of that Name in wasm would get a suffix. As a result, if an exported function FOO has the name foo, we'd export it as FOO but name the function FOO_1 which is annoying. To fix that, keep sets of all names in each scope. When mangling a name we can then only care about the relevant scope, EXCEPT for local names, which must also not conflict with function names. That is, this would be bad: function foo(bar) { var bar = 0; } function bar() { .. It's not ok to call a parameter "bar" if there is a function by that name (well, it could be if it isn't called in that scope). So when mangling the Local scope, also check the Top one as well. The test output changes are due to non-overlapping scopes, specifically Local and Label. It's fine to have foo : while(1) { var foo = 5; } Those "foo"s do not conflict. Fixes emscripten-core/emscripten#11743
* Fix build for win32 (#3001)Max Graey2020-07-291-2/+2
| | | | | Check for x64 before using a non-32bit operation. See #2955 for context.
generated by cgit v1.2.3 (git 2.39.1) at 2025-07-03 01:14:05 +0000