benchmarks/JetStream2/wasm/TSF/tsf_internal.h (7 lines): - line 419: /* FIXME: open_remote() should attempt to connect just to see which - line 1092: /* FIXME: The remainder of this function could be an out-of-line slow path. */ - line 1136: /* FIXME: The remainder of this function could be an out-of-line slow path. */ - line 1170: /* FIXME: The remainder of this function could be an out-of-line slow path. */ - line 1267: /* FIXME: The remainder of this function could be an out-of-line slow path. */ - line 1320: /* FIXME: The remainder of this function could be an out-of-line slow path. */ - line 1367: /* FIXME: The remainder of this function could be an out-of-line slow path. */ benchmarks/wasm-misc/embenchen/src/run.py (5 lines): - line 31: # TODO: Annotate results with - and +, derived from - line 34: # TODO: catch any exception from the subprocess and print the log if - line 37: # TODO: Also check the output for other arguments than the default. - line 40: # TODO: In several cases below we'd like to check the entire output, - line 43: # TODO: We might like for the output not to contain any other lines than benchmarks/wasm-misc/embenchen/src/wasm_skinning.js (4 lines): - line 685: // TODO: use TextDecoder - line 1375: // TODO: In i64 mode 1, resign the two parts separately and safely - line 1565: // TODO: avoid this copy, by avoiding such static init writes - line 1566: // TODO: in shorter term, just copy up to the last static init write benchmarks/wasm-misc/embenchen/src/wasm_fannkuch.js (4 lines): - line 685: // TODO: use TextDecoder - line 1375: // TODO: In i64 mode 1, resign the two parts separately and safely - line 1565: // TODO: avoid this copy, by avoiding such static init writes - line 1566: // TODO: in shorter term, just copy up to the last static init write benchmarks/wasm-misc/embenchen/src/wasm_fasta.js (4 lines): - line 685: // TODO: use TextDecoder - line 1375: // TODO: In i64 mode 1, resign the two parts separately and safely - line 1565: // TODO: avoid this copy, by avoiding such static init writes - line 1566: // TODO: in shorter term, just copy up to the last static init write benchmarks/wasm-misc/embenchen/src/wasm_linpack_float.c.js (4 lines): - line 685: // TODO: use TextDecoder - line 1375: // TODO: In i64 mode 1, resign the two parts separately and safely - line 1565: // TODO: avoid this copy, by avoiding such static init writes - line 1566: // TODO: in shorter term, just copy up to the last static init write benchmarks/JetStream2/WSL/SpecWork/WSL.g4 (4 lines): - line 23: // TODO: what to do about floats that are too big or too small to represent? - line 24: // TODO: what is the default precision? double? - line 205: // TODO: Not sure at all how useful it is, I also still have to double check that it introduces no ambiguity. - line 212: // TODO: I tried to mimic https://en.cppreference.com/w/cpp/language/operator_precedence with regards to assignment and ternary conditionals, but it still needs some testing benchmarks/wasm-misc/embenchen/src/wasm_zlib.c.js (4 lines): - line 685: // TODO: use TextDecoder - line 1375: // TODO: In i64 mode 1, resign the two parts separately and safely - line 1565: // TODO: avoid this copy, by avoiding such static init writes - line 1566: // TODO: in shorter term, just copy up to the last static init write benchmarks/JetStream2/WSL/SPIRVCodegen.js (3 lines): - line 79: // FIXME: Implement this. - line 252: // FIXME: Figure out if we can use the Simple memory model instead of the GLSL memory model. - line 285: // FIXME: There are probably more annotations that are required than just location. benchmarks/JetStream2/WSL/ReturnChecker.js (3 lines): - line 142: // FIXME: This seems like it's missing things. For example, we need to be smart about this: - line 158: // FIXME: The fact that we need this demonstrates the need for CFG analysis. - line 183: // FIXME: This seems wrong. Consider a loop like: benchmarks/octane/crypto.js (2 lines): - line 1440: // TODO: allow reseeding after first request - line 1632: // TODO: re-calculate any missing CRT params benchmarks/JetStream2/WSL/Rewriter.js (2 lines): - line 27: // FIXME: This should have sensible behavior when it encounters definitions that it cannot handle. Right - line 283: // FIXME: This doesn't seem right. benchmarks/JetStream2/RexBench/OfflineAssembler/LowLevelInterpreter32_64.asm (2 lines): - line 1592: # FIXME: This could be massively optimized. - line 1845: muli sizeof SimpleJumpTable, t3 # FIXME: would be nice to peephole this! benchmarks/JetStream2/WSL/Lexer.js (2 lines): - line 119: // FIXME: Make this handle exp-form literals like 1e1. - line 123: // FIXME: Make this do Unicode. benchmarks/JetStream2/Octane/raytrace.js (2 lines): - line 619: /* TODO: dynamically include other scripts */ - line 759: /* TODO */ benchmarks/octane/raytrace.js (2 lines): - line 623: /* TODO: dynamically include other scripts */ - line 763: /* TODO */ benchmarks/JetStream2/Octane/crypto.js (2 lines): - line 1435: // TODO: allow reseeding after first request - line 1627: // TODO: re-calculate any missing CRT params benchmarks/JetStream2/WSL/SPIR-V.js (2 lines): - line 58: // FIXME: I'm not actually sure that Ids should be unsigned. - line 137: // FIXME: Handle parameters that require their own parameters benchmarks/JetStream2/ARES-6/Air/allocate_stack.js (2 lines): - line 188: // FIXME: https://bugs.webkit.org/show_bug.cgi?id=150454 - line 203: // FIXME: This may produce addresses that aren't valid if we end up with a ginormous stack frame. benchmarks/JetStream2/RexBench/OfflineAssembler/LowLevelInterpreter32_64.js (2 lines): - line 1598: # FIXME: This could be massively optimized. - line 1851: muli sizeof SimpleJumpTable, t3 # FIXME: would be nice to peephole this! benchmarks/JetStream2/wasm/TSF/gpc.h (2 lines): - line 864: * FIXME: this macro is wrong. the nh copy opcodes take one - line 1027: * FIXME: this really should not be a separate analysis. it should be just benchmarks/JetStream2/WSL/LateChecker.js (1 line): - line 39: // FIXME: Tighten these checks. For now, we should only accept int32, uint32, float32, and float64. benchmarks/JetStream2/WSL/Parse.js (1 line): - line 221: // FIXME: Need support for other literals too. benchmarks/JetStream2/wasm/TSF/tsf_primitive.c (1 line): - line 27: /* FIXME: primitive type should really just all be instances of each other. */ benchmarks/JetStream2/wasm/TSF/tsf_gpc_code_gen.c (1 line): - line 2915: /* FIXME: The caller should allocate the region for us. */ benchmarks/JetStream2/WSL/ResolveOverloadImpl.js (1 line): - line 57: // FIXME: It should be an error to declare a function that is at least as specific as a restricted function. benchmarks/JetStream2/WSL/Checker.js (1 line): - line 61: // FIXME: Relax these checks once we have implemented support for textures and samplers. benchmarks/JetStream2/wasm/TSF/tsf_define_helpers.h (1 line): - line 33: /* FIXME: add tsf_define helpers for dealing with primitive generators, benchmarks/JetStream2/wasm/TSF/tsf_fsdb.c (1 line): - line 572: fputs("TSF_FSDB",fout); /* FIXME: error checking? */ benchmarks/JetStream2/WSL/SPIRVTypeAnalyzer.js (1 line): - line 102: // FIXME: Using toString() in these functions is a hack. Instead, we should implement a proper type deduper. benchmarks/JetStream2/WSL/SPIRV.html (1 line): - line 9: benchmarks/JetStream2/wasm/TSF/tsf_serial_out_man.c (1 line): - line 32: /* FIXME: add an recovery helper mode, where after some number of bytes X benchmarks/JetStream2/RexBench/OfflineAssembler/LowLevelInterpreter64.js (1 line): - line 1829: muli sizeof SimpleJumpTable, t3 # FIXME: would be nice to peephole this! benchmarks/JetStream2/wasm/TSF/tsf_st.c (1 line): - line 6: /* FIXME: make this code check for out-of-memory errors and do the benchmarks/JetStream2/WSL/Test.js (1 line): - line 4932: // FIXME: This might become an error in future versions. benchmarks/JetStream2/RexBench/OfflineAssembler/LowLevelInterpreter64.asm (1 line): - line 1823: muli sizeof SimpleJumpTable, t3 # FIXME: would be nice to peephole this! benchmarks/JetStream2/WSL/Evaluator.js (1 line): - line 61: // FIXME: We should have a check that there is no way to drop out of a function without benchmarks/JetStream2/SunSpider/date-format-xparb.js (1 line): - line 420: // FIXME: Find a way to validate this test. benchmarks/JetStream2/worker/async-task.js (1 line): - line 182: // FIXME: Maybe we should be using the biased sample variance. benchmarks/JetStream2/wasm/TSF/tsf_stream_file_output.c (1 line): - line 89: /* FIXME: not sure if this is correct. even if we didn't create benchmarks/JetStream2/worker/bomb-subtests/string-validate-input.js (1 line): - line 91: // FIXME: Come up with a way of validating this test. benchmarks/JetStream2/ARES-6/Babylon/inspector-blob.js (1 line): - line 179: // FIXME: Need a better way to determine backend versions. Using DOM.pseudoElement because that was added after iOS 9. benchmarks/JetStream2/worker/bomb-subtests/date-format-tofte.js (1 line): - line 300: // FIXME: Find a way to validate this test. benchmarks/JetStream2/SunSpider/date-format-tofte.js (1 line): - line 301: // FIXME: Find a way to validate this test. benchmarks/JetStream2/worker/bomb-subtests/date-format-xparb.js (1 line): - line 419: // FIXME: Find a way to validate this test.