v8: update to v10.7.193.13.

[dio]

Signed-off-by: Dhi Aurrahman dio@rockybars.com

[PiotrSikora]

Those changes look good, but could you take a look at those failing tests? Both are timeouts, but it's possible that those are real regressions, so it would be good to test locally and compare against the current version of V8. Thanks!

[dio]

Will do. But seems like the test also failed on master (on my local) :(. I need to check again when the time permits.

[dio]

Interesting, it passed here: https://github.com/dio/proxy-wasm-cpp-host/actions/runs/3179609270/jobs/5182308733 (stripped out some jobs, only on v8, Linux, x86_64). Hopefully here it will be fine too 🙏🏽

[dio]

buildifier@latest is crashing? https://github.com/proxy-wasm/proxy-wasm-cpp-host/actions/runs/3188934515/jobs/5226316912#step:5:8

[PiotrSikora]

buildifier@latest is crashing? https://github.com/proxy-wasm/proxy-wasm-cpp-host/actions/runs/3188934515/jobs/5226316912#step:5:8

It's not crashing, it errors out because test/BUILD is not formatted correctly.

[dio]

It's not crashing, it errors out because test/BUILD is not formatted correctly.

Ah, sorry I didn't check that line has mode=check :(.

[PiotrSikora]

Thanks!

[PiotrSikora]

@dio sorry, it looks that I've messed this up by using the "update" button to merge with master and GitHub somehow decided to use my old @google.com email for the commit, even though it's not my primary email address in settings.

I removed that merge commit, but could you merge with master yourself?

[dio]

@PiotrSikora no worries. Done. (I failed to check this PR earlier).

[dio]

@PiotrSikora @mathetake, sorry, seems like Wasmtime on macOS is flaky (?). Should I investigate or we can retry?

[PiotrSikora]

@PiotrSikora @mathetake, sorry, seems like Wasmtime on macOS is flaky (?). Should I investigate or we can retry?

Bazel on macOS on GitHub Actions is somehow flaky, it's unrelated to Wasmtime, since it's the toolchain detection that fails.

I'll re-kick that test once V8 on macOS tests finish, since you can only re-kick failed tests after all tests complete.

[PiotrSikora]

There was an error in V8 on macOS tests:

[ RUN      ] WasmEngines/TestVm.WasmMemoryLimit/v8
TRACE from integration: [host->vm] infinite_memory()


#
# Fatal error in external/v8/src/heap/array-buffer-sweeper.cc, line 249
# Debug check failed: old_.bytes_ >= bytes (12 vs. 539951104).
#
#
#
#FailureMessage Object: 0x7ffee8155fa0
==== C stack trace ===============================

    0   runtime_test                        0x000000010ba13cce v8::base::debug::StackTrace::StackTrace() + 30
    1   runtime_test                        0x000000010ba13d05 v8::base::debug::StackTrace::StackTrace() + 21
    2   runtime_test                        0x000000010ac1adfe v8::platform::(anonymous namespace)::PrintStackTrace() + 30
    3   runtime_test                        0x000000010b9d48d8 V8_Fatal(char const*, int, char const*, ...) + 504
    4   runtime_test                        0x000000010b9d40ec v8::base::(anonymous namespace)::DefaultDcheckHandler(char const*, int, char const*) + 44
    5   runtime_test                        0x000000010b9d4947 V8_Dcheck(char const*, int, char const*) + 39
    6   runtime_test                        0x00000001087ce665 v8::internal::ArrayBufferSweeper::Detach(v8::internal::JSArrayBuffer, v8::internal::ArrayBufferExtension*) + 325
    7   runtime_test                        0x0000000108930c6c v8::internal::Heap::DetachArrayBufferExtension(v8::internal::JSArrayBuffer, v8::internal::ArrayBufferExtension*) + 60
    8   runtime_test                        0x0000000108f2886f v8::internal::JSArrayBuffer::Detach(bool) + 191
    9   runtime_test                        0x0000000109c2b6b3 v8::internal::WasmMemoryObject::Grow(v8::internal::Isolate*, v8::internal::Handle<v8::internal::WasmMemoryObject>, unsigned int) + 2867
    10  runtime_test                        0x000000010[97](https://github.com/proxy-wasm/proxy-wasm-cpp-host/actions/runs/3205125311/jobs/5243869780#step:9:98)ed7fc v8::internal::__RT_impl_Runtime_WasmMemoryGrow(v8::internal::RuntimeArgumentsWithoutHandles, v8::internal::Isolate*) + 476
    11  runtime_test                        0x00000001097ed5f0 v8::internal::Runtime_WasmMemoryGrow(int, unsigned long*, v8::internal::Isolate*) + 224
    12  runtime_test                        0x0000000[107](https://github.com/proxy-wasm/proxy-wasm-cpp-host/actions/runs/3205125311/jobs/5243869780#step:9:108)fabfb9 Builtins_CEntry_Return1_DontSaveFPRegs_ArgvOnStack_NoBuiltinExit + 57
    13  runtime_test                        0x0000000[108](https://github.com/proxy-wasm/proxy-wasm-cpp-host/actions/runs/3205125311/jobs/5243869780#step:9:109)0384fe Builtins_WasmMemoryGrow + 62
================================================================================

I've re-kicked it to make sure it's not a flake, but it looks like a real issue... maybe v8_flags.wasm_max_mem_pages got broken on macOS?

[dio]

@PiotrSikora, yup I saw that will check again locally.