[webkit-reviews] review granted: [Bug 173105] WebAssembly: add support for stream APIs : [Attachment 418585] Patch

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Wed Jan 27 15:13:34 PST 2021

Keith Miller <keith_miller at apple.com> has granted Yusuke Suzuki
<ysuzuki at apple.com>'s request for review:
Bug 173105: WebAssembly: add support for stream APIs

Attachment 418585: Patch


--- Comment #51 from Keith Miller <keith_miller at apple.com> ---
Comment on attachment 418585
  --> https://bugs.webkit.org/attachment.cgi?id=418585

View in context: https://bugs.webkit.org/attachment.cgi?id=418585&action=review

r=me with some comments.

> Source/JavaScriptCore/wasm/WasmStreamingCompiler.cpp:89
> +	   ensureWorklist().enqueue(WTFMove(plan));

Wait, why does the ChangeLog say you don't use the WasmWorklist? is
`ensureWorklist()` not the Wasm one?

> Source/JavaScriptCore/wasm/WasmStreamingCompiler.cpp:139
> +    auto* promise = std::exchange(m_promise, nullptr);

Why is this needed? Isn't the promise retained by the DeferredWorkTimer?

> Source/JavaScriptCore/wasm/WasmStreamingPlan.h:53
> +    // For some reason friendship doesn't extend to parent classes...

Heh, I read this and thought to myself... This doesn't seem like something
Yusuke would right. Turns out I wrote it and you copied it here, lol

> Source/WebCore/bindings/js/JSDOMGlobalObject.cpp:319
> +    bool isResponseCorsSameOrigin = inputResponse->type() ==
ResourceResponse::Type::Basic || inputResponse->type() ==
ResourceResponse::Type::Cors || inputResponse->type() ==

I feel like this should be a member function on JSFetchResponse. It's a bit
magical here but wouldn't need a comment if the condition of the if was `if

> Source/WebCore/bindings/js/JSDOMGlobalObject.cpp:326
> +    auto contentType =

Ditto but less important.

> Source/WebCore/bindings/js/JSDOMGlobalObject.cpp:347
> +	   inputResponse->consumeBodyReceivedByChunk([globalObject, compiler =
WTFMove(compiler)](auto&& result) mutable {

Does this happen on the main thread out of curiosity or can it happen

More information about the webkit-reviews mailing list