Local count too large

[fils]

Description

bug-report
In the example "showcase" and a few others (like "dashboard") I can not get them to run. Others like the custom_components work fine

Expected Results

to be able to load the page....

Actual Results

Uncaught (in promise) CompileError: AsyncCompilation: Compiling wasm function "wasm-function[42950]" failed: local count too large @+11247460

Context (Environment)

• Rust: rustc 1.35.0-nightly (26b4cb484 2019-03-09)

• yew: v0.7.0

• target:

• cargo-web: cargo-web 0.6.23

• browser if relevant
  I get the error in both Chrome and Firefox Dev Edition.
  Chrome: Version 72.0.3626.121 (Official Build) (64-bit) Linux
  FF: 66.0b14 (64-bit) Linux

Any help or guidance appreciated!
Any further info I can provide to help with this let me know!

[mokeyish]

I got this problem too.

[iammichiel]

Same here. Total guess but I think it would be some kind of recursion problem.
My code is not an example sample but I can single it out to a single line.

self.task = Some(self.fetch_service.fetch(get_request, handler));

The code is quite explicit and partially copied from https://github.com/DenisKolodin/yew/blob/a6a73948331ae50a9b6a196c721f21b4a39b52d1/examples/dashboard/src/lib.rs

Once the line is commented, every works fine/compiles.
Will try to dig a little further if I can discover something.

[nbrr]

See #455, at the moment the fix is to compile with --release.

[Th3Whit3Wolf]

It seems it would be beneficial to add this to the README in the root of this repo or in a README in the example directory that require this to function.

[matiu2]

For me it prints this in the console in the brave browser:

Uncaught (in promise) CompileError: WebAssembly.compile(): Compiling function #4181:"_ZN4http6header4name9parse_hdr17h241b60d2c73f294fE" failed: local count too large @+1056340

After some inspector cluseuing I think the fault lies with this function in the http library:

https://github.com/hyperium/http/blob/master/src/header/name.rs#L1061

It has a lot of heavy macro usage, and a big match statement.

[matiu2]

I ran it through: cargo expand < name.rs and it has a bunch of lines like this, which I expect will be the cause:

this:

if eq!(b == b'c' b'o' b'n' b't' b'e' b'n' b't' b'-' b's' b'e' b'c' b'u' b'r' b'i' b't' b'y' b'-' b'p' b'o' b'l' b'i' b'c' b'y' b'-' b'r' b'e' b'p' b'o' b'r' b't' b'-' b'o' b'n' b'l' b'y') { Ok(ContentSecurityPolicyReportOnly.into())
.. expands to something like this:

                    if b[0] == b'c'
                        && b[0 + 1] == b'o'
                        && b[0 + 1 + 1] == b'n'
                        && b[0 + 1 + 1 + 1] == b't'
                        && b[0 + 1 + 1 + 1 + 1] == b'e'
                        && b[0 + 1 + 1 + 1 + 1 + 1] == b'n'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1] == b't'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1] == b'-'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1] == b's'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1] == b'e'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1] == b'c'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1] == b'u'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1] == b'r'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1] == b'i'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1] == b't'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1] == b'y'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1]
                            == b'-'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1]
                            == b'p'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1]
                            == b'o'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1]
                            == b'l'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1]
                            == b'i'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1]
                            == b'c'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1]
                            == b'y'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1]
                            == b'-'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1]
                            == b'r'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1]
                            == b'e'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1]
                            == b'p'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1]
                            == b'o'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1]
                            == b'r'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1]
                            == b't'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1]
                            == b'-'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1]
                            == b'o'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1]
                            == b'n'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1]
                            == b'l'
                        && b[0 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1]
                            == b'y'
                    {
                        Ok(ContentSecurityPolicyReportOnly.into())

I guess that with optimisation turned off it doesn't pre-calculate all the addition and makes wasm do it.

[matiu2]

I have created a pull request with the http project that fixes this: hyperium/http#324

Until this all gets accepted and merged, you can use my temporary fork of yew in debug mode: https://github.com/matiu2/yew

[matiu2]

This landed in http trunk now. I guess once they do a new stable release, it'll be fixed for everyone here too: hyperium/http#324

[matiu2]

I can confirm this is working now since about 5 days ago, when http released https://github.com/hyperium/http/releases/tag/v0.1.18

To test, just run the dashboard example in non-release mode and download something through http.

This issue can be closed.

[jstarry]

Great, thanks @matiu2!