On Thu, Apr 23, 2020 at 3:10 PM キャロウ マーク <[email protected]> wrote:
> [ Sorry about the previous empty message. Finger slip.] > > Hi Alon, > > On Apr 23, 2020, at 11:09, Alon Zakai <[email protected]> wrote: > > It might be good to look at how it's used in the test suite, > > > https://github.com/emscripten-core/emscripten/blob/9fd8eb088fa6795502e9bd2a2c92ff33390ef49d/tests/test_core.py#L7286 > > > Thanks. I tried to find this by working back from the test link given in > the webidl_binder documentation but failed. > > > Looks like it just does python and then the path to the binder. I agree > that emsdk integration could be better, seems like right now you need to > give it the full path into the subdirectory of the emsdk where it exists. > > > I filed https://github.com/emscripten-core/emsdk/issues/478 about this. > > > I also agree those temp files shouldn't be emitted. I'm not sure offhand > why that happens, it could be due to the IDL parser library we use, which > we didn't write ourselves. Worth filing an issue, hopefully someone can > look into that. > > And https://github.com/emscripten-core/emscripten/issues/10986 about this. > > I also filed https://github.com/emscripten-core/emscripten/issues/10987 about > the very, very low signal to noise ratio in webidl_binder’s output when > reporting a syntax error. > > ... > > >> The binder is emitting the following code into the .cpp file: >> >> void array_bounds_check(const int array_size, const int array_idx) { >> if (array_idx < 0 || array_idx >= array_size) { >> EM_ASM({ >> throw 'Array index ' + $0 + ' out of bounds: [0,' + $1 + ')'; >> }, array_idx, array_size); >> } >> } >> >> emcc barfs on " '$' in identifier [-Wdollar-in-identifier-extension]”. >> WebIDL-Binder.html says nothing about either EM_ASM or >> ‘array_bounds_check’. And I don’t see any flag to emcc relevant to enabling >> ASM_JS. How do I fix this? >> >> > How can I stop this? the test runner doesn’t seem to set any special emcc > options relevant to this. > > I think we can fix that by avoiding an EM_ASM. Currently EM_ASMs need $'s in their params. Instead, we could rewrite that to use an EM_JS which avoids that. I opened a PR for one of them here: https://github.com/emscripten-core/emscripten/pull/10988 But there are others which are less trivial. Maybe if you have time you can take a look, a PR would be very welcome. Or, meanwhile you can disable that warning on that file maybe. > I could turn off warnings as errors so I’d just get noise but I do not > want to do that. At the moment this is blocking me from further progress. > > Regards > > -Mark > > > > > -- > You received this message because you are subscribed to the Google Groups > "emscripten-discuss" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To view this discussion on the web visit > https://groups.google.com/d/msgid/emscripten-discuss/5AE37AE3-6D21-4241-9E2F-8C573995E945%40callow.im > <https://groups.google.com/d/msgid/emscripten-discuss/5AE37AE3-6D21-4241-9E2F-8C573995E945%40callow.im?utm_medium=email&utm_source=footer> > . > -- You received this message because you are subscribed to the Google Groups "emscripten-discuss" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/emscripten-discuss/CAEX4NpQYBmxKeRZm1pBuo%2BJXBKkcLXdfrG%2BnO6QtEwp3M%3DpNjA%40mail.gmail.com.
