Cool, glad that was it. Btw, the --separate-asm option might help here, it emits the asm.js in another file. Another option, if speed is not that important, is to disable asm.js validation (-s ASM_JS=0).
On Wed, Aug 29, 2018 at 5:07 AM Joe Bibbo <[email protected]> wrote: > Alon, > > Brilliant! You are right. I was looking at the output after Webpack got a > hold of it. > > The output(.js) out of the emscripten tool chain produces clean > asmjs code. It is a problem with my packaging tools. > > Now, I just a need to figure out how to get Webpack to ignore the embedded > asmjs code when merging it. But obviously not an emscripten issue. > > Thanks for spotting my problem. > > Regards, > Joe > > > On Tuesday, August 28, 2018 at 6:21:47 PM UTC-5, Alon Zakai wrote: >> >> What options did you compile with? And did you run anything on the JS >> afterwards? >> >> That code sample looks a little like what closure compiler or another JS >> minifier might do. In emscripten, building with --closure 2 might lead to >> similar stuff - but we would disable asm.js validation in that case. If you >> run a minifier yourself on the output, without disabling that, then an >> error like this could occur. >> >> >> On Tue, Aug 28, 2018 at 2:22 PM Joe Bibbo <[email protected]> wrote: >> >>> >>> After successfully building and verifying the functionality of a large >>> C++ library, I was noticing that the code was running slower than expected. >>> The browser was putting up a warning about >>> the emscripten-generated asm.js "Invalid asm.js: Undefined global >>> variable" I tracked it down to the asm.js code, and any place that the >>> toolchain >>> had decided to use use a 'true' or 'false' constant. The browser would >>> complain ( chrome, firefox ). >>> >>> . >>> . >>> . >>> c[(F + 48 | 0) + 8 >> 2] = 674; >>> Kk(0, 0, 184863, F + 48 | 0); >>> Eb() >>> } >>> if (true ? (c[m + 8 >> 2] & 32 | 0) == 0 : 0) >>> if ((d[m + 4 >> 0] | 0) < 2) >>> a = 36; >>> else >>> a = 37; >>> else >>> a = 36; >>> >>> . >>> >>> >>> >>> >>> Building on mac, >>> emcc (Emscripten gcc/clang-like replacement) 1.38.5 (commit >>> a4474e59db658cea570c78254fa71119cf688db5) >>> >>> Any ideas on what I must do to avoid this type of generated code? >>> >>> Thanks in advance for any help or advice. >>> >>> Regards, >>> Joe >>> >>> -- >>> >>> -- > 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]. > For more options, visit https://groups.google.com/d/optout. > -- 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]. For more options, visit https://groups.google.com/d/optout.
