Please ignore last comment - now appears to be building fine on windows, macos and linux with sdk-incoming-64bit, I just had to rebuild everything from scratch my end.
On Fri, Dec 29, 2017 at 6:52 PM, Mark Sibly <[email protected]> wrote: > Ok, can't quite build with incoming 64bit on Windows, I get this error > after linking: > > em++ -s USE_SDL=2 -s TOTAL_MEMORY=268435456 -s > DISABLE_EXCEPTION_CATCHING=1 -O2 -s ASSERTIONS=2 --preload-file > "D:/dev/monkey2/modules/mojo3d/tests/ducks.buildv1.1. > 09/emscripten_debug/assets/@/assets" -s BINARYEN=1 -o > "D:/dev/monkey2/modules/mojo3d/tests/ducks.products/Emscripten/Ducks.js" > @tmp/lnkFiles1.txt > > warning: unresolved symbol: glDrawBuffer > warning: unresolved symbol: glReadBuffer > Failed opening 'D:/dev/monkey2/modules/mojo3d/tests/ducks.products/ > Emscripten/Ducks.js.mem' > > Traceback (most recent call last): > > File "D:\devtools\emsdk-portable-64bit\emscripten\incoming\\em++", line > 16, in <module> > > emcc.run() > > File "D:\devtools\emsdk-portable-64bit\emscripten\incoming\emcc.py", > line 1876, in run > > wasm_text_target, misc_temp_files, optimizer) > > File "D:\devtools\emsdk-portable-64bit\emscripten\incoming\emcc.py", > line 2348, in do_binaryen > > subprocess.check_call(cmd) > > File "D:\depot_tools\win_tools-2_7_6_bin\python\bin\lib\subprocess.py", > line 540, in check_call > > raise CalledProcessError(retcode, cmd) > > subprocess.CalledProcessError: Command '['C:\\Users\\The > Razorback\\.emscripten_ports\\binaryen\\binaryen-version_40\\bin\\asm2wasm', > 'D:/dev/monkey2/modules/mojo3d/tests/ducks.products/Emscripten/Ducks.temp.asm.js', > '--total-memory=268435456', '--trap-mode=allow', '-O2', > '--mem-init=D:/dev/monkey2/modules/mojo3d/tests/ducks. > products/Emscripten/Ducks.js.mem', '--mem-base=1024', '--wasm-only', > '-o', > 'D:/dev/monkey2/modules/mojo3d/tests/ducks.products/Emscripten/Ducks.wasm']' > returned non-zero exit status 1 > > Note: The glDrawBuffer and glReadBuffer warnings are kind of weird. These > are extensions in WebGL, and as far a I can work out, you just 'use them' > if they're in the GL_EXTENSIONS string without having to GetProcAddress > them or anything. No idea how this works, but it usually also works find in > wasm so I've just been ignoring these warnings. > > I can build on macos and linux with incoming 64 bit fine though. > > > > On Thu, Dec 28, 2017 at 10:15 PM, Mark Sibly <[email protected]> > wrote: > >> Yay, macos is finally going. >> >> I ended up having to install 'incoming', which meant a build from source >> which took a while but worked fine. >> >> As far as I can work out, all the precompiled sdks are borked due to the >> llvm-ar issue. I tried copying llvm-ar around between versions but gave up >> in the end. >> >> Incoming fixes the EXTRA_EXPORTED_RUNTIME_METHODS issue too, and of >> course has fixed TRAP_MODE and I assume lots else. Might be time to release >> a new precompiled SDK? >> >> On Thu, Dec 28, 2017 at 6:25 PM, Mark Sibly <[email protected]> >> wrote: >> >>> Is there perhaps another emsdk version with working llvm-ar? >>> >>> >>> On Thu, Dec 28, 2017 at 6:20 PM, Mark Sibly <[email protected]> >>> wrote: >>> >>>> Hi, >>>> >>>> Ok I can confirm macos doesn't like: >>>> >>>> EXTRA_EXPORTED_RUNTIME_METHODS="['Pointer_stringify']" >>>> >>>> is has to be: >>>> >>>> "EXTRA_EXPORTED_RUNTIME_METHODS=['Pointer_stringify']" >>>> >>>> Next problem is I seem to have hit this issue: >>>> >>>> https://github.com/kripken/emscripten/issues/5418 >>>> >>>> The fix seems to be to use 'homebrew' llvm-ar instead. >>>> >>>> Any alternative to this? >>>> >>>> >>>> On Thu, Dec 28, 2017 at 5:25 PM, Mark Sibly <[email protected]> >>>> wrote: >>>> >>>>> Cool bananas. >>>>> >>>>> Am gonna have another go at installing emscripten on macos, have never >>>>> had much luck with that in the past... >>>>> >>>>> >>>>> On Thu, Dec 28, 2017 at 5:24 PM, Alon Zakai <[email protected]> >>>>> wrote: >>>>> >>>>>> I think that's a quoting issue, which >>>>>> >>>>>> https://github.com/kripken/emscripten/pull/5992 >>>>>> >>>>>> gives a better error for, and adds documentation about. I merged that >>>>>> in now. >>>>>> >>>>>> On Wed, Dec 27, 2017 at 8:19 PM, Mark Sibly <[email protected]> >>>>>> wrote: >>>>>> >>>>>>> Well, this fixes it for me, but another user on mac is getting the >>>>>>> following error now when building: >>>>>>> >>>>>>> Traceback (most recent call last): >>>>>>> File “/Users/jesse/emsdk-portable/emscripten/1.37.27/em++”, line >>>>>>> 16, in <module> >>>>>>> emcc.run() >>>>>>> File “/Users/jesse/emsdk-portable/emscripten/1.37.27/emcc.py”, line >>>>>>> 912, in run >>>>>>> setattr(shared.Settings, key, eval(value)) >>>>>>> File “<string>”, line 1, in <module> >>>>>>> NameError: name ‘Pointer_stringify’ is not defined >>>>>>> >>>>>>> On Thu, Dec 28, 2017 at 2:45 PM, Mark Sibly <[email protected]> >>>>>>> wrote: >>>>>>> >>>>>>>> > It adds a link to the FAQ at the top there, and the entry it >>>>>>>> links to is more detailed. How does it look now? >>>>>>>> >>>>>>>> Much better! >>>>>>>> >>>>>>>> Will add -O3 for link too, thanks. >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> On Thu, Dec 28, 2017 at 1:57 PM, Alon Zakai <[email protected]> >>>>>>>> wrote: >>>>>>>> >>>>>>>>> Yeah, that's the right fix - sorry about this, we've made changes >>>>>>>>> there recently, and the docs should be better. I opened >>>>>>>>> >>>>>>>>> https://github.com/kripken/emscripten/pull/5994 >>>>>>>>> >>>>>>>>> now with some improvements, and pushed it live at >>>>>>>>> >>>>>>>>> http://kripken.github.io/emscripten-site/docs/api_reference/ >>>>>>>>> module.html >>>>>>>>> >>>>>>>>> It adds a link to the FAQ at the top there, and the entry it links >>>>>>>>> to is more detailed. How does it look now? >>>>>>>>> >>>>>>>>> About your flags, they look mostly good, but make sure you have >>>>>>>>> -O3 for link as well as compiling object files. >>>>>>>>> >>>>>>>>> Btw, another option you might want to try is >>>>>>>>> BINARYEN_IGNORE_IMPLICIT_TRAPS, which can help a little with >>>>>>>>> size/speed, but like TRAP_MODE it may not work on all code (so it's >>>>>>>>> off by >>>>>>>>> default). >>>>>>>>> >>>>>>>>> On Wed, Dec 27, 2017 at 4:43 PM, Mark Sibly <[email protected] >>>>>>>>> > wrote: >>>>>>>>> >>>>>>>>>> Yay, sort of fixed it, ended up adding this to my 'linker' >>>>>>>>>> options (where linker=ec++): >>>>>>>>>> >>>>>>>>>> -s EXTRA_EXPORTED_RUNTIME_METHODS=['Pointer_stringify'] >>>>>>>>>> >>>>>>>>>> Is this the 'right' fix? >>>>>>>>>> >>>>>>>>>> Actually, I have a bunch of build/link switches now that I'm not >>>>>>>>>> really sure about. For compiling source code I use: >>>>>>>>>> >>>>>>>>>> -std=c++11 -O3 -DNDEBUG -s USE_SDL=2 -s TOTAL_MEMORY=268435456 -s >>>>>>>>>> DISABLE_EXCEPTION_CATCHING=1 >>>>>>>>>> >>>>>>>>>> Are all these necessary or are some only needed for linking? For >>>>>>>>>> linking I use: >>>>>>>>>> >>>>>>>>>> -s USE_SDL=2 -s TOTAL_MEMORY=268435456 -s >>>>>>>>>> DISABLE_EXCEPTION_CATCHING=1 -s BINARYEN=1 -s >>>>>>>>>> BINARYEN_TRAP_MODE='allow' -s EXTRA_EXPORTED_RUNTIME_METHODS >>>>>>>>>> =['Pointer_stringify'] >>>>>>>>>> >>>>>>>>>> Ditto, are these all necessary? >>>>>>>>>> >>>>>>>>>> The goal here is to create the *fastest* code! >>>>>>>>>> >>>>>>>>>> On Thursday, December 28, 2017 at 1:32:43 PM UTC+13, Mark Sibly >>>>>>>>>> wrote: >>>>>>>>>>> >>>>>>>>>>> Hello, >>>>>>>>>>> >>>>>>>>>>> My wasm apps have started producing the following error with a >>>>>>>>>>> fresh install of emsdk 'latest' on Windows 10: >>>>>>>>>>> >>>>>>>>>>> > 'Pointer_stringify' was not exported. add it to >>>>>>>>>>> EXTRA_EXPORTED_RUNTIME_METHODS (see the FAQ) >>>>>>>>>>> >>>>>>>>>>> This error didn't used to happen before on a different install - >>>>>>>>>>> sorry but I can't remember the version. >>>>>>>>>>> >>>>>>>>>>> em++ -v gives me: >>>>>>>>>>> >>>>>>>>>>> emcc (Emscripten gcc/clang-like replacement + linker emulating >>>>>>>>>>> GNU ld) 1.37.26 >>>>>>>>>>> clang version 4.0.0 (emscripten 1.37.26 : 1.37.26) >>>>>>>>>>> Target: x86_64-pc-windows-msvc >>>>>>>>>>> Thread model: posix >>>>>>>>>>> InstalledDir: D:\devtools\emsdk-portable-64b >>>>>>>>>>> it\clang\e1.37.26_64bit >>>>>>>>>>> INFO:root:(Emscripten: Running sanity checks) >>>>>>>>>>> >>>>>>>>>>> I did find this page: >>>>>>>>>>> >>>>>>>>>>> http://kripken.github.io/emscripten-site/docs/api_reference/ >>>>>>>>>>> module.html?highlight=extra_exported_runtime_methods#overrid >>>>>>>>>>> ing-execution-environment >>>>>>>>>>> >>>>>>>>>>> Which mentions exporting functions using >>>>>>>>>>> EXTRA_EXPORTED_RUNTIME_METHODS but I have no idea what that means. >>>>>>>>>>> >>>>>>>>>>> I would greatly appreciate any help here. >>>>>>>>>>> >>>>>>>>>>> Bye, >>>>>>>>>>> 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]. >>>>>>>>>> For more options, visit https://groups.google.com/d/optout. >>>>>>>>>> >>>>>>>>> >>>>>>>>> -- >>>>>>>>> You received this message because you are subscribed to a topic in >>>>>>>>> the Google Groups "emscripten-discuss" group. >>>>>>>>> To unsubscribe from this topic, visit >>>>>>>>> https://groups.google.com/d/topic/emscripten-discuss/m-GobZu >>>>>>>>> 3Tm0/unsubscribe. >>>>>>>>> To unsubscribe from this group and all its topics, 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. >>>>>>> >>>>>> >>>>>> -- >>>>>> You received this message because you are subscribed to a topic in >>>>>> the Google Groups "emscripten-discuss" group. >>>>>> To unsubscribe from this topic, visit https://groups.google.com/d/to >>>>>> pic/emscripten-discuss/m-GobZu3Tm0/unsubscribe. >>>>>> To unsubscribe from this group and all its topics, 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.
