First issue:
emscripten/system/lib/libcxx contains a symbols file containing the symbols 
of the library. How was this generated?

On Monday, July 17, 2017 at 10:30:49 PM UTC+2, Dirk Vanden Boer wrote:
>
> I'll focus on the upgrade, I'm way too unfamiliar with the code base to 
> make proper decisions about how to split off libcxx.
>
> I've looked into upgrading libcxx. I found the build_libcxx function in 
> tools/system_libs.py, I suppose this gets called when invoking 
> ./embuilder.py build libcxx.
>
> So if I update the libcxx code and make sure "./embuilder.py build libcxx 
> libcxxabi" builds fine and the tests succeed, I can consider the upgrade 
> successful?
>
> On Sunday, July 16, 2017 at 7:02:04 AM UTC+2, Alon Zakai wrote:
>>
>> I think we can put this in a separate repo using the ports system, the 
>> same mechanism that binaryen is. Tthe binaryen code is in 
>> tools/ports/binaryen.py , pretty simple.
>>
>> I think the one complication we ran into last time we discussed this was 
>> that libc++ and other system libraries have headers that are needed to 
>> compile, so we'd need to fetch it even before linking. I don't remember if 
>> we decided anything about that.
>>
>>
>> On Fri, Jul 14, 2017 at 5:04 PM, 'Derek Schuff' via emscripten-discuss <
>> [email protected]> wrote:
>>
>>> We try to minimize them but I expect there are a few local changes (grep 
>>> for EMSCRIPTEN in system/lib/libcxx). And I'd love to move to an 
>>> out-of-tree repo, although currently we don't have the mechanism in place 
>>> for that, and I'd hesitate to make someone who just wants to update libcxx 
>>> invent one. (or do we? Or I guess we could just do what we do for fastcomp 
>>> and let whoever builds emscripten worry about that? Or just a git 
>>> submodule?) Did we ever get a github organization for emscripten? That 
>>> would be nice to keep the repos together.
>>>
>>>
>>> On Fri, Jul 14, 2017 at 3:08 PM Dirk Vanden Boer <[email protected]> 
>>> wrote:
>>>
>>>> I'm willing to spend some time on this. I'll first try to get current 
>>>> tests running. You use vanilla libcxx or do you have local changes?
>>>>
>>>>
>>>> On Friday, July 14, 2017 at 8:07:33 PM UTC+2, Alon Zakai wrote:
>>>>
>>>>> If someone has time to look into this, I'd be happy to help with 
>>>>> guidance. Mostly it's just updating the in-tree code and running the test 
>>>>> suite. (However, we should move it to an out-of-tree repo, next update 
>>>>> might be a good time for that. That shouldn't be much work either.)
>>>>>
>>>>> On Thu, Jul 13, 2017 at 2:23 PM, Dirk Vanden Boer <[email protected]> 
>>>>> wrote:
>>>>>
>>>>>> I'm currently experimenting with emscripten and webassembly. I'm 
>>>>>> compiling a recent project that uses c++17 std::variant.
>>>>>> The latest released version of libcxx has support for std::variant.
>>>>>>
>>>>>> Does anyone know when this new libcxx version is expected to be used 
>>>>>> in emscripten?
>>>>>>
>>>>>> Thanks
>>>>>>
>>>>>> -- 
>>>>>> 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.
>>>>
>>> -- 
>>> 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.

Reply via email to