[Openembedded-architecture] Bringing clang to OE-Core

[Khem Raj]

> On Apr 7, 2016, at 1:13 PM, Richard Purdie <richard.purdie at linuxfoundation.org> wrote:
> 
> On Thu, 2016-04-07 at 10:25 -0700, Khem Raj wrote:
>> There has been steady demand for clang infrastructure in OE from
>> multiple users and developers. I have therefore opened a feature
>> enhancement
>> request, its being used primarily in SDKs by some users. Its in good
>> shape for most of architectures which are supported in OE-Core.
>> 
>> https://bugzilla.yoctoproject.org/show_bug.cgi?id=9417
>> 
>> to bring meta-clang into OE-core as additional compiler option.
>> 
>> It will be an add on and default will remain same with possibility to
>> make it default or compiler particular recipes with one or other
>> and eSDK should be made able to include both the compilers to co
>> -exist.
>> 
>> Thoughts/flames ?
> 
> Is this a request to merge the layer in, or add it to the testing
> matrix, or both?

Its for merging the layer in and if someone is seriously going to use it
I would expect them to provide testing coverage for their own good, but
on OE-Core using same testing framework

> 
> I ask since testing already takes around 8 hours per cycle and I'm
> doing all I can to try and reduce that and be more responsive to
> patches. We have requests to add musl for all arches, now clang for all
> arches and so on yet there is both a physical cost to this which the
> project needs to pick up and an increase in testing times.
> 
> I'm not against the layer merge, I do want to understand how people
> feel about it though. The testing matrix is a growing concern for me.
> How big is the demand for this?

I understand the demands, and thats why its a secondary toolchain and
not to interfere with existing ways.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 204 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <http://lists.openembedded.org/pipermail/openembedded-architecture/attachments/20160407/7a3eabfd/attachment-0002.sig>