llvm dev - Oct 2014 - [LLVMdev] [lld][ELF] obj2yaml vs normalized input files (similar to macho)

Hi,

After looking at the Normalized Readers in mach-o, I feel it would be 
nice if the Gnu flavor to adopt this design what mach-o did in addition 
to using obj2yaml.

Its so much easier to test in the context of lld with normalized 
readers, IMO.

Opinions ?

Thanks

Shankar Easwaran

-- 
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by the
Linux Foundation

On Mon, Oct 20, 2014 at 5:25 AM, Shankar Easwaran
<shankare at codeaurora.org> wrote:
> Hi,
>
> After looking at the Normalized Readers in mach-o, I feel it would be nice
> if the Gnu flavor to adopt this design what mach-o did in addition to using
> obj2yaml.
>
> Its so much easier to test in the context of lld with normalized readers,
> IMO.
>
> Opinions ?
>
> Thanks
>
> Shankar Easwaran
>
> --
> Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted
by
> the Linux Foundation
>
> _______________________________________________
> LLVM Developers mailing list
> LLVMdev at cs.uiuc.edu         http://llvm.cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev
There's a non-trivial private memory overhead with this approach. I
also don't see how it would make testing lld any easier. I would
rather we embed yaml2obj/obj2yaml into lld.

- Michael Spencer

On 10/20/2014 4:28 PM, Michael Spencer wrote:
> On Mon, Oct 20, 2014 at 5:25 AM, Shankar Easwaran
> <shankare at codeaurora.org> wrote:
>> Hi,
>>
>> After looking at the Normalized Readers in mach-o, I feel it would be
nice
>> if the Gnu flavor to adopt this design what mach-o did in addition to
using
>> obj2yaml.
>>
>> Its so much easier to test in the context of lld with normalized
readers,
>> IMO.
>>
>> Opinions ?
>>
>> Thanks
>>
>> Shankar Easwaran
>>
>> --
>> Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
hosted by
>> the Linux Foundation
>>
>> _______________________________________________
>> LLVM Developers mailing list
>> LLVMdev at cs.uiuc.edu         http://llvm.cs.uiuc.edu
>> http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev
> There's a non-trivial private memory overhead with this approach. I
> also don't see how it would make testing lld any easier. I would
> rather we embed yaml2obj/obj2yaml into lld.
You could just embed yaml in tests similar to what mach-o does and 
create in memory objects, this would make it much simpler right ?

yaml2obj/obj2yaml would need to be made as a library then if I 
understand you correctly, no ?

Shankar Easwaran

-- 
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by the
Linux Foundation