[PATCH] D29273: [ELF] - Added partial support for --emit-relocs (no --gc-section case)

[George Rimar via llvm-commits]

>> Index: test/ELF/emit-relocs.s
>> ===================================================================
>> --- test/ELF/emit-relocs.s
>> +++ test/ELF/emit-relocs.s
>> @@ -0,0 +1,98 @@
>> +# REQUIRES: x86
>> +# RUN: llvm-mc -filetype=obj -triple=x86_64-unknown-linux %s -o %t1.o
>> +# RUN: llvm-mc -filetype=obj -triple=x86_64-unknown-linux %p/Inputs/emit-relocs.s -o %t2.o
>
>You can write this test with one file by having relocations against weak
>and undefined symbols, no? Is there any advantage to using two?

Initially I used two files because supposed it worth to check/show that we can have at least 2 different .rela.text sections
and able to concatenate it correctly with fixup of values to output section.

For SHT_REL[A] sections the are currently 3 logical branches when them are met:
 -relocatable, common way and --emit-relocs. 
And that is the only testcase which has at least 2 RELA sections we have atm with -emit-relocs.

>Cheers,
>Rafael