[PATCH] D70961: [LLD][ELF] add support for PT_GNU_PROPERTY

[Peter Smith via Phabricator via llvm-commits]

peter.smith marked 5 inline comments as done.
peter.smith added a comment.

Thanks for the comments I'll upload a new version shortly.



================
Comment at: lld/test/ELF/aarch64-feature-bti.s:66
+# BTISO-NEXT:    10360: bti     c
+# BTISO-NEXT:    10364: stp     x16, x30, [sp, #-16]!
+# BTISO-NEXT:    10368: adrp    x16, #131072
----------------
MaskRay wrote:
> Recently I think we can probably keep the address for the first instruction of a basic block and omit addresses for the rest of instructions.
> 
> ```
> 10360: bti c
>        stp x16, x30, [sp, #-16]!
>        adrp    x16, #131072
> ```
> 
> Addresses of instructions that are neither the first instructions nor jump targets are not useful, and can make address updating annoying.
I think that is a good idea. I've updated the tests to display only the address of the first instruction. 


================
Comment at: lld/test/ELF/aarch64-pt-gnu-property.s:1
+# REQUIRES: aarch64
+# RUN: llvm-mc -filetype=obj -triple=aarch64-linux-gnu %s -o %t.o
----------------
MaskRay wrote:
> grimar wrote:
> > I think you can use x86 for this test.
> Eventually we may need both. I think we also need a similar x86-64-pt-gnu-property.s. 
I've changed to use x86. I think the benefit of having both x86 and aarch64 is marginal right now if we are just testing that PT_GNU_PROPERTY describes the range of .note.gnu.property. I can add a new one if preferred.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D70961/new/

https://reviews.llvm.org/D70961