[llvm] [AArch64] Add support for SHF_AARCH64_PURECODE ELF section flag (1/3) (PR #125687)

[James Henderson via llvm-commits]

=?utf-8?q?Csanád_Hajdú?= <csanad.hajdu at arm.com>,
=?utf-8?q?Csanád_Hajdú?= <csanad.hajdu at arm.com>,
=?utf-8?q?Csanád_Hajdú?= <csanad.hajdu at arm.com>,
=?utf-8?q?Csanád_Hajdú?= <csanad.hajdu at arm.com>
Message-ID:
In-Reply-To: <llvm.org/llvm/llvm-project/pull/125687 at github.com>


================
@@ -110,3 +110,15 @@ Symbols:
 # ARM-NEXT:  L (link order), O (extra OS processing required), G (group), T (TLS),
 # ARM-NEXT:  C (compressed), x (unknown), o (OS specific), E (exclude),
 # ARM-NEXT:  R (retain), y (purecode), p (processor specific)
+
+## For an EM_AARCH64 target we print "y" for the SHF_AARCH64_PURECODE section flag.
----------------
jh7370 wrote:

I think we can avoid some of the duplication with the ARM case here, since the only difference is the machine name. I'd do it like this:

1) Update the ARM comment to something like the following:
> For an EM_ARM or EM_AARCH64 target, we print "y" for the PURECODE section flag.

2) Move the AARCH64 RUN lines to immediately after the RUN lines for the ARM case (with a blank line to separate them).
3) Change the AARCH64 --check-prefix to be `--check-prefix=ARM`.
4) Optionally, change the prefix name to something more generic, but since AARCH64 is just 64-bit ARM, I personally think ARM is fine to use for both cases.
5) Drop the old AARCH64 checks, since they aren't needed anymore.

https://github.com/llvm/llvm-project/pull/125687