[llvm] [llvm-objdump] Implement decoding auxiliary header for xcoff with llvm-objdump --private-headers (PR #105682)

[James Henderson via llvm-commits]

================
@@ -0,0 +1,128 @@
+## Test that `the llvm-objdump --private-headers` print out the auxiliary header of XCOFF object file.
+
+# RUN: yaml2obj %s -o %t1
+# RUN: llvm-objdump --private-headers %t1 | \
+# RUN:   FileCheck %s --check-prefix=CHECK32 --match-full-lines --strict-whitespace
+# RUN: cp %t1 %t1_err1
+# RUN: python -c "with open(r'%t1_err1', 'r+b') as input: input.seek(17); input.write(b'\x45')"
+# RUN: python -c "with open(r'%t1_err1', 'r+b') as input: input.seek(4); input.write(b'\x00')"
----------------
jh7370 wrote:

Is there a reason these two lines can't be folded together into one python statement? You may also wish to look at using split-file, to allow you to have a proper python script, rather than these inline snippets.

https://github.com/llvm/llvm-project/pull/105682