[PATCH] D65680: [yaml2obj][tests] Fix overly restrictive od output check

[Fangrui Song via Phabricator via llvm-commits]

MaskRay added inline comments.


================
Comment at: test/tools/yaml2obj/elf-header-sh-fields.yaml:50
+# NEWSIZE: {{^[^[:blank:]]+}} 01
+# OLDSIZE: {{^[^[:blank:]]+}} 40
 
----------------
hubert.reinterpretcast wrote:
> grimar wrote:
> > Previously it tested that `e_shentsize` in ELF header at offset 0x72 contained value 64 (0x40) and with this YAML we changed it to 0x01.
> > But your version just test that somewhere in the file there are bytes 0x01 and 0x40 I think. Am I missing something?
> > But your version just test that somewhere in the file there are bytes 0x01 and 0x40 I think.
> My version tests that there is a byte whose value is 0x01 or 0x40 (for `NEWSIZE` and `OLDSIZE`, respectively) at offset 0x3a (octal 072).
I think it may be fine to just delete `{{^[^[:blank:]]+}}`. `-j 0x3a -N 1` implies there is only one byte.

@hubert.reinterpretcast What does `od -t x1 -v -j 0x3a -N 1 %t2 -A x` (note `-A x`) do on AIX?

If the address is still different, I would prefer `# OLDSIZE: 40` (I'm happy to change if @grimar has a different opinion)


Repository:
  rL LLVM

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D65680/new/

https://reviews.llvm.org/D65680