[PATCH] D46462: [sanitizer] Allow Fuchsia symbolizer to be reused by Myriad RTEMS
Walter Lee via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Tue May 8 15:48:23 PDT 2018
waltl marked an inline comment as done.
waltl added a comment.
In https://reviews.llvm.org/D46462#1088561, @mcgrathr wrote:
> I'd prefer to keep all the markup string constants together. Moving them all to a header would be fine, but I think it should be a separate sanitizer_symbolizer_fuchsia.h so that non-Fuchsia builds aren't using symbolizer_fuchsia.h, which is about OS-specific stuff distinct from the symbolizer markup format. But nothing in this patch actually needs the constant to be visible outside sanitizer_symbolizer_fuchsia.cc, so I'd like to see some explanation of the need for that.
>
> I'd be very glad to see a more thorough refactoring to make the symbolizer markup an option independent of OS. I think renaming symbolizer_fuchsia* to symbolizer_markup* probably makes sense. Ideally we'd nicely support both the Fuchsia/RTEMS configs where we want to support only markup and not have OS dependencies for things like files and subprocesses the other symbolizer code has; and also make markup a runtime selectable option for symbolizer style for the other OS builds that today have several runtime selectable choices.
I've done the basic refactoring as suggested. Please take a look.
Repository:
rL LLVM
https://reviews.llvm.org/D46462
More information about the llvm-commits
mailing list