More robust optimisation of `ignore`

Treat `ignore` as any other function during source-level optimisation,
to avoid having its warning-suppression effects cancelled by repeated
passes.  Instead, define a custom code generation function.

* lisp/emacs-lisp/byte-opt.el (byte-optimize-form-code-walker):
Don't treat `ignore' specially here.
(side-effect-free-fns): Don't mark `ignore` as side-effect-free
or error-free (although it is), since that would allow the optimiser
to elide calls.
* lisp/emacs-lisp/bytecomp.el (ignore, byte-compile-ignore):
Define and register a code-gen function.

1 files changed, 3 insertions, 8 deletions

diff --git a/lisp/emacs-lisp/byte-opt.el b/lisp/emacs-lisp/byte-opt.el
index 0c30d83f065..23c5a566cea 100644
--- a/lisp/emacs-lisp/byte-opt.el
+++ b/lisp/emacs-lisp/byte-opt.el
@@ -562,13 +562,6 @@ Same format as `byte-optimize--lexvars', with shared structure and contents.")
          `(catch ,(byte-optimize-form tag nil)
             . ,(byte-optimize-body exps for-effect))))
 
-      (`(ignore . ,exps)
-       ;; Don't treat the args to `ignore' as being
-       ;; computed for effect.  We want to avoid the warnings
-       ;; that might occur if they were treated that way.
-       ;; However, don't actually bother calling `ignore'.
-       `(progn ,@(mapcar #'byte-optimize-form exps) nil))
-
       ;; Needed as long as we run byte-optimize-form after cconv.
       (`(internal-make-closure . ,_)
        ;; Look up free vars and mark them to be kept, so that they
@@ -1419,7 +1412,9 @@ See Info node `(elisp) Integer Basics'."
 	 fixnump floatp following-char framep
 	 get-largest-window get-lru-window
 	 hash-table-p
-	 identity ignore integerp integer-or-marker-p interactive-p
+         ;; `ignore' isn't here because we don't want calls to it elided;
+         ;; see `byte-compile-ignore'.
+	 identity integerp integer-or-marker-p interactive-p
 	 invocation-directory invocation-name
 	 keymapp keywordp
 	 list listp