GNU bug report logs - #39709
28.0.50; [PATCH] substring and the byte compiler

Previous Next

To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 39709 in the body.
You can then email your comments to 39709 AT debbugs.gnu.org in the normal way.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: Mark Oteiza <mvoteiza <at> udel.edu>
To: bug-gnu-emacs <at> gnu.org
Subject: 28.0.50; [PATCH] substring and the byte compiler
Date: Fri, 21 Feb 2020 01:22:06 -0500

Hi,

I don't really understand the byte compiler, but I noticed in
disassembly that substring, despite having an opcode, is getting
call'd:

10      constant  substring
11      stack-ref 3
12      constant  0
13      stack-ref 7
15      call      3

when really I'd expect to see something like

57      stack-ref 2
58      stack-ref 7
60      min       
61      stack-ref 7
63      substring 

The only relevant change I found is bug#33807.  Enlightenment welcome.
Naive patch below.

diff --git a/lisp/emacs-lisp/bytecomp.el b/lisp/emacs-lisp/bytecomp.el
index fce5e4aed6..ec340cd451 100644
--- a/lisp/emacs-lisp/bytecomp.el
+++ b/lisp/emacs-lisp/bytecomp.el
@@ -3506,6 +3506,7 @@ byte-defop-compiler
 					(0-1 . byte-compile-zero-or-one-arg)
 					(1-2 . byte-compile-one-or-two-args)
 					(2-3 . byte-compile-two-or-three-args)
+					(1-3 . byte-compile-one-to-three-args)
 					)))
 			   compile-handler
 			   (intern (concat "byte-compile-"
@@ -3690,6 +3691,13 @@ byte-compile-two-or-three-args
 	  ((= len 4) (byte-compile-three-args form))
 	  (t (byte-compile-subr-wrong-args form "2-3")))))
 
+(defun byte-compile-one-to-three-args (form)
+  (let ((len (length form)))
+    (cond ((= len 2) (byte-compile-one-arg form))
+          ((= len 3) (byte-compile-two-args form))
+          ((= len 4) (byte-compile-three-args form))
+	  (t (byte-compile-subr-wrong-args form "1-3")))))
+
 (defun byte-compile-noop (_form)
   (byte-compile-constant nil))
 
diff --git a/lisp/emacs-lisp/byte-opt.el b/lisp/emacs-lisp/byte-opt.el
index fe0930c684..d3c84335ec 100644
--- a/lisp/emacs-lisp/byte-opt.el
+++ b/lisp/emacs-lisp/byte-opt.el
@@ -1511,11 +1511,11 @@ byte-compile-side-effect-and-error-free-ops
 (defconst byte-compile-side-effect-free-ops
   (nconc
    '(byte-varref byte-nth byte-memq byte-car byte-cdr byte-length byte-aref
-     byte-symbol-value byte-get byte-concat2 byte-concat3 byte-sub1 byte-add1
-     byte-eqlsign byte-gtr byte-lss byte-leq byte-geq byte-diff byte-negate
-     byte-plus byte-max byte-min byte-mult byte-char-after byte-char-syntax
-     byte-buffer-substring byte-string= byte-string< byte-nthcdr byte-elt
-     byte-member byte-assq byte-quo byte-rem)
+     byte-symbol-value byte-get byte-substring byte-concat2 byte-concat3
+     byte-sub1 byte-add1 byte-eqlsign byte-gtr byte-lss byte-leq byte-geq
+     byte-diff byte-negate byte-plus byte-max byte-min byte-mult byte-char-after
+     byte-char-syntax byte-buffer-substring byte-string= byte-string<
+     byte-nthcdr byte-elt byte-member byte-assq byte-quo byte-rem)
    byte-compile-side-effect-and-error-free-ops))
 
 ;; This crock is because of the way DEFVAR_BOOL variables work.

Message #8 received at 39709 <at> debbugs.gnu.org (full text, mbox):

From: Mattias Engdegård <mattiase <at> acm.org>
To: Mark Oteiza <mvoteiza <at> udel.edu>
Cc: 39709 <at> debbugs.gnu.org, Stefan Monnier <monnier <at> iro.umontreal.ca>
Subject: #39709: 28.0.50; [PATCH] substring and the byte compiler
Date: Fri, 21 Feb 2020 12:48:01 +0100

[Message part 1 (text/plain, inline)]

The missing 1-3 case seems to be a simple omission -- thanks for spotting it.

+(defun byte-compile-one-to-three-args (form)
+  (let ((len (length form)))
+    (cond ((= len 2) (byte-compile-one-arg form))
+          ((= len 3) (byte-compile-two-args form))
+          ((= len 4) (byte-compile-three-args form))
+	  (t (byte-compile-subr-wrong-args form "1-3")))))

Not sure how this would work. Don't you need to add the missing arguments?

Cleaned-up patch attached.
(Unless, of course, there is evidence that the byte op has never been emitted --- then we have an opportunity to drop it instead.)

[0001-Generate-substring-byte-op-bug-39709.patch (application/octet-stream, attachment)]

Message #13 received at 39709-done <at> debbugs.gnu.org (full text, mbox):

From: Mattias Engdegård <mattiase <at> acm.org>
To: Mark Oteiza <mvoteiza <at> udel.edu>
Cc: 39709-done <at> debbugs.gnu.org, Stefan Monnier <monnier <at> iro.umontreal.ca>
Subject: Re: #39709: 28.0.50; [PATCH] substring and the byte compiler
Date: Tue, 25 Feb 2020 16:42:28 +0100

No objection arrived -- pushed to master.

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.