Re: trying to make things a bit faster...

[Sven Wischnowsky <wischnow@informatik.hu-berlin.de>]

Peter Stephenson wrote:

> [ about p-mem-1 ]
> 
> I had this same idea.  I also found out it didn't make that much
> difference.  Actually, it wasn't quite the same:  it also skipped heaps
> which had less than a small amount of memory (16 or 32 bytes or something)
> left.  But it still didn't seem to make much difference.

But since it's really quite safe, why shouldn't we...

> [ about p-mem-2 ]
> 
> I only use the zsh memory routines on one machine, so I can't test this all
> that well.  (I'm still worried something nasty may be happening here with
> --enable-lfs, too, but in a compiler-dependent way.)

I've been testing most of these patches with some heavy uses of the
new completion system. In some cases it reduced the number of brk()s and 
sbrk()s from several hundred to less than fifty.
The sizes of the limits may be a bit high, but zsh can get really
hungry for memory.

> 
> > p-exec-1
> >   Several things:
> 
> I think we just have to put all this in and try it.

The ones in p-exec are all quite safe, btw.

> [ about p-exec-2 ]
> 
> I put in a DPUTS at the end of simplifyright() and it doesn't seem to get
> triggered for me either, so far.  I thought it was for really simple
> commands, like a single command given on the command line?  But they same
> to arrive with l->right null anyway.  There's no other case where
> simplification matters so much, so I suggest we just delete it.

Yes, I just didn't dare to do that without asking first.

> ...
> 
> It could be really ancient pfalstad stuff where he wanted to track how
> things we getting executed.  They're all cases where that seems to be the
> only possible use.

Again, yes. In the calling functions the modified structures are not
used again and when the trees were correctly duplicated, they are
never used again.

Bye
 Sven

diff -u os/exec.c Src/exec.c
--- os/exec.c	Tue Jun  8 15:23:06 1999
+++ Src/exec.c	Thu Jun 10 16:48:16 1999
@@ -717,7 +717,6 @@
 	/* Reset donetrap:  this ensures that a trap is only *
 	 * called once for each sublist that fails.          */
 	donetrap = 0;
-	simplifyright(list);
 	slist = list->left;
 
 	/* Loop through code followed by &&, ||, or end of sublist. */
diff -u os/text.c Src/text.c
--- os/text.c	Tue Jun  8 15:23:10 1999
+++ Src/text.c	Thu Jun 10 16:49:38 1999
@@ -169,7 +169,6 @@
 	    if (_List(n)->type & Z_DISOWN)
 		taddstr("|");
 	}
-	simplifyright(_List(n));
 	if (_List(n)->right) {
 	    if (tnewlins)
 		taddnl();
diff -u os/utils.c Src/utils.c
--- os/utils.c	Tue Jun  8 15:23:10 1999
+++ Src/utils.c	Thu Jun 10 16:48:07 1999
@@ -2075,30 +2075,6 @@
     return 0;
 }
 
-/* see if the right side of a list is trivial */
-
-/**/
-void
-simplifyright(List l)
-{
-    Cmd c;
-
-    /*---- when can this be useful? */
-
-    if (l == &dummy_list || !l->right)
-	return;
-    if (l->right->right || l->right->left->right ||
-	l->right->left->flags || l->right->left->left->right ||
-	l->left->flags)
-	return;
-    c = l->left->left->left;
-    if (c->type != SIMPLE || (c->args && nonempty(c->args)) ||
-	(c->redir && nonempty(c->redir)) || (c->vars && nonempty(c->vars)))
-	return;
-    l->right = NULL;
-    return;
-}
-
 /* the ztypes table */
 
 /**/

--
Sven Wischnowsky                         wischnow@informatik.hu-berlin.de