From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <zsh-workers-return-30184-mason-zsh=primenet.com.au@zsh.org>
Received: (qmail 3596 invoked by alias); 6 Feb 2012 16:53:00 -0000
Mailing-List: contact zsh-workers-help@zsh.org; run by ezmlm
Precedence: bulk
X-No-Archive: yes
List-Id: Zsh Workers List <zsh-workers.zsh.org>
List-Post: <mailto:zsh-workers@zsh.org>
List-Help: <mailto:zsh-workers-help@zsh.org>
X-Seq: 30184
Received: (qmail 26010 invoked from network); 6 Feb 2012 16:52:58 -0000
X-Spam-Checker-Version: SpamAssassin 3.3.2 (2011-06-06) on f.primenet.com.au
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 required=5.0 tests=BAYES_00,RCVD_IN_DNSWL_NONE
	autolearn=ham version=3.3.2
Received-SPF: none (ns1.primenet.com.au: domain at closedmail.com does not designate permitted sender hosts)
From: Bart Schaefer <schaefer@brasslantern.com>
Message-id: <120206085229.ZM9099@torch.brasslantern.com>
Date: Mon, 06 Feb 2012 08:52:29 -0800
In-reply-to: <120206082008.ZM8593@torch.brasslantern.com>
Comments: In reply to Bart Schaefer <schaefer@brasslantern.com>
 "Re: [PATCH] Use access instead of stat in hashdir" (Feb  6,  8:20am)
References: <20120206131302.GA46184@Xye>
	<120206082008.ZM8593@torch.brasslantern.com>
X-Mailer: OpenZMail Classic (0.9.2 24April2005)
To: Zsh workers <zsh-workers@zsh.org>
Subject: Re: [PATCH] Use access instead of stat in hashdir
MIME-version: 1.0
Content-type: text/plain; charset=us-ascii

On Feb 6,  8:20am, Bart Schaefer wrote:
>
> On Feb 6,  6:43pm, Raghavendra D Prabhu wrote:
> } 
> } I found that using access instead of two stat calls results in 
> } faster rehash when it is done. I came across this when I noticed 
> } too many stat calls while 'strace -c' 
> 
> }              if (unset(HASHEXECUTABLESONLY) ||
> } -                   (stat(pathbuf, &statbuf) == 0 &&
> } -                    S_ISREG(statbuf.st_mode) && (statbuf.st_mode & S_IXUGO)))
> } +                   !access(pathbuf,X_OK))
> }                  add = 1;
> }          }

In Src/exec.c the function iscom():

int
iscom(char *s)
{
    struct stat statbuf;
    char *us = unmeta(s);

    return (access(us, X_OK) == 0 && stat(us, &statbuf) >= 0 &&
	    S_ISREG(statbuf.st_mode));
}

So let's see if we can get some speed out of this:

Index: Src/hashtable.c
===================================================================
diff -c -r1.20 Src/hashtable.c
--- hashtable.c	1 Jun 2011 06:40:00 -0000	1.20
+++ hashtable.c	6 Feb 2012 16:49:24 -0000
@@ -663,8 +663,10 @@
 		 * This is the same test as for the glob qualifier for
 		 * executable plain files.
 		 */
-		if (stat(pathbuf, &statbuf) == 0 &&
-		    S_ISREG(statbuf.st_mode) && (statbuf.st_mode & S_IXUGO))
+		if (unset(HASHEXECUTABLESONLY) ||
+		    (access(pathbuf, X_OK) == 0 &&
+		     stat(pathbuf, &statbuf) == 0 &&
+		     S_ISREG(statbuf.st_mode) && (statbuf.st_mode & S_IXUGO)))
 		    add = 1;
 	    }
 	    if (add) {

The S_IXUGO test may now be redundant, but hardly worth skipping.