Re: [PATCH 3/6] refactored output function

From: Dave Jones <davej@redhat.com>
To: Ildar Muslukhov <ildarm@google.com>
Cc: trinity@vger.kernel.org
Subject: Re: [PATCH 3/6] refactored output function
Date: Thu, 10 Oct 2013 14:20:37 -0400	[thread overview]
Message-ID: <20131010182037.GA7939@redhat.com> (raw)
In-Reply-To: <1381267615-9826-3-git-send-email-ildarm@google.com>

Since applying this patch, coverity is picking up an out of bounds write.
It looks like it can't happen, but I'm wondering why this code was
done this way..

hand-editted diff for clarity:

 >  	/* copy buffer, sans ANSI codes */
 >  	len = strlen(outputbuf);
 > -	for (i = 0, j = 0; i < len; i++) {
 > +	for (i = 0, j = 0; (i < len) && (i + 2 < BUFSIZE) && (j < BUFSIZE); i++) {
 >  		if (outputbuf[i] == '^[') {
 >  			if (outputbuf[i + 2] == '1')
 >  				i += 6;	// ANSI_COLOUR
 >  			else
 >  				i += 3;	// ANSI_RESET
 >  		} else {
 >  			monobuf[j] = outputbuf[i];
 >  			j++;
 >    		}
 >  	}
 >  	monobuf[j] = '\0';

What's the intent behind this ?
It seems redundant, as everything seems to work fine without this change.

	Dave