sane-project-website/old-archive/1999-08/0230.html

<!-- received="Wed Aug 18 11:45:38 1999 PDT" -->
<!-- sent="Wed, 18 Aug 1999 20:04:56 +0200" -->
<!-- name="Andreas Beck" -->
<!-- email="becka@rz.uni-duesseldorf.de" -->
<!-- subject="Re: Mustek Paragon 1200SP" -->
<!-- id="" -->
<!-- inreplyto="Pine.LNX.4.05.9908181315260.31167-100000@obastrlx.combox.de" -->
<title>sane-devel: Re: Mustek Paragon 1200SP</title>
<h1>Re: Mustek Paragon 1200SP</h1>
<b>Andreas Beck</b> (<a href="mailto:becka@rz.uni-duesseldorf.de"><i>becka@rz.uni-duesseldorf.de</i></a>)<br>
<i>Wed, 18 Aug 1999 20:04:56 +0200</i>
<p>
<ul>
<li> <b>Messages sorted by:</b> <a href="date.html#230">[ date ]</a><a href="index.html#230">[ thread ]</a><a href="subject.html#230">[ subject ]</a><a href="author.html#230">[ author ]</a>
<!-- next="start" -->
<li> <b>Next message:</b> <a href="0231.html">Nick Lamb: "Re: SANE V2"</a>
<li> <b>Previous message:</b> <a href="0229.html">Nick Lamb: "Re: SANE V2"</a>
<!-- nextthread="start" -->
<!-- reply="end" -->
</ul>
<!-- body="start" -->
<i>&gt; &gt; In case that does not fit your problem, please have a look at sane-mustek(1),</i><br>
<i>&gt; &gt; which discribes two options for mustek.conf that might help you. Well actually</i><br>
<i>&gt; &gt; one of them might help. The lineart fix isn't really helpful for you.</i><br>
<i>&gt; If you could send me the fix, I would test it tonight....</i><br>
<p>
O.K. - here you go:<br>
<p>
--- mustek.c.orig	Fri Oct 30 07:56:12 1998<br>
+++ mustek.c	Tue Aug  3 21:27:28 1999<br>
@@ -1767,6 +1767,9 @@<br>
 	 the green buffer must be able to hold up to 1*dy lines. */<br>
       s-&gt;ld.buf[RED] = malloc (3 * dy * (long) bpc);<br>
       s-&gt;ld.buf[GRN] = s-&gt;ld.buf[RED] + 2 * dy * bpc;<br>
+      memset(s-&gt;ld.buf[RED], 0, 3 * dy * (long) bpc);<br>
+<br>
+      DBG(2, "fix_line_distance_mfs: malloc and clear. \n");<br>
     }<br>
 <br>
   /* restore the red and green lines from the previous buffer: */<br>
@@ -1833,19 +1836,19 @@<br>
   /* save red and green lines: */<br>
   for (y = 0; y &lt; 2*dy; ++y)<br>
     {<br>
-      if (num_lines - 2*dy + y &gt;= 0)<br>
+      if (num_lines + y &gt;= 2*dy)<br>
 	src = raw + (num_lines - 2*dy + y)*bpl;<br>
-      else<br>
+      else <br>
 	src = s-&gt;ld.buf[RED] + (y + num_lines)*bpc;<br>
-      memcpy (s-&gt;ld.buf[RED] + y*bpc, src, bpc);<br>
+      memmove (s-&gt;ld.buf[RED] + y*bpc, src, bpc);<br>
     }<br>
   for (y = 0; y &lt; 1*dy; ++y)<br>
     {<br>
-      if (num_lines - 1*dy + y &gt;= 0)<br>
+      if (num_lines + y &gt;= 1*dy)<br>
 	src = raw + (num_lines - 1*dy + y)*bpl + bpc;<br>
       else<br>
 	src = s-&gt;ld.buf[GRN] + (y + num_lines)*bpc;<br>
-      memcpy (s-&gt;ld.buf[GRN] + y*bpc, src, bpc);<br>
+      memmove (s-&gt;ld.buf[GRN] + y*bpc, src, bpc);<br>
     }<br>
 }<br>
 <br>
And NO: The patch is not a NOP. <br>
<p>
The memset is just more clean and doesn't leave you with garbage in the <br>
top few lines (which are wrong anyway due to the lack of correction data). <br>
<p>
The memcpy-&gt;memmove is not required and should probably not be applied for <br>
the final version. It's a relic from testing.<br>
<p>
The important one is the rephrasing of the arithmetics. Though both<br>
expressions are mathematically equivalent, they are not in C, as there is a<br>
subtle problem with automatic type promotion.<br>
<p>
CU, ANdy<br>
<p>
<pre>
-- 
= Andreas Beck                    |  Email :  &lt;<a href="mailto:andreas.beck@ggi-project.org">andreas.beck@ggi-project.org</a>&gt; =
<p>
<pre>
--
Source code, list archive, and docs: <a href="http://www.mostang.com/sane/">http://www.mostang.com/sane/</a>
To unsubscribe: echo unsubscribe sane-devel | mail <a href="mailto:majordomo@mostang.com">majordomo@mostang.com</a>
</pre>
<!-- body="end" -->
<p>
<ul>
<!-- next="start" -->
<li> <b>Next message:</b> <a href="0231.html">Nick Lamb: "Re: SANE V2"</a>
<li> <b>Previous message:</b> <a href="0229.html">Nick Lamb: "Re: SANE V2"</a>
<!-- nextthread="start" -->
<!-- reply="end" -->
</ul>