<?xml version="1.0" encoding="utf-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
	>
<channel>
	<title>Comments on: Codewarrior Command Line Compiler</title>
	<atom:link href="http://tewha.net/2004/12/codewarrior-command-line-compiler/feed/" rel="self" type="application/rss+xml" />
	<link>http://tewha.net/2004/12/codewarrior-command-line-compiler/</link>
	<description>A blog on technical stuff by Steven Fisher. Cat pictures not included.</description>
	<pubDate>Tue, 06 Jan 2009 09:24:09 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.7</generator>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
		<item>
		<title>By: Objects At Rest &#187; Blog Archive &#187; CodeWarrior&#8217;s command line build</title>
		<link>http://tewha.net/2004/12/codewarrior-command-line-compiler/comment-page-1/#comment-37</link>
		<dc:creator>Objects At Rest &#187; Blog Archive &#187; CodeWarrior&#8217;s command line build</dc:creator>
		<pubDate>Fri, 12 Aug 2005 16:59:05 +0000</pubDate>
		<guid isPermaLink="false">http://palmdev.objectsatrest.com/?p=9#comment-37</guid>
		<description>[...] So I finally worked up the courage to try to add scripted building to my project again. I originally wanted to use the command line tools, but that didn&#8217;t work. So I decided this time to try the IDE. [...]</description>
		<content:encoded><![CDATA[<p>[...] So I finally worked up the courage to try to add scripted building to my project again. I originally wanted to use the command line tools, but that didn&#8217;t work. So I decided this time to try the IDE. [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Steven Fisher</title>
		<link>http://tewha.net/2004/12/codewarrior-command-line-compiler/comment-page-1/#comment-6</link>
		<dc:creator>Steven Fisher</dc:creator>
		<pubDate>Thu, 03 Mar 2005 08:15:09 +0000</pubDate>
		<guid isPermaLink="false">http://palmdev.objectsatrest.com/?p=9#comment-6</guid>
		<description>You might be fair when you say it isn't completely unusable. However, in my view, when I can't trust a compiler to at least generate a warning when code might be invalid... well, trust is the first thing I require out of a compiler. I'd actually much rather have the crash.</description>
		<content:encoded><![CDATA[<p>You might be fair when you say it isn&#8217;t completely unusable. However, in my view, when I can&#8217;t trust a compiler to at least generate a warning when code might be invalid&#8230; well, trust is the first thing I require out of a compiler. I&#8217;d actually much rather have the crash.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Ben Combee</title>
		<link>http://tewha.net/2004/12/codewarrior-command-line-compiler/comment-page-1/#comment-2</link>
		<dc:creator>Ben Combee</dc:creator>
		<pubDate>Thu, 13 Jan 2005 05:22:11 +0000</pubDate>
		<guid isPermaLink="false">http://palmdev.objectsatrest.com/?p=9#comment-2</guid>
		<description>Stephen, I'm curious why you call the 9.2 68K code completely unusable.  IIRC, there were two 68K code gen bug fixes in 9.3:

&lt;blockquote&gt;A major bug affecting inline assembly in the 68K C/C++ compilers was
fixed.  In V9.0 through V9.2, some inline assembly opcodes would be
output incorrectly.

A code generation bug would cause comparisons to fail if they
immediately followed the load from memory of a data item of a
different size, such as reading a char into an int, then comparing
that int to 0.&lt;/blockquote&gt;

I'm really sorry that I messed up the error routines -- that shouldn't have happened, but we didn't have tests of building bad code with the command line tools, just code gen tests for correctness.

I have talked to MWRon about this bug, BTW, but I don't have access to the source or tools needed to fix this myself.  The problem was a bad define in one of the compiler's headers that made all the error message numbers off-by-lots, and it only affected the command line tools because those had error messages bound into the executable using a different mechanism than the plugin compiler.</description>
		<content:encoded><![CDATA[<p>Stephen, I&#8217;m curious why you call the 9.2 68K code completely unusable.  IIRC, there were two 68K code gen bug fixes in 9.3:</p>
<blockquote><p>A major bug affecting inline assembly in the 68K C/C++ compilers was<br />
fixed.  In V9.0 through V9.2, some inline assembly opcodes would be<br />
output incorrectly.</p>
<p>A code generation bug would cause comparisons to fail if they<br />
immediately followed the load from memory of a data item of a<br />
different size, such as reading a char into an int, then comparing<br />
that int to 0.</p></blockquote>
<p>I&#8217;m really sorry that I messed up the error routines &#8212; that shouldn&#8217;t have happened, but we didn&#8217;t have tests of building bad code with the command line tools, just code gen tests for correctness.</p>
<p>I have talked to MWRon about this bug, BTW, but I don&#8217;t have access to the source or tools needed to fix this myself.  The problem was a bad define in one of the compiler&#8217;s headers that made all the error message numbers off-by-lots, and it only affected the command line tools because those had error messages bound into the executable using a different mechanism than the plugin compiler.</p>
]]></content:encoded>
	</item>
</channel>
</rss>
