<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
</head>
<body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space;" class="">
Hello again,
<div class=""><br class="">
</div>
<div class="">thank you Clément and Eliot for your reactions on my simulation issues. For time reasons, I didn’t get it running yet, and try to build my demo without it for now. Here is a demo of my current work:</div>
<div class=""><br class="">
</div>
<div class=""><a href="https://www.youtube.com/watch?v=RN_2ggE3YQk" class="">https://www.youtube.com/watch?v=RN_2ggE3YQk</a></div>
<div class=""><br class="">
</div>
<div class="">I have built a debugger that allows us to debug multiple language levels within the same debugger. In this video-walkthrough,</div>
<div class="">I am in an application and debugging the issue that minSmallInt // -1 does return minSmallInt.[1] I step through the application, AST, and Smalltalk Bytecode. On the lower levels, I have a feature to debug the translators and finally fix the bug
in the JIT-Compiler right in the debugger.</div>
<div class=""><br class="">
</div>
<div class="">This is why I try to make the VMMaker run the Slang and JIT simulation: To get the result of a primitive with some given arguments and step through its execution. So far, I got the Slang up and running. What is missing to enable my debugger stepping
through the Bochs-Simulated JIT, is that I didn’t get the enilopmarts and stack setup right and wasn’t really sure where to fetch my evaluation result.</div>
<div class=""><br class="">
</div>
<div class="">In that process, I created some general wrapping and unwrapping methods which I needed. I may try to clean up my changes to the VMMaker (e.g. general wrapping/unwrapping) to feed them back to the trunk, if desired.</div>
<div class=""><br class="">
</div>
<div class="">I also experimented with using this debugger to debug a Smalltalk-hosted interpreter of a LISP-dialect.[2] The walkthrough at the top is going to be the core of my master thesis. I will speak about it on SAC in April.[3]</div>
<div class=""><br class="">
</div>
<div class="">So guys, what do you think about this? Are you full of question marks or did the video make sense to you? Just drop a short email to me with your current thoughts. I really appreciate your feedback :)</div>
<div class=""><br class="">
</div>
<div class="">Best regards,</div>
<div class="">Basti</div>
<div class=""><br class="">
</div>
<div class="">[1] This was an actual bug in Cogit: <a href="https://github.com/OpenSmalltalk/opensmalltalk-vm/issues/6" class="">https://github.com/OpenSmalltalk/opensmalltalk-vm/issues/6</a></div>
<div class=""><span style="text-indent: -42.66666793823242px;" class=""><br class="">
</span></div>
<div class=""><span style="text-indent: -42.66666793823242px;" class="">[2] Kruck, B. et al. Multi-Level Debugging For Interpreter Developers (WIP-Paper at LaMod), Demo: </span><a href="https://www.youtube.com/watch?v=6KfQ6Vpe-hU&t=2m48s" style="text-indent: -42.66666793823242px;" class="">https://www.youtube.com/watch?v=6KfQ6Vpe-hU&t=2m48s</a><span style="text-indent: -42.66666793823242px;" class=""> </span>
<p style="margin-left:32pt;text-indent:-32.0pt" class="">[3] Kruck, B. et al. Crossing Abstraction Barriers When Debugging in Dynamic Languages (To appear in SAC17)</p>
</div>
</body>
</html>