<html><head></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; "><br><div><div>On May 3, 2012, at 11:10 AM, Chandler Carruth wrote:</div><br class="Apple-interchange-newline"><blockquote type="cite"><div class="gmail_quote">On Thu, May 3, 2012 at 10:47 AM, Chad Rosier <span dir="ltr"><<a href="mailto:mcrosier@apple.com" target="_blank">mcrosier@apple.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div style="word-wrap:break-word"><div><div class="h5"><br><div><div>On May 3, 2012, at 10:37 AM, Chandler Carruth wrote:</div><br><blockquote type="cite"><div class="gmail_quote">On Thu, May 3, 2012 at 10:07 AM, Chad Rosier <span dir="ltr"><<a href="mailto:mcrosier@apple.com" target="_blank">mcrosier@apple.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Author: mcrosier<br>
Date: Thu May 3 12:07:55 2012<br>
New Revision: 156079<br>
<br>
URL: <a href="http://llvm.org/viewvc/llvm-project?rev=156079&view=rev" target="_blank">http://llvm.org/viewvc/llvm-project?rev=156079&view=rev</a><br>
Log:<br>
[driver crash diagnostics] Remove more flags that reference absolute paths that<br>
aren't necessary to reproduce the clang crash.<br></blockquote><div><br></div><div>Out of curiosity, is it easy to test this by triggering a crash in the test suite, and then FileCheck-ing the resulting repro stuff?</div>
</div>
</blockquote></div><br></div></div><div>In r155205 (<a href="http://llvm.org/viewvc/llvm-project?view=rev&revision=155205" target="_blank">http://llvm.org/viewvc/llvm-project?view=rev&revision=155205</a>) I added a simple hack to cause the compiler to crash on demand. We could add a test case to cause a crash, but what were checking is the contents of the generated script file, which has a non-deterministic name. You'd have to grep the name from the clang output and then FileCheck the output script. Doesn't sound too difficult, but since this is off the critical path I honestly haven't spent a lot of time trying to get that to work. I can do so if you (or whoever) really think it's necessary.</div>
</div></blockquote><div><br></div><div>It's not a big deal... I was more thinking if we're going to keep tweaking this, maybe we should get some testing coverage of it.</div></div></blockquote><div><br></div><div>I expect the noise to die down shortly, but test cases are never a bad thing.</div><br><blockquote type="cite"><div class="gmail_quote"><div><br></div><div>If you're interested in testing it, I would just add an environment variable much like the one you added to force the crash (maybe the same one?) that forces the output to a particular location.</div></div></blockquote><div><br></div><div>I like this idea.</div><br><blockquote type="cite"><div class="gmail_quote">
<div><br></div><div>Actually, now that I think about it, I have a very distinct use case for exactly such an environment variable... maybe I can write up a patch and send it your way?</div></div>
</blockquote></div><div><br></div><div>Sounds good to me. Thanks, Chandler.</div><div><br></div><div> Chad</div></body></html>