<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=us-ascii">
<meta name="Generator" content="Microsoft Exchange Server">
<!-- converted from text --><style><!-- .EmailQuote { margin-left: 1pt; padding-left: 4pt; border-left: #800000 2px solid; } --></style>
</head>
<body>
<meta content="text/html; charset=UTF-8">
<style type="text/css" style="">
<!--
p
{margin-top:0;
margin-bottom:0}
-->
</style>
<div dir="ltr">
<div id="x_divtagdefaultwrapper" style="font-size:12pt; color:#000000; background-color:#FFFFFF; font-family:Calibri,Arial,Helvetica,sans-serif">
<p>Hi,</p>
<p><br>
</p>
<p>The error message in log you linked to [1] indicates an assertion failure, but the failing assertion is one that the commit below *removed*. It's what you would see if somehow the bot picked up the change this commit made to file
<span>llvm/trunk/test/Transforms/Inline/inline-funclets.ll</span> without also picking up the change it made to file
<span>llvm/trunk/lib/Transforms/Utils/InlineFunction.cpp.</span></p>
<div><span><br>
</span></div>
<div><span>Please let me know if you think there's more to it than a bad sync.</span></div>
<div><span><br>
</span></div>
<div><span>Thanks,</span></div>
<div><span>-Joseph</span></div>
<div><span><br>
</span></div>
<div><span>[1] - <span id="x_ms-rterangepaste-start"></span><span>http://lab.llvm.org:8011/builders/clang-cmake-armv7-a15/builds/14952/steps/ninja%20check%201/logs/FAIL%3A%20LLVM%3A%3Ainline-funclets.ll</span><span id="x_ms-rterangepaste-end"></span></span></div>
<p><br>
</p>
</div>
<hr tabindex="-1" style="display:inline-block; width:98%">
<div id="x_divRplyFwdMsg" dir="ltr"><font face="Calibri, sans-serif" color="#000000" style="font-size:11pt"><b>From:</b> Renato Golin <renato.golin@linaro.org><br>
<b>Sent:</b> Monday, September 5, 2016 6:07:50 PM<br>
<b>To:</b> Joseph Tremoulet<br>
<b>Cc:</b> LLVM Commits<br>
<b>Subject:</b> Re: [llvm] r280610 - Fix inliner funclet unwind memoization</font>
<div> </div>
</div>
</div>
<font size="2"><span style="font-size:10pt;">
<div class="PlainText">On 4 September 2016 at 02:23, Joseph Tremoulet via llvm-commits<br>
<llvm-commits@lists.llvm.org> wrote:<br>
> Author: josepht<br>
> Date: Sat Sep 3 20:23:20 2016<br>
> New Revision: 280610<br>
><br>
> URL: <a href="https://na01.safelinks.protection.outlook.com/?url=http%3a%2f%2fllvm.org%2fviewvc%2fllvm-project%3frev%3d280610%26view%3drev&data=02%7c01%7cjotrem%40microsoft.com%7c2d506e4ebf484105f7ae08d3d5d91465%7c72f988bf86f141af91ab2d7cd011db47%7c1%7c0%7c636087100736429912&sdata=lwg2ZaaFR0hQMqf%2f1gS3UgCS9DmroNDNgh5ajJvEigU%3d">
https://na01.safelinks.protection.outlook.com/?url=http%3a%2f%2fllvm.org%2fviewvc%2fllvm-project%3frev%3d280610%26view%3drev&data=02%7c01%7cjotrem%40microsoft.com%7c2d506e4ebf484105f7ae08d3d5d91465%7c72f988bf86f141af91ab2d7cd011db47%7c1%7c0%7c636087100736429912&sdata=lwg2ZaaFR0hQMqf%2f1gS3UgCS9DmroNDNgh5ajJvEigU%3d</a><br>
> Log:<br>
> Fix inliner funclet unwind memoization<br>
><br>
> Summary:<br>
> The inliner may need to determine where a given funclet unwinds to,<br>
> and this determination may depend on other funclets throughout the<br>
> funclet tree. The code that performs this walk in getUnwindDestToken<br>
> memoizes results to avoid redundant computations. In the case that<br>
> a funclet's unwind destination is derived from its ancestor, there's<br>
> code to walk back down the tree from the ancestor updating the memo<br>
> map of its descendants to record the unwind destination. This change<br>
> fixes that code to account for the case that some descendant has a<br>
> different unwind destination, which can happen if that unwind dest<br>
> is a descendant of the EHPad being queried and thus didn't determine<br>
> its unwind destination.<br>
><br>
> Also update test inline-funclets.ll, which is supposed to cover such<br>
> scenarios, to include a case that fails an assertion without this fix<br>
> but passes with it.<br>
><br>
> Fixes PR29151.<br>
><br>
><br>
> Reviewers: majnemer<br>
><br>
> Subscribers: llvm-commits<br>
><br>
> Differential Revision: <a href="https://na01.safelinks.protection.outlook.com/?url=https%3a%2f%2freviews.llvm.org%2fD24117&data=02%7c01%7cjotrem%40microsoft.com%7c2d506e4ebf484105f7ae08d3d5d91465%7c72f988bf86f141af91ab2d7cd011db47%7c1%7c0%7c636087100736429912&sdata=RQAHVAy%2fu7hcIUS8W4y5PM6SfPRFaGVrGd7ukNVDwDE%3d">
https://na01.safelinks.protection.outlook.com/?url=https%3a%2f%2freviews.llvm.org%2fD24117&data=02%7c01%7cjotrem%40microsoft.com%7c2d506e4ebf484105f7ae08d3d5d91465%7c72f988bf86f141af91ab2d7cd011db47%7c1%7c0%7c636087100736429912&sdata=RQAHVAy%2fu7hcIUS8W4y5PM6SfPRFaGVrGd7ukNVDwDE%3d</a><br>
><br>
> Modified:<br>
> llvm/trunk/lib/Transforms/Utils/InlineFunction.cpp<br>
> llvm/trunk/test/Transforms/Inline/inline-funclets.ll<br>
<br>
Hi,<br>
<br>
This patch is very likely to have broken this test:<br>
<br>
<a href="https://na01.safelinks.protection.outlook.com/?url=http%3a%2f%2flab.llvm.org%3a8011%2fbuilders%2fclang-cmake-armv7-a15%2fbuilds%2f14952%2fsteps%2fninja%2520check%25201%2flogs%2fFAIL%253A%2520LLVM%253A%253Ainline-funclets.ll&data=02%7c01%7cjotrem%40microsoft.com%7c2d506e4ebf484105f7ae08d3d5d91465%7c72f988bf86f141af91ab2d7cd011db47%7c1%7c0%7c636087100736429912&sdata=l3zTvhBzJWs3rXXNCEb7jdmysIjWquUolDM1rZnzRoc%3d">https://na01.safelinks.protection.outlook.com/?url=http%3a%2f%2flab.llvm.org%3a8011%2fbuilders%2fclang-cmake-armv7-a15%2fbuilds%2f14952%2fsteps%2fninja%2520check%25201%2flogs%2fFAIL%253A%2520LLVM%253A%253Ainline-funclets.ll&data=02%7c01%7cjotrem%40microsoft.com%7c2d506e4ebf484105f7ae08d3d5d91465%7c72f988bf86f141af91ab2d7cd011db47%7c1%7c0%7c636087100736429912&sdata=l3zTvhBzJWs3rXXNCEb7jdmysIjWquUolDM1rZnzRoc%3d</a><br>
<br>
Unfortunately, our bots didn't detect because of a change in the<br>
builder which made all bots not report any errors.<br>
<br>
If you can't fix it quickly, please revert and we'll investigate that<br>
later, as right now, we're dealing with a large number of concurrent<br>
failures and we won't have time to investigate further.<br>
<br>
Sorry.<br>
--renato<br>
</div>
</span></font>
</body>
</html>