<html>
<head>
<base href="https://bugs.llvm.org/">
</head>
<body><table border="1" cellspacing="0" cellpadding="8">
<tr>
<th>Bug ID</th>
<td><a class="bz_bug_link
bz_status_NEW "
title="NEW - litsupport: support for test prefix"
href="https://bugs.llvm.org/show_bug.cgi?id=33160">33160</a>
</td>
</tr>
<tr>
<th>Summary</th>
<td>litsupport: support for test prefix
</td>
</tr>
<tr>
<th>Product</th>
<td>Test Suite
</td>
</tr>
<tr>
<th>Version</th>
<td>trunk
</td>
</tr>
<tr>
<th>Hardware</th>
<td>PC
</td>
</tr>
<tr>
<th>OS</th>
<td>All
</td>
</tr>
<tr>
<th>Status</th>
<td>NEW
</td>
</tr>
<tr>
<th>Severity</th>
<td>normal
</td>
</tr>
<tr>
<th>Priority</th>
<td>P
</td>
</tr>
<tr>
<th>Component</th>
<td>lit
</td>
</tr>
<tr>
<th>Assignee</th>
<td>unassignedbugs@nondot.org
</td>
</tr>
<tr>
<th>Reporter</th>
<td>llvm@meinersbur.de
</td>
</tr>
<tr>
<th>CC</th>
<td>daniel@zuster.org, llvm-bugs@lists.llvm.org
</td>
</tr></table>
<p>
<div>
<pre>When test-suite program use the prefix argument, such as
llvm_singlesource(PREFIX prefix)
the data collector in litsupport do not collect the correct data. Prefixes are
not required to end with a "." as the source seems to expect. Sometime a hyphen
was used, which unfortunately introduced some ambiguity as some tests include a
hyphen in their name. Since r286276, no separator is forced anymore.
Unfortunately during this change of PREFIX,
litsupport/compiletime.py
litsupport/stats.py
litsupport/timeit.py
have not been adapted to this change and therefore collect wrong results.
It seems to not be possible to query the PREFIX value from inside python. My
try to get correct values again is shown below (for stats.py). If feels like a
very hacky approach, so I don't submit it as a patch. I would be glad if
collecting data would work again without my local changes.
def _getStats(context):
# We compile multiple benchmarks in the same directory in SingleSource
# mode. Only look at compiletime files starting with the name of our test.
prefix = ""
if context.config.single_source:
prefix = "%s." % os.path.basename(context.executable)
stats = defaultdict(lambda: 0.0)
dir = os.path.dirname(context.test.getFilePath())
while True:
for path, subdirs, files in os.walk(dir):
for file in files:
if file.endswith('.stats') and file.startswith(prefix):
fullpath = os.path.join(path, file)
_mergeStats(stats, fullpath)
# It is possible that a prefix has been added to the executable; search
# again without the prefix. Because we don't know what the prefix is, we
# peel off one character after the other
if len(stats) == 0 and prefix:
prefix = prefix[1:]
continue
if len(stats) == 0:
print("No stats for
'{name}'".format(name=context.test.getFullName()))
logging.warning("No stats for '%s'", context.test.getFullName())
if context.config.single_source and len(stats) != 1:
print("Found {n} stats for single-source target
'{name}'".format(n=len(stats), name=context.test.getFullName()))
result = dict()
for key, value in stats.iteritems():
result[key] = lit.Test.toMetricValue(value)
return result</pre>
</div>
</p>
<hr>
<span>You are receiving this mail because:</span>
<ul>
<li>You are on the CC list for the bug.</li>
</ul>
</body>
</html>