<?xml version="1.0" encoding="UTF-8" standalone="yes" ?>
<!DOCTYPE bugzilla SYSTEM "https://bugs.webkit.org/page.cgi?id=bugzilla.dtd">

<bugzilla version="5.0.4.1"
          urlbase="https://bugs.webkit.org/"
          
          maintainer="admin@webkit.org"
>

    <bug>
          <bug_id>82122</bug_id>
          
          <creation_ts>2012-03-24 01:43:05 -0700</creation_ts>
          <short_desc>[Chromium] LayoutTests: css3/selector3/xml tests are flaky / have wrong expectations on Snow Leopard</short_desc>
          <delta_ts>2013-04-09 17:05:48 -0700</delta_ts>
          <reporter_accessible>1</reporter_accessible>
          <cclist_accessible>1</cclist_accessible>
          <classification_id>1</classification_id>
          <classification>Unclassified</classification>
          <product>WebKit</product>
          <component>New Bugs</component>
          <version>528+ (Nightly build)</version>
          <rep_platform>Unspecified</rep_platform>
          <op_sys>Unspecified</op_sys>
          <bug_status>RESOLVED</bug_status>
          <resolution>WONTFIX</resolution>
          
          
          <bug_file_loc></bug_file_loc>
          <status_whiteboard></status_whiteboard>
          <keywords></keywords>
          <priority>P2</priority>
          <bug_severity>Normal</bug_severity>
          <target_milestone>---</target_milestone>
          
          
          <everconfirmed>1</everconfirmed>
          <reporter name="Pavel Feldman">pfeldman</reporter>
          <assigned_to name="Ojan Vafai">ojan</assigned_to>
          <cc>jchaffraix</cc>
    
    <cc>kbr</cc>
          

      

      

      

          <comment_sort_order>oldest_to_newest</comment_sort_order>  
          <long_desc isprivate="0" >
    <commentid>587047</commentid>
    <comment_count>0</comment_count>
    <who name="Pavel Feldman">pfeldman</who>
    <bug_when>2012-03-24 01:43:05 -0700</bug_when>
    <thetext>At first it seemed like the expectations dependency graph glitch, but I saw more and more = TEXT results as I was marking tests as SNOWLEOPARD = TEXT in the expectations. Could they actually be flaky?</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>587956</commentid>
    <comment_count>1</comment_count>
    <who name="Ojan Vafai">ojan</who>
    <bug_when>2012-03-26 11:45:39 -0700</bug_when>
    <thetext>Looks like ~120 of the css3/selectors3/xml tests are flaky. The text dump in some cases has two extra height==0, anonymous blocks that have empty RenderText&apos;s in them.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>703457</commentid>
    <comment_count>2</comment_count>
    <who name="Kenneth Russell">kbr</who>
    <bug_when>2012-08-23 15:25:59 -0700</bug_when>
    <thetext>These tests are failing in the same way on the Chromium 10.7 debug bot, e.g.:

http://build.chromium.org/p/chromium.webkit/builders/Webkit%20Mac10.7%20%28dbg%29/builds/71

Though interestingly not on the 10.7 release bot, e.g.:

http://build.chromium.org/p/chromium.webkit/builders/Webkit%20Mac10.7/builds/10973

Why would these differences in the render tree be occurring mainly in debug mode?

I&apos;m going to generalize these expected results from &quot;SNOWLEOPARD&quot; to &quot;MAC&quot;. Simply adding &quot;LION&quot; seems like the wrong approach since we might add a Mountain Lion bot soon.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>703472</commentid>
    <comment_count>3</comment_count>
    <who name="Kenneth Russell">kbr</who>
    <bug_when>2012-08-23 15:37:10 -0700</bug_when>
    <thetext>Committed r126487: &lt;http://trac.webkit.org/changeset/126487&gt;</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>703503</commentid>
    <comment_count>4</comment_count>
    <who name="Ojan Vafai">ojan</who>
    <bug_when>2012-08-23 16:06:32 -0700</bug_when>
    <thetext>Could we restrict these lines to DEBUG only? They don&apos;t seem to fail on release.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>703635</commentid>
    <comment_count>5</comment_count>
    <who name="Kenneth Russell">kbr</who>
    <bug_when>2012-08-23 18:34:10 -0700</bug_when>
    <thetext>Yes, we could definitely tighten up the test expectations. All of the tests fail only on the debug bots except the following two:

css3/selectors3/xml/css3-modsel-101.xml fails on 10.6 Release but passes (with occasional flake) on 10.6 (dbg)
css3/selectors3/xml/css3-modsel-124b.xml fails on 10.6 Release but apparently passes on 10.6 (dbg)

Let me try to update the expectations now.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>703679</commentid>
    <comment_count>6</comment_count>
    <who name="Kenneth Russell">kbr</who>
    <bug_when>2012-08-23 20:01:19 -0700</bug_when>
    <thetext>Committed r126529: &lt;http://trac.webkit.org/changeset/126529&gt;</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>703684</commentid>
    <comment_count>7</comment_count>
    <who name="Ojan Vafai">ojan</who>
    <bug_when>2012-08-23 20:30:21 -0700</bug_when>
    <thetext>Thanks! This way we at least have test coverage for these tests in release.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>872393</commentid>
    <comment_count>8</comment_count>
    <who name="Stephen Chenney">schenney</who>
    <bug_when>2013-04-09 17:05:48 -0700</bug_when>
    <thetext>Marked LayoutTest bugs, bugs with Chromium IDs, and some others as WontFix. Test failure bugs still are trackable via TestExpectations or disabled unit tests.</thetext>
  </long_desc>
      
      

    </bug>

</bugzilla>