<?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>29696</bug_id>
          
          <creation_ts>2009-09-23 15:12:18 -0700</creation_ts>
          <short_desc>[CHROMIUM] Paths fail to render in SVG with large viewbox</short_desc>
          <delta_ts>2009-09-29 14:11:59 -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>SVG</component>
          <version>528+ (Nightly build)</version>
          <rep_platform>PC</rep_platform>
          <op_sys>Windows XP</op_sys>
          <bug_status>RESOLVED</bug_status>
          <resolution>FIXED</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="Stephen White">senorblanco</reporter>
          <assigned_to name="Stephen White">senorblanco</assigned_to>
          <cc>eric</cc>
    
    <cc>rolandsteiner</cc>
          

      

      

      

          <comment_sort_order>oldest_to_newest</comment_sort_order>  
          <long_desc isprivate="0" >
    <commentid>149518</commentid>
    <comment_count>0</comment_count>
    <who name="Stephen White">senorblanco</who>
    <bug_when>2009-09-23 15:12:18 -0700</bug_when>
    <thetext>SVG paths with large coordinate values (and correspondingly large viewbox) are not rendered in Chrome.

See http://crbug.com/21174.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>149519</commentid>
    <comment_count>1</comment_count>
      <attachid>40021</attachid>
    <who name="Stephen White">senorblanco</who>
    <bug_when>2009-09-23 15:13:39 -0700</bug_when>
    <thetext>Created attachment 40021
Fix for SVG/skia coordinate issue</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>149593</commentid>
    <comment_count>2</comment_count>
      <attachid>40021</attachid>
    <who name="Eric Seidel (no email)">eric</who>
    <bug_when>2009-09-23 17:51:51 -0700</bug_when>
    <thetext>Comment on attachment 40021
Fix for SVG/skia coordinate issue

I&apos;m just gonna trust you that this is safe.  The Skia seat-belt is so lame. :(  If you want this committed by the commit-queue you&apos;ll need to set commit-queue=?.  I can&apos;t remember if you&apos;re a committer or not.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>149744</commentid>
    <comment_count>3</comment_count>
    <who name="Stephen White">senorblanco</who>
    <bug_when>2009-09-24 06:50:37 -0700</bug_when>
    <thetext>(In reply to comment #2)
&gt; (From update of attachment 40021 [details])
&gt; I&apos;m just gonna trust you that this is safe.

I ran it by the layout trybots on all three platforms, if that makes you feel better.

&gt; The Skia seat-belt is so lame. :( 

The best solution for that (IMHO) is to write floating point (or even better, SSE2) versions of the pixel loops, so there are no fixed-point limitations to worry about.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>149822</commentid>
    <comment_count>4</comment_count>
      <attachid>40021</attachid>
    <who name="Eric Seidel (no email)">eric</who>
    <bug_when>2009-09-24 13:49:42 -0700</bug_when>
    <thetext>Comment on attachment 40021
Fix for SVG/skia coordinate issue

According to http://trac.webkit.org/browser/trunk/WebKitTools/Scripts/modules/committers.py Stephen is not a committer, but it looks like he was invited August 4th.  So I&apos;ll assume he just hasn&apos;t added himself to that list yet and leave this for him to commit.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>149843</commentid>
    <comment_count>5</comment_count>
    <who name="Stephen White">senorblanco</who>
    <bug_when>2009-09-24 14:18:53 -0700</bug_when>
    <thetext>(In reply to comment #4)
&gt; (From update of attachment 40021 [details])
&gt; According to
&gt; http://trac.webkit.org/browser/trunk/WebKitTools/Scripts/modules/committers.py
&gt; Stephen is not a committer, but it looks like he was invited August 4th.  So
&gt; I&apos;ll assume he just hasn&apos;t added himself to that list yet and leave this for
&gt; him to commit.

Actually, I did commit it, in the end.. is there something I should do so that appears here?</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>150268</commentid>
    <comment_count>6</comment_count>
    <who name="Eric Seidel (no email)">eric</who>
    <bug_when>2009-09-26 01:28:56 -0700</bug_when>
    <thetext>If you committed this bug, than you should update the bug with the revision number that you committed and close it.  Searching the commit logs, I don&apos;t see that this has been committed though.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>150447</commentid>
    <comment_count>7</comment_count>
    <who name="Roland Steiner">rolandsteiner</who>
    <bug_when>2009-09-27 19:24:27 -0700</bug_when>
    <thetext>(In reply to comment #3)
&gt; The best solution for that (IMHO) is to write floating point (or even better,
&gt; SSE2) versions of the pixel loops, so there are no fixed-point limitations to
&gt; worry about.

I agree that this is probably the best interim fix for this bug. 

However, it means that the paths that are not rendered due to my original patch are (potentially) rendered wrong instead. Until the fixed-point implementation in Skia is changed, there doesn&apos;t seem to be much that one could do about it, though.

Very minor nit: in currentPathInLocalCoordinates(): one can wait with copying the patch until the matrix inversion has run, in case that fails. Since that&apos;s an exceptional case however, there shouldn&apos;t be much real impact one way or the other.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>151046</commentid>
    <comment_count>8</comment_count>
    <who name="Eric Seidel (no email)">eric</who>
    <bug_when>2009-09-29 14:11:59 -0700</bug_when>
    <thetext>This was landed as r48726 and can be closed.</thetext>
  </long_desc>
      
          <attachment
              isobsolete="0"
              ispatch="1"
              isprivate="0"
          >
            <attachid>40021</attachid>
            <date>2009-09-23 15:13:39 -0700</date>
            <delta_ts>2009-09-24 13:49:42 -0700</delta_ts>
            <desc>Fix for SVG/skia coordinate issue</desc>
            <filename>dom_path_units_webkit.patch</filename>
            <type>text/plain</type>
            <size>2334</size>
            <attacher name="Stephen White">senorblanco</attacher>
            
              <data encoding="base64">SW5kZXg6IFdlYkNvcmUvQ2hhbmdlTG9nCj09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09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</data>
<flag name="review"
          id="21179"
          type_id="1"
          status="+"
          setter="eric"
    />
          </attachment>
      

    </bug>

</bugzilla>