<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <div class="moz-cite-prefix"><br>
      Hi Erik,<br>
      <br>
      You mean like this?<br>
      <br>
      <a moz-do-not-send="true"
        href="http://cr.openjdk.java.net/%7Erkennke/8171853/webrev.04.diff/">http://cr.openjdk.java.net/~rkennke/8171853/webrev.04.diff/</a><br>
      <br>
      Full webrev here:<br>
      <a moz-do-not-send="true"
        href="http://cr.openjdk.java.net/%7Erkennke/8171853/webrev.04/">http://cr.openjdk.java.net/~rkennke/8171853/webrev.04/</a><br>
      <br>
      Thanks,<br>
      Roman<br>
      <br>
      <span class="removed"></span><span class="removed"></span></div>
    <blockquote type="cite"
      cite="mid:872910c6-a17b-d3df-bc80-fa850b9738d9@oracle.com">Hello
      Roman,
      <br>
      <br>
      In hotspot.m4, I believe the check on line 328 (pre changes) is
      still relevant for just the zero case.
      <br>
      <br>
      Otherwise build changes look good to me.
      <br>
      <br>
      /Erik
      <br>
      <br>
      <br>
      On 2017-10-16 00:00, Roman Kennke wrote:
      <br>
      <blockquote type="cite">
        <br>
        Ok, I fixed all the comments you mentioned.
        <br>
        <br>
        Differential (against webrev.01):
        <br>
        <a class="moz-txt-link-freetext" href="http://cr.openjdk.java.net/~rkennke/8171853/webrev.03.diff/">http://cr.openjdk.java.net/~rkennke/8171853/webrev.03.diff/</a>
        <a class="moz-txt-link-rfc2396E" href="http://cr.openjdk.java.net/%7Erkennke/8171853/webrev.03.diff/"><http://cr.openjdk.java.net/%7Erkennke/8171853/webrev.03.diff/></a>
        <br>
        Full webrev:
        <br>
        <a class="moz-txt-link-freetext" href="http://cr.openjdk.java.net/~rkennke/8171853/webrev.03/">http://cr.openjdk.java.net/~rkennke/8171853/webrev.03/</a>
        <a class="moz-txt-link-rfc2396E" href="http://cr.openjdk.java.net/%7Erkennke/8171853/webrev.03/"><http://cr.openjdk.java.net/%7Erkennke/8171853/webrev.03/></a>
        <br>
        <br>
        Roman
        <br>
        <br>
        <blockquote type="cite">Just spotted this:
          <br>
          <br>
          ./hotspot/jtreg/compiler/whitebox/CompilerWhiteBoxTest.java:
          /** {@code CompLevel::CompLevel_full_optimization} -- C2 or
          Shark */
          <br>
          <br>
          David
          <br>
          <br>
          On 16/10/2017 7:25 AM, David Holmes wrote:
          <br>
          <blockquote type="cite">On 16/10/2017 7:01 AM, Roman Kennke
            wrote:
            <br>
            <blockquote type="cite">Hi David,
              <br>
              <br>
              thanks!
              <br>
              <br>
              I'm uploading a 2nd revision of the patch that excludes
              the generated-configure.sh part, and adds a smallish
              Zero-related fix.
              <br>
              <br>
              <a class="moz-txt-link-freetext" href="http://cr.openjdk.java.net/~rkennke/8171853/webrev.01/">http://cr.openjdk.java.net/~rkennke/8171853/webrev.01/</a>
              <a class="moz-txt-link-rfc2396E" href="http://cr.openjdk.java.net/%7Erkennke/8171853/webrev.01/"><http://cr.openjdk.java.net/%7Erkennke/8171853/webrev.01/></a>
              <br>
            </blockquote>
            <br>
            Can you point me to the exact change please as I don't want
            to re-examine it all. :)
            <br>
            <br>
            I'll pull this in and do a test build run internally.
            <br>
            <br>
            Thanks,
            <br>
            David
            <br>
            <br>
            <blockquote type="cite">Thanks, Roman
              <br>
              <br>
              <br>
              <blockquote type="cite">Hi Roman,
                <br>
                <br>
                The build changes must be reviewed on build-dev - now
                cc'd.
                <br>
                <br>
                Thanks,
                <br>
                David
                <br>
                <br>
                On 15/10/2017 8:41 AM, Roman Kennke wrote:
                <br>
                <blockquote type="cite">The JEP to remove the Shark
                  compiler has received exclusively positive feedback
                  (JDK-8189173) on zero-dev. So here comes the big patch
                  to remove it.
                  <br>
                  <br>
                  What I have done:
                  <br>
                  <br>
                  grep -i -R shark src
                  <br>
                  grep -i -R shark make
                  <br>
                  grep -i -R shark doc
                  <br>
                  grep -i -R shark doc
                  <br>
                  <br>
                  and purged any reference to shark. Almost everything
                  was straightforward.
                  <br>
                  <br>
                  The only things I wasn't really sure of:
                  <br>
                  <br>
                  - in globals.hpp, I re-arranged the KIND_* bits to
                  account for the gap that removing KIND_SHARK left. I
                  hope that's good?
                  <br>
                  - in relocInfo_zero.hpp I put a ShouldNotCallThis() in
                  pd_address_in_code(), I am not sure it is the right
                  thing to do. If not, what *would* be the right thing?
                  <br>
                  <br>
                  Then of course I did:
                  <br>
                  <br>
                  rm -rf src/hotspot/share/shark
                  <br>
                  <br>
                  I also went through the build machinery and removed
                  stuff related to Shark and LLVM libs.
                  <br>
                  <br>
                  Now the only references in the whole JDK tree to shark
                  is a 'Shark Bay' in a timezone file, and 'Wireshark'
                  in some tests ;-)
                  <br>
                  <br>
                  I tested by building a regular x86 JVM and running
                  JTREG tests. All looks fine.
                  <br>
                  <br>
                  - I could not build zero because it seems broken
                  because of the recent Atomic::* changes
                  <br>
                  - I could not test any of the other arches that seemed
                  to reference Shark (arm and sparc)
                  <br>
                  <br>
                  Here's the full webrev:
                  <br>
                  <br>
                  <a class="moz-txt-link-freetext" href="http://cr.openjdk.java.net/~rkennke/8171853/webrev.00/">http://cr.openjdk.java.net/~rkennke/8171853/webrev.00/</a>
<a class="moz-txt-link-rfc2396E" href="http://cr.openjdk.java.net/%7Erkennke/8171853/webrev.00/"><http://cr.openjdk.java.net/%7Erkennke/8171853/webrev.00/></a>
                  <br>
                  <br>
                  Can I get a review on this?
                  <br>
                  <br>
                  Thanks, Roman
                  <br>
                  <br>
                </blockquote>
              </blockquote>
              <br>
            </blockquote>
          </blockquote>
        </blockquote>
        <br>
      </blockquote>
      <br>
    </blockquote>
    <p><br>
    </p>
  </body>
</html>