<html>
  <head>
    <meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    Please see inline.<br>
    <br>
    Tao<br>
    <br>
    On 10/8/13 2:03 PM, Jon Masamitsu wrote:
    <blockquote cite="mid:52547332.10409@oracle.com" type="cite">
      <meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
      <pre>   product(double, CMSSmallSplitSurplusPercent, 1.10,                        \
<span class="removed">-          "CMS: the factor by which to inflate estimated demand of small"   \</span>
<span class="removed">-          " block sizes to prevent splitting to supply demand for smaller"  \</span>
<span class="removed">-          " blocks")                                                        \</span>
<span class="new">+          "CMS: the factor by which to inflate estimated demand of small "  \</span>
<span class="new">+          "block sizes to prevent splitting to supply demand for smaller"   \</span>
<span class="new">+          "blocks")

Second line of next text, space after "smaller".</span></pre>
    </blockquote>
    will do.<br>
    <blockquote cite="mid:52547332.10409@oracle.com" type="cite">
      <pre><span class="new">

</span>   product(uintx, CMS_SweepWeight, 75,                                       \
<span class="removed">-          "Percentage (0-100) used to weight the current sample when "      \</span>
<span class="new">+          "Percentage (0-100) used to weigh the current sample when "      

I think the original is correct.

</span></pre>
      weight<br>
      <span class="main-fl"><em>transitive verb</em></span>
      <div class="ld_on_collegiate">
        <p class="bottom_entry">: to put a weight on (something) to make
          it heavier or to keep it from moving</p>
      </div>
      <div class="sblk">
        <div class="snum">4</div>
        <div class="scnt"><span class="ssens"> <strong>:</strong>  to
            assign a statistical weight to<br>
            <br>
            <br>
            What do you think?<br>
            <br>
          </span></div>
      </div>
    </blockquote>
    Umm...you are right.<br>
    <blockquote cite="mid:52547332.10409@oracle.com" type="cite">
      <div class="sblk">
        <div class="scnt"><span class="ssens"> </span><br>
          <pre>   manageable(intx, CMSAbortablePrecleanWaitMillis, 100,                     \
<span class="removed">-          "(Temporary, subject to experimentation)"                         \</span>
<span class="removed">-          " Time that we sleep between iterations when not given"           \</span>
<span class="removed">-          " enough work per iteration")                                     \</span>
<span class="new">+          "(Temporary, subject to experimentation) "                        \</span>
<span class="new">+          "Time that we sleep between iterations when not given"            \</span>
<span class="new">+          "enough work per iteration") 

Second line of new text, space after "given"</span></pre>
        </div>
      </div>
    </blockquote>
    will do.<br>
    <blockquote cite="mid:52547332.10409@oracle.com" type="cite">
      <div class="sblk">
        <div class="scnt">
          <pre><span class="new">

 
</span>   notproduct(bool, PrintFlagsWithComments, false,                           \
<span class="removed">-         "Print all VM flags with default values and descriptions and exit")\</span>
<span class="new">+          "Print all VM flags with default values and descriptions and exit") \

Not sure on this.  Sometimes I don't see the right number of
blanks but do you want a separate line so  "\" line up?
</span></pre>
        </div>
      </div>
    </blockquote>
    Okay, doable.<br>
    <blockquote cite="mid:52547332.10409@oracle.com" type="cite">
      <div class="sblk">
        <div class="scnt">
          <pre><span class="new">
Thanks all.

Jon
</span></pre>
          <span class="ssens"><br>
          </span></div>
      </div>
      <div class="moz-cite-prefix">On 10/4/2013 5:08 PM, Tao Mao wrote:<br>
      </div>
      <blockquote cite="mid:524F5873.8050309@oracle.com" type="cite">Hi
        all, <br>
        <br>
        A new webrev responding to Jon's recent review <br>
        <a moz-do-not-send="true" class="moz-txt-link-freetext"
          href="http://cr.openjdk.java.net/%7Etamao/8010506/webrev.05/">http://cr.openjdk.java.net/~tamao/8010506/webrev.05/</a>
        <br>
        <br>
        Thanks. <br>
        Tao <br>
        <br>
        On 10/4/13 4:51 PM, Tao Mao wrote: <br>
        <blockquote type="cite">Forwarded Jon's review comments to the
          main thread. <br>
          <br>
          Tao <br>
          <br>
          On 10/4/13 2:53 PM, Jon Masamitsu wrote: <br>
          <blockquote type="cite">If my comments conflict with any other
            comments, ignore <br>
            might. <br>
            <br>
               product(bool, MaxFDLimit,
            true,                                           \ <br>
            -          "Bump the number of file descriptors to max in
            solaris.")         \ <br>
            +          "Bump the number of file descriptors to max in
            solaris")          \ <br>
            <br>
            I would prefer "maximum" instead of "max".  There are
            examples of both in the <br>
            file.  If we're standarizing on spelling, I prefer t he
            complete word. <br>
            <br>
            <br>
             492   product(bool, JavaMonitorsInStackTrace,
            true,                             \ <br>
             493           "Print info about Java monitor locks when the
            stacks are dumped") \ <br>
            <br>
            Spell out "info".  Should be "information". <br>
            <br>
            <br>
               product(bool, UseSerialGC,
            false,                                         \ <br>
            -          "Use the serial garbage
            collector")                               \ <br>
            +          "Use the Serial garbage collector") <br>
            <br>
            I think "serial" is more correct that "Serial". <br>
            <br>
            <br>
            <br>
            <br>
            I would have changed this <br>
            <br>
             584           "Shows oopmap generation") <br>
            <br>
            to <br>
            <br>
            <br>
             584           "Shows oop map generation") <br>
            <br>
            I don't know what the phase "Show OopMapGeneration" means. <br>
            <br>
            +          "Show message about safepoint synch") <br>
            Change "synch" to "synchronization"                        \
            <br>
            <br>
            <br>
            +          "directory) of the dump file (defaults to
            java_pid<pid>.hprof "   \ <br>
            <br>
            Why the extra space at the end of " ...
            java_pid<pid>.hprof " <br>
            <br>
            <br>
            <br>
            <br>
            <br>
            +          "Maximum compaction in the Parallel Old garbage
            collector for "   \ <br>
                       "a system GC") <br>
            <br>
            <br>
            "Maximum compaction" -> "Use maximum compaction" <br>
            <br>
            <br>
            <br>
            <br>
            +          "Always tenure objects in eden. (Parallel GC
            only)")              \ <br>
            <br>
            "Parallel GC" does not start a sentence so I would use <br>
            "parallel GC". <br>
            <br>
            <br>
            <br>
            +          "Use adaptive free Lists in the CMS
            generation")                  \ <br>
            <br>
            "Lists" -> "lists" <br>
            <br>
            -          "Verbose output for parallel
            GC.")                                \ <br>
            +          "Verbose output for parallel
            gc")                                 \ <br>
            <br>
            I've seen "parallel GC" before in the changes.  I <br>
            prefer "parallel GC" to "parallel gc" <br>
            <br>
            <br>
            +          "it is disabled because this currently conflicts
            with "           \ <br>
            +          "parallel card scanning under certain
            conditions.")               \ <br>
            <br>
            Did you intend the period after "conditions"? <br>
            <br>
            <br>
            +          "Minimum percentage (0-100) of the CMS
            incremental duty cycle "   \ <br>
            +          "when CMSIncrementalPacing is enable") <br>
            <br>
            prefer "used whenCMSIncrementalPacing is
            enable"                     \ <br>
            <br>
            <br>
            <br>
            <br>
            +          "Use Mark-Sweep-Compact algorithm at full
            collections") <br>
            <br>
            Prefer "mark-sweek-compact" to "Mark-Sweep-Compact" <br>
            <br>
            <br>
            <br>
            <br>
            +          "Whether we should simulate frequent marking
            stack / work queue " \ <br>
            +          "overflow") <br>
            <br>
            <br>
            "Whether we should simulate"  -> "Simulate" <br>
            More consistent with other changes. <br>
            <br>
            <br>
            <br>
               notproduct(uintx, CMSMarkStackOverflowInterval,
            1000,                     \ <br>
            -          "An `interval' counter that determines how
            frequently"            \ <br>
            -          " we simulate overflow; a smaller number
            increases frequency")    \ <br>
            +          "An \"interval\" counter that determines how
            frequently "         \ <br>
            +          "we simulate overflow; a smaller number increases
            frequency") <br>
            <br>
            "we simulate overflow" -> "to simulate overflow" <br>
            <br>
            <br>
               product(bool, CMSConcurrentMTEnabled,
            true,                               \ <br>
            -          "Whether multi-threaded concurrent work enabled
            (if ParNewGC)")   \ <br>
            +          "Whether multi-threaded concurrent work enabled
            "                 \ <br>
            +          "(effective only if ParNewGC)") <br>
            <br>
            "Whether multi-threaded concurrent work enabled" <br>
              -> <br>
            "Enablemulti-threaded concurrent work  " <br>
            <br>
            <br>
            <br>
                product(uintx, CMSPrecleanThreshold,
            1000,                                \ <br>
            -          "Don't re-iterate if #dirty cards less than
            this")                \ <br>
            +          "Do not re-iterate if number of dirty cards is
            less than this")   \ <br>
            <br>
            <br>
            "Do not re-iterate  ..." -> "Do not iterate again ..." <br>
            <br>
            <br>
               product(bool, UseAdaptiveSizePolicyWithSystemGC,
            false,                   \ <br>
            -          "Use statistics from System.GC for adaptive size
            policy")         \ <br>
            +          "Use statistics from System.gc() for adaptive
            size policy") <br>
            <br>
            "Use statistics" -> "Include statistics" <br>
            <br>
            <br>
               diagnostic(bool, VerifySilently,
            false,                                   \ <br>
            -          "Don't print print the verification
            progress")                    \ <br>
            +          "Do not print print the verification
            progress")                   \ <br>
            <br>
            "print print" -> "print" <br>
            <br>
            <br>
               product(bool, DisableExplicitGC,
            false,                                   \ <br>
            -          "Tells whether calling System.gc() does a full
            GC")               \ <br>
            +          "Tell whether calling System.gc() does a full
            GC")                \ <br>
            <br>
            "Tell whether calling" -> "Disable (ignore) calls to" <br>
            <br>
            <br>
            <br>
            -          "Print the time taken by each parallel old gc
            phase."             \ <br>
            -          "PrintGCDetails must also be
            enabled.")                           \ <br>
            +          "Print the time taken by each ParallelOldGC phase
            "               \ <br>
            +          "(PrintGCDetails must also be enabled)") <br>
            <br>
            I prefer the original with "gc" -> "GC" <br>
            <br>
            or <br>
            <br>
            "Print the time taken by each ParallelOldGC phase" <br>
             -> <br>
            "Print the time taken by each phase with UseParallelOldGC" <br>
            <br>
            <br>
               develop(bool, TraceParallelOldGCMarkingPhase,
            false,                      \ <br>
            -          "Trace parallel old gc marking
            phase")                            \ <br>
            +          "Trace ParallelOldGC marking phase") <br>
            <br>
            Prefer original with "gc" -> "GC" <br>
            <br>
            or <br>
            <br>
            "Trace parallel old gc marking phase" <br>
            <br>
            -> <br>
            "Trace marking phases with UseParallelOldGC" <br>
            <br>
            <br>
            Similarly with <br>
            <br>
               develop(bool, TraceParallelOldGCSummaryPhase,
            false,                      \ <br>
            -          "Trace parallel old gc summary
            phase")                            \ <br>
            +          "Trace ParallelOldGC summary
            phase")                              \ <br>
                                                                                        

            \ <br>
               develop(bool, TraceParallelOldGCCompactionPhase,
            false,                   \ <br>
            -          "Trace parallel old gc compaction
            phase")                         \ <br>
            +          "Trace ParallelOldGC compaction
            phase")                           \ <br>
                                                                                        

            \ <br>
               develop(bool, TraceParallelOldGCDensePrefix,
            false,                       \ <br>
            -          "Trace parallel old gc dense prefix
            computation")                 \ <br>
            +          "Trace ParallelOldGC dense prefix
            computation")                   \ <br>
            <br>
            <br>
            <br>
            <br>
            +          "Rotatingly use gclog files (for long running
            applications). "    \ <br>
            +          "It requires -Xloggc:<filename>") <br>
            <br>
            <br>
            "Rotatingly use gclog" -> "Rotate gclog" <br>
            <br>
            <br>
            <br>
               product(bool, UseCompiler,
            true,                                          \ <br>
            -          "use
            compilation")                                               
            \ <br>
            +          "Use compilation") <br>
            <br>
            "Use compilation" -> "Use Just-In-Time compilation" <br>
            <br>
            <br>
            <br>
            -          "how many entries we'll try to leave on the stack
            during "        \ <br>
            -          "parallel
            GC")                                                    \ <br>
            +          "Number of entries we will try to leave on the
            stack "            \ <br>
            +          "during parallel gc" <br>
            <br>
            I prefer "GC" to "gc". <br>
            <br>
            I will probably not have time to respond to any reply <br>
            you have until next week. <br>
            <br>
            <br>
            Jon </blockquote>
          <br>
          <br>
          On 10/3/13 7:58 PM, Jesper Wilhelmsson wrote: <br>
          <blockquote type="cite">Looks good! <br>
            Thanks for doing this! <br>
            /Jesper <br>
            <br>
            Tao Mao skrev 3/10/13 11:27 PM: <br>
            <blockquote type="cite">Hi all, <br>
              <br>
              Mikael V. and Jesper's points are taken. There's turned
              out to be a bunch of <br>
              changes. <br>
              <br>
              Here's the new webrev: <br>
              <a moz-do-not-send="true" class="moz-txt-link-freetext"
                href="http://cr.openjdk.java.net/%7Etamao/8010506/webrev.04/">http://cr.openjdk.java.net/~tamao/8010506/webrev.04/</a>
              <br>
              <br>
              CR: <br>
              <a moz-do-not-send="true" class="moz-txt-link-freetext"
                href="https://bugs.openjdk.java.net/browse/JDK-8010506">https://bugs.openjdk.java.net/browse/JDK-8010506</a>
              <br>
              <br>
              Please review it so that I can get it in before ZBB. <br>
              <br>
              Thanks. <br>
              Tao <br>
              <br>
              On 10/2/13 4:54 PM, Jesper Wilhelmsson wrote: <br>
              <blockquote type="cite">Hi Tao, <br>
                <br>
                I noticed that in a few places there is a space between
                the description string <br>
                and the right parenthesis. E.g. <br>
                <br>
                product(intx, SyncVerbose, 0, "(Unstable)" ) <br>
                <br>
                Search for '" )' and you will find a few. <br>
                <br>
                There is also a capital S in "LAB'S" in the description
                of <br>
                CMSParPromoteBlocksToClaim. <br>
                <br>
                <br>
                I didn't mention these before since I really want this
                to be pushed asap, but <br>
                if you decide to act on Mikael's comments about Print vs
                Prints maybe you can <br>
                fix these minor things as well? <br>
                <br>
                I don't think a new webrev is needed for these changes.
                <br>
                <br>
                Thanks, <br>
                /Jesper <br>
                <br>
                <br>
                Mikael Vidstedt skrev 3/10/13 1:40 AM: <br>
                <blockquote type="cite"> <br>
                  Tao, <br>
                  <br>
                  I sampled the changes and this is definitely a great
                  clean up! <br>
                  <br>
                  Question: Do you want to attack inconsistencies like
                  "Print" vs "Prints" etc. <br>
                  since you're changing so much of this anyway? <br>
                  <br>
                  For example I see that you changed this: <br>
                  <br>
                  product(bool, ProfileIntervals,
                  false,                                    \ <br>
                  -          "Prints profiles for each interval (see
                  ProfileIntervalsTicks)")  \ <br>
                  +          "Print profiles for each interval (see
                  ProfileIntervalsTicks)")   \ <br>
                  <br>
                  <br>
                  But there's still this: <br>
                  <br>
                      develop(bool, TraceClearedExceptions,
                  false,                              \ <br>
                              "Prints when an exception is forcibly
                  cleared")                   \ <br>
                  <br>
                  <br>
                  and this: <br>
                  <br>
                  3237   develop(intx, TraceBytecodesAt, <br>
                  0,                                        \ <br>
                  3238           "Traces bytecodes starting with
                  specified bytecode <br>
                  number")       \ <br>
                  <br>
                  3596   develop(bool, PerfTraceDataCreation, <br>
                  false,                               \ <br>
                  3597           "Trace creation of Performance Data <br>
                  Entries")                     \ <br>
                  <br>
                  Cheers, <br>
                  Mikael <br>
                  <br>
                  On 2013-10-01 18:00, Tao Mao wrote: <br>
                  <blockquote type="cite">Hi all, <br>
                    <br>
                    Jesper's suggestion is integrated. <br>
                    <br>
                    new webrev: "Frame" seems to be the best way to
                    review it. <br>
                    <br>
                    <a moz-do-not-send="true"
                      class="moz-txt-link-freetext"
                      href="http://cr.openjdk.java.net/%7Etamao/8010506/webrev.03/">http://cr.openjdk.java.net/~tamao/8010506/webrev.03/</a>
                    <br>
                    <br>
                    Please review it so that I can get it in before ZBB.
                    <br>
                    <br>
                    Thanks. <br>
                    Tao <br>
                    <br>
                    On 9/24/13 12:47 PM, Jesper Wilhelmsson wrote: <br>
                    <blockquote type="cite">Hi Tao, <br>
                      <br>
                      The change looks good. Thanks for fixing this! <br>
                      <br>
                      Purely esthetic's: <br>
                      Lines 1820, 1821, 2221 and 2308 in your new
                      globals.hpp needs their \ aligned. <br>
                      <br>
                      Since you are touching pretty much the whole file
                      there are an inconsistency <br>
                      that would be nice to get rid of once and for all.
                      The following flags have <br>
                      their descriptions end with a period although they
                      are only one sentence. <br>
                      Most descriptions do not have the period at the
                      end. <br>
                      <br>
                      TracePageSizes <br>
                      InlineMathNatives <br>
                      ShowSafepointMsgs <br>
                      MaxFDLimit <br>
                      PredictedLoadedClassCount <br>
                      ParallelOldDeadWoodLimiterMean <br>
                      ParallelOldDeadWoodLimiterStdDev <br>
                      ScavengeWithObjectsInToSpace <br>
                      UseParNewGC <br>
                      ParallelGCVerbose <br>
                      ParallelGCBufferWastePct <br>
                      ParallelGCRetainPLAB <br>
                      PLABWeight <br>
                      CMSParPromoteBlocksToClaim <br>
                      OldPLABWeight <br>
                      AlwaysPreTouch <br>
                      CMSExpAvgFactor <br>
                      CMS_FLSWeight <br>
                      CMS_FLSPadding <br>
                      CMS_SweepPadding <br>
                      CMSDumpAtPromotionFailure <br>
                      CMSPrintChunksInDump <br>
                      CMSPrintObjectsInDump <br>
                      CMSVerifyReturnedBytes <br>
                      ExecuteInternalVMTests <br>
                      ConcGCYieldTimeout <br>
                      PrintParallelOldGCPhaseTimes <br>
                      ObjectCountCutOffPercent <br>
                      AbortVMOnExceptionMessage <br>
                      MaxBCEAEstimateLevel <br>
                      MaxBCEAEstimateSize <br>
                      CodeCacheMinimumFreeSpace <br>
                      CodeCacheMinBlockLength <br>
                      ExitOnFullCodeCache <br>
                      Tier0InvokeNotifyFreqLog <br>
                      Tier2InvokeNotifyFreqLog <br>
                      Tier3InvokeNotifyFreqLog <br>
                      Tier0BackedgeNotifyFreqLog <br>
                      Tier2BackedgeNotifyFreqLog <br>
                      Tier3BackedgeNotifyFreqLog <br>
                      Tier3CompileThreshold <br>
                      Tier4CompileThreshold <br>
                      IncreaseFirstTierCompileThresholdAt <br>
                      UsePerfData <br>
                      PerfDataMemorySize <br>
                      DumpSharedSpaces <br>
                      <br>
                      <br>
                      Thanks! <br>
                      /Jesper <br>
                      <br>
                      <br>
                      Tao Mao skrev 27/6/13 7:05 PM: <br>
                      <blockquote type="cite">new webrev: "Frame" seems
                        to be the best way to review it. <br>
                        <br>
                        <a moz-do-not-send="true"
                          class="moz-txt-link-freetext"
                          href="http://cr.openjdk.java.net/%7Etamao/8010506/webrev.02/">http://cr.openjdk.java.net/~tamao/8010506/webrev.02/</a>
                        <br>
                        <br>
                        Suggestion are taken mostly from Thomas. <br>
                        <br>
                        Thanks. <br>
                        Tao <br>
                        <br>
                        On 5/14/13 2:46 PM, Jon Masamitsu wrote: <br>
                        <blockquote type="cite"> <br>
                          On 4/30/13 4:58 PM, Tao Mao wrote: <br>
                          <blockquote type="cite">new webrev <br>
                            <a moz-do-not-send="true"
                              class="moz-txt-link-freetext"
                              href="http://cr.openjdk.java.net/%7Etamao/8010506/webrev.01/">http://cr.openjdk.java.net/~tamao/8010506/webrev.01/</a>
                            <br>
                            <br>
                            It needs reviews. <br>
                          </blockquote>
                          <br>
                          Tao, <br>
                          <br>
                          Changes you made look good.   I did not see a
                          list of the <br>
                          suggested changes in the CR <br>
                          <br>
                          <a moz-do-not-send="true"
                            class="moz-txt-link-freetext"
                            href="https://jbs.oracle.com/bugs/browse/JDK-8010506">https://jbs.oracle.com/bugs/browse/JDK-8010506</a>
                          <br>
                          <br>
                          so don't have any comment about what you
                          decided not <br>
                          to change. <br>
                          <br>
                          <blockquote type="cite"> <br>
                            Changeset: <br>
                            Thank you for the thorough report of all
                            these typos/errros, Thomas. <br>
                            <br>
                            I changed most places according to your
                            suggestions (for the rest few, I <br>
                            did <br>
                            not change because it may slightly change
                            the meaning of the definition). <br>
                            Also, I went through the file to change some
                            similar common errors that <br>
                            occurred. <br>
                            <br>
                            Several issues: <br>
                            <br>
                            (1) parallel gc should be lower case or
                            upper case? <br>
                            As I know and, perhaps, someone told me, the
                            term "parallel gc" sometimes <br>
                            refers to a more generic terminology rather
                            than ParallelGC or <br>
                            ParallelOldGC, <br>
                            if it doesn't imply these two. Basically,
                            any GC algorithms that can be <br>
                            parallelized could fall into this category.
                            Right now, we probably refer <br>
                            "parallel gc" to ParallelGC or
                            ParallelOldGC. But, we reserve the right to
                            <br>
                            change it to something else. That's why
                            "fuzziness" exists here. I guess <br>
                            the <br>
                            same way to deal with "concurrent gc". So, I
                            didn't make these words upper <br>
                            cased. <br>
                            <br>
                            (2) Identification: <br>
                            The general rule is to follow the example
                            (product()'s identification). The <br>
                            whole file checked and space adjusted
                            accordingly. <br>
                            <br>
                            (3) Line breaks <br>
                            A space should be provided at the line
                            breaks if multiple lines is needed. <br>
                            Some lines that are slightly longer than
                            normal length is kept untouched as <br>
                            long as they read well. <br>
                            <br>
                            *(4) Why does the VM option hashCode have
                            lower case initial? <br>
                            The direct answer is "I don't know". (But it
                            is understandable that <br>
                            hashCode <br>
                            is "legacy" such that it has it own style)
                            Anyone knows the answer? <br>
                          </blockquote>
                          <br>
                          I don't know for sure either but might be
                          because of the Java code <br>
                          <br>
                          public int*hashCode*() <br>
                          <br>
                          <br>
                          Jon <br>
                          <br>
                          <blockquote type="cite"> <br>
                            Thanks. <br>
                            Tao <br>
                            <br>
                            On 3/21/13 3:09 PM, Tao Mao wrote: <br>
                            <blockquote type="cite">8010506 Typos and
                              errors in gc-related flags in globals.hpp
                              <br>
                              <a moz-do-not-send="true"
                                class="moz-txt-link-freetext"
                                href="https://jbs.oracle.com/bugs/browse/JDK-8010506">https://jbs.oracle.com/bugs/browse/JDK-8010506</a>
                              <br>
                              <br>
                              webrev: <br>
                              <a moz-do-not-send="true"
                                class="moz-txt-link-freetext"
                                href="http://cr.openjdk.java.net/%7Etamao/8010506/webrev.00/">http://cr.openjdk.java.net/~tamao/8010506/webrev.00/</a>
                              <br>
                              <br>
                              Changeset: <br>
                              NewRatio should defined as <br>
                              product(uintx, NewRatio, <br>
                              2,                                              

                              \ <br>
                                  "Ratio of old/new generation
                              sizes")                              \ <br>
                              instead of <br>
                              product(uintx, NewRatio, <br>
                              2,                                              

                              \ <br>
                                  "Ratio of new/old generation
                              sizes")                              \ <br>
                              <br>
                              PLUS other typos found by Jesper. <br>
                            </blockquote>
                          </blockquote>
                          <br>
                        </blockquote>
                      </blockquote>
                    </blockquote>
                  </blockquote>
                  <br>
                </blockquote>
              </blockquote>
            </blockquote>
          </blockquote>
        </blockquote>
      </blockquote>
      <br>
    </blockquote>
  </body>
</html>