Flexible and Safe GCOV Format duplicated in projects/cfg.html

Flexible and Safe GCOV Format duplicated in projects/cfg.html

[Richard Almquist]

There are two sections titled "Flexible and Safe GCOV Format" with 
nearly identical text in cfg.html.

<<<<<<<<<<<<<< lines 290-323 <<<<<<<<<<<<<<<<<
<h3>Flexible and Safe GCOV Format</h3>
 
<p>This work is done by <a
href="mailto:bim@atrey.karlin.mff.cuni.cz">Pavel Nejedly</a>.</p>
 
<h4>Theory</h4>
 
<p>Currently the profile is fragile, since there is no verification
that the compiled program matches the profiled data.  Since the
profiler eliminates the redundancy in data, the mismatch is often not
discovered at all causing completely nonsensical results.
nonsense.</p>
 
<h4>Implementation in GCC</h4>
 
<p>We plan to calculate a checksum (CRC) of each graph when it is
constructed.  Also the format will be extended so that more
information can be recorded, such as histograms for the number of
iterations for loops that can then be used for loop peeling and loop
unrolling.</p>
 
<p>We also want to add versioning and further information to make it
look like a real file format.</p>
 
<p>There are a few references to much more advanced profiling systems
in <a href="#3">[3]</a>.</p>
 
<h4>Status</h4>
 
<p>First version is included in the branch. Due to the ability to compute
overall summary of counters it allows detection of hot spots in program
and improves to detect hot spots better reducing code size and improving
performance occasionally.</p>

 >>>>>>>>>>>>lines 339-370>>>>>>>>>>>>>>
<h3>Flexible and Safe GCOV Format</h3>

<p>This work is done by <a
href="mailto:jh@suse.cz">Jan Hubicka</a>.</p>
 
<h4>Theory</h4>
 
<p>Currently the profile is fragile, since there is no verification
that the compiled program matches the profiled data.  Since the
profiler eliminates the redundancy in data, the mismatch is often not
discovered at all thereby making results to be completely
nonsense.</p>
 
<h4>Implementation in GCC</h4>
 
<p>We plan to calculate a checksum (CRC) of each graph when it is
constructed.  Also the format will be extended so that more
information can be recorded, such as histograms for the number of
iterationunrolling.</p>
 
<p>We also want to add versioning and further information to make it
look like a real file format.</p>
 
<p>There are a few references to much more advanced profiling systems
in <a href="#3">[3]</a>.</p>
 
<h4>Status</h4>
 
<p>First version is included in the branch. Due to ability to compute
overall summary of counters it allows detection of hot spots in program
and improves to detect hot spots better reducing code size and improving
performance occasionally.</p>

Re: Flexible and Safe GCOV Format duplicated in projects/cfg.html

[Andreas Jaeger]

Richard Almquist <ralmquist@ssi-corp.com> writes:

> There are two sections titled "Flexible and Safe GCOV Format" with
> nearly identical text in cfg.html.

Thanks, I've committed the appended patch to the page,
Andreas

Index: htdocs/projects/cfg.html
===================================================================
RCS file: /cvs/gcc/wwwdocs/htdocs/projects/cfg.html,v
retrieving revision 1.7
diff -u -p -r1.7 cfg.html
--- htdocs/projects/cfg.html	11 Feb 2002 08:08:30 -0000	1.7
+++ htdocs/projects/cfg.html	11 Apr 2002 13:17:49 -0000
@@ -334,40 +334,6 @@ into different parts.
 A first version has been written and has been installed on the branch.
 It adds about 1% to spec2000 runs with profile feedback.</p>
 
-<h3>Flexible and Safe GCOV Format</h3>
-
-<p>This work is done by <a
-href="mailto:jh@suse.cz">Jan Hubicka</a>.</p>
-
-<h4>Theory</h4>
-
-<p>Currently the profile is fragile, since there is no verification
-that the compiled program matches the profiled data.  Since the
-profiler eliminates the redundancy in data, the mismatch is often not
-discovered at all thereby making results to be completely
-nonsense.</p>
-
-<h4>Implementation in GCC</h4>
-
-<p>We plan to calculate a checksum (CRC) of each graph when it is
-constructed.  Also the format will be extended so that more
-information can be recorded, such as histograms for the number of
-iterations for loops that can then be used for loop peeling and loop
-unrolling.</p>
-
-<p>We also want to add versioning and further information to make it
-look like a real file format.</p>
-
-<p>There are a few references to much more advanced profiling systems
-in <a href="#3">[3]</a>.</p>
-
-<h4>Status</h4>
-
-<p>First version is included in the branch. Due to ability to compute
-overall summary of counters it allows detection of hot spots in program
-and improves to detect hot spots better reducing code size and improving
-performance occasionally.</p>
-
 <h3>MIDlevel RTL</h3>
 
 <p>This work is done by <a

-- 
 Andreas Jaeger
  SuSE Labs aj@suse.de
   private aj@arthur.inka.de
    http://www.suse.de/~aj