<?xml version="1.0" encoding="utf-8"?>
<feed xml:lang="en" xmlns="http://www.w3.org/2005/Atom"><title>Recent changes to feature-requests</title><link href="https://sourceforge.net/p/kcachegrind/feature-requests/" rel="alternate"/><link href="https://sourceforge.net/p/kcachegrind/feature-requests/feed.atom" rel="self"/><id>https://sourceforge.net/p/kcachegrind/feature-requests/</id><updated>2016-01-17T18:49:57.228000Z</updated><subtitle>Recent changes to feature-requests</subtitle><entry><title>#11 Allow reading callgrind data from STDIN </title><link href="https://sourceforge.net/p/kcachegrind/feature-requests/11/?limit=25#2822" rel="alternate"/><published>2016-01-17T18:49:57.228000Z</published><updated>2016-01-17T18:49:57.228000Z</updated><author><name>Arnout Engelen</name><uri>https://sourceforge.net/u/raboofje/</uri></author><id>https://sourceforge.net55939b7398cc832669d88c2d82d807ea553d89a3</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;We could use the convention of allowing '-' as a filename parameter to indicate kcachegrind should read from stdin?&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>#11 Allow reading callgrind data from STDIN </title><link href="https://sourceforge.net/p/kcachegrind/feature-requests/11/?limit=25#202b" rel="alternate"/><published>2016-01-17T09:30:13.766000Z</published><updated>2016-01-17T09:30:13.766000Z</updated><author><name>Josef Weidendorfer</name><uri>https://sourceforge.net/u/weidendo/</uri></author><id>https://sourceforge.net5d63cfe7e17203bf03f3927ee25b681c4cd2a2f6</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;The feature itself may be interesting, but loading currently blocks the GUI, and waiting&lt;br/&gt;
for stdin would result in a not-responsive window with a naive implementation.&lt;/p&gt;
&lt;p&gt;But you can write a short shell script doing to same: write stdin to a random file,&lt;br/&gt;
and call kcachegrind then on this random file, e.g. using the bash $RANDOM feature.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Allow reading callgrind data from STDIN </title><link href="https://sourceforge.net/p/kcachegrind/feature-requests/11/" rel="alternate"/><published>2016-01-16T22:09:08.310000Z</published><updated>2016-01-16T22:09:08.310000Z</updated><author><name>Arnout Engelen</name><uri>https://sourceforge.net/u/raboofje/</uri></author><id>https://sourceforge.net189a3d8ea75662bd754319fad2e015838316b24d</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;It would be really nice if kcachegrind could take the callgrind data from stdin, so you can collect data and start kcachegrind in one go.&lt;/p&gt;
&lt;p&gt;The '-r' flag does this for 'regular' cachegrind use, but that won't work for other tools that can be converted to callgrind format.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Allow custom aggregate functions for events / costs</title><link href="https://sourceforge.net/p/kcachegrind/feature-requests/10/" rel="alternate"/><published>2015-02-04T13:32:24.738000Z</published><updated>2015-02-04T13:32:24.738000Z</updated><author><name>Hagen Graf</name><uri>https://sourceforge.net/u/hcgraf/</uri></author><id>https://sourceforge.net78c13bdb2dd721e9119d977b822f8fc46c4a7997</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;I was trying to use KCachegrind's great visualization capabilities for a kind of data-flow analysis, when I unfortunately hit a limit…&lt;br /&gt;
However, it might be easy to overcome, and it might be a generally useful feature for the program.&lt;/p&gt;
&lt;h4 id="what-im-trying-to-visualize"&gt;What I'm trying to visualize&lt;/h4&gt;
&lt;p&gt;I have an algorithm which makes use of floating-point variables. I'm trying to figure out if some parts of the algorithm use only a limited range of values, in order to optimize those parts by using fixed-point variables or smaller floating-point formats.&lt;br /&gt;
I know that this kind of optimization doesn't make sense in desktop computers, but it's planned to be used for porting algorithms to Microcontrollers or even FPGAs or ASICs.&lt;/p&gt;
&lt;h5 id="fictive-kcachegrind-input-file"&gt;Fictive KCachegrind Input File&lt;/h5&gt;
&lt;p&gt;I created the following input file for KCachegrind manually (i.e. the corresponding program doesn't exist).&lt;br /&gt;
It can be opened and visualized, &lt;strong&gt;but&lt;/strong&gt; the "events" minExp, maxExp and range are &lt;strong&gt;accumulated&lt;/strong&gt;, which doesn't make sense.&lt;/p&gt;
&lt;div class="codehilite"&gt;&lt;pre&gt;&lt;span class="n"&gt;positions&lt;/span&gt;&lt;span class="o"&gt;:&lt;/span&gt; &lt;span class="n"&gt;instr&lt;/span&gt;

&lt;span class="n"&gt;event&lt;/span&gt;&lt;span class="o"&gt;:&lt;/span&gt; &lt;span class="n"&gt;flops&lt;/span&gt; &lt;span class="o"&gt;:&lt;/span&gt; &lt;span class="n"&gt;Floating&lt;/span&gt;&lt;span class="o"&gt;-&lt;/span&gt;&lt;span class="n"&gt;Point&lt;/span&gt; &lt;span class="n"&gt;Operations&lt;/span&gt;
&lt;span class="n"&gt;event&lt;/span&gt;&lt;span class="o"&gt;:&lt;/span&gt; &lt;span class="n"&gt;minExp&lt;/span&gt; &lt;span class="o"&gt;:&lt;/span&gt; &lt;span class="n"&gt;Minimum&lt;/span&gt; &lt;span class="n"&gt;Exponent&lt;/span&gt;
&lt;span class="n"&gt;event&lt;/span&gt;&lt;span class="o"&gt;:&lt;/span&gt; &lt;span class="n"&gt;maxExp&lt;/span&gt; &lt;span class="o"&gt;:&lt;/span&gt; &lt;span class="n"&gt;Maximum&lt;/span&gt; &lt;span class="n"&gt;Exponent&lt;/span&gt;
&lt;span class="n"&gt;event&lt;/span&gt;&lt;span class="o"&gt;:&lt;/span&gt; &lt;span class="n"&gt;range&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="n"&gt;maxExp&lt;/span&gt; &lt;span class="o"&gt;-&lt;/span&gt; &lt;span class="n"&gt;minExp&lt;/span&gt; &lt;span class="o"&gt;:&lt;/span&gt; &lt;span class="n"&gt;Dynamic&lt;/span&gt; &lt;span class="n"&gt;Range&lt;/span&gt;
&lt;span class="n"&gt;events&lt;/span&gt;&lt;span class="o"&gt;:&lt;/span&gt; &lt;span class="n"&gt;flops&lt;/span&gt; &lt;span class="n"&gt;minExp&lt;/span&gt; &lt;span class="n"&gt;maxExp&lt;/span&gt;

&lt;span class="n"&gt;fl&lt;/span&gt;&lt;span class="o"&gt;=&lt;/span&gt;&lt;span class="n"&gt;file1&lt;/span&gt;&lt;span class="o"&gt;.&lt;/span&gt;&lt;span class="na"&gt;c&lt;/span&gt;
&lt;span class="n"&gt;fn&lt;/span&gt;&lt;span class="o"&gt;=&lt;/span&gt;&lt;span class="n"&gt;main&lt;/span&gt;
&lt;span class="mi"&gt;16&lt;/span&gt; &lt;span class="mi"&gt;20&lt;/span&gt; &lt;span class="mi"&gt;10&lt;/span&gt; &lt;span class="mi"&gt;80&lt;/span&gt;
&lt;span class="n"&gt;cfn&lt;/span&gt;&lt;span class="o"&gt;=&lt;/span&gt;&lt;span class="n"&gt;func1&lt;/span&gt;
&lt;span class="n"&gt;calls&lt;/span&gt;&lt;span class="o"&gt;=&lt;/span&gt;&lt;span class="mi"&gt;1&lt;/span&gt; &lt;span class="mi"&gt;50&lt;/span&gt;
&lt;span class="mi"&gt;17&lt;/span&gt; &lt;span class="mi"&gt;400&lt;/span&gt; &lt;span class="mi"&gt;15&lt;/span&gt; &lt;span class="mi"&gt;70&lt;/span&gt;
&lt;span class="n"&gt;cfl&lt;/span&gt;&lt;span class="o"&gt;=&lt;/span&gt;&lt;span class="n"&gt;file2&lt;/span&gt;&lt;span class="o"&gt;.&lt;/span&gt;&lt;span class="na"&gt;c&lt;/span&gt;
&lt;span class="n"&gt;cfn&lt;/span&gt;&lt;span class="o"&gt;=&lt;/span&gt;&lt;span class="n"&gt;func2&lt;/span&gt;
&lt;span class="n"&gt;calls&lt;/span&gt;&lt;span class="o"&gt;=&lt;/span&gt;&lt;span class="mi"&gt;3&lt;/span&gt; &lt;span class="mi"&gt;20&lt;/span&gt;
&lt;span class="mi"&gt;18&lt;/span&gt; &lt;span class="mi"&gt;400&lt;/span&gt; &lt;span class="mi"&gt;10&lt;/span&gt; &lt;span class="mi"&gt;80&lt;/span&gt;
&lt;span class="mi"&gt;19&lt;/span&gt; &lt;span class="mi"&gt;10&lt;/span&gt; &lt;span class="mi"&gt;10&lt;/span&gt; &lt;span class="mi"&gt;90&lt;/span&gt;

&lt;span class="n"&gt;fn&lt;/span&gt;&lt;span class="o"&gt;=&lt;/span&gt;&lt;span class="n"&gt;func1&lt;/span&gt;
&lt;span class="mi"&gt;51&lt;/span&gt; &lt;span class="mi"&gt;100&lt;/span&gt; &lt;span class="mi"&gt;15&lt;/span&gt; &lt;span class="mi"&gt;70&lt;/span&gt;
&lt;span class="n"&gt;cfl&lt;/span&gt;&lt;span class="o"&gt;=&lt;/span&gt;&lt;span class="n"&gt;file2&lt;/span&gt;&lt;span class="o"&gt;.&lt;/span&gt;&lt;span class="na"&gt;c&lt;/span&gt;
&lt;span class="n"&gt;cfn&lt;/span&gt;&lt;span class="o"&gt;=&lt;/span&gt;&lt;span class="n"&gt;func2&lt;/span&gt;
&lt;span class="n"&gt;calls&lt;/span&gt;&lt;span class="o"&gt;=&lt;/span&gt;&lt;span class="mi"&gt;2&lt;/span&gt; &lt;span class="mi"&gt;20&lt;/span&gt;
&lt;span class="mi"&gt;52&lt;/span&gt; &lt;span class="mi"&gt;300&lt;/span&gt; &lt;span class="mi"&gt;15&lt;/span&gt; &lt;span class="mi"&gt;70&lt;/span&gt;

&lt;span class="n"&gt;fl&lt;/span&gt;&lt;span class="o"&gt;=&lt;/span&gt;&lt;span class="n"&gt;file2&lt;/span&gt;&lt;span class="o"&gt;.&lt;/span&gt;&lt;span class="na"&gt;c&lt;/span&gt;
&lt;span class="n"&gt;fn&lt;/span&gt;&lt;span class="o"&gt;=&lt;/span&gt;&lt;span class="n"&gt;func2&lt;/span&gt;
&lt;span class="mi"&gt;20&lt;/span&gt; &lt;span class="mi"&gt;700&lt;/span&gt; &lt;span class="mi"&gt;10&lt;/span&gt; &lt;span class="mi"&gt;80&lt;/span&gt;
&lt;/pre&gt;&lt;/div&gt;
&lt;h4 id="my-proposed-solution"&gt;My proposed solution&lt;/h4&gt;
&lt;p&gt;As mentioned in the title, a flexible solution would be the use of custom &lt;a class="" href="http://en.wikipedia.org/wiki/Aggregate_function" rel="nofollow"&gt;Aggregate Functions&lt;/a&gt; borrowed from databases.&lt;br /&gt;
This function would be used to calculate the Inclusive- and Self-Cost, if multiple cost lines are given.&lt;br /&gt;
In my case, the aggregate function for flops would be sum(), for minExp min(), for maxExp max(), and for range also max().&lt;br /&gt;
Maybe a possible syntax could be&lt;/p&gt;
&lt;div class="codehilite"&gt;&lt;pre&gt;&lt;span class="n"&gt;event&lt;/span&gt;&lt;span class="o"&gt;:&lt;/span&gt; &lt;span class="n"&gt;flops&lt;/span&gt; &lt;span class="o"&gt;:&lt;/span&gt; &lt;span class="n"&gt;Floating&lt;/span&gt;&lt;span class="o"&gt;-&lt;/span&gt;&lt;span class="n"&gt;Point&lt;/span&gt; &lt;span class="n"&gt;Operations&lt;/span&gt;
&lt;span class="n"&gt;event&lt;/span&gt;&lt;span class="o"&gt;:&lt;/span&gt; &lt;span class="n"&gt;minExp&lt;/span&gt;&lt;span class="o"&gt;(&lt;/span&gt;&lt;span class="n"&gt;min&lt;/span&gt;&lt;span class="o"&gt;)&lt;/span&gt; &lt;span class="o"&gt;:&lt;/span&gt; &lt;span class="n"&gt;Minimum&lt;/span&gt; &lt;span class="n"&gt;Exponent&lt;/span&gt;
&lt;span class="n"&gt;event&lt;/span&gt;&lt;span class="o"&gt;:&lt;/span&gt; &lt;span class="n"&gt;maxExp&lt;/span&gt;&lt;span class="o"&gt;(&lt;/span&gt;&lt;span class="n"&gt;max&lt;/span&gt;&lt;span class="o"&gt;)&lt;/span&gt; &lt;span class="o"&gt;:&lt;/span&gt; &lt;span class="n"&gt;Maximum&lt;/span&gt; &lt;span class="n"&gt;Exponent&lt;/span&gt;
&lt;span class="n"&gt;event&lt;/span&gt;&lt;span class="o"&gt;:&lt;/span&gt; &lt;span class="n"&gt;range&lt;/span&gt;&lt;span class="o"&gt;(&lt;/span&gt;&lt;span class="n"&gt;max&lt;/span&gt;&lt;span class="o"&gt;)&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="n"&gt;maxExp&lt;/span&gt; &lt;span class="o"&gt;-&lt;/span&gt; &lt;span class="n"&gt;minExp&lt;/span&gt; &lt;span class="o"&gt;:&lt;/span&gt; &lt;span class="n"&gt;Dynamic&lt;/span&gt; &lt;span class="n"&gt;Range&lt;/span&gt;
&lt;/pre&gt;&lt;/div&gt;
&lt;p&gt;Of course, &lt;strong&gt;sum()&lt;/strong&gt; should be the default aggregate function.&lt;/p&gt;
&lt;p&gt;For aggregate functions other than &lt;strong&gt;sum()&lt;/strong&gt;, the relative cost view of KCachegrind doesn't make a lot of sense, however, it wouldn't do harm… (values &amp;gt;100% would be possible)&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Save as a CSV file</title><link href="https://sourceforge.net/p/kcachegrind/feature-requests/9/" rel="alternate"/><published>2013-05-04T02:42:07.376000Z</published><updated>2013-05-04T02:42:07.376000Z</updated><author><name>bglyne</name><uri>https://sourceforge.net/u/bglyne/</uri></author><id>https://sourceforge.net784a2abd65196ce606d66e828f9af71e7bfbd8e2</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;It would be handy if it's possible to save the contents of listview control as a comma-separated values file, or copy it into the clipboard in CSV format.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>re-centering selected code/assembly line</title><link href="https://sourceforge.net/p/kcachegrind/feature-requests/8/" rel="alternate"/><published>2012-12-07T11:16:14Z</published><updated>2012-12-07T11:16:14Z</updated><author><name>Simon Fe</name><uri>https://sourceforge.net/u/abcthruzxyz/</uri></author><id>https://sourceforge.net9528f11b392db852cb7d8fc549da39f92cd0789f</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;When using the "source code" or "machine code" windows, I often sort by Cycle Estimate to identify the hot spots, select the line of code/instruction and then go back to sorting by source line/address to see the context.   Currently, that involves scrolling the window until the highligthed line becomes visible. It would be *really* handy if the system defaulted to automatically putting the selected line in the centre of the window.&lt;/p&gt;
&lt;p&gt;Thanks&lt;br /&gt;
S&lt;br /&gt;
PS: Fantastic tool BTW.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Filtering profile</title><link href="https://sourceforge.net/p/kcachegrind/feature-requests/7/" rel="alternate"/><published>2012-07-24T11:19:51Z</published><updated>2012-07-24T11:19:51Z</updated><author><name>Wojciech Waga</name><uri>https://sourceforge.net/u/wojek/</uri></author><id>https://sourceforge.netfa774295d3a9498d44d6de29f2ecafab1d1be53e</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Hi, it would be very handy to put a "filter" button maybe instead of the "search". This could work like grep. I.e. someone types vector and he gets a list of all functions / objects with "vector" in name. Maybe some special value to search for constructors (destructors are easy '::~') to check numbers of both.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Generate "Callee Map" graphic as png</title><link href="https://sourceforge.net/p/kcachegrind/feature-requests/6/" rel="alternate"/><published>2012-06-04T14:45:43Z</published><updated>2012-06-04T14:45:43Z</updated><author><name>BRULE Herman</name><uri>https://sourceforge.net/u/alphaonex86bis/</uri></author><id>https://sourceforge.netdb73598fdc17d15b5dc5b0370b35dedc77ee3fef</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Hello, can you generate "Callee Map" graphic as png (via the right click menu), and ask before save the dimension of image generated?&lt;br /&gt;
Thanks.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Use other objdump than default</title><link href="https://sourceforge.net/p/kcachegrind/feature-requests/5/" rel="alternate"/><published>2011-10-18T08:15:33Z</published><updated>2011-10-18T08:15:33Z</updated><author><name>Anonymous</name><uri>https://sourceforge.net/u/userid-None/</uri></author><id>https://sourceforge.net4787e5236b2637536e006127636418250c522dab</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;I would really like to be able to specify which objdump to use since I frequently use KCachegrind to profile ARM assembly code. I managed to accomplish this in KCachegrind 0.4.6 (the one that came with Kubuntu) by creating a symbolic link to arm-none-linux-gnueabi-objdump in the current directory and then adding . to PATH, but if I do the same in KCachegrind 0.6.0 (the one that comes with Kubuntu 11.04) I do not get the disassembled instructions (all disassembled text is shown as ...). Can this be since I am running KCachegrind on a 64-bit machine while the arm code runs on a 32-bit machine?&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Grouping several lines</title><link href="https://sourceforge.net/p/kcachegrind/feature-requests/4/" rel="alternate"/><published>2011-05-11T16:31:36Z</published><updated>2011-05-11T16:31:36Z</updated><author><name>Alex V Breger</name><uri>https://sourceforge.net/u/sbreg/</uri></author><id>https://sourceforge.net04051755a2a8e89472271e91ca58656fd94297f8</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Hello&lt;/p&gt;
&lt;p&gt;I wish to be able to group several lines form "source view" and "assembler view" together in terms of their Ir counts (callgrind dump-instr + tracejump run analysis). E.g. I have some hot-spot with several lines:&lt;/p&gt;
&lt;p&gt;for(i=0;i&amp;lt;100000000;i++)&lt;br /&gt;
{&lt;br /&gt;
statement1&lt;br /&gt;
statement2&lt;br /&gt;
}&lt;/p&gt;
&lt;p&gt;For both statements, the Ir count is same in my case. I want to select both lines and to get summary Ir count for this fragment.&lt;/p&gt;
&lt;p&gt;The same is for asm view.&lt;/p&gt;&lt;/div&gt;</summary></entry></feed>