[Insight-developers] itkMath doxygen missing?

Arnaud Gelas arnaudgelas at gmail.com
Fri Jun 15 02:34:35 EDT 2012


Brad,

AFAIK doxygen is a single threaded process; however the generation of 
graphs is multithreaded (dot instances are created in parallel). With 
the current doxygen configuration file, all threads are meant to be used 
in this process (which occurs towards the end of the html generation).

Note that having more CPUs significantly reduced the documentation 
generation. Also, I am not sure if there any speed improvement in the 
latest doxygen version (?)

---

Matt,

After upgrading doxygen, you should also upgrade the configuration file 
to take full benefit from new features (and potential enhancement); else 
default values are used.

You only to run

     doxygen -u <config_file>

     see http://www.stack.nl/~dimitri/doxygen/doxygen_usage.html

Each parameters are well documented in the configuration file, but you 
can also have a look at the following page:

     http://www.stack.nl/~dimitri/doxygen/config.html

Finally, once you upgrade the configuration file, feel free to add me as 
reviewer when you submit the patch on gerrit.

Cheers,
Arnaud

On 06/15/2012 06:10 AM, Matt McCormick wrote:
> Hi,
>
> I bumped the version of Doxygen, so we will see how that affects the
> build tomorrow.
>
> The repository is stored on an SSD without a contention.  Maybe that
> has something to do with it.
>
> Unfortunately, I probably won't make the TCon tomorrow.
>
> Matt
>
> On Thu, Jun 14, 2012 at 2:55 PM, Bradley Lowekamp
> <blowekamp at mail.nih.gov>  wrote:
>> Hello,
>>
>> I tried out a configuration to build doxygen. It's on a VM I have which I
>> only allocated 2 vCPUs and 8GB of memory on our VM server which has 2-quad
>> core 3.47Ghz Xeon processors ( x2 hyper-threaded). ( If you are keeping
>> track at home all the machines labeled lhcp-*vm are running on the same
>> physical box. ) As doxygen is a single threaded process this seems like the
>> ideal way to allocate resources to this process, since there is no need to
>> utilize a whole machine.
>>
>> However, it is reporting that it takes ~12 hour to run. And there are some
>> errors which I assume are cause by an older version of doxygen.
>>
>> Hopefully, there will be some to discuss the best way to deal with doxygen
>> at a TCON.
>>
>> Brad
>>
>> On Jun 12, 2012, at 11:33 PM, Matt McCormick wrote:
>>
>> Hey Brad,
>>
>> The build is running on a 2nd Gen Intel Core 2 (i7) processor at 2.8
>> GHz, which I think the limiting factor since it runs single-threaded
>> for most of the time.  It still takes 12 hours, which is not nearly as
>> good as Arnaud's 8 hours!
>>
>> Arnaud has some ideas to modularize, which might make it faster...
>>
>> The nightly xml can now be found here:
>>
>>   http://public.kitware.com/pub/itk/itk-doxygen-xml.tar.gz
>>
>> Along with the Doxygen tag file:
>>
>>   http://public.kitware.com/pub/itk/InsightDoxygen.tag.gz
>>
>> I still have to work with Luis to figure out how to update the nightly html.
>>
>> If you want to make a link on the front page, that sounds good.
>>
>> Thanks,
>> Matt
>>
>> On Tue, Jun 12, 2012 at 9:37 PM, Bradley Lowekamp
>> <blowekamp at mail.nih.gov>  wrote:
>>
>> Hello Matt,
>>
>>
>> I see you are running the doxygen on the dashboard.
>>
>>
>> This is good you are able to run it, in your basement. When I have run it
>> recently it has taken me over 24hrs for some odd reason. My virtual machine
>> server is really getting taxed.  I don't know what resources the process
>> needs to run efficiently.
>>
>>
>> I also see that you have the XML generation turned on. Is there a place
>> where I can download the XML? I would like to mine it for some documentation
>> for SimpleITK. Perhaps we should have it linked of the front page of the
>> Doxygen?
>>
>>
>> Brad
>>
>>
>>
>> On Jun 7, 2012, at 2:44 AM, Matt McCormick wrote:
>>
>>
>> Hi Arnaud,
>>
>>
>> I have setup my home box to generate Doxygen:
>>
>>
>>   http://open.cdash.org/buildSummary.php?buildid=2328774
>>
>>
>> I'll contact you and Luis off list to get the results uploaded to the
>>
>> appropriate locations.
>>
>>
>> Thanks,
>>
>> Matt
>>
>>
>> On Wed, Jun 6, 2012 at 10:58 AM, Arnaud Gelas<arnaudgelas at gmail.com>  wrote:
>>
>> Hi Brad,
>>
>>
>> I can't anymore build the doxygen documentation every night on my machine.
>>
>> It takes more and more time to generate it for the whole toolkit.
>>
>> It would be great if the documentation could be also modularized, it will
>>
>> really decrease the time...
>>
>>
>> I was planing to send an email on the list, I could generate it for each rc,
>>
>> or once a week; unless someone else has a machine to do this job?
>>
>>
>> Arnaud
>>
>>
>>
>> On 06/05/2012 10:05 PM, Bradley Lowekamp wrote:
>>
>>
>> Hello David,
>>
>>
>> The other interesting bit is the generated date at the bottom of the page:
>>
>>
>> "Generated on Sat Jun 2 2012 02:34:31 for ITK"
>>
>>
>> Looks like the systems had some kind of hiccup.
>>
>>
>> Brad
>>
>>
>> On Jun 5, 2012, at 3:56 PM, David Doria wrote:
>>
>>
>> I noticed that this page is not found:
>>
>> http://www.itk.org/Doxygen/html/classitk_1_1Math.html
>>
>>
>> but everything else seems to be working (e.g.
>>
>> http://www.itk.org/Doxygen/html/classitk_1_1Vector.html)
>>
>>
>> Does anyone know why?
>>
>>
>> David
>>
>> _______________________________________________
>>
>> Powered by www.kitware.com
>>
>>
>> Visit other Kitware open-source projects at
>>
>> http://www.kitware.com/opensource/opensource.html
>>
>>
>> Kitware offers ITK Training Courses, for more information visit:
>>
>> http://kitware.com/products/protraining.php
>>
>>
>> Please keep messages on-topic and check the ITK FAQ at:
>>
>> http://www.itk.org/Wiki/ITK_FAQ
>>
>>
>> Follow this link to subscribe/unsubscribe:
>>
>> http://www.itk.org/mailman/listinfo/insight-developers
>>
>>
>>
>> ========================================================
>>
>>
>> Bradley Lowekamp
>>
>>
>> Medical Science and Computing for
>>
>>
>> Office of High Performance Computing and Communications
>>
>>
>> National Library of Medicine
>>
>>
>> blowekamp at mail.nih.gov
>>
>>
>>
>>
>>
>>
>>
>> _______________________________________________
>>
>> Powered by www.kitware.com
>>
>>
>> Visit other Kitware open-source projects at
>>
>> http://www.kitware.com/opensource/opensource.html
>>
>>
>> Kitware offers ITK Training Courses, for more information visit:
>>
>> http://kitware.com/products/protraining.php
>>
>>
>> Please keep messages on-topic and check the ITK FAQ at:
>>
>> http://www.itk.org/Wiki/ITK_FAQ
>>
>>
>> Follow this link to subscribe/unsubscribe:
>>
>> http://www.itk.org/mailman/listinfo/insight-developers
>>
>>
>>
>>
>> _______________________________________________
>>
>> Powered by www.kitware.com
>>
>>
>> Visit other Kitware open-source projects at
>>
>> http://www.kitware.com/opensource/opensource.html
>>
>>
>> Kitware offers ITK Training Courses, for more information visit:
>>
>> http://kitware.com/products/protraining.php
>>
>>
>> Please keep messages on-topic and check the ITK FAQ at:
>>
>> http://www.itk.org/Wiki/ITK_FAQ
>>
>>
>> Follow this link to subscribe/unsubscribe:
>>
>> http://www.itk.org/mailman/listinfo/insight-developers
>>
>>
>>
>>
>> ========================================================
>>
>> Bradley Lowekamp
>>
>> Medical Science and Computing for
>>
>> Office of High Performance Computing and Communications
>>
>> National Library of Medicine
>>
>> blowekamp at mail.nih.gov
>>
>>
>>
>>



More information about the Insight-developers mailing list