[LITMUS^RT] Compiling the video src from Proceedings of the 32nd IEEE Real-Time Systems Symposium, pp. 93-103, December 2011
Christopher Kenna
cjk at cs.unc.edu
Sun Sep 29 06:04:32 CEST 2013
Hi Debo,
I do not have a newer version of the video player plugin. I'm not sure
which APIs changed, but I can tell you that the FFMpeg documentation was
surprisingly good when I was working on this. You can find it here:
http://www.ffmpeg.org/doxygen/trunk/index.html
Off of the list, you asked me about implementing a task graph model that
takes into account the precedence of tasks. I think that Cong is probably
your best resource for this type of question. Fortunately for you, he's at
your university now (congrats Cong!):
http://www.utdallas.edu/~cxl137330/pub.html
-- Chris
On Wed, Sep 25, 2013 at 11:25 AM, Glenn Elliott <gelliott at cs.unc.edu> wrote:
> Looking at the code in rtspin, it appears the file format is a csv with
> the following properties:
> 1) Execution time may be expressed in floating point.
> 2) Time is in milliseconds by default, but this can be scaled with
> rtspin's "-s" option.
> 3) The user specifies the column in the csv that contains execution times
> for the task with "-o". I believe the first column of the csv is selected
> by default.
> 4) Each row in the csv denotes the execution time for a job. So, the
> first row is for job_0, second row is for job_1, etc.
> 5) The rtspin will execute N jobs before terminating, where N is the
> number of rows. There is no "looping back around". I believe this implies
> that the "duration" argument to rtspin is ignored.
> 6) Only job execution times are read from the file.
>
> Pertinent lines of code:
> 1) Reading the file:
> https://github.com/LITMUS-RT/liblitmus/blob/85cc826d76ad8bcfeb675710c49c410583cc86a9/bin/rtspin.c#L78
> 2) Column selection:
> https://github.com/LITMUS-RT/liblitmus/blob/85cc826d76ad8bcfeb675710c49c410583cc86a9/bin/rtspin.c#L248
> 3) Time unit scaling:
> https://github.com/LITMUS-RT/liblitmus/blob/85cc826d76ad8bcfeb675710c49c410583cc86a9/bin/rtspin.c#L374
>
> rtspin's file format is a bit restrictive. It was coded up for a
> particular paper where job execution times had to vary from job-to-job in a
> deterministic/repeatable way. I don't know if this format was put in place
> for the video paper. If so, it was probably done to allow the
> specification of different execution times for I-, P-, and B-frames. It's
> not meant to be an alternative method of general use. You may want to
> explore using input redirection to send regular command line arguments to
> rtspin (http://linuxcommand.org/lts0060.php). Of course, you could also
> modify rtspin to do what you need. The code is relatively straightforward.
>
> -Glenn
>
>
> On Sep 25, 2013, at 12:57 PM, ayoade gbadebo <gbaduz at gmail.com> wrote:
>
> Hello,
> Please what is the file input format for rtspin application.
> I want to generate a set of random task using file and not command line
> parameters
>
>
> On Tue, Sep 24, 2013 at 9:38 AM, Björn Brandenburg <bbb at mpi-sws.org>wrote:
>
>>
>> On Sep 24, 2013, at 4:07 PM, ayoade gbadebo <gbaduz at gmail.com> wrote:
>>
>> > I tried comiling the video src from this paper but it complained of
>> deprecated fuctions, do you have any new source I can use ?
>> >
>>
>> That code is not part of the LITMUS^RT project and not maintained by us
>> (the project). That said, you could try asking Chris (the first author),
>> who wrote most of the video code if I recall correctly.
>>
>> - Björn
>>
>>
>> _______________________________________________
>> litmus-dev mailing list
>> litmus-dev at lists.litmus-rt.org
>> https://lists.litmus-rt.org/listinfo/litmus-dev
>>
>
> _______________________________________________
> litmus-dev mailing list
> litmus-dev at lists.litmus-rt.org
> https://lists.litmus-rt.org/listinfo/litmus-dev
>
>
>
> _______________________________________________
> litmus-dev mailing list
> litmus-dev at lists.litmus-rt.org
> https://lists.litmus-rt.org/listinfo/litmus-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.litmus-rt.org/pipermail/litmus-dev/attachments/20130928/73b99a13/attachment.html>
More information about the litmus-dev
mailing list