GStreamer-devel

Time taken by gst_pad_push function

Classic

List

Threaded

4 messages Options

Deepak T.Y

Time taken by gst_pad_push function

Hi all,
       I am trying to optimize my Gstreamer based application. I have a gstreamer pulgin
which pushs around 35-40k data to next element in the pipeline. The following are my
observations:

1. Time taken across gst_pad_push() function is around 50,000 usec. i.e
    gettimeofday(&t1,NULL);
    gst_pad_push(pad,buf);
    gettimeofday(&t2,NULL);
    Why gst_pad_push() function is taking so much time?

2. The time taken between start of gst_pad_push() and start of the chain function of next
   element in the pipeline is around 4,000 usec. Compared to time taken by gst_pad_push()
   to return to the calling fucnction, 4000 usec is less.

Please can anyone help me in improving the timing of gst_pad_push()?

Thanking you,
Deepak T Y

This message (including any attachment) is confidential and may be legally privileged. Access to this message by anyone other than the intended recipient(s) listed above is unauthorized. If you are not the intended recipient you are hereby notified that any disclosure, copying, or distribution of the message, or any action taken or omission of action by you in reliance upon it, is prohibited and may be unlawful. Please immediately notify the sender by reply e-mail and permanently delete all copies of the message if you have received this message in error.

ved kpl

Re: Time taken by gst_pad_push function

Hi,

That depends on the pipeline you are using. The gst_pad_push() by the
element COULD BE chained right upto downmost element, resulting in
gst_pad_push() by the element eventually calling _chain() of sink
element (with few elements in between).

But, why do you want to profile gst_pad_push()?

Ved

On Fri, Jan 16, 2009 at 11:51 AM, Deepak T.Y <[hidden email]> wrote:

> Hi all,
> I am trying to optimize my Gstreamer based application. I have a
> gstreamer pulgin
> which pushs around 35-40k data to next element in the pipeline. The
> following are my
> observations:
>
> 1. Time taken across gst_pad_push() function is around 50,000 usec. i.e
> gettimeofday(&t1,NULL);
> gst_pad_push(pad,buf);
> gettimeofday(&t2,NULL);
> Why gst_pad_push() function is taking so much time?
>
> 2. The time taken between start of gst_pad_push() and start of the chain
> function of next
> element in the pipeline is around 4,000 usec. Compared to time taken by
> gst_pad_push()
> to return to the calling fucnction, 4000 usec is less.
>
> Please can anyone help me in improving the timing of gst_pad_push()?
>
>
> Thanking you,
> Deepak T Y
>
>
>
>
>
>
>
> This message (including any attachment) is confidential and may be legally
> privileged. Access to this message by anyone other than the intended
> recipient(s) listed above is unauthorized. If you are not the intended
> recipient you are hereby notified that any disclosure, copying, or
> distribution of the message, or any action taken or omission of action by
> you in reliance upon it, is prohibited and may be unlawful. Please
> immediately notify the sender by reply e-mail and permanently delete all
> copies of the message if you have received this message in error.
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by:
> SourcForge Community
> SourceForge wants to tell your story.
> http://p.sf.net/sfu/sf-spreadtheword
> _______________________________________________
> gstreamer-devel mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gstreamer-devel
>
>

Deepak T.Y

Re: Time taken by gst_pad_push function

Hi Ved,
Thanks for your reply.
The element(lets say X) where I am calling gst_pad_push(), pumps data
to next element. Untill gst_pad_push returns back, element X doesn't
start processing for the next push. Effectively my gstreamer pipeline
has become sequential.

1. Currently I dont have any "queue" element in my pipeline.
Will it help If I introduce "queue" between every element in my pipeline?

2. I guess my pipeline is a downstream one. Will it the performance of
my application improve if I make it upstream?

With Regards,
Deepak T Y

-----Original Message-----
From: ved kpl [mailto:[hidden email]]
Sent: Fri 1/16/2009 12:41 PM
To: Discussion of the development of GStreamer
Subject: Re: [gst-devel] Time taken by gst_pad_push function

Hi,

That depends on the pipeline you are using. The gst_pad_push() by the
element COULD BE chained right upto downmost element, resulting in
gst_pad_push() by the element eventually calling _chain() of sink
element (with few elements in between).

But, why do you want to profile gst_pad_push()?

Ved

On Fri, Jan 16, 2009 at 11:51 AM, Deepak T.Y <[hidden email]> wrote:

------------------------------------------------------------------------------
This SF.net email is sponsored by:
SourcForge Community
SourceForge wants to tell your story.
http://p.sf.net/sfu/sf-spreadtheword
_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gstreamer-devel

This message (including any attachment) is confidential and may be legally privileged. Access to this message by anyone other than the intended recipient(s) listed above is unauthorized. If you are not the intended recipient you are hereby notified that any disclosure, copying, or distribution of the message, or any action taken or omission of action by you in reliance upon it, is prohibited and may be unlawful. Please immediately notify the sender by reply e-mail and permanently delete all copies of the message if you have received this message in error.

------------------------------------------------------------------------------
This SF.net email is sponsored by:
SourcForge Community
SourceForge wants to tell your story.
http://p.sf.net/sfu/sf-spreadtheword
_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gstreamer-devel

winmail.dat (5K) Download Attachment

ved kpl

Re: Time taken by gst_pad_push function

On Fri, Jan 16, 2009 at 2:58 PM, Deepak T.Y <[hidden email]> wrote:
> Hi Ved,
> Thanks for your reply.
> The element(lets say X) where I am calling gst_pad_push(), pumps data
> to next element. Untill gst_pad_push returns back, element X doesn't
> start processing for the next push. Effectively my gstreamer pipeline
> has become sequential.

>
> 1. Currently I dont have any "queue" element in my pipeline.
> Will it help If I introduce "queue" between every element in my pipeline?

You can use queue if you want to decouple two elements. Queue
starts a different thread to push data downstream. It entirely depends
on your pipeline design. Queue accepts data form the thread above it
self and pushes the data downstream in a different thread.

>
> 2. I guess my pipeline is a downstream one. Will it the performance of
> my application improve if I make it upstream?

Sorry, I havent understood the question. What do you mean by
making the pipeline upstream?

>
> With Regards,
> Deepak T Y
>
>
> -----Original Message-----
> From: ved kpl [mailto:[hidden email]]
> Sent: Fri 1/16/2009 12:41 PM
> To: Discussion of the development of GStreamer
> Subject: Re: [gst-devel] Time taken by gst_pad_push function
>
> Hi,
>
> That depends on the pipeline you are using. The gst_pad_push() by the
> element COULD BE chained right upto downmost element, resulting in
> gst_pad_push() by the element eventually calling _chain() of sink
> element (with few elements in between).
>
> But, why do you want to profile gst_pad_push()?
>
> Ved
>
> On Fri, Jan 16, 2009 at 11:51 AM, Deepak T.Y <[hidden email]> wrote:
>> Hi all,
>> I am trying to optimize my Gstreamer based application. I have a
>> gstreamer pulgin
>> which pushs around 35-40k data to next element in the pipeline. The
>> following are my
>> observations:
>>
>> 1. Time taken across gst_pad_push() function is around 50,000 usec. i.e
>> gettimeofday(&t1,NULL);
>> gst_pad_push(pad,buf);
>> gettimeofday(&t2,NULL);
>> Why gst_pad_push() function is taking so much time?
>>
>> 2. The time taken between start of gst_pad_push() and start of the chain
>> function of next
>> element in the pipeline is around 4,000 usec. Compared to time taken by
>> gst_pad_push()
>> to return to the calling fucnction, 4000 usec is less.
>>
>> Please can anyone help me in improving the timing of gst_pad_push()?
>>
>>
>> Thanking you,
>> Deepak T Y
>>
>>
>>
>>
>>
>>
>>
>> This message (including any attachment) is confidential and may be legally
>> privileged. Access to this message by anyone other than the intended
>> recipient(s) listed above is unauthorized. If you are not the intended
>> recipient you are hereby notified that any disclosure, copying, or
>> distribution of the message, or any action taken or omission of action by
>> you in reliance upon it, is prohibited and may be unlawful. Please
>> immediately notify the sender by reply e-mail and permanently delete all
>> copies of the message if you have received this message in error.
>>
>> ------------------------------------------------------------------------------
>> This SF.net email is sponsored by:
>> SourcForge Community
>> SourceForge wants to tell your story.
>> http://p.sf.net/sfu/sf-spreadtheword
>> _______________________________________________
>> gstreamer-devel mailing list
>> [hidden email]
>> https://lists.sourceforge.net/lists/listinfo/gstreamer-devel
>>
>>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by:
> SourcForge Community
> SourceForge wants to tell your story.
> http://p.sf.net/sfu/sf-spreadtheword
> _______________________________________________
> gstreamer-devel mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gstreamer-devel
>
>
>
> This message (including any attachment) is confidential and may be legally privileged. Access to this message by anyone other than the intended recipient(s) listed above is unauthorized. If you are not the intended recipient you are hereby notified that any disclosure, copying, or distribution of the message, or any action taken or omission of action by you in reliance upon it, is prohibited and may be unlawful. Please immediately notify the sender by reply e-mail and permanently delete all copies of the message if you have received this message in error.
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by:
> SourcForge Community
> SourceForge wants to tell your story.
> http://p.sf.net/sfu/sf-spreadtheword
> _______________________________________________
> gstreamer-devel mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gstreamer-devel
>
>