public inbox for ecos-discuss@sourceware.org
 help / color / mirror / Atom feed
* [ECOS] eCos or microITRON
@ 2003-08-27  4:02 Satish Kumar
  2003-08-27 12:00 ` Alex Schuilenburg
  0 siblings, 1 reply; 5+ messages in thread
From: Satish Kumar @ 2003-08-27  4:02 UTC (permalink / raw)
  To: ecos-discuss

Hi,
     i m interested in knowing the performance of eCos with respect
to microITRON rtos..can anybody give me some tips or suggestions
regarding their performance, code size..etc..

thanks
satish
-- 
Satish Kumar <satish@bvt.sc.sanyo.co.jp>


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [ECOS] eCos or microITRON
  2003-08-27  4:02 [ECOS] eCos or microITRON Satish Kumar
@ 2003-08-27 12:00 ` Alex Schuilenburg
  2003-08-28  1:13   ` Satish Kumar
  0 siblings, 1 reply; 5+ messages in thread
From: Alex Schuilenburg @ 2003-08-27 12:00 UTC (permalink / raw)
  To: Satish Kumar; +Cc: ecos-discuss

Satish Kumar wrote:

> Hi,
>      i m interested in knowing the performance of eCos with respect
> to microITRON rtos..can anybody give me some tips or suggestions
> regarding their performance, code size..etc..

Are you referring to the microITRON native OS and the eCos native OS, or 
are you referring to the microITRON API layer provided with eCos?

I understand from our customers using the eCos uITRON layer that 
performance of the eCos uITRON is roughly the same as the native uITRON 
with minor differences in certain areas.  eCos uTIRON is slightly slower 
and larger, as could be expected running a compatability layer on top of 
eCos, but the timing differences were negligible to our customers anyway.

I would therefore expect eCos native applications to be smaller and 
faster then uITRON native applications (given the removal of the uITRON 
API layer) but have not done any direct comparisons and am not aware of 
any other comparisons between uITRON and eCos.

eCos is freely available for download, including tools, so please try 
eCos and compare against uITRON for yourself.  Also feel free to publish 
the results here :-)

Cheers
-- Alex

> 
> thanks
> satish


-- 
Managing Director / CEO                           eCosCentric Limited
http://www.ecoscentric.com/              The eCos and RedBoot experts



-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [ECOS] eCos or microITRON
  2003-08-27 12:00 ` Alex Schuilenburg
@ 2003-08-28  1:13   ` Satish Kumar
  2003-08-28  8:14     ` Daniel Morris
  2003-08-28 13:06     ` Grant Edwards
  0 siblings, 2 replies; 5+ messages in thread
From: Satish Kumar @ 2003-08-28  1:13 UTC (permalink / raw)
  To: Alex Schuilenburg; +Cc: ecos-discuss

well, i m referring to the microITRON native OS and the eCos native OS.
i see that the eCos code size is very big..it goes in Mb`s attimes..this
is with out including ITRON API package..i was wondering if i develop
some ITRON applications on eCos by including the ITRON package..will the
performance be very low when compared to the ITRON native applications.

  basically i m more curious about the eCos kernel size and its
performance when compared to the miroITRON native OS. how can i research
on this..can somebody direct me some fast approach to work on this 
problem.

      my customer is interested to know, whether eCos kernel can be 
used as the base kernel and develop some ITRON and POSIX applications on
top of it. we are not sure if this approach could degrade the
performance when compared to developing the same applications on ITRON
native OS.

can some one share their experince on the above..

thanks
satish


On Wed, 27 Aug 2003 13:00:51 +0100
Alex Schuilenburg <alexs@ecoscentric.com> wrote:

> Satish Kumar wrote:
> 
> > Hi,
> >      i m interested in knowing the performance of eCos with respect
> > to microITRON rtos..can anybody give me some tips or suggestions
> > regarding their performance, code size..etc..
> 
> Are you referring to the microITRON native OS and the eCos native OS, or 
> are you referring to the microITRON API layer provided with eCos?
> 
> I understand from our customers using the eCos uITRON layer that 
> performance of the eCos uITRON is roughly the same as the native uITRON 
> with minor differences in certain areas.  eCos uTIRON is slightly slower 
> and larger, as could be expected running a compatability layer on top of 
> eCos, but the timing differences were negligible to our customers anyway.
> 
> I would therefore expect eCos native applications to be smaller and 
> faster then uITRON native applications (given the removal of the uITRON 
> API layer) but have not done any direct comparisons and am not aware of 
> any other comparisons between uITRON and eCos.
> 
> eCos is freely available for download, including tools, so please try 
> eCos and compare against uITRON for yourself.  Also feel free to publish 
> the results here :-)
> 
> Cheers
> -- Alex
> 
> > 
> > thanks
> > satish
> 
> 
> -- 
> Managing Director / CEO                           eCosCentric Limited
> http://www.ecoscentric.com/              The eCos and RedBoot experts
> 
> 
> 
> -- 
> Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
> and search the list archive: http://sources.redhat.com/ml/ecos-discuss

-- 
Satish Kumar <satish@bvt.sc.sanyo.co.jp>


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [ECOS] eCos or microITRON
  2003-08-28  1:13   ` Satish Kumar
@ 2003-08-28  8:14     ` Daniel Morris
  2003-08-28 13:06     ` Grant Edwards
  1 sibling, 0 replies; 5+ messages in thread
From: Daniel Morris @ 2003-08-28  8:14 UTC (permalink / raw)
  To: Satish Kumar; +Cc: Alex Schuilenburg, ecos-discuss

On Thu, Aug 28, 2003 at 10:13:54AM +0900, Satish Kumar wrote:
> well, i m referring to the microITRON native OS and the eCos native OS.
> i see that the eCos code size is very big..it goes in Mb`s attimes..this

How are you determining this size?

> is with out including ITRON API package..i was wondering if i develop
> some ITRON applications on eCos by including the ITRON package..will the
> performance be very low when compared to the ITRON native applications.
> 
>   basically i m more curious about the eCos kernel size and its

Crudely, I'd expect it to be of the order of 20-30 KiB when built
from a default configuration (ie there is significant scope to 
reduce this depending on the features that you wish to have the
kernel provide and the drivers the image will include).

> performance when compared to the miroITRON native OS. how can i research
> on this..can somebody direct me some fast approach to work on this 
> problem.

You can run through the results of tm_basic and compare those with
any published figures for your ITRON native OS - please be sure to
read the section in the reference guide regarding how the Real Time
Characterisation tests are implemented - just to be sure that you
are comparing apples with apples, as such measurements are always
subject to the interpretation of the test implementer.

> 
>       my customer is interested to know, whether eCos kernel can be 
> used as the base kernel and develop some ITRON and POSIX applications on
> top of it. we are not sure if this approach could degrade the
> performance when compared to developing the same applications on ITRON
> native OS.

As already discussed, there is bound to be some tradeoff of speed of
execution & size when calling through a library against portability. As 
Alex said, we've seen (from our customers) that the timing differences 
have been marginal. 

It sounds like you need to look at the most used features of ITRON from
your current application and come up with some realistic tests to run on 
both ITRON and eCos+POSIX or eCos+ITRON. Perhaps the key design decision 
here is whether the ITRON or POSIX layers are needed at all, if performance 
is that critical.

 Daniel


> 
> can some one share their experince on the above..
> 
> thanks
> satish
> 
> 
> On Wed, 27 Aug 2003 13:00:51 +0100
> Alex Schuilenburg <alexs@ecoscentric.com> wrote:
> 
> > Satish Kumar wrote:
> > 
> > > Hi,
> > >      i m interested in knowing the performance of eCos with respect
> > > to microITRON rtos..can anybody give me some tips or suggestions
> > > regarding their performance, code size..etc..
> > 
> > Are you referring to the microITRON native OS and the eCos native OS, or 
> > are you referring to the microITRON API layer provided with eCos?
> > 
> > I understand from our customers using the eCos uITRON layer that 
> > performance of the eCos uITRON is roughly the same as the native uITRON 
> > with minor differences in certain areas.  eCos uTIRON is slightly slower 
> > and larger, as could be expected running a compatability layer on top of 
> > eCos, but the timing differences were negligible to our customers anyway.
> > 
> > I would therefore expect eCos native applications to be smaller and 
> > faster then uITRON native applications (given the removal of the uITRON 
> > API layer) but have not done any direct comparisons and am not aware of 
> > any other comparisons between uITRON and eCos.
> > 
> > eCos is freely available for download, including tools, so please try 
> > eCos and compare against uITRON for yourself.  Also feel free to publish 
> > the results here :-)
> > 
> > Cheers
> > -- Alex
> > 
> > > 
> > > thanks
> > > satish
> > 
> > 
> > -- 
> > Managing Director / CEO                           eCosCentric Limited
> > http://www.ecoscentric.com/              The eCos and RedBoot experts
> > 
> > 
> > 
> > -- 
> > Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
> > and search the list archive: http://sources.redhat.com/ml/ecos-discuss
> 
> -- 
> Satish Kumar <satish@bvt.sc.sanyo.co.jp>
> 
> 
> -- 
> Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
> and search the list archive: http://sources.redhat.com/ml/ecos-discuss

-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [ECOS] eCos or microITRON
  2003-08-28  1:13   ` Satish Kumar
  2003-08-28  8:14     ` Daniel Morris
@ 2003-08-28 13:06     ` Grant Edwards
  1 sibling, 0 replies; 5+ messages in thread
From: Grant Edwards @ 2003-08-28 13:06 UTC (permalink / raw)
  To: Satish Kumar; +Cc: Alex Schuilenburg, ecos-discuss

On Thu, Aug 28, 2003 at 10:13:54AM +0900, Satish Kumar wrote:

> i see that the eCos code size is very big..it goes in Mb`s
> attimes..

Wha?

Where did you get the "MB's" from?

I've got an eCos configuration for ARM that includes TCP/IP,
SNMP, Ethernet driver, serial drivers for 32 ports, real time
clock and watchdog timer drivers. The total for OS and all
drivers isn't more that about 256KB (text).  If I add all my
application code and the GoAhead web server, SNMP server, and
some miscellaneous library routines I'm still under 512K.

I can add SSL and still be under 1MB.


>   basically i m more curious about the eCos kernel size

Basic kernel w/o networking is probably in the 50KB-100KB size
depending on options and architecture.  The single biggest
factor is network.  You probably need to figure 128KB of text
and another 128-256KB of data for one of the BSD TCP/IP stacks.

-- 
Grant Edwards
grante@visi.com

-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2003-08-28 13:06 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2003-08-27  4:02 [ECOS] eCos or microITRON Satish Kumar
2003-08-27 12:00 ` Alex Schuilenburg
2003-08-28  1:13   ` Satish Kumar
2003-08-28  8:14     ` Daniel Morris
2003-08-28 13:06     ` Grant Edwards

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).