Experimental Physics and
Industrial Control System

"Feng, Shuchen" <[email protected]> · Thu, 24 Mar 2005 07:51:36 -0500

Hi all,

A while ago, wnolan wrote:

> I was curious what sort or rates are being observed using the catime
> utility for RTEMS installations. 

I finally got around to compare the CA connect rate and  catime between
RTEMS-mvme5500 and vxWorks-mvme5500.  The conclusion is that EPICS/RTEMS
and EPICS/vxWorks is competitive to each other while tested under the same
network setup.  

Many thanks to Mark Heron and Steve Singleton of Diamond Light Source
for collaborations in obtaining data for EPICS/vxWorks/mvme5500.

Till Straumann wrote :
> a) epics/rtems frequently uses 'malloc_free_space' which can be
>    a very expensive operation under RTEMS (designed to be a
>    diagnostic). AFAIK, this has been addressed and a more efficient
>    implementation is available with the development version (4.7).
>    An unofficial patch is available for 4.6 but my usual workaround
>    is to hack
>    libCom/osi/os/RTEMS/osdPoolStatus.c:osiSufficientSpaceInPool()
>    to always return 1.

>            RTEMS-svgm  vxworks-svgm    RTEMS-svgm  vxworks-svgm
>               (1000 channels)              (10000 channels)
>search       4.8 Mbps     3.0 Mbps      14.5 Mbps    11.1 Mbps
>pend       215.7 It/s   215.8 It/s     211.1 It/s   206.2 It/s
>async put    5.5 Mbps     6.2 Mbps      11.0 Mbps    11.5 Mbps
>async get   12.0 Mbps    10.6 Mbps      25.6 Mbps    19.7 Mbps
>sync  get    0.7 Mbps     0.4 Mbps       0.2 Mbps     0.2 Mbps

The average result of mvme5500 is attached at the end.  Table 1 is
that of the workaround, and Table 2 is that of the patched
malloc_free_space() taken on a different day.  I am happy with
the result of EPICS/RTEMS-mvme5500.

Regards,
Kate

--------------------------------------------------------------------------
			catime performance         (March,2005)
			    between  
		EPICS3.14.6/RTEMS: 4.6.0/mvme5500 BSP
			     and
	         EPIICS3.14.x/vxWorks/mvme5500 
----------------------------------------------------------------------------
--
Table 1: RTEMS/osdPoolStatus.c: osiSufficentSpaceInPool() to always
return 1 without calling  malloc_free_space():

            RTEMS-m5500  vxworks-m5500    RTEMS-m5500  vxworks-m5500
                (1000 channels)              (10000 channels)
search       6.2  Mbps     5.9  Mbps      30.3  Mbps    19.2  Mbps
pend       454.9K It/s   455.3K It/s     450.1K It/s   425.5K It/s
async put   16.8  Mbps    16.6  Mbps      30.6  Mbps    22.0  Mbps
async get   21.7  Mbps    23.0  Mbps      54.0  Mbps    34.5  Mbps
sync  get    1.6  Mbps     0.9  Mbps       1.0  Mbps     0.3  Mbp

----------------------------------------------------------------------------
--

Table 2: RTEMS/osdPoolStatus.c :osiSufficentSpaceInPool() calls
malloc_free_space(), which was patched with a more efficient
algorithm implemented by Joel Sherrill <[email protected]>:

            RTEMS-m5500  vxworks-m5500    RTEMS-m5500  vxworks-m5500
                (1000 channels)              (10000 channels)
search       5.8  Mbps     5.8  Mbps      27.1  Mbps    18.2  Mbps
pend       413.6K It/s   470.3K It/s     445.2K It/s   447.6K It/s
async put   15.4  Mbps    13.0 Mbps       30.4  Mbps    22.0  Mbps
async get   19.1  Mbps    15.0 Mbps       51.4  Mbps    32.5  Mbps
sync  get    1.4  Mbps     0.8 Mbps        0.4  Mbps     0.3  Mbp
--------------------------------------------------------------------

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System