hi, all,
As we know, message is the only kind of IPC method that realizes in the QNX
kernel.But i have a question here: is these method faster than the shared
memory on the local node?
regards,
ycao
Hi.
As far as I know shared memory is faster than message passing. We mainly
use shared memory because we have a large number of processes which need to
access large amounts of memory.
On the other hand, shared memory is local to the node and message passing is
not. But I guess if you go across the node, memory access will be very slow
compared to memory access on the node.
Augie
“ycao” <ycao@mail.ipp.ac.cn> wrote in message
news:am62cf$mt2$1@inn.qnx.com…
hi, all,
As we know, message is the only kind of IPC method that realizes in the
QNX
kernel.But i have a question here: is these method faster than the shared
memory on the local node?regards,
ycao
\
ycao <ycao@mail.ipp.ac.cn> wrote:
hi, all,
As we know, message is the only kind of IPC method that realizes in the QNX
kernel.But i have a question here: is these method faster than the shared
memory on the local node?
regardless rate, message passing is recommended anyway. Andy
regards,
ycao
“ycao” <ycao@mail.ipp.ac.cn> wrote in message
news:am62cf$mt2$1@inn.qnx.com…
hi, all,
As we know, message is the only kind of IPC method that realizes in the
QNX
kernel.But i have a question here: is these method faster than the shared
memory on the local node?
No they are not, but if you Send() big chunks of data is almost as fast as a
memcpy. Of course shared memory will always be faster because data is not
copied!
regards,
ycao
\
In a past fast application I had success choosing shared memory rather than
ipc messaging to have access to a common set of variables among processes
within a node. I only had to use semaphores to keep the data integrity.
Oscar E. Soto
Control Adaptable SA de CV
Tel: +52 (81) 8340·0006
Fax: +52 (81) 8340·0014
“ycao” <ycao@mail.ipp.ac.cn> escribió en el mensaje
news:am62cf$mt2$1@inn.qnx.com…
hi, all,
As we know, message is the only kind of IPC method that realizes in the
QNX
kernel.But i have a question here: is these method faster than the shared
memory on the local node?regards,
ycao
\