Re: [Bug 1910148] Re: Broadcast address detection is unreliable on Linux

["Johnson, Andrew N. via Core-talk" <core-talk at aps.anl.gov>]

Hi Chris,

On Jan 7, 2021, at 7:21 PM, Chris Johns via Core-talk <core-talk at aps.anl.gov> wrote:

On 6/1/21 2:47 am, mdavidsaver via Core-talk wrote:

Do you mean your patch to be a proof of concept, or to be accepted as
is?  If the later, compatibility with the embedded RTOS targets (RTEMS
4.x and vxWorks) needs to be maintained.  These targets do not provide
getifaddrs(), and would fail to compile with getifaddrs.patch.

Just thought I would mention RTEMS 5 provides getifaddrs. It is present in
source for the RTEMS legacy networking code and the newer rtems-libbsd
networking stack we now provide. As indicated elsewhere in this thread there is
no defines to indicate it is available.

Thanks, that’s good to know, and shouldn’t be a problem. Once we merge the RTEMS-5 support code into the epics-base/7.0 branch we will have 2 sets of sources for interfacing to RTEMS, one set for RTEMS-4 and one for RTEMS-5 (plus others that can be built for both). The way that Michael R has structured the latest version of this PR that means the new RTEMS-5 set should compile the new getifaddrs() version of the code, while the RTEMS-4 set will continue to use the original.

I think it’s likely that this PR will get merged before the RTEMS-5 port is ready, but work on that seems to be heating up.

- Andrew

-- 

Complexity comes for free, simplicity you have to work for.