Details
-
Type: New Feature
-
Status: Closed
-
Priority: Minor
-
Resolution: No Change Required
-
Affects Version/s: None
-
Fix Version/s: None
-
Labels:None
-
Account:SPC EthernetIp Slave (SPCETHERNE)
Description
This ticket was cloned from the linked ticket since I am not happy with what has been implemented and I think this should be checked again.
There are at least two resource stakeholders for multicast groups:
1. The EIP stack (different demands fro adapter and scanner)
2. The socket API
There is a global pool of multcast groups in LWIP which we have to dimension. The size of this pool must be at least the sum of at 1 & 2. This does not seem to be the case. Please investigate. This should not work or is at least unclear.
A shared pool of multicast addresses for multiple users was/is not supported by the EIP core, additional measures are required for proper system dimensioning and partitioning of the globbal pool between the users. Implemented concept is not clear to me.
I also doubt that it is was such a good idea to duplicate the default initializers from g_tLwip_PoolConfig_Small as magic values. Deep lwIP knowledge is required to understand these parameters and I see no benefit of having the hardcoded defaults copied.
Attachments
Issue Links
- clones
-
PSEISV5-449 Allow Socket API applications to use IP multicast communication by joining and leaving Multicast groups
- Closed