hack: try to avoid "buffer full" problem when pushing DL traffic

This is hack-level development.

With this commit you can do UDP DL traffic of say 100Mb/s over
a 5MHz link with one connected UE and the eNB should not crash
because of memory exhaustion. Of course on the receiver side
you won't get 100Mb/s and many many lost packets. But the system
should not crash. 1Gb/s does not work. So in any case try to
remain within some reasonable limits. There is no reason to
push more than twice the maximum achievable throughput of
the link.

This work is based on a patch proposed by Francesco Gringoli.
This commit is contained in:
Cedric Roux
2018-03-09 10:18:12 +01:00
parent 963805820c
commit 71e7f9713a
4 changed files with 130 additions and 1 deletions
+12
View File
@@ -108,6 +108,18 @@ int itti_send_broadcast_message(MessageDef *message_p);
**/
int itti_send_msg_to_task(task_id_t task_id, instance_t instance, MessageDef *message);
/* TODO: this is a hack. Almost no caller of itti_send_msg_to_task checks
* the return value so it has been changed to crash the program in case
* of failure instead of returning -1 as the documentation above says.
* The RLC UM code may receive too much data when doing UDP at a higher
* throughput than the link allows and so for this specific case we need
* a version that actually returns -1 on failure.
*
* This needs to be cleaned at some point.
*/
/* look for HACK_RLC_UM_LIMIT for others places related to the hack. Please do not remove this comment. */
int itti_try_send_msg_to_task(task_id_t task_id, instance_t instance, MessageDef *message);
/** \brief Add a new fd to monitor.
* NOTE: it is up to the user to read data associated with the fd
* \param task_id Task ID of the receiving task