Neels Hofmeyr has posted comments on this change. ( 
https://gerrit.osmocom.org/12180 )

Change subject: rsl: Send PDCH ACT NACK if TCH chan is still active
......................................................................


Patch Set 1: Code-Review+1

(3 comments)

the fixing logic looks sane, but I'm still having trouble wrapping my mind 
around the entire issue. Also the comments sound a bit weird.

https://gerrit.osmocom.org/#/c/12180/1//COMMIT_MSG
Commit Message:

https://gerrit.osmocom.org/#/c/12180/1//COMMIT_MSG@19
PS1, Line 19: rsl.c:2103 (bts=0,trx=0,ts=3,ss=0) Tx PDCH DEACT ACK
one thing I don't understand yet in the overall failure cause: is it an 
osmo-bsc bug? is it sending some PDCH while the ts is still active as TCH?

ok wait, in this log, the TS is in PDCH mode. BSC sends a PDCH DEACT. That's 
quite normal and allowed; and the bug is that osmo-bts NACKs that perfectly 
legal situation? (I somehow wonder how my dyn TS tests when implementing it 
could ever work then)


https://gerrit.osmocom.org/#/c/12180/1/src/common/rsl.c
File src/common/rsl.c:

https://gerrit.osmocom.org/#/c/12180/1/src/common/rsl.c@2207
PS1, Line 2207:                 /* TCH takes preference over PDCH so allow 
forcing PDCH DEACT,
plz explain: if a TS is in PDCH mode, sending a DEACT is not forcing anything. 
Doing a PDCH DEACT in PDCH is a normal expected message. Is the lchan->state 
not NONE in PDCH mode? I thought it is... maybe I'm confused from bsc code


https://gerrit.osmocom.org/#/c/12180/1/src/common/rsl.c@2208
PS1, Line 2208:                  * but forbid forcing PDCH ACT if lchan still 
active */
(say it without the word "forcing"?)



--
To view, visit https://gerrit.osmocom.org/12180
To unsubscribe, or for help writing mail filters, visit 
https://gerrit.osmocom.org/settings

Gerrit-Project: osmo-bts
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: I6d6d12aec10c801fe55012ca6e58d0bc8755b15d
Gerrit-Change-Number: 12180
Gerrit-PatchSet: 1
Gerrit-Owner: Pau Espin Pedrol <[email protected]>
Gerrit-Reviewer: Jenkins Builder (1000002)
Gerrit-Reviewer: Neels Hofmeyr <[email protected]>
Gerrit-Comment-Date: Thu, 06 Dec 2018 19:06:19 +0000
Gerrit-HasComments: Yes
Gerrit-HasLabels: Yes

Reply via email to