Attention is currently required from: lynxis lazus.
6 comments:
Patchset:
(summing in past +1 by others to give a +2)
Commit Message:
Patch Set #5, Line 13: b) if the HLR do a Cancel Location Procedure with reason withdraw
@neels I'm a little bit unsure, if the MSC informs the user. […]
I think your idea is correct, that we simply drop the subscriber from the VLR cache. Maybe somewhere in the specs (in the VLR state machine graphs?) they say that some other action could be performed to immediately notify the MS, but as i see it that would be just some kind of optimization.
If there is an active conn, we should probably close it with a BSSMAP Clear, which I see in the code is done, so that should be fine.
Commit Message:
Patch Set #8, Line 15: d) duplicate entries (unsure if this is valid, but keeping the code as is)
duplicate entries are very important to prevent.
These obscure cases took some deep thinking to fix back in the days.
Search for "evil twin" in MSC_Tests.ttcn. If those tests still pass, you should be fine, IIRC
File include/osmocom/vlr/vlr.h:
Patch Set #8, Line 207: enum vlr_inval_reason {
these aren't from a spec, right? If they are, a comment to link to it would be great
.
*/
(weird comment line break)
File src/libmsc/gsm_04_08.c:
Patch Set #8, Line 1587: msc_a_release_cn(msc_a);
(re my comment above, this is the release i was talking about, so it is already done properly)
To view, visit change 38488. To unsubscribe, or for help writing mail filters, visit settings.