Attention is currently required from: Hoernchen, laforge, fixeria.
12 comments:
File include/osmocom/gprs/gmm/gmm_pdu.h:
Patch Set #1, Line 16: tlv_definition
You should include `<osmocom/gsm/tlv.h>` defining this struct.
Ack
Patch Set #1, Line 32: uint8_t sres[4],
Why not a const pointer?
Ack
File libosmo-gprs-gmm.pc.in:
Patch Set #1, Line 9: libosmo-gprs-llc
btw, why is this needed? because of header files?
Because gmm_prim uses the llc_prim for the LLGMM API, and it uses the struct defintiions and functions to allocate the primitives.
That doesn't mean it forcefully needs the entire LLC stack in libosmo-gprs-llc to function though, since the app is the one forwarding the primitives.
File src/gmm/gmm_pdu.c:
Patch Set #1, Line 31: struct gprs_gmm_ms_net_cap {
Run `struct_endianess.py`.
Ack
`GSM_MAC_BLOCK_LEN`?
yeah, but this is all temporary to have some base upon to work with.
Patch Set #1, Line 147: rc = bitvec_unhex(&bv, "171933432b37159ef98879cba28c6621e72688b198879c00");
Makes sense to add a TODO here, since you're hard-coding the payload.
yeah, but this is all temporary to have some base upon to work with.
Patch Set #1, Line 172: msgb_put_u8(msg, sizeof(ms_net_cap_def));
Use `msgb_lv_put()`.
I'll check.
Patch Set #1, Line 274: GSM_MI_TYPE_TMSI
`GSM_MI_TYPE_IMEI`!
Ack
Patch Set #1, Line 312: (void)imeisv_requested;
What is this for?
IDK, I yet have to discover, that's why it's TODO :D but it shows up in the specs in the messages.
if you mean the (void) thing, it's to avoid having the "empty if clause" or alike warning. This is left here so that in the future we can quickly find out where to use it once we read the related specs.
File src/gmm/gmm_prim.c:
Patch Set #1, Line 107: gmm_llc_down_cb_dummy
`__func__`
I don't really expect this to change but ok.
Patch Set #1, Line 467: rc = gprs_gmm_prim_handle_unsupported(gmm_prim);
This case can fall-through to default, but not critical.
I did this on purpose since it's really the only primitive we expect to have here, so the final logic is already in place.
Patch Set #1, Line 589: llc_prim
Why would anybody pass you NULL pointer here? Just curious.
>Why would anybody pass you NULL pointer here? Just curious.
You can apply this question to literally any ASSERT checking a null pointer in the whole history of code.
To view, visit change 32021. To unsubscribe, or for help writing mail filters, visit settings.