Add missing KeyMgmt caps

[monich]

That should silence KeyMgmt SAE not known warning

[pvuorela]

With this I'm still getting the warning on sae.

Looks like there's another mapping in gsupplicant.c. Wonder why there are two of those.

[monich]

Hmm, do you have a backtrace?

[pvuorela]

Not sure does the backtrace matter much when the warning source is known, but since you asked added a breakpoint there:

Thread 1 "connmand" hit Breakpoint 1, gsupplicant_parse_bits_array (mask=0, count=3, map=0xb6b9b8ec <keymgmt_map>, value=0xbea9d950, name=0xb604b490 "KeyMgmt") at src/gsupplicant_util.c:187
187 src/gsupplicant_util.c: No such file or directory.
(gdb) bt
#0 gsupplicant_parse_bits_array (mask=0, count=3, map=0xb6b9b8ec <keymgmt_map>, value=0xbea9d950, name=0xb604b490 "KeyMgmt") at src/gsupplicant_util.c:187
#1 gsupplicant_parse_bits_array (name=name@entry=0xb604b490 "KeyMgmt", value=value@entry=0xb4c010c0, map=0xb6b9b8ec <keymgmt_map>, count=count@entry=3, mask=0) at src/gsupplicant_util.c:163
#2 0xb6b7e060 in gsupplicant_bss_parse_wpa (name=, value=, data=, name=, value=, data=) at src/gsupplicant_bss.c:489
#3 0xb6b7ddd6 in gsupplicant_dict_parse (data=0x15892cc, fn=0xb6b7e039 <gsupplicant_bss_parse_wpa>, dict=0x1) at src/gsupplicant_util.c:320
#4 gsupplicant_dict_parse (dict=, fn=, data=, dict=, fn=, data=) at src/gsupplicant_util.c:298
#5 0xb6b805ac in gsupplicant_bss_update_wpa (self=, self=) at src/gsupplicant_bss.c:531
#6 0xb6b835be in gsupplicant_bss_proxy_created (bus=, result=, data=) at src/gsupplicant_bss.c:1130
#7 0xb6ed4464 in g_task_return_now (task=0x152d400) at ../gio/gtask.c:1230
#8 0xb6ed4d06 in g_task_return (task=0x152d400, type=) at ../gio/gtask.c:1299
#9 0xb6ed5340 in g_task_return (type=G_TASK_RETURN_SUCCESS, task=) at ../gio/gtask.c:1839
#10 g_task_return_boolean (task=, result=) at ../gio/gtask.c:1843
#11 0xb6f1b1b4 in init_second_async_cb (source_object=0x158fe48, res=0x1581348, user_data=0x152d400) at ../gio/gdbusproxy.c:1776
#12 0xb6ed4464 in g_task_return_now (task=0x1581348) at ../gio/gtask.c:1230
#13 0xb6ed4d06 in g_task_return (task=0x1581348, type=) at ../gio/gtask.c:1299
#14 0xb6ed50f4 in g_task_return (type=G_TASK_RETURN_SUCCESS, task=) at ../gio/gtask.c:1726
#15 g_task_return_pointer (task=, result=, result_destroy=) at ../gio/gtask.c:1731
#16 0xb6f1aab8 in async_init_get_all_cb (connection=, res=, user_data=0x1581348) at ../gio/gdbusproxy.c:1404
#17 0xb6ed4464 in g_task_return_now (task=0xb6002480) at ../gio/gtask.c:1230
#18 0xb6ed4d06 in g_task_return (task=0xb6002480, type=) at ../gio/gtask.c:1299
#19 0xb6ed50f4 in g_task_return (type=G_TASK_RETURN_SUCCESS, task=) at ../gio/gtask.c:1726
#20 g_task_return_pointer (task=, result=, result_destroy=) at ../gio/gtask.c:1731
#21 0xb6f12e72 in g_dbus_connection_call_done (source=, result=0x152d4f0, user_data=0xb6002480) at ../gio/gdbusconnection.c:5879
#22 0xb6ed4464 in g_task_return_now (task=0x152d4f0) at ../gio/gtask.c:1230
#23 0xb6ed4498 in complete_in_idle_cb (task=0x152d4f0) at ../gio/gtask.c:1244
#24 0xb6d48f12 in g_main_dispatch (context=0x14cbac0) at ../glib/gmain.c:3417
#25 g_main_context_dispatch (context=context@entry=0x14cbac0) at ../glib/gmain.c:4135
#26 0xb6d49180 in g_main_context_iterate (context=0x14cbac0, block=block@entry=1, dispatch=dispatch@entry=1, self=) at ../glib/gmain.c:4211
#27 0xb6d49434 in g_main_loop_run (loop=0x14cb098) at ../glib/gmain.c:4411
#28 0x00023d8c in main (argc=, argv=) at src/main.c:1028

[monich]

Ah, it's parsing bss rather than interface caps. Thanks, will update the PR.

[pvuorela]

Looking more closely gsupplicant_bss_parse_wpa from the backtrace, looks like there's a third (!) mapping on gsupplicant_bss.c:484

[monich]

Right. There are three KeyMgmt maps in https://w1.fi/wpa_supplicant/devel/dbus.html and since the code is mirroring that, three sounds like the right number. Unfortunately the spec doesn't seem to be up to date, I'll have to dig through the wpa_supplicant code. Or just use the same mapping for all cases. I haven't made up my mind yet.

[monich]

I have decided to use the same mapping for all cases. Does that fix the problem?

[monich]

And reuse GSUPPLICANT_CIPHER map too.

[pvuorela]

Now the sae warning is gone, but this one getting printed a lot:
[gsupplicant] WARNING! Unexpected Group value

[monich]

I have to admit that I didn't test the changes on a real device 😐 Will do.

[monich]

Now it should become silent

[pvuorela]

Seems good now. Thanks.