Detecting Dual SIM on EM7455
Aleksander Morgado
aleksander at aleksander.es
Thu Apr 8 07:24:44 UTC 2021
Hey Jack,
>> >> Could you please test showing the failed case where, upon boot,
>> >> --uim-get-card-status shows one single SIM instead of two?
>> >>
>> > My apologies if this wasn't clear. The above case is the "failing" case. --uim-get-card-status shows two SIMs, it doesn't, however, show the ICCID for that card, which makes it appear from software as if that SIM card is not installed.
>>
>> Oh, I see. Then the issue has nothing to do with my assumption :)
>>
>> > I know there are "SIM Detect" lines on the SIM cards, however, we leave them both floating because we do not know if a SIM will be installed or not during production. We also do not have SIM slots that physically short that pin to ground if a SIM is inserted.
>>
>> If the modem reports a second SIM is available, even if it does not
>> read its ICCID, I would assume that the modem is doing some kind of
>> SIM probing of some sort on the secondary slot too. But then I'm not
>> sure; I believe I haven't seen before a case where SIM detect isn't
>> being actively used.
>
> Yeah it's not ideal but the hardware SIM detect line always reports "SIM Present" regardless of whether the SIM is there. Ideally it would probe the non-active SIM slot as well.
>>
>>
>> > Get Slot Status:
>> > root at EDG4600:~# qmicli -p -d /dev/lte/qmi_b --uim-get-slot-status
>> > [/dev/lte/qmi_b] Successfully got slots status
>> > [/dev/lte/qmi_b] 2 physical slots found:
>> > Physical slot 1:
>> > Card status: present
>> > Slot status: inactive
>> > ICCID: unknown
>>
>> Ok, present but ICCID unknown, as you said. You're actively querying
>> the modem with qmicli, so I would rule out a timing issue (e.g. I
>> don't think the SIM is taking too much to initialize or something like
>> that).
>>
>> > Physical slot 2:
>> > Card status: present
>> > Slot status: active
>> > Logical slot: 1
>> > ICCID: 89011703278374634023
>> >
>> > root at EDG4600:~# qmicli -p -d /dev/lte/qmi_b --uim-get-card-status
>> > [/dev/lte/qmi_b] Successfully got card status
>> > Provisioning applications:
>> > Primary GW: slot '1', application '1'
>> > Primary 1X: session doesn't exist
>> > Secondary GW: session doesn't exist
>> > Secondary 1X: session doesn't exist
>> > Slot [1]:
>> > Card state: 'present'
>> > UPIN state: 'not-initialized'
>> > UPIN retries: '0'
>> > UPUK retries: '0'
>> > Application [1]:
>> > Application type: 'usim (2)'
>> > Application state: 'ready'
>> > Application ID:
>> > A0:00:00:00:87:10:02:FF:FF:FF:FF:89:03:02:00:00
>> > Personalization state: 'ready'
>> > UPIN replaces PIN1: 'no'
>> > PIN1 state: 'disabled'
>> > PIN1 retries: '3'
>> > PUK1 retries: '10'
>> > PIN2 state: 'enabled-not-verified'
>> > PIN2 retries: '3'
>> > PUK2 retries: '10'
>> > Application [2]:
>> > Application type: 'isim (5)'
>> > Application state: 'detected'
>> > Application ID:
>> > A0:00:00:00:87:10:04:FF:FF:FF:FF:89:03:02:00:00
>> > Personalization state: 'unknown'
>> > UPIN replaces PIN1: 'no'
>> > PIN1 state: 'disabled'
>> > PIN1 retries: '3'
>> > PUK1 retries: '10'
>> > PIN2 state: 'not-initialized'
>> > PIN2 retries: '0'
>> > PUK2 retries: '0'
>> > Slot [2]:
>> > Card state: 'absent'
>> > UPIN state: 'not-initialized'
>> > UPIN retries: '0'
>> > UPUK retries: '0'
>> >
>> > *** Switch primary to slot 1
>> > A potentially unrelated issue is that it appears MM restarts when switching slots on a modem that needs to switch profiles as a result:
>> > Apr 7 10:56:07 EDG4600 ModemManager[15523]: <info> ModemManager (version 1.16.2) starting in system bus...
>> >
>> > I have not looked into this, only just noticed as I switched the slot...
>> >
>>
>> Oh that looks like a segfault. If you could get a backtrace and MM
>> debug log after running with "gdb --args /usr/sbin/ModemManager debug"
>> (then "r" for run, wait for it to crash, and "bt" for backtrace).
>> you'll need to make sure the system-started MM is stopped before doing
>> that also (e.g. "sudo systemctl stop ModemManager if using systemd")
>
>
> I'll look into that. We don't bundle gdb or gdbserver in our OpenWRT version but I will build those packages and see what I can get out of it.
>>
>>
>> > After changing SIM slots only slot info 1 is shown. If the profile doesn't switch, then the ICCID for both SIM's will be shown
>> (which is the example I shared previously of --uim-get-card-status):
>> > root at EDG4600:~# qmicli -p -d /dev/lte/qmi_b --uim-get-card-status
>> > [/dev/lte/qmi_b] Successfully got card status
>> > Provisioning applications:
>> > Primary GW: slot '1', application '1'
>> > Primary 1X: session doesn't exist
>> > Secondary GW: session doesn't exist
>> > Secondary 1X: session doesn't exist
>> > Slot [1]:
>> > Card state: 'present'
>> > UPIN state: 'not-initialized'
>> > UPIN retries: '0'
>> > UPUK retries: '0'
>> > Application [1]:
>> > Application type: 'usim (2)'
>> > Application state: 'ready'
>> > Application ID:
>> > A0:00:00:00:87:10:02:FF:86:FF:FF:89:FF:FF:FF:FF
>> > Personalization state: 'ready'
>> > UPIN replaces PIN1: 'no'
>> > PIN1 state: 'disabled'
>> > PIN1 retries: '3'
>> > PUK1 retries: '3'
>> > PIN2 state: 'enabled-not-verified'
>> > PIN2 retries: '3'
>> > PUK2 retries: '3'
>> > Slot [2]:
>> > Card state: 'absent'
>> > UPIN state: 'not-initialized'
>> > UPIN retries: '0'
>> > UPUK retries: '0'
>> > root at EDG4600:~# qmicli -p -d /dev/lte/qmi_b --uim-get-slot-status
>> > [/dev/lte/qmi_b] Successfully got slots status
>> > [/dev/lte/qmi_b] 2 physical slots found:
>> > Physical slot 1:
>> > Card status: present
>> > Slot status: active
>> > Logical slot: 1
>> > ICCID: 89860000502000180722
>> > Physical slot 2:
>> > Card status: present
>> > Slot status: inactive
>> > ICCID: unknown
>> >
>> > An EM7511 with dual SIM shows the ICCID for both SIM's on bootup:
>>
>> Are you using the latest EM7455 firmware available? Maybe this is an
>> issue already solved by a newer firmware.
>
> We have the following PRI's installed: Very useful feature in MM, much easier to consume the json output of this than the output of qmicli :)
You may be interested in this MR
https://gitlab.freedesktop.org/mobile-broadband/libqmi/-/merge_requests/190
:) I didn't find time yet to update qmicli to support that JSON output
in all commands, but it's doable and not really difficult.
> root at EDG4600:~# mmcli -m 1 --firmware-list
> ----------------
> Firmware | list: 02.32.11.00_ATT_002.070_000
> | current: yes
> | gobi pri unique id: 002.070_000
> | gobi modem unique id: ?_?
> | 02.24.05.06_BELL_001.005_000
> | current: no
> | gobi pri unique id: 001.005_000
> | gobi modem unique id: ?_?
> | 02.33.03.00_GENERIC_002.072_000
> | current: no
> | gobi pri unique id: 002.072_000
> | gobi modem unique id: ?_?
> | 02.32.11.00_ROGERS_001.040_000
> | current: no
> | gobi pri unique id: 001.040_000
> | gobi modem unique id: ?_?
> | 02.32.11.00_SPRINT_002.062_000
> | current: no
> | gobi pri unique id: 002.062_000
> | gobi modem unique id: ?_?
> | 02.32.11.00_TELUS_001.040_000
> | current: no
> | gobi pri unique id: 001.040_000
> | gobi modem unique id: ?_?
> | 02.33.03.00_VERIZON_002.079_001
> | current: no
> | gobi pri unique id: 002.079_001
> | gobi modem unique id: ?_?
> | 02.24.03.00_VODAFONE_001.001_000
> | current: no
> | gobi pri unique id: 001.001_000
> | gobi modem unique id: ?_?
>
> We have an AT&T and a private SIM installed so we are switching between 02.32.11.00_ATT_002.070_000 and 02.33.03.00_GENERIC_002.072_000 which I believe are the latest (the Sierra site was subjected to some sort of attack and "The Source" hasn't been back up since...)
>
Oh really? I knew it was down but didn't know anything about an attack.
> Apologies for my ignorance, but if libqmi is just querying the modem, and it is "working" (showing ICCID/IMSI) for both slots on the EM7511, could this just be a difference between the MDM9230 and MDM9250?
Could be a different between the firmware versions you're using in
both, yes. Have you tried to set it up in a way so that AUTO-SIM is
not enabled and you use 02.33.03.00_GENERIC for both slots? Does the
issue happen as well?
> I am not familiar enough with libqmi to understand how low of a level it looks when --uim-get-slot-status/--get-card-status is called.
Oh, it's very easy to understand. --uim-get-slot-status runs the "UIM
Get Slot Status" command, and prints whatever the modem replies. And
--uim-get-card-status runs the "UIM Get Card Status" command, and
prints whatever the modem replies. If you run the qmicli command with
--verbose you'll see all the messages transferred (including the UIM
client allocation/release commands).
--
Aleksander
https://aleksander.es
More information about the ModemManager-devel
mailing list