Skip to content

omamobile-modem-sit 0.3.9: the Pixel 9a's first SIM test, texts and data on its carrier's network - #356

Merged
derekross merged 7 commits into
derekross:mainfrom
W4HO-ham:modem-sit-cellular-psi
Nov 3, 2026
Merged

derekross merged 7 commits into
derekross:mainfrom
W4HO-ham:modem-sit-cellular-psi

Conversation

@W4HO-ham

@W4HO-ham W4HO-ham commented Nov 2, 2026 •

Copy link
Copy Markdown
Contributor

The Pixel 9a's first SIM test ran this afternoon, a day early, and it ends with the palmtop registered on LTE, mobile data over rmnet0, and texts both ways. Getting there took four changes to omamobile-modem-sit, each tried on its own boot, and the record is in MODEM-PLAN under the SIM runbook. No carrier name, APN or number is in the diff.

What was wrong: every SETUP_DATA_CALL answered status 36 within 150 ms, five times over five boots, with registration fine before and after. 36 is Android's REGULAR_DEACTIVATION, not "APN refused", so the runbooks that said so are corrected. A recording build showed that about a second after each refusal the modem sent three IND_REG_FAILED, two IND_NTW_STATE_CHANGED and an empty data-call list. The IND_REG_FAILED are LTE cause 10, implicitly detached (domain CS+PS, then PS twice). With a wait after the data permission (SET_MOBILE_DATA_STATE, SET_PS_SERVICE), the same indications come and go before the setup, and the setup succeeds. What triggers the detach isn't proven.

The commits:

  1. 0.3.5: no PDU session ID on cellular bearers. Stock reserves one only for a bearer over Wi-Fi and sends 0 on LTE and NR; sitd reserved one with ALLOC_PDU_SESSION. Records from 0.3.4 with an allocated ID are still released.
  2. MODEM-PLAN: status 36 is a regular deactivation, not a refused APN (tegu's and cheetah's runbooks).
  3. 0.3.6: a refused setup keeps its full record in a private file in RAM, the channel stays open five seconds logging each indication by name (a data-call list change decoded to status, CID, active, PDP type), the registration and data-call count are logged before and after the NR change, and the probe's first request waits 20 s (the CP answered it 9.4 s after ONLINE).
  4. 0.3.7: the setup request carries the APN-type bitmap at 0xf1 (DEFAULT for the internet bearer, IMS for the IMS one), as the 9a's stock builder writes it; ours sent 0 there under a comment calling it an MTU. The attach request keeps bytes 0xf1 to 0xf4 zero, as stock's does.
  5. 0.3.8: after the data permission, the setup waits until the modem has sent nothing for three seconds (at most twenty), keeping the raw indications privately, then reads the registration again and waits for it if it was lost, and only then sends the setup. This is the change that made the difference.
  6. MODEM-PLAN, DEVICES: the record, and the 9a's Modem cell.
  7. 0.3.9: the indications read during that wait and after a refusal are handed to the service's own handler too, so an SMS arriving then is stored and acknowledged, not only logged (found in review; tested for both windows). MODEM-PLAN's wording follows the evidence.

The first two code changes didn't fix it on their own but stay: they match stock's requests, checked against the 9a's own libraries (from the owner's factory vendor_a.img; layouts verified, offsets in the commit messages). Stock's handlers send the two permission requests separately (SET_PS_SERVICE's success path sleeps 500 ms) and its setup path doesn't send them; when Android's callers do isn't in the libraries. Sending them before the last registration check is a reasonable next try, with the wait kept; I'd rather land what ran first.

Tested on the palmtop (2026-11-02, the daily kernel, omamobile-modem 0.1.10, omamobile-modem-sit 0.3.8 from this branch; 0.3.9 only changes where those indications go, and its tests cover it): CP ONLINE in 9 s; probe: card state 1, a USIM and an ISIM ready, SMSC present; the RIL: registered on LTE at home, the six indications came and went in the wait, the internet bearer up (status 0, CID 1, PDP type 3) and the IMS bearer after it (CID 2, PDP type 2), AIMS "LTE, IMS voice and SMS registered", "Calls and SMS service ready"; omamobile-sit data on: rmnet0 with an address, the default route at metric 700, three pings answered (78 ms), an HTTPS GET returned 200; the carrier's welcome text, a text from another phone and a text sent from the palmtop all delivered; radio off, the RIL stopped, an attended reboot, the six NV partitions unchanged against the backup, UFS error handler 0 throughout. The Pixel 7 Pro uses the same package and hasn't run with a SIM; its stock builder fills the same fields the same way.

packages/omamobile-modem-sit/test/all (unit tests and the RIL against sit-sim, with new cases for the session ID, the bitmap, the attach bytes, the kept record, the listen and the settle) and tools/pr-check pass.

…stock

The palmtop's first SIM test registered on LTE and then got status 36 (REGULAR_DEACTIVATION) for the internet bearer, twice. Our setup differed from stock's in the PDU session ID: sitd reserved one with ALLOC_PDU_SESSION and sent it at 0xf6, but Android allocates one only for a bearer over Wi-Fi (IWLAN), and the 9a's own protocol library copies the framework's value, 0 on LTE and NR, into the request. Both bearers now send 0 and the modem picks the ID; an older ownership record with an allocated ID is still released on deactivation.

sit-sim logs the ID it receives and the simulator test checks for 0 with no ALLOC or REL. The 0xf1 comment now says what stock puts there: the APN-type bitmap, not an MTU. It still sends zero.
The modem reports Android DataFailCause numbers, which stock's RIL passes through unchanged. A refused APN is 27 or 33. 36 is REGULAR_DEACTIVATION. The 9a's runbook pointed at the APN for the status it got on 2026-11-02.
…r it, and snapshot around the NR change

A refused SETUP_DATA_CALL answered "36" three times on the palmtop on 2026-11-02 and nothing else was kept: the record beyond its first five bytes was dropped, and the two or three messages the modem sent about a second later were lost because sitd had already closed the channel. Now the whole record goes to a private file in RAM, the channel stays open five seconds while each indication is logged by name and length (a data-call list change decoded to status, CID, active and PDP type), and the registration state and data-call count are logged before and after the NR change, so the next refusal says more than a number.

The probe's first request waits 20 s: the CP answered GET_SIM_STATUS 9.4 s after ONLINE.
…L, as stock

The 983-byte setup request carries the APN's supported-types bitmap as a u32 at 0xf1: the Pixel 9a's stock builder copies ApnSetting's mask there (vendor.radio.base ApnSetting::SetSupportedTypesBitmask, member +0x160; the stream library's builders write it at 0xf1). Ours sent 0 there, under a comment calling it an MTU, and the 9a's modem answered every internet setup with status 36 (REGULAR_DEACTIVATION) on 2026-11-02, three times, with the session ID fixed in between. The internet bearer now sends DEFAULT (1) and the IMS bearer IMS (0x40).

The attach request copies the setup's first 233 bytes, so it keeps bytes 0xf1 to 0xf4 zero explicitly: stock's 251-byte attach builder leaves them zero. Its roaming PDP type at 0xf8 is now named for what it is.
…sion before the setup

On 2026-11-02 five setups in a row were answered with status 36 within 150 ms, each followed within a second by three IND_REG_FAILED, two IND_NTW_STATE_CHANGED and an empty data-call list: the modem was re-attaching around the request, whatever its cause. The setup now waits until the modem has sent nothing for three seconds after the data permission (at most twenty), logging each indication by name and keeping the raw packets in a private file for decoding, then reads the registration again and waits up to a minute for it if it was lost, and only then sends the setup.
Six CP starts over six boots, all attended: the SIM read, LTE registration every time, five data requests refused with status 36, then, with the RIL letting the modem settle after the data permission, the internet and IMS bearers up, mobile data over rmnet0, and texts both ways. The record is under the runbook; the Modem cell stays WIP with what works on the owner's kernel.
@W4HO-ham
W4HO-ham force-pushed the modem-sit-cellular-psi branch from e3f3869 to 211811a Compare November 2, 2026 22:47
…each the service

0.3.6 and 0.3.8 read indications themselves while the setup waited and after a refusal, and only logged and kept them: an SMS arriving in that window got neither stored nor acknowledged. They're now handed to the service's own handler as well, as everywhere else, with a test for an SMS in both windows. MODEM-PLAN now says what the run showed and no more: the three IND_REG_FAILED were LTE cause 10 (implicitly detached), whether the data permission triggers it isn't proven, and only byte 132 of the refused record is the IPC extension version.
@W4HO-ham

W4HO-ham commented Nov 3, 2026

Copy link
Copy Markdown
Contributor Author

For when you get to these: the phone-software side is a stack, and all of it has run on the palmtop: this PR (registration, data, texts), then #364 (an unconfirmed call release leaves the modem for an attended reboot instead of stopping it), then #367 (speakerphone call audio on the 9a, opt-in). On 2026-11-03 they ran together on a kernel built from omamobile plus the open kernel PRs (linux-omarchy-mobile#187 with #185 and #162): texts both ways, mobile data, and calls in and out with audio both ways. #363 (the modem package's dependency) is independent and lets this install without -dd.

@derekross

Copy link
Copy Markdown
Owner

Thank you for this, and for running it one change per boot with the NV checksums before and after every one: that record in MODEM-PLAN is exactly how I'd like a first SIM test written up. I went through the setup, attach and settle code against the simulator and ran the package tests on #356, #364 and #367 merged together; all pass, and tools/pr-check is clean.

From my side: my Verizon-network work on the 7 Pro (#381) sits on top of this branch, and your data-call changes (session ID 0, the bitmap, the settle and the registration re-check) went through unchanged in last night's run on the 7 Pro, where data came up too. So this changes the 7 Pro's data path deliberately and has been run there.

Two small things for a later commit, neither blocking:

  • The comment at 0xf1 in data_call.py reads as if sending 0 there caused the 36s, but your own record shows 0.3.7 alone still got 36 and the settle fixed it.
  • I believe AOSP's ApnSetting.TYPE_DEFAULT is DEFAULT|HIPRI (0x11), so stock may send 0x11 for an internet APN rather than 1. Only if you have the builder open anyway; 1 works on both phones.

Merging this first. #381 lands right after it, so #364 and #367 will need a rebase on top; details on each.

@derekross
derekross merged commit e89e8aa into derekross:main Nov 3, 2026
2 checks passed
derekross added a commit that referenced this pull request Nov 3, 2026
…s network in seconds, carry data, show the modem in the shell (#381)

* omamobile-modem-sit 0.3.5: no PDU session ID on cellular bearers, as stock

The palmtop's first SIM test registered on LTE and then got status 36 (REGULAR_DEACTIVATION) for the internet bearer, twice. Our setup differed from stock's in the PDU session ID: sitd reserved one with ALLOC_PDU_SESSION and sent it at 0xf6, but Android allocates one only for a bearer over Wi-Fi (IWLAN), and the 9a's own protocol library copies the framework's value, 0 on LTE and NR, into the request. Both bearers now send 0 and the modem picks the ID; an older ownership record with an allocated ID is still released on deactivation.

sit-sim logs the ID it receives and the simulator test checks for 0 with no ALLOC or REL. The 0xf1 comment now says what stock puts there: the APN-type bitmap, not an MTU. It still sends zero.

* MODEM-PLAN: status 36 is a regular deactivation, not a refused APN

The modem reports Android DataFailCause numbers, which stock's RIL passes through unchanged. A refused APN is 27 or 33. 36 is REGULAR_DEACTIVATION. The 9a's runbook pointed at the APN for the status it got on 2026-11-02.

* omamobile-modem-sit 0.3.6: keep a refused setup's record, listen after it, and snapshot around the NR change

A refused SETUP_DATA_CALL answered "36" three times on the palmtop on 2026-11-02 and nothing else was kept: the record beyond its first five bytes was dropped, and the two or three messages the modem sent about a second later were lost because sitd had already closed the channel. Now the whole record goes to a private file in RAM, the channel stays open five seconds while each indication is logged by name and length (a data-call list change decoded to status, CID, active and PDP type), and the registration state and data-call count are logged before and after the NR change, so the next refusal says more than a number.

The probe's first request waits 20 s: the CP answered GET_SIM_STATUS 9.4 s after ONLINE.

* omamobile-modem-sit 0.3.7: send the APN-type bitmap in SETUP_DATA_CALL, as stock

The 983-byte setup request carries the APN's supported-types bitmap as a u32 at 0xf1: the Pixel 9a's stock builder copies ApnSetting's mask there (vendor.radio.base ApnSetting::SetSupportedTypesBitmask, member +0x160; the stream library's builders write it at 0xf1). Ours sent 0 there, under a comment calling it an MTU, and the 9a's modem answered every internet setup with status 36 (REGULAR_DEACTIVATION) on 2026-11-02, three times, with the session ID fixed in between. The internet bearer now sends DEFAULT (1) and the IMS bearer IMS (0x40).

The attach request copies the setup's first 233 bytes, so it keeps bytes 0xf1 to 0xf4 zero explicitly: stock's 251-byte attach builder leaves them zero. Its roaming PDP type at 0xf8 is now named for what it is.

* omamobile-modem-sit 0.3.8: let the modem settle after the data permission before the setup

On 2026-11-02 five setups in a row were answered with status 36 within 150 ms, each followed within a second by three IND_REG_FAILED, two IND_NTW_STATE_CHANGED and an empty data-call list: the modem was re-attaching around the request, whatever its cause. The setup now waits until the modem has sent nothing for three seconds after the data permission (at most twenty), logging each indication by name and keeping the raw packets in a private file for decoding, then reads the registration again and waits up to a minute for it if it was lost, and only then sends the setup.

* device/tegu MODEM-PLAN, DEVICES: the SIM test passed on 2026-11-02

Six CP starts over six boots, all attended: the SIM read, LTE registration every time, five data requests refused with status 36, then, with the RIL letting the modem settle after the data permission, the internet and IMS bearers up, mobile data over rmnet0, and texts both ways. The record is under the runbook; the Modem cell stays WIP with what works on the owner's kernel.

* omamobile-modem-sit 0.3.9: indications seen during the setup's wait reach the service

0.3.6 and 0.3.8 read indications themselves while the setup waited and after a refusal, and only logged and kept them: an SMS arriving in that window got neither stored nor acknowledged. They're now handed to the service's own handler as well, as everywhere else, with a test for an SMS in both windows. MODEM-PLAN now says what the run showed and no more: the three IND_REG_FAILED were LTE cause 10 (implicitly detached), whether the data permission triggers it isn't proven, and only byte 132 of the refused record is the IPC extension version.

* omamobile-modem-sit 0.3.16, omamobile-mmd 0.1.1: register on Verizon's network in seconds, carry data, and show the modem in the shell

Tested on a Pixel 7 Pro with a Verizon-network MVNO SIM on 2026-11-02. On top of #356 (0.3.9).

- [network] attach_apn: some networks attach only on their IMS APN.
  Verizon's never registered while the internet APN was the attach APN,
  and registered once the IMS APN was. Empty means "same as apn".
- [profiles]: every modem-managed data profile in stock's
  SET_DATA_PROFILE layout (IDs, APN-type bits, PDP types), so the modem
  has the carrier's full table before radio on.
- Stock's order: attach APN, profiles and NR mode, then radio on once
  with no AT+CFUN cycle. NR mode is set right after radio on when the
  modem refuses it with the radio off (error 2). Registration went from
  53 s to 10+ min down to 3.3 s after radio on.
- omamobile-sit-data finds the rmnet device that carries the bearer by
  probing; it isn't rmnet<CID-1> once the attach is on another APN. A stop
  during the probe puts every candidate back.
- /run/omamobile-sit stays 0710 root:wheel (SIT clients no longer chmod
  it to 0700), so the shell's tile can read state.json again.
- IMS: P-CSCF extension decoding, the P-CSCF request in an IMS-APN
  attach, and [network] ims_pcscf_required (yes keeps today's behaviour).
- [network] carrier_manifest: answers the modem's own carrier-config
  request (0x4602) with one OEM_SET_CARRIER_CONFIG_MANIFEST naming a
  manifest from the owner's image, and holds the IMS stack start until
  the modem's status. Empty by default; nothing reads SIM identities.
  extract-vendor-chroot copies vendor/firmware/carrierconfig for it.
- Logging: runtime registration changes, IND_REG_FAILED causes,
  carrier-config status, IND_PHONE_RESET bytes and registration time.
- Radio state 10 (Android's RadioState ON) is "on", and the LTE signal
  comes from IND_SIGNAL_STRENGTH. omamobile-mmd exports both, so the
  Mobile tile and the bar show the network.

IMS still doesn't register without the carrier profile persisting in
the modem's NV; that is the NV overlay's PR.
device/cheetah/MODEM-PROVIDER-PROFILES.md records stock's databases,
order and APNs for this SIM type.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d

* omamobile-modem-sit 0.3.19: the carrier profile before radio on, the ISIM's P-CSCF, IMS failure logging

Tested on the reference Pixel 7 Pro on 2026-11-02 (Verizon-network SIM).

- With carrier_manifest set, the radio goes on only after the modem's
  carrier config status (0x4601), as stock orders it. Before, the status
  came 11 s after radio on and the modem stopped searching; now it
  registers in 3.3 s with the profile in place.
- [network] ims_pcscf_from_isim (default yes): when the IMS bearer's PCO
  has no P-CSCF, read the ISIM's EF_PCSCF (6F09) through SIM_IO. The
  transport allows SIM_IO only for GET RESPONSE / READ RECORD of that one
  file on the ISIM AID; nothing else of the ISIM is read or logged. FQDN
  entries are resolved over the IMS bearer, then the host resolver; IPv6
  results fill ADD_PDN's list. On this network the entry is an FQDN that
  resolves only on the carrier's IMS DNS, which the host can't reach (the
  modem owns that bearer), so IMS still doesn't register.
- omamobile-sitd.service also allows AF_INET/AF_INET6, for that lookup.
- IND_AIMS_REGISTRATION is logged (state, features, status, the stack's
  reason text; never the URI).
- NR mode is set right after radio on when the modem refuses it with the
  radio off (error 2).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d

* omamobile-modem-sit 0.3.20: request the IMS bearer as stock does, and IMS-SHANNON.md

From a read of stock's IMS service (ShannonIms) on the owner's image,
recorded in device/cheetah/IMS-SHANNON.md (request layouts, the order of
AIMS requests, the registration state and reason values):

- The IMS bearer takes the home IP type of the [profiles] entry typed ims
  for ims_apn (IPv4v6 on Verizon's network: 0xda=3, 0xdb=3), else IPv6
  as before.
- ADD_PDN carries IP type 3, the IPv4 address, IPv4 P-CSCFs and IPv4 DNS
  for an IPv4v6 bearer; an IPv6-only bearer gives the same bytes as 0.3.19.
- Stock's OMA-DM hidden-menu item goes out before ADD_PDN.
- The ISIM's P-CSCF FQDN is also resolved through the internet bearer's
  DNS, as stock resolves it on the default network.
- Registration indications log the state and reason names.

Stock never sends ADD_PDN for IMS without a P-CSCF, and status 5 is
IMS_SERVICE_REGISTER_FAIL, its answer to an empty list. On the 7 Pro
(2026-11-03) the IPv4v6 IMS bearer still carried no P-CSCF and the
ISIM's FQDN didn't resolve on the internet bearer's DNS, so with
ims_pcscf_required = yes sitd now skips IMS like stock. Whether the
line is provisioned for VoLTE is the open question.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d

---------

Co-authored-by: W4HO <87742609+W4HO-ham@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
derekross added a commit that referenced this pull request Nov 3, 2026
…99: SIM setup with no APN to type (#373)

* omamobile-modem-sit 0.3.12: the APNs from the SIM, and registration without one

A reseller SIM on Verizon's network needed "apn = vzwinternet" typed
into modem-sit.conf before sitd would start, and then registered only with
the IMS APN as the initial attach APN. A SIM should just work.

With [network] apn and attach_apn at "auto" (the new default; empty means
the same), sitd sends GET_IMSI once after the USIM is READY, keeps only the
MCC/MNC and looks them up in the project's apn-overrides.conf, then in the
system's mobile-broadband-provider-info (CC-PDDC, now a dependency). The
attach APN is an override's attach_apn, else a database APN typed "ia",
else the internet APN. The only override is Verizon's codes: internet
vzwinternet (Verizon's own page), attach ims (our capture on 2026-11-02).
A typed value always wins. Only the source is logged, never the IMSI, the
network code or an APN; status carries the APNs in use and the SIM's PIN1
state for the shell.

Without an APN sitd no longer stops before the radio: it registers, skips
the internet bearer and goes on to IMS, and reports data unavailable.

[data] roaming (default yes, the behaviour so far) drops the cellular route
while roaming; omamobile-sit-data reads it on every pass. The new
omamobile-sit-config, run through pkexec (polkit: the active user), changes
only apn, attach_apn and roaming, and restarts nothing: an APN applies at
the next start-up, which needs a fresh CP.

tools/sit-sim answers GET_IMSI with a made-up subscriber part and logs the
attach APN. Overlaps #356/#364/#367 (0.3.9-0.3.11) in bringup.start(); the
attach_apn key matches #356's line.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d

* omamobile-mmd 0.1.1: the APNs sitd's start-up used, not "auto"

With sitd 0.3.12 the config's APN can be "auto". The bearer's apn and
Modem3gpp.InitialEpsBearerSettings now show what sitd's status reports
(the internet APN and the attach APN), and fall back to the config's
values, with "auto" shown as no APN.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d

* omamobile-shell 0.2.93: a Mobile network page, and the SIM in setup

omamobile.mobilenetwork, opened from Setup > Mobile network (only where the
modem stack is) and the Mobile data tile's chevron: carrier, registration,
signal, Mobile data and Data roaming switches, the APN in use with a field
to type one, the attach APN under Advanced, and the SIM PIN lock. "No modem"
without the stack. omamobile-mobile-network reads sitd's public state, its
config and ModemManager, and changes settings through omamobile-sit-config
under pkexec; it never restarts the modem services, so an APN change offers
a restart of the phone.

omamobile-onboarding names the carrier and says mobile data is set up
automatically when a SIM is in and the stack runs; nothing to type.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d

* TELEPHONY-APPS-PLAN: 7.6, SIM setup with no APN to type

The design of the automatic APN and attach APN, the carrier database and
its licence, the Mobile network page, what is logged, and the phone checks.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d

* omamobile-modem-sit 0.3.12: Verizon's attach APN is vzwims, as captured

The override said "ims", which was never tried as an attach APN. The SIM on
Verizon's network registered on 2026-11-02 (20:26 and 20:35) with vzwims as
the attach APN and vzwinternet for internet, so the override, its citation,
the comments, the design notes and the tests now say vzwims.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d

* omamobile-modem-sit 0.3.29: say when GET_IMSI is sent, and an override's IMS CID is the override's fault

Review of #373. MODEM-CLIENT.md's "no identity reads, but one" paragraph
and the apn comment in modem-sit.conf said GET_IMSI goes out when apn or
attach_apn is auto, from 0.3.12. choose_apn() sends it when any of apn,
attach_apn, ims_apn, ims_first or internet_cid is auto or empty, which
with the shipped defaults is every boot, and typing apn and attach_apn
alone still sends it for the bearer order; only typing all five reads no
IMSI. Both now say so, and that it landed in 0.3.29.

override_values() accepted internet_cid = 2 (the IMS bearer's CID), and
start() then stopped with a message blaming [network] internet_cid. The
override now refuses 2 itself, as it does 16: the value is left out, its
key is listed under 'refused', and choose_apn() logs that the project's
carrier override for this SIM had it, never the config and never the
network code. A test with such an override checks CID 1 and the message.

Also from the review: carrier_apn.py and omamobile-sit-config in
MODEM-CLIENT's module table, the [profiles] comment says the override's
profiles apply only when the SIM was read (not with all five keys typed),
and TELEPHONY-APPS-PLAN 7.6 names the versions after the merge with main
(shell 0.2.99; main's 0.3.28 is #364's).

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
W4HO-ham added a commit to W4HO-ham/omarchy-mobile that referenced this pull request Nov 11, 2026
The commits under this pull request's own are on main now in their merged form (derekross#356 to derekross#372, derekross#590), so the result is today's main plus this pull request's own commits, replayed on it; only version files conflicted, and they keep main's numbers.
W4HO-ham added a commit to W4HO-ham/omarchy-mobile that referenced this pull request Nov 11, 2026
The commits under this pull request's own are on main now in their merged form (derekross#356 to derekross#372, derekross#590), so the result is today's main plus this pull request's own commits, replayed on it; only version files conflicted, and they keep main's numbers.
W4HO-ham added a commit to W4HO-ham/omarchy-mobile that referenced this pull request Nov 11, 2026
The commits under this pull request's own are on main now in their merged form (derekross#356 to derekross#372, derekross#590), so the result is today's main plus this pull request's own commits, replayed on it; only version files conflicted, and they keep main's numbers.
W4HO-ham added a commit to W4HO-ham/omarchy-mobile that referenced this pull request Nov 11, 2026
The commits under this pull request's own are on main now in their merged form (derekross#356 to derekross#372, derekross#590), so the result is today's main plus this pull request's own commits, replayed on it; only version files conflicted, and they keep main's numbers.
robertzas pushed a commit to robertzas/omarchy-mobile that referenced this pull request Nov 14, 2026
…ata on its carrier (derekross#356)

* omamobile-modem-sit 0.3.5: no PDU session ID on cellular bearers, as stock

The palmtop's first SIM test registered on LTE and then got status 36 (REGULAR_DEACTIVATION) for the internet bearer, twice. Our setup differed from stock's in the PDU session ID: sitd reserved one with ALLOC_PDU_SESSION and sent it at 0xf6, but Android allocates one only for a bearer over Wi-Fi (IWLAN), and the 9a's own protocol library copies the framework's value, 0 on LTE and NR, into the request. Both bearers now send 0 and the modem picks the ID; an older ownership record with an allocated ID is still released on deactivation.

sit-sim logs the ID it receives and the simulator test checks for 0 with no ALLOC or REL. The 0xf1 comment now says what stock puts there: the APN-type bitmap, not an MTU. It still sends zero.

* MODEM-PLAN: status 36 is a regular deactivation, not a refused APN

The modem reports Android DataFailCause numbers, which stock's RIL passes through unchanged. A refused APN is 27 or 33. 36 is REGULAR_DEACTIVATION. The 9a's runbook pointed at the APN for the status it got on 2026-11-02.

* omamobile-modem-sit 0.3.6: keep a refused setup's record, listen after it, and snapshot around the NR change

A refused SETUP_DATA_CALL answered "36" three times on the palmtop on 2026-11-02 and nothing else was kept: the record beyond its first five bytes was dropped, and the two or three messages the modem sent about a second later were lost because sitd had already closed the channel. Now the whole record goes to a private file in RAM, the channel stays open five seconds while each indication is logged by name and length (a data-call list change decoded to status, CID, active and PDP type), and the registration state and data-call count are logged before and after the NR change, so the next refusal says more than a number.

The probe's first request waits 20 s: the CP answered GET_SIM_STATUS 9.4 s after ONLINE.

* omamobile-modem-sit 0.3.7: send the APN-type bitmap in SETUP_DATA_CALL, as stock

The 983-byte setup request carries the APN's supported-types bitmap as a u32 at 0xf1: the Pixel 9a's stock builder copies ApnSetting's mask there (vendor.radio.base ApnSetting::SetSupportedTypesBitmask, member +0x160; the stream library's builders write it at 0xf1). Ours sent 0 there, under a comment calling it an MTU, and the 9a's modem answered every internet setup with status 36 (REGULAR_DEACTIVATION) on 2026-11-02, three times, with the session ID fixed in between. The internet bearer now sends DEFAULT (1) and the IMS bearer IMS (0x40).

The attach request copies the setup's first 233 bytes, so it keeps bytes 0xf1 to 0xf4 zero explicitly: stock's 251-byte attach builder leaves them zero. Its roaming PDP type at 0xf8 is now named for what it is.

* omamobile-modem-sit 0.3.8: let the modem settle after the data permission before the setup

On 2026-11-02 five setups in a row were answered with status 36 within 150 ms, each followed within a second by three IND_REG_FAILED, two IND_NTW_STATE_CHANGED and an empty data-call list: the modem was re-attaching around the request, whatever its cause. The setup now waits until the modem has sent nothing for three seconds after the data permission (at most twenty), logging each indication by name and keeping the raw packets in a private file for decoding, then reads the registration again and waits up to a minute for it if it was lost, and only then sends the setup.

* device/tegu MODEM-PLAN, DEVICES: the SIM test passed on 2026-11-02

Six CP starts over six boots, all attended: the SIM read, LTE registration every time, five data requests refused with status 36, then, with the RIL letting the modem settle after the data permission, the internet and IMS bearers up, mobile data over rmnet0, and texts both ways. The record is under the runbook; the Modem cell stays WIP with what works on the owner's kernel.

* omamobile-modem-sit 0.3.9: indications seen during the setup's wait reach the service

0.3.6 and 0.3.8 read indications themselves while the setup waited and after a refusal, and only logged and kept them: an SMS arriving in that window got neither stored nor acknowledged. They're now handed to the service's own handler as well, as everywhere else, with a test for an SMS in both windows. MODEM-PLAN now says what the run showed and no more: the three IND_REG_FAILED were LTE cause 10 (implicitly detached), whether the data permission triggers it isn't proven, and only byte 132 of the refused record is the IPC extension version.
robertzas pushed a commit to robertzas/omarchy-mobile that referenced this pull request Nov 14, 2026
…s network in seconds, carry data, show the modem in the shell (derekross#381)

* omamobile-modem-sit 0.3.5: no PDU session ID on cellular bearers, as stock

The palmtop's first SIM test registered on LTE and then got status 36 (REGULAR_DEACTIVATION) for the internet bearer, twice. Our setup differed from stock's in the PDU session ID: sitd reserved one with ALLOC_PDU_SESSION and sent it at 0xf6, but Android allocates one only for a bearer over Wi-Fi (IWLAN), and the 9a's own protocol library copies the framework's value, 0 on LTE and NR, into the request. Both bearers now send 0 and the modem picks the ID; an older ownership record with an allocated ID is still released on deactivation.

sit-sim logs the ID it receives and the simulator test checks for 0 with no ALLOC or REL. The 0xf1 comment now says what stock puts there: the APN-type bitmap, not an MTU. It still sends zero.

* MODEM-PLAN: status 36 is a regular deactivation, not a refused APN

The modem reports Android DataFailCause numbers, which stock's RIL passes through unchanged. A refused APN is 27 or 33. 36 is REGULAR_DEACTIVATION. The 9a's runbook pointed at the APN for the status it got on 2026-11-02.

* omamobile-modem-sit 0.3.6: keep a refused setup's record, listen after it, and snapshot around the NR change

A refused SETUP_DATA_CALL answered "36" three times on the palmtop on 2026-11-02 and nothing else was kept: the record beyond its first five bytes was dropped, and the two or three messages the modem sent about a second later were lost because sitd had already closed the channel. Now the whole record goes to a private file in RAM, the channel stays open five seconds while each indication is logged by name and length (a data-call list change decoded to status, CID, active and PDP type), and the registration state and data-call count are logged before and after the NR change, so the next refusal says more than a number.

The probe's first request waits 20 s: the CP answered GET_SIM_STATUS 9.4 s after ONLINE.

* omamobile-modem-sit 0.3.7: send the APN-type bitmap in SETUP_DATA_CALL, as stock

The 983-byte setup request carries the APN's supported-types bitmap as a u32 at 0xf1: the Pixel 9a's stock builder copies ApnSetting's mask there (vendor.radio.base ApnSetting::SetSupportedTypesBitmask, member +0x160; the stream library's builders write it at 0xf1). Ours sent 0 there, under a comment calling it an MTU, and the 9a's modem answered every internet setup with status 36 (REGULAR_DEACTIVATION) on 2026-11-02, three times, with the session ID fixed in between. The internet bearer now sends DEFAULT (1) and the IMS bearer IMS (0x40).

The attach request copies the setup's first 233 bytes, so it keeps bytes 0xf1 to 0xf4 zero explicitly: stock's 251-byte attach builder leaves them zero. Its roaming PDP type at 0xf8 is now named for what it is.

* omamobile-modem-sit 0.3.8: let the modem settle after the data permission before the setup

On 2026-11-02 five setups in a row were answered with status 36 within 150 ms, each followed within a second by three IND_REG_FAILED, two IND_NTW_STATE_CHANGED and an empty data-call list: the modem was re-attaching around the request, whatever its cause. The setup now waits until the modem has sent nothing for three seconds after the data permission (at most twenty), logging each indication by name and keeping the raw packets in a private file for decoding, then reads the registration again and waits up to a minute for it if it was lost, and only then sends the setup.

* device/tegu MODEM-PLAN, DEVICES: the SIM test passed on 2026-11-02

Six CP starts over six boots, all attended: the SIM read, LTE registration every time, five data requests refused with status 36, then, with the RIL letting the modem settle after the data permission, the internet and IMS bearers up, mobile data over rmnet0, and texts both ways. The record is under the runbook; the Modem cell stays WIP with what works on the owner's kernel.

* omamobile-modem-sit 0.3.9: indications seen during the setup's wait reach the service

0.3.6 and 0.3.8 read indications themselves while the setup waited and after a refusal, and only logged and kept them: an SMS arriving in that window got neither stored nor acknowledged. They're now handed to the service's own handler as well, as everywhere else, with a test for an SMS in both windows. MODEM-PLAN now says what the run showed and no more: the three IND_REG_FAILED were LTE cause 10 (implicitly detached), whether the data permission triggers it isn't proven, and only byte 132 of the refused record is the IPC extension version.

* omamobile-modem-sit 0.3.16, omamobile-mmd 0.1.1: register on Verizon's network in seconds, carry data, and show the modem in the shell

Tested on a Pixel 7 Pro with a Verizon-network MVNO SIM on 2026-11-02. On top of derekross#356 (0.3.9).

- [network] attach_apn: some networks attach only on their IMS APN.
  Verizon's never registered while the internet APN was the attach APN,
  and registered once the IMS APN was. Empty means "same as apn".
- [profiles]: every modem-managed data profile in stock's
  SET_DATA_PROFILE layout (IDs, APN-type bits, PDP types), so the modem
  has the carrier's full table before radio on.
- Stock's order: attach APN, profiles and NR mode, then radio on once
  with no AT+CFUN cycle. NR mode is set right after radio on when the
  modem refuses it with the radio off (error 2). Registration went from
  53 s to 10+ min down to 3.3 s after radio on.
- omamobile-sit-data finds the rmnet device that carries the bearer by
  probing; it isn't rmnet<CID-1> once the attach is on another APN. A stop
  during the probe puts every candidate back.
- /run/omamobile-sit stays 0710 root:wheel (SIT clients no longer chmod
  it to 0700), so the shell's tile can read state.json again.
- IMS: P-CSCF extension decoding, the P-CSCF request in an IMS-APN
  attach, and [network] ims_pcscf_required (yes keeps today's behaviour).
- [network] carrier_manifest: answers the modem's own carrier-config
  request (0x4602) with one OEM_SET_CARRIER_CONFIG_MANIFEST naming a
  manifest from the owner's image, and holds the IMS stack start until
  the modem's status. Empty by default; nothing reads SIM identities.
  extract-vendor-chroot copies vendor/firmware/carrierconfig for it.
- Logging: runtime registration changes, IND_REG_FAILED causes,
  carrier-config status, IND_PHONE_RESET bytes and registration time.
- Radio state 10 (Android's RadioState ON) is "on", and the LTE signal
  comes from IND_SIGNAL_STRENGTH. omamobile-mmd exports both, so the
  Mobile tile and the bar show the network.

IMS still doesn't register without the carrier profile persisting in
the modem's NV; that is the NV overlay's PR.
device/cheetah/MODEM-PROVIDER-PROFILES.md records stock's databases,
order and APNs for this SIM type.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d

* omamobile-modem-sit 0.3.19: the carrier profile before radio on, the ISIM's P-CSCF, IMS failure logging

Tested on the reference Pixel 7 Pro on 2026-11-02 (Verizon-network SIM).

- With carrier_manifest set, the radio goes on only after the modem's
  carrier config status (0x4601), as stock orders it. Before, the status
  came 11 s after radio on and the modem stopped searching; now it
  registers in 3.3 s with the profile in place.
- [network] ims_pcscf_from_isim (default yes): when the IMS bearer's PCO
  has no P-CSCF, read the ISIM's EF_PCSCF (6F09) through SIM_IO. The
  transport allows SIM_IO only for GET RESPONSE / READ RECORD of that one
  file on the ISIM AID; nothing else of the ISIM is read or logged. FQDN
  entries are resolved over the IMS bearer, then the host resolver; IPv6
  results fill ADD_PDN's list. On this network the entry is an FQDN that
  resolves only on the carrier's IMS DNS, which the host can't reach (the
  modem owns that bearer), so IMS still doesn't register.
- omamobile-sitd.service also allows AF_INET/AF_INET6, for that lookup.
- IND_AIMS_REGISTRATION is logged (state, features, status, the stack's
  reason text; never the URI).
- NR mode is set right after radio on when the modem refuses it with the
  radio off (error 2).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d

* omamobile-modem-sit 0.3.20: request the IMS bearer as stock does, and IMS-SHANNON.md

From a read of stock's IMS service (ShannonIms) on the owner's image,
recorded in device/cheetah/IMS-SHANNON.md (request layouts, the order of
AIMS requests, the registration state and reason values):

- The IMS bearer takes the home IP type of the [profiles] entry typed ims
  for ims_apn (IPv4v6 on Verizon's network: 0xda=3, 0xdb=3), else IPv6
  as before.
- ADD_PDN carries IP type 3, the IPv4 address, IPv4 P-CSCFs and IPv4 DNS
  for an IPv4v6 bearer; an IPv6-only bearer gives the same bytes as 0.3.19.
- Stock's OMA-DM hidden-menu item goes out before ADD_PDN.
- The ISIM's P-CSCF FQDN is also resolved through the internet bearer's
  DNS, as stock resolves it on the default network.
- Registration indications log the state and reason names.

Stock never sends ADD_PDN for IMS without a P-CSCF, and status 5 is
IMS_SERVICE_REGISTER_FAIL, its answer to an empty list. On the 7 Pro
(2026-11-03) the IPv4v6 IMS bearer still carried no P-CSCF and the
ISIM's FQDN didn't resolve on the internet bearer's DNS, so with
ims_pcscf_required = yes sitd now skips IMS like stock. Whether the
line is provisioned for VoLTE is the open question.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d

---------

Co-authored-by: W4HO <87742609+W4HO-ham@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
robertzas pushed a commit to robertzas/omarchy-mobile that referenced this pull request Nov 14, 2026
…99: SIM setup with no APN to type (derekross#373)

* omamobile-modem-sit 0.3.12: the APNs from the SIM, and registration without one

A reseller SIM on Verizon's network needed "apn = vzwinternet" typed
into modem-sit.conf before sitd would start, and then registered only with
the IMS APN as the initial attach APN. A SIM should just work.

With [network] apn and attach_apn at "auto" (the new default; empty means
the same), sitd sends GET_IMSI once after the USIM is READY, keeps only the
MCC/MNC and looks them up in the project's apn-overrides.conf, then in the
system's mobile-broadband-provider-info (CC-PDDC, now a dependency). The
attach APN is an override's attach_apn, else a database APN typed "ia",
else the internet APN. The only override is Verizon's codes: internet
vzwinternet (Verizon's own page), attach ims (our capture on 2026-11-02).
A typed value always wins. Only the source is logged, never the IMSI, the
network code or an APN; status carries the APNs in use and the SIM's PIN1
state for the shell.

Without an APN sitd no longer stops before the radio: it registers, skips
the internet bearer and goes on to IMS, and reports data unavailable.

[data] roaming (default yes, the behaviour so far) drops the cellular route
while roaming; omamobile-sit-data reads it on every pass. The new
omamobile-sit-config, run through pkexec (polkit: the active user), changes
only apn, attach_apn and roaming, and restarts nothing: an APN applies at
the next start-up, which needs a fresh CP.

tools/sit-sim answers GET_IMSI with a made-up subscriber part and logs the
attach APN. Overlaps derekross#356/derekross#364/derekross#367 (0.3.9-0.3.11) in bringup.start(); the
attach_apn key matches derekross#356's line.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d

* omamobile-mmd 0.1.1: the APNs sitd's start-up used, not "auto"

With sitd 0.3.12 the config's APN can be "auto". The bearer's apn and
Modem3gpp.InitialEpsBearerSettings now show what sitd's status reports
(the internet APN and the attach APN), and fall back to the config's
values, with "auto" shown as no APN.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d

* omamobile-shell 0.2.93: a Mobile network page, and the SIM in setup

omamobile.mobilenetwork, opened from Setup > Mobile network (only where the
modem stack is) and the Mobile data tile's chevron: carrier, registration,
signal, Mobile data and Data roaming switches, the APN in use with a field
to type one, the attach APN under Advanced, and the SIM PIN lock. "No modem"
without the stack. omamobile-mobile-network reads sitd's public state, its
config and ModemManager, and changes settings through omamobile-sit-config
under pkexec; it never restarts the modem services, so an APN change offers
a restart of the phone.

omamobile-onboarding names the carrier and says mobile data is set up
automatically when a SIM is in and the stack runs; nothing to type.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d

* TELEPHONY-APPS-PLAN: 7.6, SIM setup with no APN to type

The design of the automatic APN and attach APN, the carrier database and
its licence, the Mobile network page, what is logged, and the phone checks.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d

* omamobile-modem-sit 0.3.12: Verizon's attach APN is vzwims, as captured

The override said "ims", which was never tried as an attach APN. The SIM on
Verizon's network registered on 2026-11-02 (20:26 and 20:35) with vzwims as
the attach APN and vzwinternet for internet, so the override, its citation,
the comments, the design notes and the tests now say vzwims.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d

* omamobile-modem-sit 0.3.29: say when GET_IMSI is sent, and an override's IMS CID is the override's fault

Review of derekross#373. MODEM-CLIENT.md's "no identity reads, but one" paragraph
and the apn comment in modem-sit.conf said GET_IMSI goes out when apn or
attach_apn is auto, from 0.3.12. choose_apn() sends it when any of apn,
attach_apn, ims_apn, ims_first or internet_cid is auto or empty, which
with the shipped defaults is every boot, and typing apn and attach_apn
alone still sends it for the bearer order; only typing all five reads no
IMSI. Both now say so, and that it landed in 0.3.29.

override_values() accepted internet_cid = 2 (the IMS bearer's CID), and
start() then stopped with a message blaming [network] internet_cid. The
override now refuses 2 itself, as it does 16: the value is left out, its
key is listed under 'refused', and choose_apn() logs that the project's
carrier override for this SIM had it, never the config and never the
network code. A test with such an override checks CID 1 and the message.

Also from the review: carrier_apn.py and omamobile-sit-config in
MODEM-CLIENT's module table, the [profiles] comment says the override's
profiles apply only when the SIM was read (not with all five keys typed),
and TELEPHONY-APPS-PLAN 7.6 names the versions after the merge with main
(shell 0.2.99; main's 0.3.28 is derekross#364's).

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants