Repository navigation
omamobile-device-tegu 0.2.12: start the Pixel 9a's modem and phone service at boot, opt-in - #372
Conversation
|
Ran on the palmtop (2026-11-03, the daily kernel, so the bring-up modules path): enabled |
|
Cold boot with |
|
Thank you for this. Packaging the 9a's start with the latch before the drivers, the inhibitor taken first, and a boot unit that stays opt-in (with a test that holds it to that) is exactly the shape I want, and the cold boot you reported is great to see. Two things before it can merge:
Smaller points: the title says 0.2.10 and the version file 0.2.11, and the description still gives the old check-first order. The NV check now runs after the drivers, which TELEPHONY-PLAN §2.3 wanted the other way round: either restore a NV-only check up front or move it to the plan's open steps (the same for the transient unit, which §2.2 plans to drop). The signal-strength commit changes shared sitd, which now tells every CP, the 7 Pro's included, that the screen is on at each bring-up and never turns it off; #381 (now on main) already has the signal decode, so please drop that commit or make the screen-state part 9a-only. The in-tree driver branch hasn't run through the script yet, so please say so, and check |
… boot Derek asked for one with derekross#372: preconditions, one step per boot, Expect/Pass, stop conditions and rollback, with the cold boot, the clean shutdown that removes the latch and the UFS count recorded here.
d8be650 to
8bf2b07
Compare
|
Thanks. Done in the new top commit, rebased on main and #371:
Smaller points: the title and description are fixed. The NV-only check and dropping the transient unit stay as the plan's open items. The signal decode is dropped, and the screen-on at start is 9a-only through tegu's sitd drop-in. The description says the in-tree branch hasn't run through the script. |
…m stack at boot, opt-in (stacked on #384) (#386) * omamobile-modem-sit 0.3.21: attach as stock does on a Verizon network, IMS on the modem's attach bearer On a Verizon-network SIM, stock's RIL sends no SET_INITIAL_ATTACH_APN for the SIM's slot: the modem attaches with the carrier manifest's settings, lists its attach bearer unsolicited (CID 2, IPv6, with three P-CSCF addresses), and the IMS service announces that bearer with ADD_PDN_INFO. 0.3.20 sends an explicit attach APN, which most likely replaces the manifest's attach and its P-CSCF request, and its own IMS setup comes back without a P-CSCF, so IMS never registers. - [network] attach_apn = none skips SET_INITIAL_ATTACH_APN; the profiles and the internet bearer go out as before. - After registration, sitd waits up to 20 s for an active IPv6 or IPv4v6 bearer it didn't set up that carries a P-CSCF, reading GET_DATA_CALL_LIST and IND_DATA_CALL_LIST_CHANGED, and logs each item as CID, type, status, active, P-CSCF and DNS counts (never addresses). Found: no IMS setup of its own, ADD_PDN_INFO from that bearer. Not found: 0.3.20's path (own IMS bearer, then the ISIM fallback). - Bearers the modem brought up are recorded as modem-owned: activate() accepts them beside sitd's own and never uses their CID, deactivate() refuses them. - The shipped conf documents a Verizon-network example as comments, and IMS-SHANNON.md section 11 has the gated runbook. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d * omamobile-modem-sit 0.3.23: the IMS setup carries the IMS profile's APN types Stock's setup builder copies the APN's whole supported-type bitmask into the request (0xf1); for Verizon's IMS APN that is ims|ia (0x140), where ours sent IMS alone (0x40). The modem matches setups against its modem-cognitive profiles, and the ia bit marks the attach APN, so this is the one byte left where our IMS setup differs from stock's. With a [profiles] entry of the ims type for ims_apn, its bitmask is sent; without one nothing changes. IMS-SHANNON section 12 records the rest from libsitril: stock sends no SET_INITIAL_ATTACH_APN for a modem-cognitive attach profile (so attach_apn = none is its mode), the 0x0603 and 0x0613 layouts field by field, that stock never uses CID 1 (its attach context) for a bearer of its own, and a ranked list of further changes with a runbook. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d * omamobile-modem-sit 0.3.24: IMS registers on Verizon's network: IMS first on CID 2, the internet on CID 3 Tested on the reference Pixel 7 Pro, 2026-11-03 02:00, Verizon-network MVNO SIM, kernel 7.3.0-rc6-omamobile-g2219cc6cda77. The first IMS registration on this phone: "state 2 (VoLTE registered), features 3", then an SMS sent over IMS (AIMS_SEND_SMS accepted). Stock's RIL keeps CID 1 for the modem's attach context and never sets a bearer up on it; it sets the IMS bearer up first, on CID 2. On a network that attaches on the IMS APN, sitd's internet setup on CID 1 replaced the attach PDN, and the IMS bearer set up afterwards on CID 2 carried no P-CSCF (every variant tried: attach APN IMS or none, IPv6 or IPv4v6, P-CSCF requested, the carrier manifest, the ISIM's FQDN). With the IMS bearer first on CID 2 the setup response carries three IPv6 P-CSCFs and the modem's IMS stack registers within 2 s of ADD_PDN. - [network] internet_cid (default 1, today's behaviour; 3 here): the CID of sitd's internet bearer. IMS stays on CID 2. paths() resolves any non-IMS CID to the internet record, so omamobile-sit-data (no configuration) keeps working, and its probe found rmnet2. - [network] ims_first (default no): the IMS bearer and ADD_PDN before the internet bearer, as stock orders it; the internet comes up after the IMS setup whether or not IMS succeeded. - The IMS setup also reads the P-CSCF from the data-call list entry when the setup response has none (the modem lists the attach bearer's PCO there on some firmware), and the setup carries the IMS profile's whole APN-type bitmask (ims|ia), as stock's does (0.3.23). Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d * NOTES: 2026-11-03, IMS registered and an SMS sent on a Verizon-network SIM Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d * omamobile-modem 0.1.13: the modem stack at boot, opt-in (omamobile-modem-boot.target) Until 2026-11-03 the maintainer started the modem stack by hand every boot: omamobile-modem, then omamobile-sitd, omamobile-sit data on and omamobile-mmd. IMS registered on a Verizon-network SIM that night and an SMS went out at 02:00, the condition set for starting it at boot. Enabling the units one by one doesn't work: sitd Requires/After the modem unit, which was Type=simple, so it ran while the CP was still booting and exited ("CP is not ONLINE"). What changes, with nothing enabled by the package: - omamobile-modem-boot.target is the one switch (systemctl enable; the next release enables it once the carrier configuration is automatic, #373). It wants the four units; each unit's own After= keeps the order, a missing unit is simply not started, and DefaultDependencies=no keeps multi-user.target (and omamobile-boot-success after it) from waiting on a registration that can take minutes. No wants link of any kind is shipped, and test/all holds the package to that. - omamobile-modem-inhibit.service holds a systemd-inhibit sleep:idle block, required by and ordered before the modem unit, refused a manual stop, so it lasts until the reboot: suspend with the modem on has never run on this phone. The start script refuses to load a driver until the lock shows in systemd-inhibit --list. - omamobile-modem.service is Type=notify: ready once CPIF's ports exist and the profile's read-only check has passed, before the CP is powered on. The CP's ONLINE is waited for by the device's sitd drop-in. - The bring-up latch (modem-bringup-started) is written and synced before the first modprobe, as CONTRIBUTING 3.4 asks, not after the drivers and the check as before; a check that refuses removes it again, so a refusal doesn't cost the next boot its modem. - A second marker, modem-up, written by the watcher at ONLINE (date and boot id) and removed by the unit's ExecStopPost when it is this boot's: a boot that ends with the modem up and no stop, a hang or a reset, leaves it, and the next boot refuses until a human has looked, so a hang after ONLINE can't become a reset loop. Only under the unit (INVOCATION_ID): a start by hand would have nothing to remove it. - Nothing restarts the modem in place; the units have no Restart=. Tests: the script's order (refusals, the inhibitor, the latch, modprobe, the check with its rm, READY, exec), both refusals and --stopped and --retry run for real against scratch markers, the units' relations and the no-wants-link rule, systemd-analyze verify on the three units. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d * device-cheetah 0.1.41: sitd waits for the CP's ONLINE before it runs omamobile-modem.service (0.1.13) is ready once CPIF's ports exist, before cbd has booted the CP, 10-30 s more; a sitd started on that alone exits ("CP is not ONLINE", seen 2026-11-03 with both units started at once). This device's sitd drop-in adds ExecStartPre=omamobile-modem-cp wait-online 120, the profile's own cbd timeout: a CP that never comes up fails sitd, and the data fallback and the Phone and Messages bridge, ordered after it, stay down until the reboot. Nothing is enabled: the opt-in is omamobile-modem-boot.target, and the test checks that no wants link for it or the four units is shipped. 0.1.41 leaves 0.1.40 to #372's 10-device-modem.conf in the same directory, which moves sitd's Requires/After on the modem unit here; the two drop-ins are independent. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d * MODEM-PLAN: M4.6, the gated runbook for the modem stack at boot The one-switch form (omamobile-modem-boot.target, omamobile-modem 0.1.13, device-cheetah 0.1.41) with its reasons: install and look (B0), enable and reboot (B1), the chain in order with the inhibitor and the markers (B2), data and the bridge with one SMS (B3), the second reboot that proves the clean stop clears the modem-up marker (B4), and an optional forced reset that proves the guard refuses (B5). Expect/pass, stop and rollback per step; every reboot with the modem up costs one UFS pass. The old M4.6 line, enabling the three units one by one, is marked superseded: it started sitd before the CP was ONLINE. 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>
8bf2b07 to
1a24adb
Compare
|
Thank you, this is the shape I wanted: the latch on disk before any driver and kept until the profile's clean stop removes it, the preconditions and rollback in the README and the unit, the screen-on only through the 9a's drop-in, and the test holding the order. I read the ExecStopPost escaping and the clean-stop rule and agree it works. I'll merge it after #371 and #360 (the runbook lives in #360's §2.7). Two things, neither blocking: the boot unit's 300 s timeout can run out while |
…#360) * device/tegu TELEPHONY-PLAN: call audio and starting the modem at boot The SIM test passed with everything started by hand. This plans the two steps to a palmtop that works as a phone on its own: the call audio route from the 9a's stock audio HAL (EP5 down, EP4 up, the TAS2572 gain caveat, the AoC voice window to audit) with the safety gap in sitd's call release to close first, and an opt-in boot start for the modem with its guards and the packaging fix. * device/tegu TELEPHONY-PLAN: what has run, and which section 2 items #363 and #372 do * device/tegu TELEPHONY-PLAN: a gated runbook for enabling the modem at boot Derek asked for one with #372: preconditions, one step per boot, Expect/Pass, stop conditions and rollback, with the cold boot, the clean shutdown that removes the latch and the UFS count recorded here.
1a24adb to
073167d
Compare
|
Thanks. Rebased onto main and #371, with both points handled now:
The palmtop is back: this morning I started the boot unit by hand after boot (it was off for the overnight sleep test), and it came up registered for calls and texts, with the bridge and the agent running. The full record for §2.7 (phone-smoke, the UFS count after a shutdown with the CP up, nv-backup --check) comes at the next attended reboot. One more: |
5fc2d4f to
0b95247
Compare
|
Thank you, @W4HO-ham. Both points are in, and the comment above What's left is on my side. First I'll run the shared stop path on my 7 Pro, One heads-up: this PR and #409 both change the 9a's line in DEVICES (GNSS and |
|
Something you should know before your 7 Pro run of the stop path, since it is the same path. Nothing changes in this PR. Over 11-08 and 11-09 the palmtop did thirteen
So it is one in thirteen, cause unknown, not reproduced in one repeat, and I can't say whether the modem's stop was involved or the shutdown hung elsewhere. It's recorded in TELEPHONY-PLAN (with #590, |
|
A follow-up to the unconfirmed stop above. I went looking for it overnight and didn't find it again. A laptop loop rebooted the palmtop thirty times with the modem up (2026-11-09, 20:59 to 22:32 MST; build #100, the bring-up modules, the unit starting the modem at each boot): wait for the phone service and CP ONLINE, hold two minutes,
With the thirteen before, that is 42 confirmed stops in 43 on the 9a. The one remains unexplained, and I still can't tell whether the modem's stop had a part in it; the loop keeps pstore now, so a second one would come with its console. Recorded in TELEPHONY-PLAN with #590 ( |
|
Thank you for telling me about the unconfirmed stop before my run, and for going after it with thirty more reboots and keeping pstore this time. One in 43 with the latch doing exactly what it's for is a fair place to be, and "copy pstore first after a slow reboot" is a rule I'll follow on the 7 Pro too. Nothing has changed here since my last comment and I still need nothing from you. What's left is mine: the stop and reboot on my 7 Pro with the modem up, then the merge. Two notes on that so you aren't surprised: main has since taken omamobile-modem 0.1.17 (#568) and omamobile-modem-sit 0.3.33 (#569), so this will land as 0.1.18 and 0.3.34, with device-tegu at the next free number, and I'll write the title and body to match what the PR became. #590 will need a rebase once this is squashed. |
Main took 0.1.17 (derekross#568) and 0.3.33 (derekross#569) for other changes and omamobile-device-tegu 0.2.11 (derekross#591), so this branch's packages move to the next numbers and the README names the modem package version the stop command comes with.
|
#591's merge left this conflicting on |
|
The 7 Pro run is done, on this head ( Pixel 7 Pro,
So the shared stop path behaves on the 7 Pro as it did before, and nothing refuses. The NV overlay is off on this phone, so its capture wasn't exercised. On the 7 Pro rfsd doesn't exit during the stop (it is killed at the timeout), so the new "orderly stop" branch didn't run here; that fits what you wrote about Merging. #590 will need a rebase onto main after this; #594 is independent. |
…a's modem-cpif overlay and GNSS link driver (#594) Adds device/tegu/overlays/modem-cpif.dtso to the Pixel 9a's DTB_OVERLAYS (it needs a kernel with linux-omarchy-mobile#187), CONFIG_EXYNOS_GNSS_IF_SPI=m to the shared fragment (two modules that bind only to the 9a's nodes), and gnss_spi and gnssif to the 9a's guards, so they don't load by themselves now that the kernel has the 9a's GNSS nodes (linux-omarchy-mobile#237). Nothing here starts the modem; that stays #372's opt-in unit. MODEM-PLAN records six boots on the owner's palmtop on 2026-11-09, on the published kernel tip plus kernel #224, #236, #237 and #274's config: registration, a call, texts, data, GNSS fixes and confirmed stops, with the NV partitions unchanged. The Pixel 7 Pro gains two modules that never load. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
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.
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.
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.
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.
…m stack at boot, opt-in (stacked on derekross#384) (derekross#386) * omamobile-modem-sit 0.3.21: attach as stock does on a Verizon network, IMS on the modem's attach bearer On a Verizon-network SIM, stock's RIL sends no SET_INITIAL_ATTACH_APN for the SIM's slot: the modem attaches with the carrier manifest's settings, lists its attach bearer unsolicited (CID 2, IPv6, with three P-CSCF addresses), and the IMS service announces that bearer with ADD_PDN_INFO. 0.3.20 sends an explicit attach APN, which most likely replaces the manifest's attach and its P-CSCF request, and its own IMS setup comes back without a P-CSCF, so IMS never registers. - [network] attach_apn = none skips SET_INITIAL_ATTACH_APN; the profiles and the internet bearer go out as before. - After registration, sitd waits up to 20 s for an active IPv6 or IPv4v6 bearer it didn't set up that carries a P-CSCF, reading GET_DATA_CALL_LIST and IND_DATA_CALL_LIST_CHANGED, and logs each item as CID, type, status, active, P-CSCF and DNS counts (never addresses). Found: no IMS setup of its own, ADD_PDN_INFO from that bearer. Not found: 0.3.20's path (own IMS bearer, then the ISIM fallback). - Bearers the modem brought up are recorded as modem-owned: activate() accepts them beside sitd's own and never uses their CID, deactivate() refuses them. - The shipped conf documents a Verizon-network example as comments, and IMS-SHANNON.md section 11 has the gated runbook. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d * omamobile-modem-sit 0.3.23: the IMS setup carries the IMS profile's APN types Stock's setup builder copies the APN's whole supported-type bitmask into the request (0xf1); for Verizon's IMS APN that is ims|ia (0x140), where ours sent IMS alone (0x40). The modem matches setups against its modem-cognitive profiles, and the ia bit marks the attach APN, so this is the one byte left where our IMS setup differs from stock's. With a [profiles] entry of the ims type for ims_apn, its bitmask is sent; without one nothing changes. IMS-SHANNON section 12 records the rest from libsitril: stock sends no SET_INITIAL_ATTACH_APN for a modem-cognitive attach profile (so attach_apn = none is its mode), the 0x0603 and 0x0613 layouts field by field, that stock never uses CID 1 (its attach context) for a bearer of its own, and a ranked list of further changes with a runbook. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d * omamobile-modem-sit 0.3.24: IMS registers on Verizon's network: IMS first on CID 2, the internet on CID 3 Tested on the reference Pixel 7 Pro, 2026-11-03 02:00, Verizon-network MVNO SIM, kernel 7.3.0-rc6-omamobile-g2219cc6cda77. The first IMS registration on this phone: "state 2 (VoLTE registered), features 3", then an SMS sent over IMS (AIMS_SEND_SMS accepted). Stock's RIL keeps CID 1 for the modem's attach context and never sets a bearer up on it; it sets the IMS bearer up first, on CID 2. On a network that attaches on the IMS APN, sitd's internet setup on CID 1 replaced the attach PDN, and the IMS bearer set up afterwards on CID 2 carried no P-CSCF (every variant tried: attach APN IMS or none, IPv6 or IPv4v6, P-CSCF requested, the carrier manifest, the ISIM's FQDN). With the IMS bearer first on CID 2 the setup response carries three IPv6 P-CSCFs and the modem's IMS stack registers within 2 s of ADD_PDN. - [network] internet_cid (default 1, today's behaviour; 3 here): the CID of sitd's internet bearer. IMS stays on CID 2. paths() resolves any non-IMS CID to the internet record, so omamobile-sit-data (no configuration) keeps working, and its probe found rmnet2. - [network] ims_first (default no): the IMS bearer and ADD_PDN before the internet bearer, as stock orders it; the internet comes up after the IMS setup whether or not IMS succeeded. - The IMS setup also reads the P-CSCF from the data-call list entry when the setup response has none (the modem lists the attach bearer's PCO there on some firmware), and the setup carries the IMS profile's whole APN-type bitmask (ims|ia), as stock's does (0.3.23). Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d * NOTES: 2026-11-03, IMS registered and an SMS sent on a Verizon-network SIM Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d * omamobile-modem 0.1.13: the modem stack at boot, opt-in (omamobile-modem-boot.target) Until 2026-11-03 the maintainer started the modem stack by hand every boot: omamobile-modem, then omamobile-sitd, omamobile-sit data on and omamobile-mmd. IMS registered on a Verizon-network SIM that night and an SMS went out at 02:00, the condition set for starting it at boot. Enabling the units one by one doesn't work: sitd Requires/After the modem unit, which was Type=simple, so it ran while the CP was still booting and exited ("CP is not ONLINE"). What changes, with nothing enabled by the package: - omamobile-modem-boot.target is the one switch (systemctl enable; the next release enables it once the carrier configuration is automatic, derekross#373). It wants the four units; each unit's own After= keeps the order, a missing unit is simply not started, and DefaultDependencies=no keeps multi-user.target (and omamobile-boot-success after it) from waiting on a registration that can take minutes. No wants link of any kind is shipped, and test/all holds the package to that. - omamobile-modem-inhibit.service holds a systemd-inhibit sleep:idle block, required by and ordered before the modem unit, refused a manual stop, so it lasts until the reboot: suspend with the modem on has never run on this phone. The start script refuses to load a driver until the lock shows in systemd-inhibit --list. - omamobile-modem.service is Type=notify: ready once CPIF's ports exist and the profile's read-only check has passed, before the CP is powered on. The CP's ONLINE is waited for by the device's sitd drop-in. - The bring-up latch (modem-bringup-started) is written and synced before the first modprobe, as CONTRIBUTING 3.4 asks, not after the drivers and the check as before; a check that refuses removes it again, so a refusal doesn't cost the next boot its modem. - A second marker, modem-up, written by the watcher at ONLINE (date and boot id) and removed by the unit's ExecStopPost when it is this boot's: a boot that ends with the modem up and no stop, a hang or a reset, leaves it, and the next boot refuses until a human has looked, so a hang after ONLINE can't become a reset loop. Only under the unit (INVOCATION_ID): a start by hand would have nothing to remove it. - Nothing restarts the modem in place; the units have no Restart=. Tests: the script's order (refusals, the inhibitor, the latch, modprobe, the check with its rm, READY, exec), both refusals and --stopped and --retry run for real against scratch markers, the units' relations and the no-wants-link rule, systemd-analyze verify on the three units. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d * device-cheetah 0.1.41: sitd waits for the CP's ONLINE before it runs omamobile-modem.service (0.1.13) is ready once CPIF's ports exist, before cbd has booted the CP, 10-30 s more; a sitd started on that alone exits ("CP is not ONLINE", seen 2026-11-03 with both units started at once). This device's sitd drop-in adds ExecStartPre=omamobile-modem-cp wait-online 120, the profile's own cbd timeout: a CP that never comes up fails sitd, and the data fallback and the Phone and Messages bridge, ordered after it, stay down until the reboot. Nothing is enabled: the opt-in is omamobile-modem-boot.target, and the test checks that no wants link for it or the four units is shipped. 0.1.41 leaves 0.1.40 to derekross#372's 10-device-modem.conf in the same directory, which moves sitd's Requires/After on the modem unit here; the two drop-ins are independent. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d * MODEM-PLAN: M4.6, the gated runbook for the modem stack at boot The one-switch form (omamobile-modem-boot.target, omamobile-modem 0.1.13, device-cheetah 0.1.41) with its reasons: install and look (B0), enable and reboot (B1), the chain in order with the inhibitor and the markers (B2), data and the bridge with one SMS (B3), the second reboot that proves the clean stop clears the modem-up marker (B4), and an optional forced reset that proves the guard refuses (B5). Expect/pass, stop and rollback per step; every reboot with the modem up costs one UFS pass. The old M4.6 line, enabling the three units one by one, is marked superseded: it started sitd before the CP was ONLINE. 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>
…derekross#360) * device/tegu TELEPHONY-PLAN: call audio and starting the modem at boot The SIM test passed with everything started by hand. This plans the two steps to a palmtop that works as a phone on its own: the call audio route from the 9a's stock audio HAL (EP5 down, EP4 up, the TAS2572 gain caveat, the AoC voice window to audit) with the safety gap in sitd's call release to close first, and an opt-in boot start for the modem with its guards and the packaging fix. * device/tegu TELEPHONY-PLAN: what has run, and which section 2 items derekross#363 and derekross#372 do * device/tegu TELEPHONY-PLAN: a gated runbook for enabling the modem at boot Derek asked for one with derekross#372: preconditions, one step per boot, Expect/Pass, stop conditions and rollback, with the cold boot, the clean shutdown that removes the latch and the UFS count recorded here.
…0: the Pixel 7 Pro's modem stack on by default (derekross#398) * omamobile-modem 0.1.16, device-cheetah 0.1.43, omamobile-shell 0.2.98: the Pixel 7 Pro's modem stack on by default The maintainer's decision of 2026-11-03: the next release ships the modem working out of the box (calls, texts and data on the tested network, M4.6 through three reboots). The switch of M4.6, omamobile-modem-boot.target, is now on by default on cheetah, and a boot that can't or shouldn't start the modem costs as little as possible: - omamobile-device-cheetah 0.1.43 enables the target with a vendor wants link under multi-user.target.wants, as it does for the AoC and Wi-Fi units; the four units the target wants get no link of their own, and omamobile-modem itself still enables nothing (its test holds it to that; tegu and raven are untouched). Off: systemctl mask the target. The stack's packages (omamobile-modem-sit, omamobile-mmd) join cheetah's ROOTFS_PACKAGES so the release image carries them. - A phone that isn't provisioned costs nothing: the modem and inhibit units carry ConditionPathExists= for the chroot, the NV checksums and the board id (what tools/phone-provision puts there), so both are skipped, not failed, and the inhibitor isn't held. The inhibitor needs its own copy because the modem unit's Requires= pulls it in even when that unit's conditions skip it. sitd skips itself without /dev/umts_ipc0; sit-data now does too, through a device drop-in: without it a condition-skipped sitd counted as started for sit-data's Requires=, and the fallback would have failed after 60 s and restarted every 10 s all boot. The start script checks the same three paths first, for a start by hand. - The modem-up guard clears itself after one refused boot: a stale marker still refuses that boot and logs where to look (journalctl -b -1), then moves to modem-up.refused-<time>-<pid> (the last three kept) so the next boot starts the modem again. A phone that hangs with the modem up every time loses the modem on every other boot at most, never for good. /etc/omamobile/modem-hold restores the old hard stop; --retry still clears both markers; the bring-up latch keeps its hard stop. - Every refusal leaves one short line in /run/omamobile/modem-refused, and the shell's Mobile data tile shows it under its name instead of "No modem" (omamobile-shell 0.2.98, tile-state.sh): no new notification path, the journal has the rest. Versions: derekross#372 and derekross#380 claim device-cheetah 0.1.42, derekross#389 omamobile-modem 0.1.14 and derekross#382 0.1.15, main has omamobile-shell 0.2.97 (derekross#394); hence 0.1.43, 0.1.16 and 0.2.98. Docs: RELEASES (ROOTFS_PACKAGES, decision 12 amended, checklist step 6), the release notes draft (the modem known issue becomes what works and what still needs a hand), INSTALL, DEVICES, and MODEM-PLAN M4.6b with the runbook: install, two reboots, a refused-boot drill by planting a stale marker, the hold flag. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d * omamobile-modem 0.1.16, device-cheetah 0.1.43, omamobile-shell 0.2.100: review round 2, the modem-up guard bounded The review of derekross#398 asked for three changes and a few corrections. The self-clearing modem-up guard was unbounded: a phone that hangs because the modem is up went up, hang, refused, up, hang, refused, for ever, with only a tile line on the boots in between. Now the refused copy the script already keeps is the counter: one automatic retry, then a hold. refuse_unclean keeps the marker and refuses "(held)" when a modem-up.refused-* copy exists (no clean stop since the last refusal); --stopped on this boot's marker removes the copies, so a clean stop resets the count; --retry clears both markers and the copies. The test's "four unclean boots in a row" block becomes "the second refusal holds, a clean stop resets", and M4.6b gets drill B9b for it. systemctl is-enabled reads "disabled" for a unit started by a vendor wants link alone (it reports /etc and /run links only), so B6's pass and stop lines expected the wrong answer and would have stopped a correct install; B6 now expects "enabled" before and "disabled" after the disable, and proves the link with ls -l of the /usr/lib link and systemctl list-dependencies --plain multi-user.target. The README, the target's header and the install message say the same, so nobody reads "disabled" and skips the mask. The read-only NV checksum check (M3.5) opens B6 as well as closing the runbook (AGENTS.md section 1). B8's is-active expectation is "failed inactive inactive active" (mmd runs). omamobile-shell takes 0.2.100: main is 0.2.98 after derekross#397 and derekross#373 is taking 0.2.99. The DEVICES cheetah fingerprint cell reads "works": derekross#358 made it on by default. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d * omamobile-modem 0.1.16: the start script's test stops after the marker decision The case "the boot after a refused one starts the modem again" (and the --retry case) ran lib/omamobile-modem-start past the marker refusals and relied on the machine to stop it: on the laptop and in CI no sleep inhibitor is held, so the script refused there before any driver. On the phone, where tools/phone-install-package builds the package as an unprivileged user, the inhibitor is held, so the test went on into the real bring-up and failed on /sys/module/pcie_aspm/parameters/policy (permission denied) and then on omamobile-vendor's "run as root". The script takes a test-only OMAMOBILE_MODEM_TEST_STOP_AFTER_CHECKS=1 (documented in its header; no unit or install step sets it) and exits 0 right after the marker decision, before the link driver's sysfs read, the inhibitor wait, the first modprobe and the profile. The two cases use it and assert the decision itself: exit 0, the stale marker moved aside with its refused copy kept, no latch written, the earlier reason left for the tile (it is removed only once every refusal has passed, which the stop point comes before). The test also checks that the stop point sits between the marker decision and the sysfs read, the inhibitor and the latch, and that no unit sets the variable. Verified: packages/omamobile-modem/test/all as uid 1000, with and without a systemd-inhibit stand-in that reports the inhibitor held (the phone's case); tools/pr-check origin/main. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d * release notes: the carrier APNs come from the SIM, not by hand Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BeEcdVeSfyLCbijF8XQF1d --------- Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
…rough omamobile-mmd (derekross#371) * tegu: plan Phone and Messages integration through mmd Trace the existing app, session agent, ModemManager-compatible bridge and sitd path. Record the tegu installation and unit prerequisites still missing on main, and an attended bridge-only test that preserves the running modem stack. Document the passing offline bridge and apps package suites, activation and audio limits, and the eventual board-specific image integration. * tegu: attach Phone and Messages to an attended running sitd Keep sitd's canonical name while moving the modem dependency into device-owned policy drop-ins. Tegu checks an already running modem and CP ONLINE; cheetah and raven retain their existing starter relationship. Require the device policy capability during RIL upgrades. Add an opt-in telephony package with a bridge policy and attended helper. It checks live sitd readiness, refuses legacy unit shadows and competing managers, and starts only mmd and the graphical user's agent. The service recovery check permits live calls and registration loss; no helper enables or recovers the modem. Remove mmd's service-level conflict so a bridge start cannot stop another modem manager. Package exclusion and D-Bus name ownership remain. Retain the previously developed SIM, safe-release and audio changes in this stack. Validate the helper, dependency policies and composed units offline; run all seven affected/dependency package suites and build the new architecture-independent package without installing dependencies. * device/tegu TELEPHONY-APPS-PLAN: the 2026-11-02 run, the two ways to start the bridge, and a plan that reads as one Record the palmtop run of Phone and Messages (kernel, versions, what passed, and that opening an app didn't start the agent), say which start fits which case, cite derekross#363 instead of a branch, and drop the lines that read like a task report. * TELEPHONY-APPS-PLAN, omamobile-telephony-tegu: mmd's Conflicts= goes in 0.1.3, with this pull request The plan and the README dated the removal of the unit's Conflicts=ModemManager.service to mmd 0.1.1, and the bullet above still listed it as built; this pull request is what removes it, in 0.1.3. * omamobile-telephony-tegu: the bridge's service check accepts its own WantedBy and tmp.mount, and a refusal isn't retried With derekross#372's boot unit on and this package installed, the bridge's ExecStartPre refused every start: sitd wanted omamobile-mmd.service (the bridge's own WantedBy, once enabled) and both units wanted tmp.mount (systemd adds it). Neither pulls the RIL or the CP into a start, so both are allowed now. A refusal of --service-check exits 3, which the drop-in names in RestartPreventExitStatus=, so a refused bridge stays down instead of retrying every five seconds. * omamobile-telephony-tegu: the bridge's service check is an ExecCondition RestartPreventExitStatus= applies only to the main process, so an ExecStartPre= that refused with exit 3 was still restarted every RestartSec under Restart=on-failure. As an ExecCondition=, a refusal (any exit from 1 to 254) skips the start without a retry and without marking the bridge failed. Checked on systemd 262: ExecStartPre exiting 3 with RestartPreventExitStatus=3 restarted five times, ExecCondition exiting 3 skipped the start once. * omamobile-device-cheetah 0.1.48, device-tegu 0.2.10, modem-sit 0.3.32: versions past main after the rebase The rebase onto main dropped the two commits main already had and left these three packages changed without a version above main's. * omamobile-telephony-tegu: depend on the device-tegu that lands with it (0.2.10) * omamobile-telephony-tegu: the dependency test reads device-tegu>=0.2.10 too --------- Co-authored-by: Derek Ross <derekross@gmail.com>
…-sit 0.3.34: the Pixel 9a's modem and phone service at boot, opt-in, and a witness for the modem's clean stop (derekross#372) The Pixel 9a gets an opt-in unit that starts its modem and phone service at boot (omamobile-tegu-modem-boot.service, omamobile-tegu-modem-up). A bring-up latch keeps a boot that hung or reset during the start from repeating it, with one retry (den-kloot's change, carried here). The 9a's record is in device/tegu/TELEPHONY-PLAN.md. In the shared omamobile-modem package: - a stop witness: a stop counts as clean only when the CP's boot device itself reported OFFLINE, so a missing device is no longer taken for one; - during an orderly stop, rfsd's exit holds its job open until systemd signals the supervisor, where it used to be taken for a crash. omamobile-modem-sit follows with the phone service's side. Tested on the 9a by the contributor (42 confirmed stops in 43 reboots, the one unexplained and handled by the latch). Tested on a Pixel 7 Pro on 2026-11-10 (7.3.0-rc6-omamobile-push1-g42be9bc424d4, this head as packages): the modem comes ONLINE; systemctl stop logs "CP OFFLINE" and removes the modem-up marker in 42 s with no UFS error; a systemctl reboot with the modem up is followed by ONLINE on the next boot with no refusal. The 7 Pro's unit still ends as "timeout" (derekross#532), unchanged by this.
…a's modem-cpif overlay and GNSS link driver (derekross#594) Adds device/tegu/overlays/modem-cpif.dtso to the Pixel 9a's DTB_OVERLAYS (it needs a kernel with linux-omarchy-mobile#187), CONFIG_EXYNOS_GNSS_IF_SPI=m to the shared fragment (two modules that bind only to the 9a's nodes), and gnss_spi and gnssif to the 9a's guards, so they don't load by themselves now that the kernel has the 9a's GNSS nodes (linux-omarchy-mobile#237). Nothing here starts the modem; that stays derekross#372's opt-in unit. MODEM-PLAN records six boots on the owner's palmtop on 2026-11-09, on the published kernel tip plus kernel derekross#224, derekross#236, derekross#237 and derekross#274's config: registration, a call, texts, data, GNSS fixes and confirmed stops, with the NV partitions unchanged. The Pixel 7 Pro gains two modules that never load. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
Stacked on #371 (and through it #367 and #364); this PR's own commits are the top three.
The 9a's modem start, packaged:
omamobile-tegu-modem-upinomamobile-device-tegu, and an opt-in unit,omamobile-tegu-modem-boot, that runs it with--phoneat boot. The package doesn't enable it, and a test holds it to that. The order is:pcie_gs201_cploaded withoutallow_hw.wait-online 60.The latch stays from step 2 until the profile's unit stops cleanly at shutdown. A start that never reached ONLINE, or a boot that hung or reset with the modem up, leaves it, and the next boot refuses (TELEPHONY-PLAN 2.4). On the 9a, sitd tells the modem the screen is on at start-up so it sends signal reports. That is set only by tegu's sitd drop-in, so the 7 Pro is unchanged.
The README and the unit give the preconditions (slot B marked successful, a kernel with linux-omarchy-mobile#172, the charger for the first boots) and the rollback. TELEPHONY-PLAN 2.7 (#360) has the gated runbook for enabling it.
Not done here, and in the plan's open items: an NV-only check before any driver (TELEPHONY-PLAN 2.3) and dropping the transient unit (2.2). The in-tree driver branch of the script hasn't run yet; the in-tree drivers have only run from the bring-up runners.
On the palmtop (2026-11-02, the daily kernel
7.3.0-rc6-omamobile-tegu-s2mpg15-m1-g991042e2eff3-dirty, built as 0.2.11): a cold boot came up as a phone on its own. The first try, with the NV check before the drivers, refused ("CPIF port is not loaded"), which is why the check moved. The record with phone-smoke, the UFS count and an NV check comes with the next run, which needs the palmtop back.Tests:
test/modem-up(the order, the latch on disk before any driver and kept past ONLINE, both driver branches, the refusals includingallow_hw),tests/packagesfor device-tegu, modem-sit and telephony-tegu,tools/pr-check.