Repository navigation
device/cheetah, RELEASES: AVB footers on boot and init_boot (#11) - #104
Conversation
…s#11) A second Pixel 7 Pro, installed with tools/install-phone, never started slot B: the phone ended in Fastboot Mode with "Enter reason: BL1 requested", slot-retry-count:b 1 and no USB gadget, the symptom of issue derekross#11. Its slot A runs stock CP3A.261005.002.A1 and both slots ABL 15902555, but vbmeta_b was left from BP2A.250605.031.A2 (Android 16, June 2025). Both vbmetas chain boot and init_boot. With the descriptor's AVB_FOOTERS (stock CP3A's rollback index and security patch, AOSP's test key, as lynx does), bin/omamobile images gives those two images hash footers, and after bin/omamobile flash the phone booted, ran first boot and marked slot B successful. Nothing else changed between the two: rebuilt from the same kernel output, vendor_boot, vendor_kernel_boot and dtbo are byte for byte the same, and boot and init_boot match up to their padding and footer. The maintainer's phone (ABL 15199429) and the Pixel 7 in derekross#79 (15902555) boot footer-less images, so the trigger may be the older vbmeta_b; the NOTES entry records what is verified and what isn't. RELEASES' row for AVB_FOOTERS no longer says the Pixel 7 Pro's bootloader takes images without one. Tested: tests/all (168 ok); on the phone, the boot above and tools/phone-smoke (every check ok, cameras skipped). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01TTVics5jHkiz34wzjZWkb9
|
Thank you, Jeremy, and welcome. This is a careful first pull request: the notes separate what you verified from what you inferred, the body says plainly what you didn't test, and the smoke report is there. I rebuilt cheetah's images here with main's descriptor and with yours. I think I can answer your "what triggers it" question, and the answer changes two things in the PR. Why my phone boots without footers. I checked my own The rollback index and patch level. My slot A runs CP2A.260705.006, and so did the phone in #11. CP2A's footers carry rollback index 1783209600 and security patch 2026-07-05. The descriptor would put CP3A's 1791158400 and 2026-10-05 on every Pixel 7 Pro. Rollback indexes are stored per location for the whole device, not per slot. Once a slot is marked successful, which
You offered to try a lower index, and your phone is the right one to try it on: its CP3A slot A is newer than either value, and your Two smaller things:
Once the values are changed and the notes say what triggers it, I'll put these images on my phone and merge after it boots. Thanks again: this closes a real gap for everyone installing on a stock Pixel 7 Pro. |
…e footers From the review of derekross#104. The trigger is a vbmeta_b that chains boot and init_boot, as stock's does, not the age of this phone's vbmeta_b: libavb reads a chained partition with no footer from its first byte, finds an invalid vbmeta header and returns ERROR_INVALID_METADATA, which result_should_continue() never lets an unlocked boot past (a wrong key or hash it does; external/avb c5066a9). The maintainer's vbmeta_b lists them as hash descriptors instead, which is why that phone boots footer-less images. The descriptor comment, the NOTES entry and RELEASES' row now say so. The footers now carry CP2A.260705.006's rollback index and security patch (1783209600, 2026-07-05), the oldest stock an owner runs on slot A (review lesson L-225), matching OS_PATCH_LEVEL, instead of CP3A's. Tested on this phone, whose slot A runs CP3A (1791158400): rebuilt, boot and init_boot differ from the CP3A-footer images only in the footer, the other images are byte for byte the same, and after bin/omamobile push slot B booted (orange, no failed units, slot still successful) and tools/phone-smoke passed with cameras skipped. Tested: tests/all (168 ok), tools/pr-check. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01TTVics5jHkiz34wzjZWkb9
|
Thanks for the review, and for rebuilding the images yourself. Both changes are in the second commit,
The description is updated to match. Noted on the RELEASES row and #25. |
|
Thank you, Jeremy, for the quick and thorough follow-up. Both points are in, and the libavb reading is clearly sourced. Testing the lower index on your own phone is exactly what I hoped for: a footer index below slot A's CP3A one boots under an unlocked ABL, and that's useful to know. I checked the new head here:
I've approved the CI run for this push. My One small optional fix: the RELEASES row for Thanks again: this fixes slot B for everyone installing on a stock Pixel 7 Pro. |
|
Tested on my Pixel 7 Pro, and it turned out to be a real test. My earlier note was wrong: my phone's I rebuilt the images at Merging now, with a squash message I've written so the first commit's earlier guesses don't land. I'll fix the |
…VB footers (#174) NOTES records 2026-10-29 and 10-30's charging work: the 10-29 outage was upower's 0 % suspends and a laptop route, charging works asleep, and the charger read offline at Full because the driver required CHG_OK (fixed on the kernel side, with the fast-charge timer off). It also records how the phone stopped booting footer-less images after #104's first boot, the stock recovery that followed, and the fresh radio backup. AGENTS gains the footer rule and what to do on stock's recovery screen. The kernel config drops CONFIG_GS201_CAM_PROBE: the read-only CSIS probe was left out when the kernel was curated for publishing. Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
The matrix and the Wanted list against what has merged since #140, with each pull request's state checked on GitHub today: - Pixel 7 (panther): the GPU and haptics nodes (kernel #129, #130) were closed because they landed unchanged through kernel #134, merged with the touchscreen node and s2idle by default; device-panther 0.1.1 (#145) carries the GPU's always-on rule, the feedbackd rule and the sleep.conf drop-in. GPU and Haptics stay "testing" and now cite #134, #145 and the 2026-10-28 notes. Touch stays WIP: kernel #134's body records touches driving Hyprland and the PIN screen, but the driver fixes (kernel #133) and the kernel option with the touch notes (#149) are still in review. Suspend: "no" with the s2idle default named, since the notes say it hasn't run. - Pixel 6 Pro (raven): Haptics "no" -> "works": HAPTICS-TORCH-PLAN H5 passed (2026-10-28/29), cs40l25-vibra and the node merged as kernel #114, feedbackd's rule in device-raven 0.1.27 (#113). Display names its open kernel PR, #116. - Pixel 7a (lynx): Bluetooth "no" -> WIP: the merged WIFI-PLAN (#133) records it scanning through its own UART; kernel #98 and #98 are in review. The Wanted line says so too. - Pixel 7 Pro (cheetah) detail: Boot names the AVB footers a chained vbmeta_b needs (NOTES 2026-10-28 and 10-29, #104); Suspend describes the merged tier 1 auto-sleep (#64, off by default behind the omamobile-sleep toggle, P0-P5 on 2026-10-26) in place of the old "until S8" text. Nothing from the open #171, #172, #174-#176. - Wanted: coral's #52 is merged, so its line no longer says "in review"; flame gets a line for its native install in review (#55, #56, #58, #59, split out of #17); the OnePlus 13 (dodge) is listed with #111 and kernel #115, both open. The Pixel 9a row and the Pixel 9 family line are left to #165. Every status word still parses with the website's parseCell. Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…ing (#307) The 9a's row after today's merges, with a NOTES entry for the evening's boots that the new cells link to. - **Wi-Fi, testing to works.** `omamobile-device-tegu` 0.2.4 (#238) starts it at boot now, through the shared gs201 host and zuma PHY (kernel #178 on #156, both in review) and brcmfmac (kernel #56, in review), with only the extractor's firmware. The NOTES entry has the boot, and the three rebinds with Bluetooth up after #178's FLR commit. The kernel parts are still in review, and the cell says so, as the display cells did. - **Suspend, WIP to testing.** The power key now lights the lock screen on the first press (#268, #275). Thirty attended sleeps and a day of real use with auto-sleep on all woke (SLEEP-PLAN, #284). The cell still says that three test sleeps on 10-28 to 10-30 never woke and that sleep tests stay attended. If you'd rather keep WIP until that hang is understood, I'll revert just the word. - **Display:** the display DT is merged, and brightness, DPMS off and on and the 60/120 Hz switch now have a dated record (kernel #125's boot, which also checks the panel ID). - **Charging, Haptics, Torch, NFC:** kernel #161, #96, #104 and #105 are merged. - **USB net:** the rescue also handles a replug the gadget didn't see (#280). The NOTES entry lists each boot with its release: Wi-Fi from the package, the power-key wake, the replug, storage idle (kernel #99), the panel ID (kernel #125), the RTC without the PMIC interrupt (kernel #55), and the auto-sleep day. GNSS and Modem stay as they are until the SIM test on Tuesday. Docs only; `tools/pr-check` passes, and the diff has no identifiers. Closes #307 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…ekross#11 (derekross#104) A stock Pixel 7 Pro's vbmeta chains boot and init_boot, so ABL reads their own AVB footers. Without footers, slot B depended on the bootloader tolerating the missing descriptor. cheetah's descriptor now sets AVB_FOOTERS with CP2A's rollback index (1783209600) and security patch (2026-07-05), the values on stock CP2A's boot_a, so a footer is never above stock. The trigger is a vbmeta_b that chains boot/init_boot, as stock's does; NOTES records both owners' tests and RELEASES corrects the AVB_FOOTERS row. Tested on two Pixel 7 Pros with chaining vbmeta_b: push, slot B, orange, smoke test all ok. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…VB footers (derekross#174) NOTES records 2026-10-29 and 10-30's charging work: the 10-29 outage was upower's 0 % suspends and a laptop route, charging works asleep, and the charger read offline at Full because the driver required CHG_OK (fixed on the kernel side, with the fast-charge timer off). It also records how the phone stopped booting footer-less images after derekross#104's first boot, the stock recovery that followed, and the fresh radio backup. AGENTS gains the footer rule and what to do on stock's recovery screen. The kernel config drops CONFIG_GS201_CAM_PROBE: the read-only CSIS probe was left out when the kernel was curated for publishing. Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
The matrix and the Wanted list against what has merged since derekross#140, with each pull request's state checked on GitHub today: - Pixel 7 (panther): the GPU and haptics nodes (kernel derekross#129, derekross#130) were closed because they landed unchanged through kernel derekross#134, merged with the touchscreen node and s2idle by default; device-panther 0.1.1 (derekross#145) carries the GPU's always-on rule, the feedbackd rule and the sleep.conf drop-in. GPU and Haptics stay "testing" and now cite derekross#134, derekross#145 and the 2026-10-28 notes. Touch stays WIP: kernel derekross#134's body records touches driving Hyprland and the PIN screen, but the driver fixes (kernel derekross#133) and the kernel option with the touch notes (derekross#149) are still in review. Suspend: "no" with the s2idle default named, since the notes say it hasn't run. - Pixel 6 Pro (raven): Haptics "no" -> "works": HAPTICS-TORCH-PLAN H5 passed (2026-10-28/29), cs40l25-vibra and the node merged as kernel derekross#114, feedbackd's rule in device-raven 0.1.27 (derekross#113). Display names its open kernel PR, derekross#116. - Pixel 7a (lynx): Bluetooth "no" -> WIP: the merged WIFI-PLAN (derekross#133) records it scanning through its own UART; kernel derekross#98 and derekross#98 are in review. The Wanted line says so too. - Pixel 7 Pro (cheetah) detail: Boot names the AVB footers a chained vbmeta_b needs (NOTES 2026-10-28 and 10-29, derekross#104); Suspend describes the merged tier 1 auto-sleep (derekross#64, off by default behind the omamobile-sleep toggle, P0-P5 on 2026-10-26) in place of the old "until S8" text. Nothing from the open derekross#171, derekross#172, derekross#174-derekross#176. - Wanted: coral's derekross#52 is merged, so its line no longer says "in review"; flame gets a line for its native install in review (derekross#55, derekross#56, derekross#58, derekross#59, split out of derekross#17); the OnePlus 13 (dodge) is listed with derekross#111 and kernel derekross#115, both open. The Pixel 9a row and the Pixel 9 family line are left to derekross#165. Every status word still parses with the website's parseCell. Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…ing (derekross#307) The 9a's row after today's merges, with a NOTES entry for the evening's boots that the new cells link to. - **Wi-Fi, testing to works.** `omamobile-device-tegu` 0.2.4 (derekross#238) starts it at boot now, through the shared gs201 host and zuma PHY (kernel derekross#178 on derekross#156, both in review) and brcmfmac (kernel derekross#56, in review), with only the extractor's firmware. The NOTES entry has the boot, and the three rebinds with Bluetooth up after derekross#178's FLR commit. The kernel parts are still in review, and the cell says so, as the display cells did. - **Suspend, WIP to testing.** The power key now lights the lock screen on the first press (derekross#268, derekross#275). Thirty attended sleeps and a day of real use with auto-sleep on all woke (SLEEP-PLAN, derekross#284). The cell still says that three test sleeps on 10-28 to 10-30 never woke and that sleep tests stay attended. If you'd rather keep WIP until that hang is understood, I'll revert just the word. - **Display:** the display DT is merged, and brightness, DPMS off and on and the 60/120 Hz switch now have a dated record (kernel derekross#125's boot, which also checks the panel ID). - **Charging, Haptics, Torch, NFC:** kernel derekross#161, derekross#96, derekross#104 and derekross#105 are merged. - **USB net:** the rescue also handles a replug the gadget didn't see (derekross#280). The NOTES entry lists each boot with its release: Wi-Fi from the package, the power-key wake, the replug, storage idle (kernel derekross#99), the panel ID (kernel derekross#125), the RTC without the PMIC interrupt (kernel derekross#55), and the auto-sleep day. GNSS and Modem stay as they are until the SIM test on Tuesday. Docs only; `tools/pr-check` passes, and the diff has no identifiers. Closes derekross#307 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
What and why
A Pixel 7 Pro installed with
tools/install-phonenever started slot B, the symptom of #11: the phone ended in Fastboot Mode with "Enter reason: BL1 requested",slot-retry-count:b1,slot-successful:bno, and no USB gadget reached the laptop in four minutes. Its slot A runs stockCP3A.261005.002.A1and both slots run ABLcloudripper-17.0-15902555. Itsvbmeta_b, though, was left fromBP2A.250605.031.A2(Android 16, June 2025). Both vbmetas chainboot(rollback index location 2) andinit_boot(location 4).This adds
AVB_FOOTERSto cheetah's descriptor, asdevice/lynx/device.confdoes.bin/omamobile imagesthen givesbootandinit_boothash footers. With stock CP3A's footer values, afterbin/omamobile flashthe phone booted, ran first boot and marked slot B successful. After review, the second commit carries CP2A.260705.006's values instead (rollback index 1783209600, security patch 2026-07-05, per L-225), and the phone boots those too (below).Only the footers changed between the failed boot and the good one. Rebuilt from the same kernel output with and without the descriptor lines,
vendor_boot,vendor_kernel_bootanddtboare byte for byte the same, andbootandinit_bootmatch up to their padding and footer. The footer build is reproducible (a second build was byte-identical). Each version was flashed once.Also in this PR:
device/cheetah/NOTES.mdthat marks what is verified [V] and what is inferred [I];AVB_FOOTERSrow indocs/RELEASES.md. It said the Pixel 7 Pro's bootloader takes images without footers.What the review settled (second commit):
vbmeta_bthat chainsbootandinit_boot, as stock's does. libavb reads a chained partition with no footer from its first byte, finds an invalid vbmeta header and returnsERROR_INVALID_METADATA, whichresult_should_continue()never lets an unlocked boot past. A wrong key (ERROR_PUBLIC_KEY_REJECTED) or hash (ERROR_VERIFICATION) it does let through, andVERIFICATION_DISABLEDskips the descriptors (checked in external/avbc5066a9). This phone's oldervbmeta_bisn't the trigger: stock CP3A's vbmeta chains them the same way. The descriptor comment, the NOTES entry and the RELEASES row now state the condition that way.bootandinit_bootdiffer from the CP3A-footer images only in the footer, and the other three images are byte for byte the same.bin/omamobile pushwrote and verified slot B. The phone booted it: slot_b,verifiedbootstateorange, no failed units, slot B still successful. Slot A runs CP3A, whose footers carry 1791158400, so this unlocked ABL accepts a footer index below that one. Whether it stores indexes while unlocked is still untested.Device and build
CP3A.261005.002.A1on slot Auname -r, both commits):7.3.0-rc6-omamobile-push1-g1cb4bd36f3be(omamobile-intat1cb4bd36f3be, Android clang r596125, mkbootimg, libufdt and avb from AOSP, no LineageOS tree)fastboot flash bootloader_b), so that both slots carry the same version (REINSTALL-RUNBOOK I1). I didn't read slot B's version before that write.f5da1dfwith a local signing key. Eight packages were built fromf5da1df(keyring, device-cheetah, audio, sensors, omarchy, shell, modem, firstboot); the rest were copied from the development channel. The versions are in the smoke report below.Checks (no phone needed)
tools/pr-check origin/mainpassestests/allpasses (168 ok)tests/packages: not run, since no package changedversionbumped: none changedCo-Authored-By:trailer, noSigned-off-by:On the phone
bin/omamobile flashbooted.=== firstboot done ===.omamobile-boot-successhad run and exited (slot B successful), and ufw was already active. Before provisioning,omamobile-aoc(noaoc.bin),omamobile-wifi,omamobile-bt-addr@hci0andsystemd-firstboot(no console) had failed.persist,efs,efs_backup,modem_userdata,modem_aandmodem_bmatch the read-only backup taken before provisioning.devinfodiffers only in three bytes of ABL's own header (0x6d-0x70); its records from 0x80 are identical.vbmeta_bis from the current stock build, or on the maintainer's phone;tools/install-phonerun with the footer images (tests/images-layoutcovers it offline;bin/omamobile flashandpushran on the phone);0;omamobile-boot-update(RELEASES §5.3), which isn't in the package yet. When it lands, it will need to apply these footers on cheetah too.userdata, weakens the storage guard, or reads registers from userspace (AGENTS.md §1)tools/phone-smokereport (second commit's images, 2026-10-30T04:01Z)phone-smoke: cheetah
Kernel
7.3.0-rc6-omamobile-push1-g1cb4bd36f3be, tools at64e6d42, 2026-10-30T04:01Z.Packages: hyprland 0.56.2-5.1, libcamera 0.7.2-4.21, mesa 1:26.2.4-1.4, omacam 0.3.3-1, omamobile-apps 0.2.0-1, omamobile-audio 0.5.4-1, omamobile-device-cheetah 0.1.34-1, omamobile-fingerprint 0.1.8-1, omamobile-firstboot 0.1.2-1, omamobile-gnss 0.2.1-1, omamobile-keyring 20261026-1, omamobile-mirror 0.1.1-1, omamobile-modem 0.1.7-1, omamobile-omarchy 0.1.12-1, omamobile-owl-widgets 0.1.1-1, omamobile-sensors 0.2.1-1, omamobile-shell 0.2.32-1, omarchy 4.0.6-1, quickshell 0.3.2.r7.g2b4501a-1
🤖 Generated with Claude Code
https://claude.ai/code/session_01TTVics5jHkiz34wzjZWkb9