You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
J813 now has a native DCP display on the internal panel, running a KDE Plasma Wayland session, plus reported ten-core Linux, bounded CPU scaling, temperature readings, built-in keyboard/trackpad input, visibly confirmed keyboard backlight and native USB3 device operation. These implementations are in the pinned public working branches. Rendering is software-only, with no GPU acceleration. USB4/Thunderbolt is groundwork only. J815 has no board-specific qualification in this evidence.
Refreshed October 3, 2026 at 1:47 AM EDT. Public PR states, follow-up reports, merge destinations and source were checked. This is an evidence update, not a new physical test. Baseline: linux/aurora-wip@f9572b54204e; bootloader: m1n1/aurora-wip@049a304e87c4.
⚪ No board-specific result; does not inherit J813 qualification
Evidence key: 🟢 named hardware report · 🔵 code present in the pinned source · 🟠 partial or unqualified · ⚪ no board-specific evidence found. A merge is an integration event, not a hardware test or release qualification.
🧩 Core platform
Feature
J813 · 13-inch
Evidence and limits
Native RAM boot
🟢 Reported; 🔵 integrated
linux #159 with m1n1 #10 reports two one-CPU RAM boots to BusyBox after VDM resets. Both are merged. linux #160 extends this to ten CPUs. RAM-backed mounts only; u-boot #2 is merged into aurora-wip; its integration review reports a fresh ten-core J813 RAM boot.
SMP / scheduling / timers
🟢 10 CPUs reported
linux #160 / m1n1 #12: CPUs 0–9 online; every CPU gained 1,251 timer interrupts in five seconds; ten concurrent pinned 16 MiB SHA-256 jobs matched the expected result. Hotplug, deep sleep and KVM remain unqualified.
CPU frequency
🟢 Measured bounded scaling
linux #162 with m1n1 #13: E 300–972 MHz; P 300–3516 MHz. Two policy sweeps, checksum-verified workloads and independent timing checks; approximately 3.24× E / 11.73× P min-to-max workload speed changes. scaling_cur_freq reports requested state.
Frequency integration
🔵 Present in current working tree
linux #162 merged into codex/m5-smp; its commit is also in current aurora-wip through linux #161. Firmware voltage/PLL policy is preserved; higher states, automatic governors and sustained thermal behavior are unqualified.
AIC / console / memory
🟢 Initial reports; 🔵 source
AICv3, architectural timers, DockChannel, loader-filled memory and framebuffer. Samsung UART is described but disabled. AIC initial report: 2400/4096 IRQs, error count zero.
linux #164 adds MTP/SMC/PWM resources; linux #165 adds SPMI Type-C controllers, ATC PHY, DWC3 and USB DART topology; linux #166 adds the display DARTs and lets locked apple,t8142-dart keep firmware-owned root slots, checked page by page. The named input/USB tests do not qualify every DMA client or platform bus.
Native USB host
🟢 Named USB devices reported; 🔵 integrated
linux #165 with m1n1 #18: Keychron K2 input/reconnect and Samsung T7 Shield enumeration at 10 Gbps in both cable orientations, plus read-only raw/exFAT reads. Enumeration speed is not measured throughput; this does not qualify every device, port/role combination or storage writes.
NVMe / internal storage
🟠 Unqualified
Native USB-attached SSD reads do not establish internal NVMe operation or installation.
🎨 Display, I/O and complete-system readiness
Feature
J813 status
Boundary
Native panel display (DCP)
🟢 Reported; 🔵 integrated
linux #166 with m1n1 #19 and u-boot #4: Linux attaches to the DCP session iBoot left running (H17 / macOS 26 firmware) and never stops, resets or reloads it. 5 of 5 fresh boots of the final source started DCP with appledrmfb as console and no DART faults; three DPMS off/on cycles kept flipping; KDE Plasma 6.6 Wayland ran at 2560×1600 (notch hidden) with built-in input. No firmware or panel calibration data is shipped.
Firmware framebuffer
🟢 Registration reported
2560×1600 x2r10g10b10 / CONFIG_FB_SIMPLE in the initial boot report; the DCP driver now takes over the console.
No T8142 AGX match in the pinned working-branch driver; the desktop renders in software. Panel backlight control, cursor plane, external outputs and display suspend are outside #166.
Wi-Fi / Bluetooth / networking
🟠 Unqualified
No board-qualified networking result in this series.
Speakers / microphones / camera / codecs / ANE
🟠 Unqualified
No board-qualified audio, capture, encode/decode or inference result in this series.
Built-in keyboard / trackpad
🟢 Reported; 🔵 integrated
linux #164: 198 keyboard events and 97,458 trackpad events, without detected kernel faults. Requires externally supplied apple/tpmtfw-j813.bin; missing/truncated firmware was rejected and valid firmware subsequently started. Firmware extraction and installer integration are outside the PR. u-boot #4 now leaves MTP cold on J813: a U-Boot start/stop crashed Linux's later MTP restart in 2 of 2 boots once the display driver probed first. Keyboard input inside U-Boot stays off on J813.
Keyboard backlight
🟢 Visible full-brightness confirmation
linux #164: sysfs requests at 0/127/255 and restoration passed; the operator confirmed lit keys at full brightness. This is not a luminance calibration.
SEP / Touch ID
🟠 Unqualified
Built-in input does not qualify biometric enrollment or matching.
USB4 / Thunderbolt / PCIe
🔵 Controller/tuning groundwork; 🟠 no working tunnel qualification
linux #165 / m1n1 #18 add T8142 Type 7 handling and validated runtime tuning. No ACIO/NHI device-tree nodes are added/enabled. Firmware startup/shared SRAM, PCIe-C, real links, tunnels and hotplug remain unqualified. Native USB3 is distinct from USB3 tunneling.
Temperature reporting / thermal policy
🟢 Temperature readings; 🟠 thermal control unqualified
linux #164: two reads of 129 SMC temperature channels. These include virtual/inactive keys and are not a physical sensor count. Labels preserve key identities; no new sensor-location mapping, cooling policy or sustained-load qualification.
Battery / suspend / CPU hotplug / KVM
🟠 Unqualified
Input, temperatures, USB and DPMS display cycles do not establish system suspend, battery-management or virtualization support.
Installer / complete laptop
🟠 Unqualified
Experimental RAM boots, now including a software-rendered desktop session; no complete public installation qualification.
105 compiled legacy replay cases plus physical M6 checks
Replay does not prove physical behavior on older boards. J813 evidence remains attached to its named earlier tests.
Latest bootloader integration
u-boot #2 now reports a fresh physical J813 ten-CPU RAM boot using merged public m1n1, passing timer, pinned checksum, live configuration, framebuffer and RAM-mount checks. The combined M5/M6 bootloader passed both full ARM64 builds, 33 compiled parser/console replay cases and canonical configuration comparisons. Installed bootloaders and internal disks were preserved. The previously pinned bootloader source was u-boot/aurora-wip@5a510be0119b; current u-boot/aurora-wip@55aaa060bfab adds u-boot #4 (J813 MTP left cold). These earlier bootloader checks remain limited to their tested stack; the newer peripheral, native USB and display reports are assessed separately below.
Kernel/113 DTBs with optional USB4 fragment; five compiled NHI schema cases; matched source hashes
The report cites native USB3 hardware results, but explicitly records no new M5 reboot or physical USB4 test for the final USB4 draft. Existing warnings are outside changed drivers.
5 of 5 fresh J813 boots with DCP and appledrmfb; three DPMS cycles; KDE Plasma Wayland with built-in input; W=1 with no new warnings
A sixth boot was stopped by the test preflight before Linux because DCP had not settled. Shared-code changes are gated to H17 firmware, except a parser fix for all chips that clears a stale VRR flag after display re-reports; older chips were not re-tested for this.
Retain board identity, exact source and payload hashes, kernel configuration, per-CPU affinity/results and serial transcripts. The published procedures keep testing RAM-only and preserve the installed bootloader and disks.
PR state and integration audit · October 3, 2026 at 1:47 AM EDT
PR state, destination branch, source integration and hardware results are separate. Closed does not necessarily mean merged. Hardware reports retain their tested revisions, including when the PR head subsequently changes.
🧪 Add a reproducible hardware report
Reply here with one feature or failure per report. Identify the exact row from the configurations table.
Board ID:
SoC ID and chip variant:
Model / screen size / configuration:
Feature or SoC block:
Kernel repository / branch / commit:
m1n1 / U-Boot / Mesa commits, where relevant:
Firmware version:
Steps to reproduce:
Expected result:
Observed result and repeat count:
Public log or artifact URL:
Artifact SHA-256 and command to repeat the check:
Known limits / recovery steps:
Repository admins can update this opening post through ⋯ → Edit. Record the tested board next to each result; keep other boards unqualified until their own evidence is supplied.
Keep this product generation together. Discuss shared silicon changes here and link affected sibling products through the hardware directory. A result on one board does not change another board’s status.
hardwareBoard-specific hardware evidence and SoC indexessoc:T8142Hardware discussions for T8142 (M5)
1 participant
Heading
Bold
Italic
Quote
Code
Link
Numbered list
Unordered list
Task list
Attach files
Mention
Reference
Menu
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
MacBook Air · M5 family
🧭 Hardware directory › M5 category
Important
J813 now has a native DCP display on the internal panel, running a KDE Plasma Wayland session, plus reported ten-core Linux, bounded CPU scaling, temperature readings, built-in keyboard/trackpad input, visibly confirmed keyboard backlight and native USB3 device operation. These implementations are in the pinned public working branches. Rendering is software-only, with no GPU acceleration. USB4/Thunderbolt is groundwork only. J815 has no board-specific qualification in this evidence.
Refreshed October 3, 2026 at 1:47 AM EDT. Public PR states, follow-up reports, merge destinations and source were checked. This is an evidence update, not a new physical test. Baseline:
linux/aurora-wip@f9572b54204e; bootloader:m1n1/aurora-wip@049a304e87c4.Configurations
J813J815Evidence key: 🟢 named hardware report · 🔵 code present in the pinned source · 🟠 partial or unqualified · ⚪ no board-specific evidence found. A merge is an integration event, not a hardware test or release qualification.
🧩 Core platform
aurora-wip; its integration review reports a fresh ten-core J813 RAM boot.scaling_cur_freqreports requested state.codex/m5-smp; its commit is also in currentaurora-wipthrough linux #161. Firmware voltage/PLL policy is preserved; higher states, automatic governors and sustained thermal behavior are unqualified.apple,t8142-dartkeep firmware-owned root slots, checked page by page. The named input/USB tests do not qualify every DMA client or platform bus.🎨 Display, I/O and complete-system readiness
appledrmfbas console and no DART faults; three DPMS off/on cycles kept flipping; KDE Plasma 6.6 Wayland ran at 2560×1600 (notch hidden) with built-in input. No firmware or panel calibration data is shipped.x2r10g10b10/CONFIG_FB_SIMPLEin the initial boot report; the DCP driver now takes over the console.apple/tpmtfw-j813.bin; missing/truncated firmware was rejected and valid firmware subsequently started. Firmware extraction and installer integration are outside the PR. u-boot #4 now leaves MTP cold on J813: a U-Boot start/stop crashed Linux's later MTP restart in 2 of 2 boots once the display driver probed first. Keyboard input inside U-Boot stays off on J813.🧪 Validation and integration boundary
Latest bootloader integration
u-boot #2 now reports a fresh physical J813 ten-CPU RAM boot using merged public m1n1, passing timer, pinned checksum, live configuration, framebuffer and RAM-mount checks. The combined M5/M6 bootloader passed both full ARM64 builds, 33 compiled parser/console replay cases and canonical configuration comparisons. Installed bootloaders and internal disks were preserved. The previously pinned bootloader source was
u-boot/aurora-wip@5a510be0119b; currentu-boot/aurora-wip@55aaa060bfabadds u-boot #4 (J813 MTP left cold). These earlier bootloader checks remain limited to their tested stack; the newer peripheral, native USB and display reports are assessed separately below.Latest peripheral evidence
0c89fc1cand U-Boot5a510be0. No new physical older-chip run or suspend qualification.appledrmfb; three DPMS cycles; KDE Plasma Wayland with built-in input;W=1with no new warningsapple,j813only. A warm MTP handoff is not qualified.Current source and repeatable checks: t8142-j813-display.dtsi · t8142-j813-peripherals.dtsi · t8142-j813-usb.dtsi · j813-peripherals.sh · j813-usb4.config. The peripheral kselftest records counts without retaining key codes or pointer coordinates.
Repeatable public evidence
Retain board identity, exact source and payload hashes, kernel configuration, per-CPU affinity/results and serial transcripts. The published procedures keep testing RAM-only and preserve the installed bootloader and disks.
PR state and integration audit · October 3, 2026 at 1:47 AM EDT
aurora-wipaurora-wiphistory ·b3a55ee68bfbaurora-wipaurora-wiphistory ·f1d29b8afd4eaurora-wipaurora-wiphistory ·9845b1b9a313codex/m5-smpaurora-wiphistory ·0abc47f7ebe2aurora-wipaurora-wiphistory ·70dc750325d6aurora-wipaurora-wiphistory ·81c520ffb028aurora-wipaurora-wiphistory ·f9572b54204eaurora-wipaurora-wiphistory ·02a35260725acodex/m5-j813-native757abb9d224ccodex/m5-j813-native0f13ebbdf23dcodex/m6-cpufreqaurora-wiphistory ·b85c25d2d38eaurora-wipaurora-wiphistory ·04d657bafb8aaurora-wipaurora-wiphistory ·33ab9b9d00afaurora-wipaurora-wiphistory ·049a304e87c4aurora-wipaurora-wiphistory ·1313a520b79aaurora-wipaurora-wiphistory ·55aaa060bfabPR state, destination branch, source integration and hardware results are separate. Closed does not necessarily mean merged. Hardware reports retain their tested revisions, including when the PR head subsequently changes.
🧪 Add a reproducible hardware report
Reply here with one feature or failure per report. Identify the exact row from the configurations table.
Repository admins can update this opening post through ⋯ → Edit. Record the tested board next to each result; keep other boards unqualified until their own evidence is supplied.
Keep this product generation together. Discuss shared silicon changes here and link affected sibling products through the hardware directory. A result on one board does not change another board’s status.
All reactions