From e3ebed86182b2160a15b96cafcf2dccab11e0d51 Mon Sep 17 00:00:00 2001 From: hrx-compat lane Date: Sat, 14 Nov 2026 21:58:01 -0400 Subject: [PATCH 1/2] KNOWN-ISSUES: the BeyondMax board's dead Type-C and rear USB ports (engine #823) Engine#823 asked for the release-note line and for the observation to be recorded. Measured on this box 2026-11-16: four Strix Halo USB 3.1 xHCI controllers, the ec_su_axb35 EC module loaded with no USB controls, no UCSI ACPI device -- and /sys/class/typec completely empty. The empty typec class is the load-bearing fact: this is not a Type-C port that fails to negotiate, it is a port the kernel has no model of, so no kernel-side interface (typec class, UCSI, EC) can power or switch it. That is why this is filed as a board limitation with a workaround rather than as a driver bug to fix here. --- KNOWN-ISSUES.md | 28 ++++++++++++++++++++++++++++ 1 file changed, 28 insertions(+) diff --git a/KNOWN-ISSUES.md b/KNOWN-ISSUES.md index c3ec50d1d..66a2cd785 100644 --- a/KNOWN-ISSUES.md +++ b/KNOWN-ISSUES.md @@ -17,6 +17,34 @@ limitations under the License. # Known issues +## engine#823 — on the Bosgame BeyondMax board only the front USB-A port works under Linux + +**Status: OPEN, hardware/firmware, not reproducible in software.** On this board (AXB35-02, BIOS 1.09 05/08/2026) +only the front USB-A port enumerates a device. Any drive in the front **Type-C** port or any **rear** port is +electrically dead under Linux, on our kernel and on stock Ubuntu — while the same ports work in Windows on the same +box, with the same drives (a Realtek RTL9210 USB4/USB 3.2 enclosure and a Gigastone flash drive). + +What the kernel can see, measured 2026-11-16 on this box: + +``` +lspci: 4 x "Strix Halo USB 3.1 xHCI" (c5:00.4, c7:00.0, c7:00.3, c7:00.4) + 2 x "Strix Halo USB4 Host Router" (c7:00.5, c7:00.6) +/sys/class/typec : empty -- the kernel knows of no Type-C port at all +/sys/bus/acpi/devices USBC000|PNP0CA0: 0 matches -- no UCSI device, as engine#823 reports +lsmod : ec_su_axb35 loaded, refcount 0 -- the board EC driver has no USB controls +``` + +Empty `/sys/class/typec` is the part worth stating plainly: this is not a Type-C port that fails to negotiate, it is a +port the kernel has **no model of**. Nothing in the standard interfaces (typec class, UCSI, EC) can power it or switch +it, and the BIOS exposes no option for it, so there is no software fix in this tree. The hypothesis in the issue — port +power or retimer switching done by a Windows-only driver or an EC/PD path the BIOS does not expose — is consistent with +everything above and is the reason this is filed as a hardware limitation rather than a driver bug. + +**Affected:** every install on the Bosgame BeyondMax (AXB35-02) with BIOS 1.09 or the shipping build. One usable +external USB port. Other Strix Halo boards are not known to be affected; this is the board, not the SoC. + +**Workaround:** use the front USB-A port, or a USB-A hub on it. Release notes must carry this line for this board. + ## engine#370 — intermittent MUL_MAT f32 m=1 n=1 k=2048 error on `common.mul_mat.f32_f32_decode` **Status: CLOSED 2026-10-10 — closed as not reproducible, not as fixed at source.** The issue is closed From 96674f84de75ad5df0e0d8e08a0bca6034016620 Mon Sep 17 00:00:00 2001 From: bong-water-water-bong Date: Sat, 14 Nov 2026 22:27:38 -0400 Subject: [PATCH 2/2] KNOWN-ISSUES: #823 is D3cold autosuspend, mitigated by the shipped xHCI keep-awake rule (ports work on BIOS 1.10) --- KNOWN-ISSUES.md | 45 +++++++++++++++++++++++++-------------------- 1 file changed, 25 insertions(+), 20 deletions(-) diff --git a/KNOWN-ISSUES.md b/KNOWN-ISSUES.md index 66a2cd785..7b7c67f24 100644 --- a/KNOWN-ISSUES.md +++ b/KNOWN-ISSUES.md @@ -17,33 +17,38 @@ limitations under the License. # Known issues -## engine#823 — on the Bosgame BeyondMax board only the front USB-A port works under Linux +## engine#823 — Bosgame BeyondMax board: the USB ports need the xHCI keep-awake rule -**Status: OPEN, hardware/firmware, not reproducible in software.** On this board (AXB35-02, BIOS 1.09 05/08/2026) -only the front USB-A port enumerates a device. Any drive in the front **Type-C** port or any **rear** port is -electrically dead under Linux, on our kernel and on stock Ubuntu — while the same ports work in Windows on the same -box, with the same drives (a Realtek RTL9210 USB4/USB 3.2 enclosure and a Gigastone flash drive). - -What the kernel can see, measured 2026-11-16 on this box: +**Status: MITIGATED in 1bit OS — the ports work; the fix is a udev rule, not a driver.** On this board (AXB35-02) +every external port **works** once the xHCI controllers are kept out of D3cold. The owner's plug test on **BIOS 1.10** +(2026-11-14 22:09–22:23 ADT) exercised each one: ``` -lspci: 4 x "Strix Halo USB 3.1 xHCI" (c5:00.4, c7:00.0, c7:00.3, c7:00.4) - 2 x "Strix Halo USB4 Host Router" (c7:00.5, c7:00.6) -/sys/class/typec : empty -- the kernel knows of no Type-C port at all -/sys/bus/acpi/devices USBC000|PNP0CA0: 0 matches -- no UCSI device, as engine#823 reports -lsmod : ec_su_axb35 loaded, refcount 0 -- the board EC driver has no USB controls +front Type-C c7:00.3 5 Gbps rear Type-C c7:00.4 5 Gbps +rear USB-A c5:00.4 5 Gbps rear USB-A c7:00.0 port 4, USB 2 (a USB 2 port) ``` -Empty `/sys/class/typec` is the part worth stating plainly: this is not a Type-C port that fails to negotiate, it is a -port the kernel has **no model of**. Nothing in the standard interfaces (typec class, UCSI, EC) can power it or switch -it, and the BIOS exposes no option for it, so there is no software fix in this tree. The hypothesis in the issue — port -power or retimer switching done by a Windows-only driver or an EC/PD path the BIOS does not expose — is consistent with -everything above and is the reason this is filed as a hardware limitation rather than a driver bug. +**Cause: D3cold autosuspend.** A port whose controller has been allowed to suspend does not come back on this board's +firmware, which is why the symptom looked electrical: the port is fine, the controller is asleep and does not wake for +a hot-plug. The fix is `os/v3/puppy/embed/etc/udev/rules.d/91-1bit-xhci-keepawake.rules`, **shipped in 1bit OS** and now also installed on +strixhalo, which pins the xHCI controllers in D0. + +**What the earlier write-up here got wrong, kept so nobody re-derives it:** empty `/sys/class/typec` and the absence of +a UCSI device are both **true**, and neither is the cause. The board indeed exposes no Type-C port model to the kernel, +but that does not stop the ports working — the failure was power management, not a missing port description. The +`ec_su_axb35` refcount-0 observation is likewise true and likewise not the mechanism. + +**Not established:** **USB4 / Thunderbolt devices are untested** through these ports. The keep-awake rule covers the +xHCI controllers; a USB4 tunnel is a different path (the two `Strix Halo USB4 Host Router` functions) and nobody has put +a TB device on it. + +**One device stays USB 2:** the **Ugreen RTL9210** enclosure enumerates at USB 2 on the rear USB-A port (`c7:00.0` +port 4). That port is USB 2, and the enclosure/cable pair is the other half of it — it is not the controller. -**Affected:** every install on the Bosgame BeyondMax (AXB35-02) with BIOS 1.09 or the shipping build. One usable -external USB port. Other Strix Halo boards are not known to be affected; this is the board, not the SoC. +**Affected:** every install on the Bosgame BeyondMax (AXB35-02) **without** the keep-awake rule. With the rule, all +four ports above behave. Other Strix Halo boards are not known to be affected. -**Workaround:** use the front USB-A port, or a USB-A hub on it. Release notes must carry this line for this board. +**Release notes:** 1bit OS ships `91-1bit-xhci-keepawake.rules` for this board; without it the ports appear dead. ## engine#370 — intermittent MUL_MAT f32 m=1 n=1 k=2048 error on `common.mul_mat.f32_f32_decode`