Merge pull request 'README + danctnix-besser-pkgbuild/README: point at kernel-agent pkgrel=4 flow' (#21) from noether/readme-pkgrel4-kernel-agent-flow into main

Reviewed-on: #21

README.md

@@ -53,6 +53,9 @@ CW1200-ancestry markers in current source: same author Dmitry Tarnyagin,
 |------|------|
 | **This umbrella** | `git.reauktion.de/marfrit/besser` — patches/, scripts/, fw-analysis/, notes/ |
 | **Mobian DKMS fork** (PR target) | `git.reauktion.de/marfrit/bes2600-dkms` — branches per patch; upstream = `salsa.debian.org/Mobian-team/devices/bes2600-dkms` |
+| **DanctNIX kernel package** (ohm) | `git.reauktion.de/marfrit/marfrit-packages/arch/linux-pinetab2-danctnix-besser/` — kernel-agent-driven PKGBUILD, pkgrel=4+ |
+| **kernel-agent manifest + patches** | `git.reauktion.de/marfrit/kernel-agent` — `fleet/ohm.yaml` lists the per-patch series, `bin/ka-promote ohm` emits the cumulative the PKGBUILD consumes |
+| **Historical hand-managed PKGBUILD** | `git.reauktion.de/marfrit/besser/danctnix-besser-pkgbuild/` — pkgrel≤3, deprecated; see directory README |
 
 ## Patch series
 

danctnix-besser-pkgbuild/README.md

@@ -0,0 +1,222 @@
+# linux-pinetab2-danctnix-besser
+
+Soft-upstream fork of `linux-pinetab2` (DanctNIX kernel for PineTab2) carrying the **BESser** bes2600 staging-driver patchset.
+
+Drop-in replacement for `linux-pinetab2`. Same kernel version, same config (one toggle aside — see SCS caveat below), same modules — only the `drivers/staging/bes2600/` driver differs.
+
+---
+
+> ## ⚠️ PKGBUILD MOVED
+>
+> Starting with **pkgrel=4** (2026-05-18), the canonical PKGBUILD lives at
+> **`git.reauktion.de/marfrit/marfrit-packages/arch/linux-pinetab2-danctnix-besser/`**
+> and is driven by [kernel-agent](https://git.reauktion.de/marfrit/kernel-agent)'s
+> `ka-promote ohm` cumulative-patch flow against `fleet/ohm.yaml`.
+>
+> This directory remains for historical reference (pkgrel=1..3 hand-managed
+> flow + per-patch design notes that haven't been ported to the new home yet).
+>
+> **Use the new location** for builds going forward. See
+> [kernel-agent PR #28](https://git.reauktion.de/marfrit/kernel-agent/pulls/28)
+> and [marfrit-packages PR #28](https://git.reauktion.de/marfrit/marfrit-packages/pulls/28)
+> for the migration.
+
+---
+
+## TL;DR
+
+| | |
+|---|---|
+| **Current package** | `linux-pinetab2-danctnix-besser-7.0.danctnix1-5-aarch64.pkg.tar.zst` (built via [kernel-agent](https://git.reauktion.de/marfrit/kernel-agent)) |
+| **PKGBUILD home** | `git.reauktion.de/marfrit/marfrit-packages/arch/linux-pinetab2-danctnix-besser/` *(new — pkgrel=4 onwards)* |
+| **Patch manifest** | `git.reauktion.de/marfrit/kernel-agent` `fleet/ohm.yaml` |
+| **Cumulative b2sum** | `0eb091ddaba4a8f1c3c2a78…` (pkgrel=5, `ka-promote ohm` output, 162 704 B, 4 patches) |
+| **Module srcversion** | `BEB625FA7443171EA8D55F7` for pkgrel=4 (byte-identical to pkgrel=3 source). pkgrel=5 srcversion differs because the besser#18 fix is bundled (TBD pending build verification). |
+| **Kernel base** | DanctNIX [`linux-pinetab2`](https://codeberg.org/DanctNIX/linux-pinetab2) tag `v7.0-danctnix1` |
+| **What it fixes vs upstream** | +73 % TX throughput, the `wsm_generic_confirm 0x0007` dmesg storm (besser#1 closed), the firmware-PSM-not-honored hang, the multi-function SDIO LMAC-wedge recovery |
+| **What it adds today vs pkgrel=1** | **Patch I**: 5 GHz scan filter — `iw scan freq <single-5ghz-channel>` works, multi-channel per-band sweep refused at driver boundary to dodge firmware reject cascade. NM `band=a` profiles associate to 5 GHz cleanly. **Sustained 11.32 MB/s** download (2.54 GB factory image) on `newton` 5 GHz ch.48 — **3.6× the 2.4 GHz baseline of 3.12 MB/s** on the same source. |
+| **Source-of-truth (driver)** | `git.reauktion.de/marfrit/bes2600-dkms` — branch `cleanups` for c-stack+A+B, branch `bes2600/scan-filter-5ghz` for Patch I |
+| **Caveat** | `CONFIG_SHADOW_CALL_STACK=n` (security-hardening regression, workaround for a GCC 15.2.1 + arm_neon.h pragma issue — tracked in [besser#20](https://git.reauktion.de/marfrit/besser/issues/20), restore to `=y` when GCC is fixed) |
+
+## pkgrel history
+
+| pkgrel | Date | Flow | Notes |
+|---|---|---|---|
+| 1–3 | 2026-05-08…05-18 | hand-managed, this dir | c-stack + Patches A/B/C/D/E/F/G/H + Patch I + SCS Makefile workaround |
+| 4 | 2026-05-18 | kernel-agent (`ka-promote ohm`) | migration-only release: byte-identical source to pkgrel=3 (148 149 + 7 735 + 1 562 = 157 446 cumulative arithmetic); fixes pkgrel=3 PKGBUILD's duplicated `0003-...patch` source-array bug. Available as fallback. |
+| **5** | **2026-05-18** | **kernel-agent (`ka-promote ohm`)** | adds [besser#18](https://git.reauktion.de/marfrit/besser/issues/18) lockdep fix (pending_record_lock SOFTIRQ-safe → -unsafe inversion). 4-patch cumulative, 162 704 B, b2sum `0eb091ddaba4…`. Closes besser#18 + besser#1. |
+
+---
+
+## What's in the patchset
+
+A 17-commit cumulative diff over `v7.0-danctnix1`'s in-tree `drivers/staging/bes2600/`, plus the standalone Patch I (5 GHz scan filter) and an arm64 build-environment workaround for GCC 15.
+
+Individual commits with full rationale + Phase-7 verification logs live on the **`cleanups` branch** of [`marfrit/bes2600-dkms`](https://git.reauktion.de/marfrit/bes2600-dkms/commits/branch/cleanups) and the **`bes2600/scan-filter-5ghz` branch** for Patch I. This PKGBUILD ships them squashed into separate patch files for build atomicity.
+
+| group | what it does |
+|---|---|
+| **c-stack (c5.1–c5.2.1, c6.1, c6.2, c7)** | wifi-stability fixes: scan-defer-on-firmware-reject, scan-defer-backoff-tune, LMAC recover via `mmc_hw_reset`, PM state resync, wake-state consume, firmware-doesn't-honour-PSM self-detect, multi-function SDIO `mmc_hw_reset` rescan |
+| **Patch A** | decrypt-storm fast-recover at `bes2600_rx_cb`: ≥5 `WSM_STATUS_DECRYPTFAILURE` in 5 s → `ieee80211_connection_loss(vif)`. Phase-7 confirmed N=2 (2026-05-07), storms recover ~1 s vs 109 s baseline. |
+| **Patch B** | connection-loss bus-reset: ≥3 driver-side connection-loss decisions in 60 s on the same vif → `mmc_hw_reset` instead of mac80211 reauth. Installed dormant; never tripped in production yet. |
+| **Patch C v3** | structural: drop `sdio_rx_work` workqueue relay; IRQ → bh-direct architecture (matches mainline cw1200). +73 % sustained RX. |
+| **Patch D** | `ba_lock` removed; `ba_acc/ba_cnt/ba_acc_rx/ba_cnt_rx/ba_ena` → `atomic_t`; per-RX-frame spinlock eliminated. |
+| **Patch E** | per-RX-frame `ps_state_lock` skipped when c7's `pm_unsupported` latch is on (steady-state on production firmware). |
+| **Patch F** | cw1200 mainline backports: hw_scan SKB-lifecycle UAF, `init_common` `destroy_workqueue` on error, `atomic_add(1, x) → atomic_inc(x)` cosmetic. |
+| **Patch G** | GPL-2.0 §1 attribution restoration: SPDX-License-Identifier on every file, Tarnyagin/ST-Ericsson copyright restored on cw1200-derived files. |
+| **Patch C2** | `ieee80211_rx_irqsafe → ieee80211_rx_ni` at all 6 sites (kernel.org-clean process-context API; tasklet hop removed). |
+| **Patch H** | `bh.c` hygiene cleanup: 76- and 468-line `#if 0` cw1200-ancestor fossil blocks removed; `__bes2600_irq_enable` stub removed; per-iteration `BUG_ON` → `WARN_ON_ONCE`. |
+| **Patch I** ([besser#1](https://git.reauktion.de/marfrit/besser/issues/1)) | **5 GHz scan filter.** Refuses only **multi-channel** 5 GHz scans (the per-band-sweep mac80211 issues internally) at the driver boundary with `-EOPNOTSUPP`, dodging the firmware's status-2 reject cascade. Single-channel 5 GHz scans pass through so NM/`wpa_supplicant` per-freq BSS discovery (when `802-11-wireless.band=a`) still finds and associates to 5 GHz APs. Net effect: dmesg storm gone, 5 GHz attachment works, 3.6× sustained throughput on 5 GHz HT40 vs 2.4 GHz HT20. |
+| **arm64 SCS Makefile workaround** | Adds `-ffixed-x18` explicitly for `arch/arm64/lib/xor-neon.o` when `CONFIG_SHADOW_CALL_STACK=y`. Dead code in this pkgrel (SCS is off), in place for the day SCS re-enable becomes possible. See [besser#20](https://git.reauktion.de/marfrit/besser/issues/20). |
+
+## Measured outcome
+
+- **Phase 7 (Patch I, 2026-05-18):** Pattern A `wsm_generic_confirm failed for request 0x0007` storm: 14.3/h → **0/h** over 30-min observation. 5 GHz `newton` BSSID `c0:25:06:e6:5b:33` @ 5240 MHz (ch.48), TX bitrate 150 Mbit/s MCS 7 HT40 short-GI. Internet download throughput **11.32 MB/s** (sustained 90.5 Mbit/s, ~60 % of PHY) vs 3.12 MB/s on 2.4 GHz HT20 same source.
+- **Phase 7 (Patch C v3 + F + G + D + E + C2 + H, Mobian-flavor):** N=3 stress @ 4 MB/s sender on RK3566/PineTab2 — Patch B baseline 1.36 MB/s → +73 % sustained 2.28 MB/s. Race-fix verified under stress (no `wsm_release_tx_buffer` WARN storm under load).
+- Module loads + associates cleanly; `pm_unsupported` latch fires on boot as expected.
+
+## Building (pkgrel=4+, kernel-agent flow)
+
+Builds run out of the new home:
+
+```sh
+cd ~/src/marfrit-packages/arch/linux-pinetab2-danctnix-besser
+makepkg -s
+```
+
+To refresh the cumulative patch from a new kernel-agent manifest state:
+
+```sh
+cd ~/src/kernel-agent
+./bin/ka-promote ohm
+cp build/ohm/v7.0-danctnix1/cumulative.patch \
+   ~/src/marfrit-packages/arch/linux-pinetab2-danctnix-besser/0001-bes2600-besser-kernel-agent-cumulative.patch
+cp build/ohm/v7.0-danctnix1/manifest.lock \
+   ~/src/marfrit-packages/arch/linux-pinetab2-danctnix-besser/manifest.lock
+b2sum 0001-bes2600-besser-kernel-agent-cumulative.patch  # update PKGBUILD b2sums and pkgrel
+```
+
+## Building (pkgrel ≤ 3, hand-managed flow — DEPRECATED)
+
+```sh
+cd ~/src/besser/marfrit-besser/danctnix-besser-pkgbuild/kernel
+makepkg -s
+```
+
+Produces `linux-pinetab2-danctnix-besser-<ver>-aarch64.pkg.tar.zst` plus a matching `-headers` package. Build host can be aarch64 native (recommended — no cross-toolchain setup) or x86 with an aarch64 cross-compiler.
+
+Build time: ~45–55 min on an 8-core aarch64 host (boltzmann/RPi5-class), most of it the kernel modules phase.
+
+**GCC 15.2.1 note:** This pkgrel ships with `CONFIG_SHADOW_CALL_STACK=n` because GCC 15.2.1's strict pragma validator chokes on `arm_neon.h`'s push/`target("+nothing+aes")`/pop sequences when SCS is on. The `0003-arm64-xor-neon-ffixed-x18-build-fix.patch` is a defensive Makefile-side workaround that's a no-op while SCS is off; it'll silently unblock SCS=y once GCC upstream is fixed. See [besser#20](https://git.reauktion.de/marfrit/besser/issues/20) for the re-enable plan.
+
+## Installing
+
+The package declares `provides=("linux-pinetab2=$pkgver-$pkgrel")` and `conflicts=(linux-pinetab2)`, so `pacman` will cleanly take over from upstream `linux-pinetab2`:
+
+```sh
+sudo pacman -U linux-pinetab2-danctnix-besser-7.0.danctnix1-5-aarch64.pkg.tar.zst
+```
+
+That removes the upstream `linux-pinetab2` package (if installed) and registers the BESser-flavored kernel under the same provides slot. Headers package is optional; install it if you build out-of-tree modules.
+
+The pacman `mkinitcpio` hook auto-generates `/boot/initramfs-linux-pinetab2-danctnix-besser.img`. Modules land in `/usr/lib/modules/<release>-pinetab2-danctnix-besser/`, vmlinuz at `/boot/vmlinuz-linux-pinetab2-danctnix-besser`, DTBs at `/boot/dtbs/rockchip/rk3566-pinetab2-{v0.1,v2.0}.dtb`.
+
+### Bootloader (PineTab2-specific)
+
+PineTab2 boots via U-Boot loading a script `boot.scr` (compiled from `/boot/boot.txt` via `mkscr`). After install, point the script at the new kernel + initramfs:
+
+```sh
+sudo cp /boot/boot.txt /boot/boot.txt.pre-besser
+sudo cp /boot/boot.scr /boot/boot.scr.pre-besser
+sudo sed -i \
+  -e 's|/vmlinuz-linux-pinetab2$|/vmlinuz-linux-pinetab2-danctnix-besser|' \
+  -e 's|/initramfs-linux-pinetab2\.img|/initramfs-linux-pinetab2-danctnix-besser.img|' \
+  /boot/boot.txt
+cd /boot && sudo ./mkscr
+sudo systemctl reboot
+```
+
+Backups (`*.pre-besser`) let you revert without touching the U-Boot console: `sudo cp /boot/boot.scr.pre-besser /boot/boot.scr` and reboot.
+
+## Verifying
+
+After reboot:
+
+```sh
+uname -r
+# expected: <kver>-pinetab2-danctnix-besser
+
+lsmod | grep -i bes2600
+# expected: bes2600 (loaded), bes2600_btuart (loaded if Bluetooth in use)
+
+cat /sys/module/bes2600/srcversion
+# expected: BEB625FA7443171EA8D55F7 for pkgrel=3 (and pkgrel=4 — byte-identical source)
+```
+
+`dmesg | grep bes2600` should show clean firmware load, no SDIO TX panic, no `wsm_release_tx_buffer` WARN storm under load, no `wsm_generic_confirm failed for request 0x0007` storm.
+
+For the 5 GHz fix specifically:
+```sh
+sudo iw dev wlan0 scan freq 5180
+# expected: completes, no "Operation not supported"
+
+sudo iw dev wlan0 scan freq 5180 5200 5220 5240
+# expected: "Operation not supported (-95)" — multi-channel 5 GHz refused
+```
+
+## Rolling back
+
+If the new kernel misbehaves:
+
+```sh
+sudo cp /boot/boot.scr.pre-besser /boot/boot.scr
+sudo systemctl reboot
+```
+
+That returns you to whatever kernel `boot.scr` was pointing at before the install (typically upstream `linux-pinetab2` or the previous `linux-pinetab2-danctnix-besser`). The package itself can be removed with `sudo pacman -R linux-pinetab2-danctnix-besser` and the original `linux-pinetab2` re-installed via `sudo pacman -S linux-pinetab2`.
+
+## Provenance
+
+- Mobian-flavor source-of-truth: <https://git.reauktion.de/marfrit/bes2600-dkms> (`cleanups` branch for c-stack + Patches A/B, `bes2600/scan-filter-5ghz` for Patch I)
+- Per-patch breakdown, Phase 0–7 logs, follow-up issues: <https://git.reauktion.de/marfrit/besser>
+- Upstream cw1200 mainline (architectural reference): `drivers/net/wireless/st/cw1200/` in linux-rockchip
+- Kernel base: <https://codeberg.org/DanctNIX/linux-pinetab2> tag `v7.0-danctnix1`
+- Kernel-agent mirror of the patch tree + per-host manifest: <https://git.reauktion.de/marfrit/kernel-agent>
+
+## Why it's "BESser"
+
+"Besser" = German for "better." Patch series ID across both DKMS (Mobian) and in-tree (Danctnix) trees. Single source-of-truth lives in `marfrit/bes2600-dkms`; this PKGBUILD is the danctnix-flavor consumption surface.
+
+## Soft-upstream intent
+
+Submitting this PKGBUILD to DanctNIX for review. If accepted as a replacement for `linux-pinetab2` (or sidegrade), the BESser patchset ships to all PineTab2 users via the regular danctnix package update channel. The bes2600 driver gets:
+
+- ~2× sustained RX throughput on 2.4 GHz
+- ~3.6× sustained RX throughput on 5 GHz (via Patch I + correctly using HT40)
+- Race-correctness on the hot path
+- GPL-2.0 §1 attribution compliance
+- Modern kernel API (no deprecated `from_timer`, no `_irqsafe` from process context, no `BUG_ON` in steady-state)
+
+Drop-in compatibility: same kernel version, same module names, no userspace ABI change. SCS off is the one config caveat, tracked in [besser#20](https://git.reauktion.de/marfrit/besser/issues/20).
+
+## Maintenance plan
+
+**Effective pkgrel=4+:** the per-host manifest in `marfrit/kernel-agent` (`fleet/ohm.yaml`) is the per-patch authority. `ka-promote ohm` produces the cumulative; the PKGBUILD in `marfrit/marfrit-packages` consumes it. Updates flow:
+
+- New danctnix kernel release → bump `baseline.ref` in `fleet/ohm.yaml`, re-promote, bump pkgver in marfrit-packages PKGBUILD.
+- New BESser patch → add a new series-dir in `kernel-agent/patches/driver/bes2600/`, add to `fleet/ohm.yaml` `includes:`, re-promote, refresh cumulative + b2sum in marfrit-packages PKGBUILD, bump pkgrel.
+- Both flavors continue to be maintained in lockstep via `marfrit/bes2600-dkms` source-of-truth.
+- GCC 15 SCS issue → periodically re-test build with `CONFIG_SHADOW_CALL_STACK=y` against current Arch ARM GCC. When the build succeeds, flip the config and re-deploy.
+
+## Known gaps
+
+- Cumulative diff (squashed) for the c-stack + Patches A/B; Patch I as a separate `0002-` file. Per-patch series can be regenerated if danctnix maintainers prefer.
+- Bluetooth-side `bes2600_btuart` is independent and untouched by this patchset.
+- `bes2600_switch_bt` orchestration removed (Mobian-only entry point; not used in danctnix tree).
+- Multi-band `iw scan` (no `freq` filter) still reports aborted scan because mac80211 aggregates per-band results and marks the whole scan aborted when any leg returns negative (mac80211 contract, not bes2600). Single-band scans (`iw scan freq 2462` or `iw scan freq 5180`) work normally; `nmcli connection up` with `band=bg` or `band=a` profile works normally. This is the Phase 5 reviewer's predicted residual limitation; userspace tools that need full multi-band BSS discovery should issue per-band scans.
+
+## Author
+
+Markus Fritsche <fritsche.markus@gmail.com>
+
+Built collaboratively with Claude Opus 4.7 (1M context).

danctnix-besser-pkgbuild/kernel/0002-bes2600-filter-5ghz-scan.patch

@@ -0,0 +1,168 @@
+From 093a5038b8b68f316d976b7cb69609ca7f24f322 Mon Sep 17 00:00:00 2001
+From: Markus Fritsche <fritsche.markus@gmail.com>
+Date: Mon, 18 May 2026 11:27:40 +0200
+Subject: [PATCH 1/2] bes2600: filter 5 GHz scans at the driver boundary
+ (besser#1)
+
+The BES2600 firmware refuses WSM start-scan for 5 GHz with status 2
+("rejected by policy").  This shows up in dmesg as the recurring
+
+    wsm_generic_confirm failed for request 0x0007.
+    [SCAN] Scan failed (-22).
+
+pattern (besser issue #1, ~14-16/h on ohm/PineTab2 baseline).
+
+Trace shows every reject is the second of a back-to-back pair: mac80211
+splits multi-band hw_scan requests per band when the driver does not
+set IEEE80211_HW_SINGLE_SCAN_ON_ALL_BANDS (we don't), then re-invokes
+drv_hw_scan from __ieee80211_scan_completed for each subsequent band.
+The 2.4 GHz iteration succeeds; the 5 GHz iteration is what the
+firmware rejects.  See ieee80211_prep_hw_scan in net/mac80211/scan.c
+for the loop, and the existing memory reference_bes2600_5ghz_scan_reject
+for the firmware behaviour.
+
+The 056a71a defer-on-reject patch already in this tree handles the
+BT-A2DP-coex branch and the consecutive-reject backoff, but it cannot
+prevent the per-band-loop reject: by the time defer_should_scan is
+consulted, the per-band call is already in flight, and the reject_count
+gets reset on every successful 2.4 GHz scan in between (which is
+~36% of attempts), so the threshold never trips.
+
+The fix: refuse the 5 GHz iteration upfront in bes2600_hw_scan.  The
+2.4 GHz scan still runs normally.  The 5 GHz portion is reported as
+aborted to userspace -- same outcome as today, minus the dmesg storm
+and the wsm_generic_confirm WARN cascade.
+
+5 GHz band registration is intentionally left in place: direct-BSSID
+association to a known 5 GHz AP still works (no scan is needed for
+that path), and a future firmware update that fixes the scan behaviour
+should not be foreclosed by changing band advertisement.
+
+Contract: per include/net/mac80211.h ieee80211_ops.hw_scan, a negative
+return aborts the scan without requiring ieee80211_scan_completed().
+-EOPNOTSUPP is the semantically accurate code (operation is legal,
+driver can't service it on this band today).
+
+Phase 3 evidence:
+- baseline N=3: rate ~14.3-23.6/h converged at 14.3/h (matches OP)
+- back-to-back scan gap: 6/6 rejected pairs <200us, 1/1 successful
+  pair was 114ms (single-band-only, no 5 GHz leg)
+- defer log fires: 0/9 in 30-min window (056a71a structurally bypassed)
+
+Predicted Phase 7 delta: Pattern A 14/h -> 0/h.
+---
+ bes2600/scan.c | 22 ++++++++++++++++++++++
+ 1 file changed, 22 insertions(+)
+
+diff --git a/drivers/staging/bes2600/scan.c b/drivers/staging/bes2600/scan.c
+index fb1d298..a81afb6 100644
+--- a/drivers/staging/bes2600/scan.c
++++ b/drivers/staging/bes2600/scan.c
+@@ -238,6 +238,28 @@ int bes2600_hw_scan(struct ieee80211_hw *hw,
+ 	/* Scan when P2P_GO corrupt firmware MiniAP mode */
+ 	if (priv->join_status == BES2600_JOIN_STATUS_AP)
+ 		return -EOPNOTSUPP;
++
++	/*
++	 * Firmware refuses WSM start-scan for 5 GHz with status 2 ("rejected
++	 * by policy"); see besser issue #1.  mac80211 splits multi-band
++	 * hw_scan requests per-band when the driver does not set
++	 * IEEE80211_HW_SINGLE_SCAN_ON_ALL_BANDS (we don't -- see
++	 * ieee80211_hw_set() calls in bes2600_main.c), so each per-band call
++	 * has req->channels[] from one band only (see ieee80211_prep_hw_scan
++	 * in net/mac80211/scan.c).  Refuse the 5 GHz iteration at the driver
++	 * boundary so userspace gets a clean aborted-scan for that portion
++	 * rather than waiting for the firmware reject to cascade up.  5 GHz
++	 * band registration stays intact so direct-BSSID association to a
++	 * known 5 GHz AP still works (no scan needed for that path).
++	 *
++	 * Contract: per include/net/mac80211.h struct ieee80211_ops.hw_scan
++	 * documentation, a negative return aborts the scan without requiring
++	 * ieee80211_scan_completed().
++	 */
++	if (req->n_channels > 0 &&
++	    req->channels[0]->band == NL80211_BAND_5GHZ)
++		return -EOPNOTSUPP;
++
+ #if 0
+ 	if (work_pending(&priv->offchannel_work) ||
+ 			(hw_priv->roc_if_id != -1)) {
+-- 
+2.54.0
+
+
+From 8cd10f487c8144d462a510812ba0fa717b3e24df Mon Sep 17 00:00:00 2001
+From: Markus Fritsche <fritsche.markus@gmail.com>
+Date: Mon, 18 May 2026 15:56:34 +0200
+Subject: [PATCH 2/2] bes2600: scan-filter-5ghz: allow targeted single-channel
+ scans (besser#1 follow-up)
+
+The original Patch I refused EVERY 5 GHz scan request unconditionally
+(req->n_channels > 0 && band == NL80211_BAND_5GHZ).  This eliminated
+the Pattern A storm but also broke 5 GHz association entirely:
+NM / wpa_supplicant iterates a freq_list when a connection profile
+specifies 802-11-wireless.band=a, issuing per-frequency single-channel
+scans to find the BSS before associating.  Those single-channel scans
+were also refused by our guard, so the BSS was never seen and
+'Wi-Fi network could not be found' was the only outcome.
+
+Tighten the guard: refuse only multi-channel 5 GHz scans (n_channels
+> 1), which is the per-band-sweep pattern mac80211 issues internally
+and the only one that triggers the firmware storm at the per-band
+loop boundary.  Single-channel 5 GHz scans pass through to firmware,
+which generally accepts them -- and when they happen to be rejected,
+the failure is isolated and doesn't cascade.
+
+Verified on ohm with pkgrel=3 (srcversion BEB625FA7443171EA8D55F7):
+  - Pattern A count since boot: 0 (Phase 7 prediction still holds)
+  - iw dev wlan0 scan freq 5180          -> allowed
+  - iw dev wlan0 scan freq 5180 5200 ... -> refused -EOPNOTSUPP
+  - NM 'nmcli connection up' with band=a -> associated to BSSID
+    c0:25:06:e6:5b:33 on 5240 MHz / ch.48 in ~1 second
+  - TX bitrate 150 Mbit/s MCS 7 40MHz short-GI (vs 72.2 Mbit/s
+    HT20 on 2.4 GHz) -- ~2x throughput recovered
+
+The change is a single byte (> 0 -> > 1) plus comment update; the
+test confirmation above is what motivates it.
+
+Refs: besser#1 (closed but tracked for follow-up like this), original
+Patch I sha 093a503.
+---
+ bes2600/scan.c | 16 ++++++++++++----
+ 1 file changed, 12 insertions(+), 4 deletions(-)
+
+diff --git a/drivers/staging/bes2600/scan.c b/drivers/staging/bes2600/scan.c
+index a81afb6..497523b 100644
+--- a/drivers/staging/bes2600/scan.c
++++ b/drivers/staging/bes2600/scan.c
+@@ -248,15 +248,23 @@ int bes2600_hw_scan(struct ieee80211_hw *hw,
+ 	 * has req->channels[] from one band only (see ieee80211_prep_hw_scan
+ 	 * in net/mac80211/scan.c).  Refuse the 5 GHz iteration at the driver
+ 	 * boundary so userspace gets a clean aborted-scan for that portion
+-	 * rather than waiting for the firmware reject to cascade up.  5 GHz
+-	 * band registration stays intact so direct-BSSID association to a
+-	 * known 5 GHz AP still works (no scan needed for that path).
++	 * rather than waiting for the firmware reject to cascade up.
++	 *
++	 * Only the multi-channel case is refused (n_channels > 1): that's
++	 * the per-band-sweep pattern mac80211 issues internally and the
++	 * one that triggers the firmware storm at the per-band loop
++	 * boundary.  Single-channel 5 GHz scans (BSS verification, NM's
++	 * per-freq iteration when 802-11-wireless.band=a is set) pass
++	 * through to firmware, which generally accepts them since the
++	 * storm is the back-to-back per-band issue, not a blanket 5 GHz
++	 * reject.  This preserves 5 GHz association via the
++	 * "wpa_supplicant iterates freq_list per channel" path.
+ 	 *
+ 	 * Contract: per include/net/mac80211.h struct ieee80211_ops.hw_scan
+ 	 * documentation, a negative return aborts the scan without requiring
+ 	 * ieee80211_scan_completed().
+ 	 */
+-	if (req->n_channels > 0 &&
++	if (req->n_channels > 1 &&
+ 	    req->channels[0]->band == NL80211_BAND_5GHZ)
+ 		return -EOPNOTSUPP;
+ 
+-- 
+2.54.0
+

danctnix-besser-pkgbuild/kernel/0003-arm64-xor-neon-ffixed-x18-build-fix.patch

@@ -0,0 +1,36 @@
+From: Markus Fritsche <fritsche.markus@gmail.com>
+Date: Mon, 18 May 2026 11:42:00 +0200
+Subject: [PATCH] arm64: xor-neon: restore -ffixed-x18 when SHADOW_CALL_STACK=y
+ (GCC 15+ build fix)
+
+GCC 15.2.1 enforces that -fsanitize=shadow-call-stack requires
+-ffixed-x18 inside arm_neon.h's #pragma GCC target() blocks.  The
+existing CFLAGS_REMOVE_xor-neon.o line strips the kernel-wide
+-ffixed-x18 (it's part of CC_FLAGS_NO_FPU) and CC_FLAGS_FPU does not
+restore it, so xor-neon.c fails to build on stricter GCC versions
+when CONFIG_SHADOW_CALL_STACK=y.
+
+Add an explicit -ffixed-x18 just for this object, gated on the
+SCS config so non-SCS builds are unaffected.
+
+Build environment workaround; not a kernel-runtime bug.
+---
+ arch/arm64/lib/Makefile | 4 ++++
+ 1 file changed, 4 insertions(+)
+
+diff --git a/arch/arm64/lib/Makefile b/arch/arm64/lib/Makefile
+index 1234567..2345678 100644
+--- a/arch/arm64/lib/Makefile
++++ b/arch/arm64/lib/Makefile
+@@ -9,6 +9,10 @@ ifeq ($(CONFIG_KERNEL_MODE_NEON), y)
+ obj-$(CONFIG_XOR_BLOCKS)	+= xor-neon.o
+ CFLAGS_xor-neon.o		+= $(CC_FLAGS_FPU)
+ CFLAGS_REMOVE_xor-neon.o	+= $(CC_FLAGS_NO_FPU)
++# GCC 15+ enforces that -fsanitize=shadow-call-stack requires -ffixed-x18
++# even after a #pragma GCC pop_options inside arm_neon.h.  CC_FLAGS_REMOVE
++# above strips the kernel-wide -ffixed-x18 (part of CC_FLAGS_NO_FPU); add
++# it back here so xor-neon.c still compiles when SHADOW_CALL_STACK=y.
++CFLAGS_xor-neon.o		+= $(if $(CONFIG_SHADOW_CALL_STACK),-ffixed-x18)
+ endif
+ 
+ lib-$(CONFIG_ARCH_HAS_UACCESS_FLUSHCACHE) += uaccess_flushcache.o

danctnix-besser-pkgbuild/kernel/PKGBUILD

@@ -0,0 +1,236 @@
+# Maintainer: Markus Fritsche <fritsche.markus@gmail.com>
+# Forked from: linux-pinetab2 by Danct12 <danct12@disroot.org>
+# Original Contributor: Jan Alexander Steffens (heftig) <heftig@archlinux.org>
+#
+# linux-pinetab2-danctnix-besser: linux-pinetab2 + the BESser
+# bes2600 driver patchset (race-fix, lock-removal, attribution-restore,
+# fossil-cleanup; +73% throughput vs the in-tree baseline).  Soft-upstream
+# fork of linux-pinetab2 — drop-in replacement, same kernel version, only
+# the bes2600 staging driver differs.  See git.reauktion.de/marfrit/besser
+# and git.reauktion.de/marfrit/bes2600-dkms for full provenance.
+
+pkgbase=linux-pinetab2-danctnix-besser
+pkgver=7.0.danctnix1
+pkgrel=3
+pkgdesc='PineTab2 (BESser bes2600 driver patchset)'
+_srcname=linux-pinetab2
+_srctag=v${pkgver%.*}-${pkgver##*.}
+arch=(aarch64)
+_url_git="https://codeberg.org/DanctNIX/${_srcname}"
+url="${_url_git}/commits/tag/$_srctag"
+license=(GPL-2.0-only)
+makedepends=(
+  bc
+  cpio
+  gettext
+  git
+  libelf
+  pahole
+  perl
+  python
+  tar
+  xz
+)
+options=(
+  !debug
+  !strip
+)
+source=(
+  https://cdn.kernel.org/pub/linux/kernel/v${pkgver%%.*}.x/linux-${pkgver%.*}.tar.{xz,sign}
+  ${_url_git}/releases/download/${_srctag}/${_srctag}.patch.zst{,.sig}
+  0001-bes2600-besser-cumulative-series.patch
+  0002-bes2600-filter-5ghz-scan.patch
+  0003-arm64-xor-neon-ffixed-x18-build-fix.patch
+  0003-arm64-xor-neon-ffixed-x18-build-fix.patch
+  config  # the main kernel config file
+)
+validpgpkeys=(
+  ABAF11C65A2970B130ABE3C479BE3E4300411886  # Linus Torvalds
+  647F28654894E3BD457199BE38DBBDC86092693E  # Greg Kroah-Hartman
+  F09A933C0FE0331E558CA4E166CAB7EAA45DD781  # Danct12
+)
+b2sums=('3d9795083c8938f80f480de0d10bfd9c525640e59d5c7f22983de3f12ee42c84c31be902cafb05579ddb1c32bac5ed06b0d4953f9705450be185bd2d9ab08f89'
+        'SKIP'
+        '71fe98221e802b315e54b4b10d3e8c8f376695a36bae3541d876e5776a37f3fa33c8f8dfa6e51fcbd6f5396add02e5166634165f2351836a0ea0453c172fe56c'
+        'SKIP'
+        'fca0a5badf762d5dbc085261cccc07ddeef96384d2ae0a426fb0412acd7a180e068cabd59f01342b7575d41889afc0f47dfbc9256801ab809f746278e6dab510'
+        '396acbdcf570eada62533c0b8f505ed18077e8432249bab5b8ac8d1107cabc9489bdb91a5780446237ec4fd9ba5fc57a49dff34c16ddab60dc30513fc535f00f'
+        '2714e3c0cd8ec978ce9431418f44f578220886fcabb738c9a0c43fc3c043753960b7c47ae96e1780154d8b266a2add6098407de4ffa7aee40d77ce17e8c70df9'
+        '2714e3c0cd8ec978ce9431418f44f578220886fcabb738c9a0c43fc3c043753960b7c47ae96e1780154d8b266a2add6098407de4ffa7aee40d77ce17e8c70df9'
+        '656a998ab40cb85ee4c00f087b071a91632a6c091da2c84b0f74236b51d2dea6e9db6886625f80ad81dc249d8494ec47cd79d6dd9ea4f5e44f3cde857f861e10')
+
+export KBUILD_BUILD_HOST=archlinux
+export KBUILD_BUILD_USER=$pkgbase
+export KBUILD_BUILD_TIMESTAMP="$(date -Ru${SOURCE_DATE_EPOCH:+d @$SOURCE_DATE_EPOCH})"
+
+prepare() {
+  cd linux-${pkgver%.*}
+
+  echo "Setting version..."
+  echo "-$pkgrel" > localversion.10-pkgrel
+  echo "${pkgbase#linux}" > localversion.20-pkgname
+
+  local src
+  for src in "${source[@]}"; do
+      src="${src%%::*}"
+      src="${src##*/}"
+      src="${src%.zst}"
+      [[ $src = *.patch ]] || continue
+      echo "Applying patch: $src..."
+      patch -Np1 < "../$src"
+  done
+
+  echo "Setting config..."
+  cp ../config .config
+  make olddefconfig
+  diff -u ../config .config || :
+
+  make -s kernelrelease > version
+  echo "Prepared $pkgbase version $(<version)"
+}
+
+build() {
+  cd linux-${pkgver%.*}
+  make DTC_FLAGS="-@" all
+  make -C tools/bpf/bpftool vmlinux.h feature-clang-bpf-co-re=1
+}
+
+_package() {
+  pkgdesc="The $pkgdesc kernel and modules"
+  depends=(
+    coreutils
+    kmod
+    mkinitcpio
+  )
+  optdepends=(
+    'wireless-regdb: to set the correct wireless channels of your country'
+    'linux-firmware: firmware images needed for some devices'
+  )
+  provides=(
+    KSMBD-MODULE
+    WIREGUARD-MODULE
+    "linux-pinetab2=$pkgver-$pkgrel"
+  )
+  conflicts=(linux-pinetab2)
+  replaces=(
+    wireguard-arch
+  )
+
+  cd linux-${pkgver%.*}
+  local modulesdir="$pkgdir/usr/lib/modules/$(<version)"
+
+  echo "Installing boot image..."
+  # systemd expects to find the kernel here to allow hibernation
+  # https://github.com/systemd/systemd/commit/edda44605f06a41fb86b7ab8128dcf99161d2344
+  install -Dm644 "$(make -s image_name)" "$modulesdir/vmlinuz"
+
+  # Used by mkinitcpio to name the kernel
+  echo "$pkgbase" | install -Dm644 /dev/stdin "$modulesdir/pkgbase"
+
+  echo "Installing modules..."
+  ZSTD_CLEVEL=19 make INSTALL_MOD_PATH="$pkgdir/usr" INSTALL_MOD_STRIP=1 \
+    DEPMOD=/doesnt/exist modules_install  # Suppress depmod
+
+  echo "Installing device trees..."
+  make INSTALL_DTBS_PATH="$pkgdir/boot/dtbs" dtbs_install
+
+  # Removing unnecessary device trees (keep only pinetab2 variants).
+  # Use find -delete instead of a bash for-loop: the previous for-loop
+  # silently no-op'd in the makepkg environment, leaving 234 unrelated
+  # board DTBs in the package.  find is robust to nullglob/cwd quirks.
+  find "$pkgdir"/boot/dtbs/rockchip/ -mindepth 1 -maxdepth 1 -type f \
+       ! -name 'rk3566-pinetab2-*' -delete
+
+  # remove build link
+  rm "$modulesdir"/build
+}
+
+_package-headers() {
+  pkgdesc="Headers and scripts for building modules for the $pkgdesc kernel"
+  depends=(pahole)
+
+  cd linux-${pkgver%.*}
+  local builddir="$pkgdir/usr/lib/modules/$(<version)/build"
+
+  echo "Installing build files..."
+  install -Dt "$builddir" -m644 .config Makefile Module.symvers System.map \
+    localversion.* version vmlinux tools/bpf/bpftool/vmlinux.h
+  install -Dt "$builddir/kernel" -m644 kernel/Makefile
+  install -Dt "$builddir/arch/arm64" -m644 arch/arm64/Makefile
+  cp -t "$builddir" -a scripts
+
+  # required when DEBUG_INFO_BTF_MODULES is enabled
+  install -Dt "$builddir/tools/bpf/resolve_btfids" tools/bpf/resolve_btfids/resolve_btfids
+
+  echo "Installing headers..."
+  cp -t "$builddir" -a include
+  cp -t "$builddir/arch/arm64" -a arch/arm64/include
+  install -Dt "$builddir/arch/arm64/kernel" -m644 arch/arm64/kernel/asm-offsets.s
+
+  install -Dt "$builddir/drivers/md" -m644 drivers/md/*.h
+  install -Dt "$builddir/net/mac80211" -m644 net/mac80211/*.h
+
+  # https://bugs.archlinux.org/task/13146
+  install -Dt "$builddir/drivers/media/i2c" -m644 drivers/media/i2c/msp3400-driver.h
+
+  # https://bugs.archlinux.org/task/20402
+  install -Dt "$builddir/drivers/media/usb/dvb-usb" -m644 drivers/media/usb/dvb-usb/*.h
+  install -Dt "$builddir/drivers/media/dvb-frontends" -m644 drivers/media/dvb-frontends/*.h
+  install -Dt "$builddir/drivers/media/tuners" -m644 drivers/media/tuners/*.h
+
+  # https://bugs.archlinux.org/task/71392
+  install -Dt "$builddir/drivers/iio/common/hid-sensors" -m644 drivers/iio/common/hid-sensors/*.h
+
+  echo "Installing KConfig files..."
+  find . -name 'Kconfig*' -exec install -Dm644 {} "$builddir/{}" \;
+
+  echo "Removing unneeded architectures..."
+  local arch
+  for arch in "$builddir"/arch/*/; do
+    [[ $arch = */arm64/ ]] && continue
+    echo "Removing $(basename "$arch")"
+    rm -r "$arch"
+  done
+
+  echo "Removing documentation..."
+  rm -r "$builddir/Documentation"
+
+  echo "Removing broken symlinks..."
+  find -L "$builddir" -type l -printf 'Removing %P\n' -delete
+
+  echo "Removing loose objects..."
+  find "$builddir" -type f -name '*.o' -printf 'Removing %P\n' -delete
+
+  echo "Stripping build tools..."
+  local file
+  while read -rd '' file; do
+    case "$(file -Sib "$file")" in
+      application/x-sharedlib\;*)      # Libraries (.so)
+        strip -v $STRIP_SHARED "$file" ;;
+      application/x-archive\;*)        # Libraries (.a)
+        strip -v $STRIP_STATIC "$file" ;;
+      application/x-executable\;*)     # Binaries
+        strip -v $STRIP_BINARIES "$file" ;;
+      application/x-pie-executable\;*) # Relocatable binaries
+        strip -v $STRIP_SHARED "$file" ;;
+    esac
+  done < <(find "$builddir" -type f -perm -u+x ! -name vmlinux -print0)
+
+  echo "Stripping vmlinux..."
+  strip -v $STRIP_STATIC "$builddir/vmlinux"
+
+  echo "Adding symlink..."
+  mkdir -p "$pkgdir/usr/src"
+  ln -sr "$builddir" "$pkgdir/usr/src/$pkgbase"
+}
+
+pkgname=(
+  "$pkgbase"
+  "$pkgbase-headers"
+)
+for _p in "${pkgname[@]}"; do
+  eval "package_$_p() {
+    $(declare -f "_package${_p#$pkgbase}")
+    _package${_p#$pkgbase}
+  }"
+done

notes/bug5-rx-degradation-phase0-2026-05-07.md

@@ -0,0 +1,108 @@
+# Bug #5 RX-degradation campaign — Phase 0
+
+**Date:** 2026-05-07
+**Module under test:** v3 + F (`bes2600.ko` srcversion `371C6606B73AF19299228CA`)
+**Hardware:** ohm (PineTab2, RK3566 + BES2600 SDIO), wired enu1 fallback path live.
+
+---
+
+## Research question (locked)
+
+> **Why does the bes2600 RX path collapse from ~2 MB/s sustained @ fresh-chip uptime to ~180 B/s @ ~28-min uptime, with periodic `wsm_generic_confirm failed for request 0x0007` + `ieee80211 phy0: [SCAN] Scan failed (-22)` every 300 s in the intervening window?**
+
+Reproduces on Patch B, Patch F, and Patch C v3 alike — independent of the relay/race issues v3 addressed. Side-effect that was masked by the throughput floor while v2's race was the dominant variable.
+
+## Predecessor data (reference, not anchor)
+
+| source | observation |
+|---|---|
+| Patch C v3 N=3 (uptime 200/391/582 s) | mean 2.352 MB/s @ 4 MB/s sender |
+| v3 single rep at uptime ~28 min (rep 2 of 2026-05-07 22:23) | 180 KB / 5 min = 600 B/s, sender saw "Connection reset by peer" |
+| v3 single rep at uptime ~47 min (N=3 first attempt 22:42) | 55 KB / 5 min = 180 B/s, sender timed out (exit 124) |
+| dmesg pattern observed at 47-min uptime | scan failures every 301-302 s starting at uptime 778 s (~13 min) |
+
+The shape: **fresh chip → linear data flow at ~2 MB/s sustained → sometime around 13 min uptime, NetworkManager-triggered scans start failing → sometime around 28 min uptime, data throughput collapses to <1 KB/s while link still shows associated.**
+
+Predecessor data is reference. Phase 0 will re-anchor at N=1 long-trace + 5 in-window stress probes; if the pattern doesn't reproduce, that's the campaign result.
+
+## Mechanism candidates (Phase 4 will discriminate)
+
+1. **Firmware-side resource exhaustion.** Per-scan or per-WSM-event accumulation in chip-side state. Scan-failed -22 (EINVAL) suggests firmware refusing the request — possibly out of scan handles, scan-buffer slots, or some other limit.
+2. **NetworkManager scan-fail recovery loop.** Each failed scan triggers NM retry. If retry overhead dominates the bh thread, data path starves. Verifiable by suppressing NM scans.
+3. **AP-side rate limiting.** Newton (AVM) AP could be applying QoS / fairness / probation after sustained 4 MB/s burst. Verifiable by Fritz!Box log access (Markus has it) or by switching to a different AP.
+4. **PSM state machine deadlock.** c7's `pm_unsupported` self-detect was supposed to handle this, but the latch state could become stale if a real PM_IND arrives mid-operation. Verifiable by `chip_pm_state` debugfs read at degradation onset.
+5. **SDIO bus clock degradation / mmc retune.** SDIO retune with `retune_protected` flag interacts with bes2600's data path. Verifiable by ftrace `mmc/mmc_request_*` event correlation with throughput drop.
+6. **Power-management busy-event accumulation.** `bes2600_pwr_set_busy_event` counters might leak — busy events not cleared lock the chip awake (no PSM) but also exhaust event capacity. Verifiable by `bes2600_pwr_busy_event_record` dump.
+
+## Phase 0 measurement protocol (rig armed 2026-05-07 23:18:58 CEST, T0=1778188738)
+
+Capturing for 35 minutes from fresh boot. All capture lives in `/root/bes2600-samples/run-20260507-bug5-degradation-rig/` on ohm.
+
+### Always-on streams
+
+| stream | tool | output |
+|---|---|---|
+| ftrace events | per-event `enable=1` | `trace.log` (via `trace_pipe`) |
+| cfg80211 events | `iw event -t -f` | `iw-event.log` |
+| kernel printks | `dmesg -wT` | `dmesg.log` |
+| netdev counters | per-30s shell loop | `snap.log` |
+
+### ftrace event set
+
+- `workqueue/workqueue_execute_start` — work dispatches
+- `workqueue/workqueue_queue_work` — work submissions
+- `mac80211/api_beacon_loss` — driver beacon-loss events
+- `mac80211/api_connection_loss` — driver-side conn-loss
+- `mac80211/api_disconnect` — driver-side disconnect
+- `mac80211/drv_hw_scan` — mac80211 → driver scan dispatch
+- `mac80211/drv_set_key` — key state changes
+- `cfg80211/rdev_assoc` — assoc requests
+- `cfg80211/rdev_deauth` — deauth requests
+- `cfg80211/rdev_disassoc` — disassoc requests
+- `cfg80211/cfg80211_assoc_comeback` — AP-side assoc-busy throttling
+- `cfg80211/cfg80211_send_auth_timeout` — auth timeouts
+- `cfg80211/cfg80211_scan_done` — scan completions
+- `power/suspend_resume` — PM transitions
+- `mmc/mmc_request_start` / `mmc_request_done` — bus-level transactions
+
+### Scheduled stress probes
+
+Sender on boltzmann (`/tmp/bug5-probe-loop.sh`) fires `pv -L 4m | nc ohm 12345` for 30 s at T+5/10/15/20/25 min. Each probe brackets uptime, RX-bytes pre, RX-bytes post, elapsed. Throughput-vs-uptime curve falls out of the snap.log + probe boundaries.
+
+Probe markers logged via `logger -t bes2600-bug5 PROBE_N_START/END` so they appear in dmesg.log timeline.
+
+## Anti-theatre receipts (must tick before claiming Phase 0 done)
+
+- [ ] In-session baseline: long-capture across degradation window, N=1 for now; re-run if anomalous
+- [ ] ftrace events actually firing (verify by tail of trace.log mid-capture)
+- [ ] dmesg captures the scan-failure pattern timestamp (expected ~uptime 778 s)
+- [ ] Probes actually transferred data at fresh chip (T+5 should be > 1 MB/s)
+- [ ] At least one probe in-window after scan-failure onset (expected: T+15 or T+20)
+- [ ] Snap.log shows monotonic counter behaviour (no rx_bytes going backwards)
+
+## Phase 1 hypothesis (provisional, refine after Phase 3 data)
+
+Metric candidate: **probe throughput as function of uptime, with state-transition markers (first `wsm_generic_confirm 0x0007 failed`, first `[SCAN] Scan failed (-22)`, first NetworkManager-deauth-and-reassociate)**.
+
+Discriminator question: does throughput collapse abruptly at the first scan failure, or gradually over a window? Abrupt = single-event causation; gradual = accumulator.
+
+## Phase 4 candidates (post-Phase-3)
+
+Depending on which mechanism (1-6) Phase 3 surfaces:
+- (1) firmware resource exhaustion: report to upstream; possibly disable NetworkManager scans pending firmware fix.
+- (2) NM scan-fail loop: configure `wpa_supplicant` to skip scans; or add scan-failure handling in driver to dampen retry cascade.
+- (3) AP-side: switch APs for testing; report to AVM if reproducible.
+- (4) PSM deadlock: extend c7 latch with timeout-or-progress recovery.
+- (5) SDIO retune: ftrace correlation guides the lock-ordering fix.
+- (6) PWR busy-event leak: audit set/clear pairs; add a warning-when-stale.
+
+## Out-of-scope
+
+- Patch C v3 closure (PR #5 merged, Phase 7 done).
+- Patch C2 (`ieee80211_rx_list` batch) — gated on Task #19 kerneldoc.
+- Patch D / E independent.
+- Reproduction at higher rates (8 MB/s ramp) — defer to Phase 4 once mechanism identified.
+
+---
+
+*Phase 0 plan written 2026-05-07 23:21 CEST by Claude (noether), at the close of Patch C v3 Phase 7.  Rig armed; long capture in flight; probes scheduled at T+5/10/15/20/25 min.  Post-capture analysis will populate Phase 3 results before Phase 4 plan branches off.*

notes/patch-c-v3-phase4-plan-2026-05-07.md

@@ -0,0 +1,127 @@
+# Patch C v3 — Phase 4 Plan: drop sdio_rx_work, match cw1200 architecture
+
+**Author:** Claude (noether)
+**Status:** Phase 4 v3 — supersedes v2 (PR #10) after cw1200 mainline survey showed the race-free path is structural, not lock-based.
+**Decision:** drop the `sdio_rx_work` workqueue entirely; SDIO IRQ wakes `bh_wq`; bh thread does the SDIO read inline.  Restores single-writer-from-bh invariant on `hw_bufs_used` *by construction*.  No `atomic_t` prep needed.
+
+---
+
+## §0 Why v3 supersedes v2
+
+PR #10's plan was: convert `hw_bufs_used` etc. to `atomic_t` (prep), then direct-deliver from `sdio_rx_work` (structural).  That was a workaround for the race that *only existed because of the relay*.
+
+The cw1200 mining (`~/src/linux-rockchip`, 228 cw1200 commits) showed the upstream answer: there is no relay.  cw1200's IRQ handler bumps `bh_rx` and wakes the bh thread; the bh thread does the SDIO read itself inside `cw1200_bh_rx_helper` (`drivers/net/wireless/st/cw1200/bh.c:233`).  Single thread = single writer for `hw_bufs_used` = no race.  Same `int hw_bufs_used` as bes2600, never atomic_t'd in 16 years upstream because it never needed to be.
+
+Patch C v3 brings bes2600 into that shape.  The structural simplification is bigger than v2's diff but lands the right architecture in one move.
+
+## §1 Goal
+
+Same as Patch C v2 §1: ≥ 1 MB/s sustained receive @ 4 MB/s sender, < 15 % `_raw_spin_unlock_irqrestore` CPU%, no 30-min cascade to link-death.  Stretch toward Phase 1's full 2 MB/s once Patch C2 (rx_list batch) lands separately.
+
+## §2 Situation
+
+- Cleanups branch is at Patch F merged (commit `b717251`).  All Phase 5 reviews of the F series merged via PR #4.
+- ohm rebooted with F module live (srcversion `A9438692D6A8698F92AEEA1`) — F is the new baseline for Patch C v3 Phase 7 comparison.
+- Wired path `enu1` at `192.168.88.80` survives bes2600 wedges; lmcp `ohm` still goes through wlan0.  Phase 7 telemetry collection over enu1.
+- Reboot-permission override active (ohm dev-allocated; I can `sudo reboot` directly — `feedback_user_pushes_reboot_button` override clause).
+
+## §3 Baseline measurements
+
+Carry forward from `run-20260507-patchC-preflight/baseline.tsv` (N=1, F-less Patch B module):
+
+| metric | value |
+|---|---|
+| observed receive @ 4 MB/s | 1.362 MB/s |
+| sdio_rx_work dispatches | 86.4/s = 90.3 per 1000 RX packets |
+| sdio_tx_work dispatches | 276.1/s |
+| bes2600_bh_work redispatches | 0 (single long-lived) |
+
+**Phase 6 prereq:** capture an N=3 baseline ON THE F MODULE before Patch C v3 code lands.  Same instrumentation, same stress ramp.  This is the post-F / pre-v3 reference.  Without it, Phase 7's delta is C+F vs B+nothing — confounded.
+
+## §4 Plan v3
+
+### §4.1 What gets eliminated
+
+- **`sdio_rx_work` (bes2600_sdio.c:829)** — function deleted.  No longer queued, no longer runs.
+- **`self->rx_work` work_struct** — field deleted from `struct sbus_priv`.  `INIT_WORK` removed.
+- **`self->rx_queue` + `self->rx_queue_lock`** — fields deleted.  `skb_queue_head_init` removed.  No SKB ever queued there.
+- **`bes2600_sdio_pipe_read`** — function deleted.  No callers after this patch.
+- **`sbus_ops->pipe_read`** — sbus op slot deleted (or kept and stubbed; tx_loop.c also implements it for the test-loop bus, has to stay if test-loop is preserved).
+- **`queue_work(self->sdio_wq, &self->rx_work)`** at the 3 call sites in `bes2600_sdio.c` (lines 416, 941, 1199) — removed.
+
+### §4.2 What gets added
+
+- **A new `bes2600_bh_handle_rx_skb()`** in bh.c (same shape as Patch C added, same contract block; no longer needs to also wake the bh thread because we ARE the bh thread).
+- **A new helper `bes2600_sdio_read_rx_batch()`** in bes2600_sdio.c, exported, that does what `sdio_rx_work` used to do MINUS the queuing: lock → read ctrl_reg → memcpy_fromio → packets_check → for-each-frame extract+deliver.  Called from bh.
+
+### §4.3 What gets rewired
+
+- **`bes2600_gpio_irq_handler`** in bes2600_sdio.c:413 (the GPIO-IRQ path used when CONFIG_BES2600_USE_GPIO_IRQ is set):  drop `queue_work(self->sdio_wq, &self->rx_work)`; instead call `self->irq_handler(self->irq_priv)` directly (which is `bes2600_irq_handler` in bh.c, bumps `bh_rx` + wakes `bh_wq`).  Matches cw1200_sdio_irq_handler shape.
+- **`bes2600_bh_rx_helper`** (bh.c:961, BES_SDIO_RX_MULTIPLE_ENABLE branch): instead of `pipe_read`-ing one SKB from the (now-gone) rx_queue, call the new `bes2600_sdio_read_rx_batch()` which does the SDIO read AND delivers each frame inline via `bes2600_bh_handle_rx_skb()`.  Returns count delivered, or negative on error.
+- **`bes2600_bh()` outer loop**:  after a successful rx_batch read, the helper signals whether to continue draining (more frames pending) — same shape as today's `BH_RX_CONT_LIMIT=3` outer loop.
+- **`bes2600_gpio_wakeup_mcu(SDIO_RX)`** + **`bes2600_gpio_allow_mcu_sleep(SDIO_RX)`** brackets:  currently called inside sdio_rx_work.  Move into bh thread around the `bes2600_sdio_read_rx_batch()` call.  Same wake-flag bracketing, just from a different thread.
+- **`sdio_wq` workqueue**:  keeps `tx_work` and (briefly) `scan_work`.  Renamed or kept — cosmetic.  Don't touch in this patch.
+
+### §4.4 What stays untouched
+
+- TX path (`sdio_tx_work`, `bes2600_bh_tx_helper`, `wsm_alloc_tx_buffer`).  Independent.
+- WSM protocol layer (`wsm.c`, `wsm_handle_rx`).  Same callees, just from bh thread now.
+- mac80211 RX delivery (`ieee80211_rx_irqsafe`).  That's Patch C2.
+- `BES2600_RX_IN_BH` ifdef gate.  Stays defined; the gated branch is now the only RX path.
+- Symptom-shaped artifacts (asm nop, BUG_ON in hot path) — still deferred, see task #24 post-cleanup.
+
+## §5 Shared-state delta table (the v2 lesson, applied)
+
+Every field `bes2600_bh_handle_rx_skb` mutates directly or transitively, with the v3 protection:
+
+| field | written by (today) | written by (after v3) | concurrency | required action |
+|---|---|---|---|---|
+| `hw_priv->hw_bufs_used` | bh thread (TX submit + RX confirm), main.c init | **bh thread only** (RX moves into bh) | single-writer | none — `int` is fine, race-free by construction |
+| `hw_priv->hw_bufs_used_vif[i]` | bh thread (TX vif submit + RX vif confirm), main.c init | **bh thread only** | single-writer | none |
+| `hw_priv->wsm_rx_seq[i]` | sdio_rx_work today | bh thread | single-writer | none — moves cleanly between contexts |
+| `hw_priv->wsm_tx_pending[i]` | bh thread (inc on TX submit), bh+sdio_rx_work (dec on RX confirm) | **bh thread only** | single-writer | none |
+| `hw_priv->lmac_mon_timer` / `mcu_mon_timer` | mod_timer / del_timer_sync from bh + sdio_rx_work | bh thread only | timer API safe anyway | none |
+| `hw_priv->wsm_cmd.lock` | spinlock taken inside wsm_handle_rx | same | already protected | none |
+| `priv->bh_evt_wq` wake-up | wsm_release_tx_buffer when count→0 | same | wake_up is concurrency-safe | none |
+| `bes_pwr.lock` (inside bes2600_pwr_clear_busy_event) | bh thread (today) | bh thread | already protected | none |
+| `self->rx_data_cnt` etc. (sbus_priv stats) | sdio_rx_work | bh thread | single-writer | none |
+
+**Zero fields require new locking.** The architectural pivot eliminates the race v2's atomic_t was working around.
+
+## §6 Risks
+
+1. **bh thread now holds the SDIO bus mutex during read** (currently held by sdio_rx_work).  TX work in the same bh thread is unaffected (sdio_tx_work runs on a separate workqueue and shares the same mutex anyway).  The sdio_lock contention pattern doesn't change.
+2. **Loss of "parallelism" between sdio_rx_work and bh TX**:  sdio_rx_work and bh thread *appeared* to run in parallel today, but both serialize through `bes2600_sdio_lock(self)` for the actual bus operations.  The parallelism was illusory.  Net throughput should not regress.
+3. **bh thread CPU-busy-time per RX batch increases**:  inline SDIO read is the same cost, just charged to bh instead of sdio_wq's worker.  Mitigation: the per-IRQ workqueue dispatch cost (~86/s) is what we trade for it.  Net: -86 dispatches/s, +0 µs per frame.
+4. **Multi-RX coalescing (BES_SDIO_RX_MULTIPLE_NUM=16)** stays.  bes2600_sdio_extract_packets parses the multi-frame buffer same as before, just inline now.  No functional change to chip-side behaviour.
+5. **GPIO wake-flag bracketing**:  `bes2600_gpio_wakeup_mcu(SDIO_RX)` and `bes2600_gpio_allow_mcu_sleep(SDIO_RX)` currently bracket sdio_rx_work.  Move them to bracket the new bh-side read.  If the wake-flag accounting is sub-system-scoped (it is — flag bits per subsystem), this is a clean move.
+6. **IRQ re-enable in bh thread**:  cw1200's bh re-enables IRQ via `__cw1200_irq_enable(priv, 1)` after each round.  bes2600 has the analogous `__bes2600_irq_enable(0/1)` (commented out as the `asm volatile("nop")` symptom in `bh.c:1518-1520`).  This patch does NOT re-engage the commented-out re-enable — that's still task #24's call.  But if the IRQ stays disabled across rounds, we'd never receive the next IRQ.  **Investigate before Phase 6 lands**: where does IRQ re-enable happen in the current bes2600 hot path?  The sdio_func IRQ may be auto-managed by sdio core differently.  Block Phase 6 on this audit.
+7. **Phase 7 wedge resilience**:  if v3 has a different bug shape than v2's race (which it shouldn't, since the race is gone by construction), the wired path lets us collect telemetry from a wedged ohm.
+
+## §7 Phase 5 / 6 / 7
+
+- **Phase 5**: PR on `git.reauktion.de/marfrit/besser` with this artifact.  Specifically request reviewer focus on §6 risk #6 (IRQ re-enable mechanism).
+- **Phase 6**: branch off cleanups (post-F): `bes2600/sdio-rx-no-relay`.  Implement the file changes per §4.  Build, install, smoke-test.
+- **Phase 7**: 
+  - First: N=3 stress-ramp **on F module** (post-F pre-v3 baseline).  10 min @ 1, 30 min @ 2, 30 min @ 4 MB/s.  Use wired path for telemetry.
+  - Then: install v3 module, identical N=3 ramp.  Compare deltas.
+  - Predicted: sdio_rx_work dispatch rate → 0/s (was 86/s).  observed receive lifts toward ≥ 1.0 MB/s sustained.  `_raw_spin_unlock_irqrestore` drops by the rx_queue lock contribution (was 1914/s acquires).
+
+## §8 What gets dropped from v2 plan
+
+- atomic_t prep refactor (`hw_bufs_used` → `atomic_t`): not needed.  Single-writer invariant preserved structurally.  Still a defensible standalone hardening patch *if mainlining bes2600 ever requires defense-in-depth*, but not on the Bug-#5 critical path.
+- `wsm_tx_pending[]` decrement-decision race (v2 risk #2): also moots.  Both sides single-thread under v3.
+- v2 Phase 7's "C-prep should show zero delta" gate: replaced by "v3 should match cw1200's structural shape" gate.
+
+## §9 Open question for reviewer
+
+The big one is §6 risk #6 — IRQ re-enable. cw1200 explicitly does `__cw1200_irq_enable(priv, 1)` from bh after each round; bes2600 has the call **commented out** with an `asm volatile("nop")` placeholder.  Either:
+
+(a) bes2600's SDIO IRQ is level-triggered + auto-acked by SDIO core, so re-enable isn't needed (that would explain the nop).
+(b) The current code happens to work because sdio_rx_work is queued by the IRQ regardless of whether IRQ is "enabled" by the driver-side flag.  After v3 we have to manually re-enable like cw1200 does.
+
+Need to confirm (a) vs (b) before Phase 6 lands.  Plan to grep for `__bes2600_irq_enable` callsites and trace back to whether it's load-bearing.
+
+---
+
+*Plan written 2026-05-07 by Claude (noether), after Patch F merged and Patch C v2 (PR #10) was superseded by the cw1200 architectural mining finding.  Phase 5 review on PR.  Don't curate.*

notes/patch-c2-phase4-plan-2026-05-08.md

@@ -0,0 +1,171 @@
+# Patch C2 — Phase 4 Plan: migrate ieee80211_rx_irqsafe → ieee80211_rx_list
+
+**Author:** Claude (noether)
+**Status:** Phase 4 — pending Phase 5 PR review before any Phase 6 code.
+**Predecessor:** Patch C v3 (PR #5 merged, +73% throughput, no-relay architecture); Patch D + E + F + G also landed. Cleanups branch tip = 42fd0ce.
+**Task #19 contract**: `ieee80211_rx_list` callable from process context, **requires `local_bh_disable()` + `rcu_read_lock()` wrap**, **cannot mix with `ieee80211_rx_irqsafe()` for the same hardware** → all 6 sites convert in one shot.
+
+---
+
+## §0 Substrate
+
+After Patch C v3:
+- bh thread is the sole RX-delivery context (no relay, no sdio_rx_work)
+- Per-frame work runs in process context (sleepable)
+- Single-writer-from-bh invariant covers `hw_bufs_used` and friends
+
+`ieee80211_rx_irqsafe` is currently called from process context. Per kerneldoc (`include/net/mac80211.h:5399-5411`):
+
+> **Like ieee80211_rx() but can be called in IRQ context** (internally defers to a tasklet.)
+
+The tasklet hop is the cost we pay today for delivering each RX frame from process context. `ieee80211_rx_list` is the process-context replacement.
+
+## §1 Goal
+
+Per-frame: skip the tasklet hop. Batch: process multiple SKBs from one SDIO read inside a single `local_bh_disable()`/`rcu_read_lock()` window.
+
+Phase 1 metric: **RX throughput @ 4 MB/s sender**, with v3 N=3 baseline = 2.352 MB/s. Hypothesis: small to moderate uplift (<10%) from removing the tasklet deferral. Larger improvement would be surprising — if observed, that's a finding to investigate.
+
+## §2 Situation
+
+- 6 call sites in bes2600 currently use `ieee80211_rx_irqsafe`:
+  - `ap.c:96` (AP-mode link-id RX queue drain)
+  - `sta.c:1487` (link-id rx_queue drain in ?)
+  - `txrx.c:1960` (early-data + pm_unsupported branch — Patch E added)
+  - `txrx.c:1967` (early-data + LINK_SOFT-not-set branch)
+  - `txrx.c:1971` (normal RX path)
+  - `wsm.c:2415` (beacon SKB delivery from `bes2600_beacon_handler`?)
+- All 6 must convert together (kerneldoc: cannot mix per hardware)
+- bh thread is single-writer post-v3 → `_rx_list`'s "calls must be synchronized" satisfied trivially
+- bh thread is process context → `_rx_list` callable
+
+## §3 Baseline (carry forward)
+
+From `notes/phase7-v3-2026-05-07.md` (v3 N=3 ramp, Phase 7 closed):
+
+| metric | v3 fresh-chip N=3 |
+|---|---|
+| RX throughput @ 4 MB/s | mean 2.352 MB/s, min 2.102, max 2.590 |
+| sdio_rx_work dispatches | 0/s |
+| bh_work redispatches | 0 |
+
+Phase 7 of C2 will compare against this baseline.
+
+## §4 Plan
+
+### §4.1 Conversion shape
+
+Per call site:
+```c
+ieee80211_rx_irqsafe(priv->hw, skb);
+```
+becomes:
+```c
+ieee80211_rx_list(priv->hw, NULL, skb, &priv->rx_list);
+```
+
+Where `priv->rx_list` is a `struct list_head` initialized once.
+
+**Wrap requirement:** `local_bh_disable()` + `rcu_read_lock()` must be held across the call. Per the kerneldoc, that's also needed for batch correctness.
+
+### §4.2 Wrap placement (the design decision)
+
+**Option A — per-call wrap.** Wrap each individual `ieee80211_rx_list()` call. Simple but loses the batch benefit (each call's wrap+unwrap costs as much as the avoided tasklet defer).
+
+**Option B — per-batch wrap.** Wrap the OUTER frame-iteration loop (e.g., the `for` in `bes2600_sdio_extract_packets`). All 16 SKBs from one SDIO read get delivered inside one wrap. This is the upstream-idiomatic pattern (mt76, iwl_pcie do this).
+
+Choosing **Option B**. Concrete shape:
+
+- `bes2600_sdio_read_rx_batch` (the per-SDIO-batch entry point added in Patch C v3) wraps the read+extract+deliver phase:
+  ```c
+  rcu_read_lock();
+  local_bh_disable();
+  // existing read + extract_packets that calls bh_handle_rx_skb per frame
+  local_bh_enable();
+  rcu_read_unlock();
+  ```
+- Inside `bes2600_bh_handle_rx_skb`, the single `ieee80211_rx_irqsafe` swap becomes `ieee80211_rx_list(priv->hw, NULL, skb, &priv->rx_list)`.
+- The OTHER 5 call sites (in `ap.c`, `sta.c`, `txrx.c`'s branches, `wsm.c`) need the same treatment, but they're called from the bh thread (post-v3) so they're already in the right context. Each gets its own narrow wrap (Option A applied selectively because those paths process one frame at a time, not a batch).
+
+### §4.3 The `rx_list` field
+
+Add `struct list_head rx_list` to either `struct bes2600_common` (driver-wide) or `struct bes2600_vif` (per-vif). Per-vif is cleaner because the existing `priv->hw` parameter implies vif scope.
+
+`INIT_LIST_HEAD(&priv->rx_list)` at vif setup; no teardown needed (mac80211 owns the SKBs once handed off).
+
+**Open question for reviewer:** does the `rx_list` need to be drained explicitly after the batch (e.g., via a `list_for_each_entry_safe` + `netif_receive_skb_list_internal`)? Looking at mainline mt76 / iwl_pcie usage will clarify. Phase 6 must answer this before code lands.
+
+### §4.4 What will NOT be touched
+
+- The 6 call sites change atomically (all-or-nothing per kerneldoc) — no per-site progressive migration
+- `wsm.c:2415` beacon path: same conversion shape, but beacon delivery is once-per-beacon-interval (not hot path); could stay `_irqsafe` if upstream allows mixing per-SKB-type. Re-read kerneldoc carefully — it says "per hardware", not per-call-site, so we can't keep _irqsafe even on the slow paths.
+- bh thread structure (Patch C v3 stands)
+- atomic_t counters from Patch D
+- `pm_unsupported` lock-skip from Patch E
+- mac80211 batch-delivery semantics (mainline owns this; we just call the API)
+
+### §4.5 Predicted delta in Phase 3 units
+
+| metric | predicted |
+|---|---|
+| `rx_irqsafe` tasklet schedule rate | → 0 (function no longer called) |
+| RX throughput @ 4 MB/s sustained | 2.352 → +5-15% (medium confidence) |
+| `_raw_spin_unlock_irqrestore` CPU% | small drop (no tasklet schedule lock contribution) |
+
+**Honest acknowledgment:** I don't have data on how much the tasklet hop actually costs. The improvement might be smaller than predicted if tasklet defer was already cheap on this kernel. If <2%, Phase 7 says "marginal but no regression" and we ship anyway for upstream-cleanliness.
+
+### §4.6 Risks
+
+1. **`ieee80211_rx_list` semantics surprise.** mainline drivers I have access to (mt76, iwl_pcie) use this via NAPI infrastructure. bes2600 doesn't have NAPI; we're doing process-context-direct. The kerneldoc says callable that way but we should verify a few mainline drivers actually do it. **Phase 6 contract-cite from at least one upstream caller** before code lands.
+
+2. **`rx_list` lifetime in cross-batch / cross-vif scenarios.** Multiple vifs (P2P_MULTIVIF=y in Makefile) might race on the same hw's `rx_list`. The kerneldoc says "for a single hardware" — the list is per-call destination, which means each call appends to its argument list. Per-vif `rx_list` per-call is the natural shape. No per-hw aggregator needed.
+
+3. **`local_bh_disable` cost in batch wrap.** Not free. If the batch is small (1-2 SKBs), the wrap might dominate. Estimated breakeven: 2-3 SKBs per wrap. Phase 7 should look at SKB-per-batch distribution to confirm.
+
+4. **`rcu_read_lock` across SDIO read.** SDIO read can take multi-ms (multi-block transfers). RCU reader-cs across that is fine (no preemption blocked) but it's a longer reader-cs than typical. Verifiable but not a blocker — kerneldoc requires it.
+
+5. **wsm.c:2415 (beacon) is a different SKB lifecycle** — `hw_priv->beacon` is owned by hw_priv, not allocated per-call. After `_rx_list` consumes it (by passing ownership to mac80211), `hw_priv->beacon` is dangling. **Phase 6 must verify the beacon path either reallocates after delivery or wasn't actually transferring ownership.** Risk #5 is the biggest open question.
+
+### §4.7 Phase 5 review handover
+
+PR on `git.reauktion.de/marfrit/besser` with this artifact. Specifically request reviewer focus on:
+- §4.2 wrap-placement choice (Option B vs A)
+- §4.3 rx_list scoping (per-vif)
+- §4.6 risks #1 (mainline-caller verification) and #5 (beacon path SKB ownership)
+
+Don't curate.
+
+### §4.8 Phase 6 implementation order
+
+1. Branch off cleanups: `bes2600/rx-list-batch-delivery`
+2. Add `struct list_head rx_list` to `struct bes2600_vif`, `INIT_LIST_HEAD` in vif setup
+3. Convert all 6 call sites: `ieee80211_rx_irqsafe(...)` → `ieee80211_rx_list(...)`
+4. Wrap `bes2600_sdio_read_rx_batch` outer loop with `rcu_read_lock + local_bh_disable / local_bh_enable + rcu_read_unlock`
+5. For the non-bh-thread call sites (ap.c, sta.c, wsm.c beacon): per-call narrow wrap
+6. Verify beacon path in wsm.c:2415 (Risk #5)
+7. Build, install, smoke-test
+8. Phase 7 N=3 stress ramp — compare to v3 baseline
+
+### §4.9 Phase 7 protocol (per `feedback_phase7_stress_ramp`)
+
+- N=3 reps, 30s each at 4 MB/s, fresh-chip (uptime <15 min)
+- Use wired path (`ssh mfritsche@192.168.88.80`) for telemetry
+- Fresh nc listener per rep (per `feedback_rig_failure_is_finding`)
+- Compare: throughput delta + tasklet schedule rate (ftrace `irq:tasklet_*` events)
+- If predicted delta met → close C2 + memory entry
+- If NO delta → marginal patch but no regression; ship for upstream-cleanliness
+
+## §5 Out of scope
+
+- Patch D / E already shipped (PR #7, #8 merged)
+- Patch G already shipped (PR #6 merged)
+- bh.c `#if 0` graveyard removal (Task #24 hygiene)
+- Allwinner `sw_mci_check_r1_ready` (Task #25)
+
+## §6 Summary
+
+C2 is a 6-site mechanical migration with ONE design decision (per-batch wrap), TWO open questions for the reviewer (rx_list draining + beacon path SKB ownership), and SMALL expected throughput delta (<15%). Risk-low, upstream-prep-high. Worth shipping for the kernel.org submission story even if the throughput delta is marginal.
+
+---
+
+*Plan written 2026-05-08 by Claude (noether). Phase 5 review on PR. Phase 6 contingent on review passing.*

notes/phase7-c2-2026-05-08.md

@@ -0,0 +1,63 @@
+# Patch C2 Phase 7 — N=3 ramp results
+
+**Date:** 2026-05-08
+**Module:** `bes2600.ko` srcversion `619A51E61BF5479AAC146E6` (cleanups + F + G + D + E + C2)
+**Rig:** ohm fresh boot, wired enu1 path for control, wlan0 for data probes
+**Stress:** netcat sender, `pv -L 4m`, 30 s per rep
+
+---
+
+## Results table
+
+| rep | uptime (s) | rate (MB/s) |
+|---:|---:|---:|
+| 1 | 544 | **2.289** |
+| 2 | 716 | **2.165** |
+| 3 | 750 | **2.376** |
+
+**N=3:** mean 2.277, median 2.289, min 2.165, max 2.376
+
+## Comparison to baselines
+
+| series | mean MB/s | Δ vs Patch B | Δ vs v3 |
+|---|---:|---:|---:|
+| Patch B (run-20260507-patchC-preflight, N=1) | 1.362 | — | -42% |
+| Patch C v3 N=3 (run-20260507-N3v3-rep*) | 2.352 | +73% | — |
+| Patch C v3 + F + G + D + E + C2 N=3 (this rep set) | 2.277 | +67% | -3% |
+
+Δ vs v3 is **within rep variance** (v3 N=3 had min 2.102, max 2.590 → spread ±20%; this set's spread is similar). Statistically indistinguishable.
+
+## Verdict: no measurable C2 throughput delta
+
+The tasklet hop in `ieee80211_rx_irqsafe` was apparently cheap on this kernel. Migrating 6 sites from `_irqsafe` to `_rx_ni` (synchronous-from-process-context, internal `local_bh_disable` wrap) preserves throughput but doesn't measurably improve it.
+
+**This was a predicted outcome.** The C2 Phase 4 plan §4.5 said:
+> "If <2%, Phase 7 says 'marginal but no regression' and we ship anyway for upstream-cleanliness."
+
+Observed: -3% (within noise) → falls into the "marginal but no regression" bucket. Ship for the kernel.org submission story (no `_irqsafe` from process context = upstream-idiomatic) even though performance is unchanged.
+
+## Receipts checklist
+
+- [x] N=3 reps captured at fresh-chip uptime (544/716/750 s — within first 13 min, before scan-failure-cadence onset)
+- [x] All reps under same conditions: same fresh boot, same nc listener, same AP (newton, BSSID c0:25:06:e6:61:b0 on chan 1)
+- [x] No WARN/BUG/oops on any rep
+- [x] dmesg pattern: only the pre-existing wsm_generic_confirm 0x0007 noise — same on Patch B / Patch F / Patch C v3 / D / E / C2 (firmware-side, independent of all our patches)
+- [x] Wired-rig telemetry collection — would have caught any wedge that wlan0 ate
+- [x] Rig-failure-is-finding: an early "0-throughput" set of reps was rig artifact (nc-loop race, port-binding state from a prior session) — caught and discounted per `feedback_rig_failure_is_finding`. The recovered N=3 reps used setsid-detached listener + post-reboot fresh state.
+
+## Phase 8 lesson
+
+**Drop-in replacements with the right kerneldoc reading still need Phase 7 measurement.** I expected +5-15% from removing the tasklet schedule. Got -3% (noise). The cost we were saving was already amortised by something else (NAPI infra? per-CPU softirq scheduling?). The kerneldoc-correctness story stands; the perf story does not.
+
+**Memory entry:** the perf-vs-correctness distinction is worth keeping. `_irqsafe → _rx_ni` is a CORRECTNESS / API-cleanliness move, not a performance optimization. Don't oversell predicted deltas without baseline measurement.
+
+## Out-of-scope follow-ups
+
+- Patch C v3 architectural win is the durable +73%. C / D / E / C2 / F / G are smaller cleanups that don't compound visibly.
+- Bug #5 RX-degradation campaign already closed (hypothesis falsified).
+- Task #24 (post-cleanup observation of bh.c symptom-shaped artifacts): mostly answered.
+- Task #25 (Allwinner sw_mci_check_r1_ready measurement): can be done during any future stress run; not on critical path.
+
+---
+
+*Phase 7 captured 2026-05-08 by Claude (noether). Patch C2 closes the post-Bug-#5 cleanup track. Throughput ceiling on this hardware = ~2.4 MB/s sustained @ 4 MB/s sender, fresh chip; further improvement would need firmware-side fixes (the wsm_generic_confirm 0x0007 path), not driver-side.*

notes/phase7-v3-2026-05-07.md

@@ -0,0 +1,94 @@
+# Patch C v3 Phase 7 — N=3 verification results
+
+**Date:** 2026-05-07
+**Module:** `bes2600.ko` srcversion `371C6606B73AF19299228CA` (cleanups+F+v3)
+**Rig:** ohm (PineTab2, RK3566 + BES2600 SDIO), wired enu1 path for telemetry
+**Stress:** netcat sender from boltzmann, `pv -L 4m` rate cap (4 MB/s), 3-min window per rep
+**Boot:** fresh — uptime 200 s / 391 s / 582 s at rep 1/2/3 starts (all within fresh-chip window before the ~13-min Bug #5 RX-degradation point)
+
+---
+
+## Results table
+
+| rep | elapsed (s) | RX bytes | RX MB | MB/s | sdio_rx_work | sdio_tx_work | bes2600_bh_work redispatches |
+|---:|---:|---:|---:|---:|---:|---:|---:|
+| 1 | 180.72 | 447,758,333 | 427.0 | **2.363** | 0 | 368 | 0 |
+| 2 | 180.67 | 490,669,836 | 467.9 | **2.590** | 0 | 20  | 0 |
+| 3 | 180.69 | 398,224,992 | 379.8 | **2.102** | 0 | 39  | 0 |
+
+**N=3 stats:** mean 2.352 MB/s · median 2.363 MB/s · min 2.102 MB/s · max 2.590 MB/s
+
+## Comparison to baselines
+
+### vs Patch B baseline (`run-20260507-patchC-preflight`, N=1, 5 min @ 4 MB/s, fresh chip)
+
+| | Patch B | v3 mean | Δ |
+|---|---:|---:|---:|
+| throughput | 1.362 MB/s | 2.352 MB/s | **+73%** |
+
+### vs original Bug #5 baseline (`run-20260506-0659-fresh`, N=3, decay over time)
+
+Bug #5 anchor was 725 / 663 / **75** KB/s — rep 3 saw link-death at ~9 min.
+
+| | Bug #5 floor (rep 3) | v3 floor (rep 3) | Δ |
+|---|---:|---:|---:|
+| throughput | 0.075 MB/s | 2.102 MB/s | **28× improvement** |
+
+### vs Phase 4 v3 plan §4.5 predictions
+
+| metric | predicted | observed | verdict |
+|---|---|---|---|
+| sdio_rx_work dispatch rate | → 0/s (high confidence) | 0/s all 3 reps | ✅ |
+| `bes2600_bh_work` redispatches | → 0 (high confidence) | 0 all 3 reps | ✅ |
+| observed RX @ 4 MB/s | floor lifts toward ≥ 1 MB/s sustained (medium) | 2.10 MB/s floor | ✅ exceeds prediction |
+| `_raw_spin_unlock_irqrestore` CPU% | 20% → 12-15% (medium) | not measured | deferred — perf-record run can confirm |
+
+## Workqueue dispatch rate collapse
+
+Patch B baseline (per `run-20260507-patchC-preflight`):
+- sdio_rx_work: 86.4/s
+- sdio_tx_work: 276.1/s
+- bes2600_bh_work redispatches: 0
+
+v3 N=3 mean:
+- **sdio_rx_work: 0.0/s** (function deleted)
+- **sdio_tx_work: 0.8/s** (post-tx queue_work → self->irq_handler call; the chip-side TX driver no longer needs to wake a separate workqueue)
+- bes2600_bh_work redispatches: 0 (preserved invariant; bh thread still single long-lived work item)
+
+The 99.7% reduction in `sdio_tx_work` dispatch rate is a side-effect of v3's IRQ→bh-direct rewiring: the post-TX `queue_work(self->sdio_wq, &self->rx_work)` call I replaced with `self->irq_handler()` was actually firing more often than I'd assumed (276/s on Patch B). Folding it into the bh wake-up cuts 275/s of workqueue dispatches that weren't doing anything useful.
+
+## Risks observed
+
+- **Bug #5 RX-degradation after ~13-min uptime is independent of v3.** Same scan-failure pattern observed (`wsm_generic_confirm failed for request 0x0007` + `[SCAN] Scan failed (-22)` every 300s) on v3 as on Patch B. v3 did NOT fix Bug #5; it fixed the v2-race that was ALSO present. RX-degradation is firmware-side, likely needs a separate campaign.
+- **N=3 reps were 3 minutes each instead of 5** to fit within the fresh-chip window. Direct comparison with Patch B's 5-min baseline is approximate; chip-side throughput in 3-min vs 5-min should be similar given the bug fires on uptime, not on transferred-bytes.
+- **No regression observed in 3×3 min = 9 min of stress.** The v2 race that wedged Patch C v1 within 13 s did NOT reproduce. v3's structural fix held.
+
+## Phase 8 — lesson distilled
+
+**The cw1200 mining was decisive.** Patch C v2 (atomic_t prep + direct-deliver on top of relay, PR #10 closed) would have worked correctly but kept the structural relay that was the source of the race. v3 removed the relay entirely — restoring single-writer-from-bh invariant by construction, no atomic_t needed, and delivering a 73% throughput improvement as side benefit.
+
+Without the cw1200 history mine (`~/src/linux-rockchip`, 228 cw1200 commits over 16 years), v2's atomic_t prep would have shipped. The structural fix is upstream-grade because it matches the reference driver. v2's atomic_t wrapper would have been bes2600-specific bookkeeping with no upstream parallel — defensible as a fix, but worse to maintain.
+
+**Memory entry:** *When you have an upstream-ancestral driver still in the kernel tree, mine its bug-fix history before patching the inherited fork. The architectural answer may already be there; you just have to look.*
+
+## Receipts checklist (Phase 7 done)
+
+- [x] N=3 reps captured at fresh-chip uptime (200/391/582 s)
+- [x] Same instrumentation pre/post (workqueue ftrace + rx_packets/rx_bytes counters)
+- [x] Predicted delta matched (sdio_rx_work → 0; bh redispatches → 0; throughput ≥ 1 MB/s sustained)
+- [x] No WARN/BUG/oops during stress on any rep
+- [x] Wired-rig telemetry collection (would have caught a wedge if v3 had one)
+- [x] Receiver `nc` listener restarted fresh per rep (avoiding rep-2-style TCP race)
+- [x] Stress-ramp memory honored: not steady-state low-rate; saw 4 MB/s saturate
+
+## Out-of-scope follow-ups
+
+- Patch C2 — `ieee80211_rx_list` batch delivery — gated on Task #19 kerneldoc verification.
+- Patch D — ba_lock atomicization — independent.
+- Patch E — ps_state_lock skip when pm_unsupported — independent.
+- Bug #5 RX-degradation after 13-min uptime — separate campaign, scan-failure pattern is the entry point.
+- Task #24 — observe whether `bh.c` `asm volatile("nop")` / commented-out `__bes2600_irq_enable(1)` / BUG_ON in hot path are still load-bearing post-v3. Already partially answered: `__bes2600_irq_enable` is a stub (PR #11 comment). The other artifacts can be re-read fresh.
+
+---
+
+*Phase 7 results captured 2026-05-07 by Claude (noether). v3 (PR #5) closes Patch C campaign with structural improvement + race fix + measurable throughput win.*