diff options
| author | Ching-Te Ku <ku920601@realtek.com> | 2026-07-24 21:56:36 +0800 |
|---|---|---|
| committer | Ping-Ke Shih <pkshih@realtek.com> | 2026-07-30 10:37:33 +0800 |
| commit | 068daf7086fccf4f4297609cd0ec792460b4a3f2 (patch) | |
| tree | 0364dc042ec116518948e5fdb0b726be5b174612 /tools/perf/scripts/python/stackcollapse.py | |
| parent | b23378a5645417720dc54bfc94603be91af4f4e5 (diff) | |
wifi: rtw89: coex: Add BTC report version 8 support for BT sub-reports
RTL8922A (FW >= 0.35.111) and RTL8922D (FW >= 0.35.94) set fcxbtver,
fcxbtscan and fcxbtafh to 8, but the handler in _chk_btc_report only
had branches for version 1 and 7. When version 8 arrived pfinfo was
left NULL and pcinfo->req_len was left at zero, so the length check at
validation stage rejected the report and bt->ver_info.fw was never
written, causing BT_FW:0x0 in the BTC dump.
BT-scan and BT-afh version 8 hit the goto err path for the same reason,
making all BT sub-reports silently broken on these chips.
The structural change in version 8 is that the previously reserved
second byte in each struct is now bt_id (0 = BT0, 1 = BT1), allowing
firmware to send separate reports for each Bluetooth device. All three
structs are otherwise layout-compatible with version 7.
Add rtw89_btc_fbtc_btver_v8, rtw89_btc_fbtc_btscan_v8 and
rtw89_btc_fbtc_btafh_v8 structs with the bt_id field, extend the
corresponding unions, add version 8 branches to _chk_btc_report, and
update _update_bt_report to route each report to BT0 or BT1 according
to BT ID.
Signed-off-by: Ching-Te Ku <ku920601@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20260724135640.3195044-11-pkshih@realtek.com
Diffstat (limited to 'tools/perf/scripts/python/stackcollapse.py')
0 files changed, 0 insertions, 0 deletions
