ath10k-ct: fix max antenna gain unit
authorSven Eckelmann <sven@narfation.org>
Tue, 11 Jun 2019 12:53:12 +0000 (14:53 +0200)
committerChristian Lamparter <chunkeey@gmail.com>
Fri, 14 Jun 2019 22:40:25 +0000 (00:40 +0200)
Most of the txpower for the ath10k firmware is stored as twicepower (0.5 dB
steps). This isn't the case for max_antenna_gain - which is still expected
by the firmware as dB.

The firmware is converting it from dB to the internal (twicepower)
representation when it calculates the limits of a channel. This can be seen
in tpc_stats when configuring "12" as max_antenna_gain. Instead of the
expected 12 (6 dB), the tpc_stats shows 24 (12 dB).

Tested on QCA9888 and IPQ4019 with firmware 10.4-3.5.3-00057.

Signed-off-by: Sven Eckelmann <sven@narfation.org>
package/kernel/ath10k-ct/patches/204-ath10k-fix-max-antenna-gain-unit.patch [new file with mode: 0644]

diff --git a/package/kernel/ath10k-ct/patches/204-ath10k-fix-max-antenna-gain-unit.patch b/package/kernel/ath10k-ct/patches/204-ath10k-fix-max-antenna-gain-unit.patch
new file mode 100644 (file)
index 0000000..9fafaa0
--- /dev/null
@@ -0,0 +1,49 @@
+From: Sven Eckelmann <seckelmann@datto.com>
+Date: Tue, 11 Jun 2019 13:58:35 +0200
+Subject: ath10k: fix max antenna gain unit
+
+Most of the txpower for the ath10k firmware is stored as twicepower (0.5 dB
+steps). This isn't the case for max_antenna_gain - which is still expected
+by the firmware as dB.
+
+The firmware is converting it from dB to the internal (twicepower)
+representation when it calculates the limits of a channel. This can be seen
+in tpc_stats when configuring "12" as max_antenna_gain. Instead of the
+expected 12 (6 dB), the tpc_stats shows 24 (12 dB).
+
+Tested on QCA9888 and IPQ4019 with firmware 10.4-3.5.3-00057.
+
+Fixes: 02256930d9b8 ("ath10k: use proper tx power unit")
+Signed-off-by: Sven Eckelmann <seckelmann@datto.com>
+
+Forwarded: https://patchwork.kernel.org/patch/10986723/
+
+--- a/ath10k-4.19/mac.c
++++ b/ath10k-4.19/mac.c
+@@ -1185,7 +1185,7 @@ static int ath10k_monitor_vdev_start(str
+       arg.channel.min_power = 0;
+       arg.channel.max_power = channel->max_power * 2;
+       arg.channel.max_reg_power = channel->max_reg_power * 2;
+-      arg.channel.max_antenna_gain = channel->max_antenna_gain * 2;
++      arg.channel.max_antenna_gain = channel->max_antenna_gain;
+       reinit_completion(&ar->vdev_setup_done);
+@@ -1627,7 +1627,7 @@ static int ath10k_vdev_start_restart(str
+       arg.channel.min_power = 0;
+       arg.channel.max_power = chandef->chan->max_power * 2;
+       arg.channel.max_reg_power = chandef->chan->max_reg_power * 2;
+-      arg.channel.max_antenna_gain = chandef->chan->max_antenna_gain * 2;
++      arg.channel.max_antenna_gain = chandef->chan->max_antenna_gain;
+       /* CT Firmware can support 32+ VDEVS, but can only support
+        * beacon-ing devs with dev ids 0 - 31 due to firmware limitations.
+@@ -3799,7 +3799,7 @@ static int ath10k_update_channel_list(st
+                       ch->min_power = 0;
+                       ch->max_power = channel->max_power * 2;
+                       ch->max_reg_power = channel->max_reg_power * 2;
+-                      ch->max_antenna_gain = channel->max_antenna_gain * 2;
++                      ch->max_antenna_gain = channel->max_antenna_gain;
+                       ch->reg_class_id = 0; /* FIXME */
+                       /* FIXME: why use only legacy modes, why not any