A number of the SSID aliases linked to amp1 when they should have linked
to amp2, this would manifest as both amplifiers playing the same audio
channel.
Signed-off-by: Simon Trimmer <simont@opensource.cirrus.com>
Different distributions package the parallel command in ways that
generate differing output from --version. Adjust the grep to look
'gnu parallel' case insensitive to account for these differences.
Reported-by: Genes Lists <lists@sapience.com>
Signed-off-by: Josh Boyer <jwboyer@kernel.org>
While RTL8723CS is able to start with RTL8723BS config, it doesn't
actually work. Powering on device with "power on" in bluetoothctl
and enabling scan via "scan on" doesn't find a single device.
Unfortunately, RealtekBluetoothUartPackage_2.17.119.0525 doesn't have
a config for RTL8723CS, so we have to resort to a known-working config
from armbian firmware repository [1]. Their commit history doesn't mention
its origin.
Closer inspection of the config reveals that indeed 8723CS uses different
entry offsets, e.g. 8723BT:
55ab 2387 # Magic
3a00 # Total len
2b00 06 0000 00aa bbcc # Offset = 0x002b, len = 0x06
f400 03 0100 00 # Offset = 0x00f4, len = 0x03
f800 02 8500 # Offset = 0x00f8, len = 0x02
2700 01 67 # Offset = 0x0027, len = 0x01
0c00 10 0280 9204 50c5 ea19 e11b f1af 5f01 a40b # Offset = 0x000c, len = 0x10
df01 01 01 # Offset = 0x01df, len = 0x01
fe00 01 01 # Offset = 0x00fe, len = 0x01
e301 01 00 # Offset = 0x01e3, len = 0x01
0000 00 # Offset = 0x0000, len = 0x00
8723CS:
55ab 2387 # Magic
3900 # Total len
f400 01 01 # Offset = 0x00f4, len = 0x01
f600 02 8100 # Offset = 0x00f6, len = 0x02
fa00 02 1280 # Offset = 0x00fa, len = 0x02
0c00 10 0280 9204 50c5 ea19 e11b fdaf 5f01 a40b # Offset = 0x000c, len = 0x10
2700 01 63 # Offset = 0x0027, len = 0x1
0001 01 01 # Offset = 0x0100, len = 01
5b01 04 0b0b0b0a # Offset = 0x015b, len = 0x04
4200 02 4150 # Offset = 0x0042, len = 0x02
e000 01 42 # Offset = 0x00e0, len = 0x01
The only known devices with mainline Linux support that use 8723CS are
Pine64's PineTab, PinePhone and Pinebook, all of them work with this config.
[1] https://github.com/armbian/firmware
Signed-off-by: Vasily Khoruzhick <anarsoul@gmail.com>
Revert back to internal commit:
97ae5dc916b1b174cc87c1f1502a19d2c044e6e6
The newer firmware can result in timeouts after extended
periods of encoding.
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Revert back to internal commit:
97ae5dc916b1b174cc87c1f1502a19d2c044e6e6
The newer firmware can result in timeouts after extended
periods of encoding.
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Revert back to internal commit:
97ae5dc916b1b174cc87c1f1502a19d2c044e6e6
The newer firmware can result in timeouts after extended
periods of encoding.
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3842
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Revert back to internal commit:
97ae5dc916b1b174cc87c1f1502a19d2c044e6e6
The newer firmware can result in timeouts after extended
periods of encoding.
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3842
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Revert back to internal commit:
97ae5dc916b1b174cc87c1f1502a19d2c044e6e6
The newer firmware can result in timeouts after extended
periods of encoding.
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3842
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This reduces the install-zst target time from 100 s to 25 s on my test
system, a 400% speed improvement.
* Makefile (NUM_JOBS): New variable.
(install, install-xz, install-zst): Use it.
* copy-firmware.sh (num_jobs): New variable.
(has_gnu_parallel): New procedure.
<-j>: Parse new option, and use it along GNU parallel to parallelize
firmware copying and compression/symlink creation.
Signed-off-by: Maxim Cournoyer <maxim.cournoyer@gmail.com>
The URL returns a 404. I'm also not sure what the GNOME Build API has
to do with a Makefile designed to install Linux firmware files.
Signed-off-by: Maxim Cournoyer <maxim.cournoyer@gmail.com>
Some systems do not have /usr/bin/python3, e.g. Guix System and NixOS.
Look python3 from PATH.
Signed-off-by: Maxim Cournoyer <maxim.cournoyer@gmail.com>
In general text that comes after "License:" but before any other tags
(or section separators) should be assumed to be part of the
license. However, in some cases random information is placed there
that really doesn't have anything to do with the license. Add an
"Info:" tag before this extra bit of information to make it easy for
tools to find the license from the WHENCE.
Signed-off-by: Douglas Anderson <dianders@chromium.org>
This adds firmware rtl8125bp-2 for rtl8125bp rev.b.
Signed-off-by: ChunHao Lin <hau@realtek.com>
[ml: change execute bit]
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
QCS615 uses the same video core as sc7180, so update the firmware binary
which can be used for both the SoCs.
Delete "venus_s6.mbn" since no SoC is currently using this.
Signed-off-by: Dikshita Agarwal <quic_dikshita@quicinc.com>
Acked-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Josh Boyer <jwboyer@kernel.org>
This laptop is a SoundWire cs42l43-bridge system and needs a slightly
different filename format that uses l and r rather than 1 and 2.
Signed-off-by: Simon Trimmer <simont@opensource.cirrus.com>
This patch adds and updates the firmware files for a number of products
using the Cirrus CS35L54 and CS35L56 smart amplifiers.
Signed-off-by: Simon Trimmer <simont@opensource.cirrus.com>
This adds firmware rtl8125d-2 for rtl8125d rev.b.
Signed-off-by: ChunHao Lin <hau@realtek.com>
[ML: adjust execute bit]
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
