- 06 Oct, 2023 10 commits
-
-
Cezary Rojewski authored
While AudioDSP drivers assign streams exclusively of HOST or LINK type, nothing blocks a user to attempt to assign a COUPLED stream. As supplied substream instance may be a stub, what is the case when code-loading, such scenario ends with null-ptr-deref. Signed-off-by:
Cezary Rojewski <cezary.rojewski@intel.com> Link: https://lore.kernel.org/r/20231006102857.749143-2-cezary.rojewski@intel.comSigned-off-by:
Takashi Iwai <tiwai@suse.de>
-
Rob Herring authored
asm/prom.h should not be included directly as it no longer contains anything drivers need. Drivers should include of.h and/or other headers which were getting implicitly included. Signed-off-by:
Rob Herring <robh@kernel.org> Link: https://lore.kernel.org/r/20231003163209.770750-1-robh@kernel.orgSigned-off-by:
-
Ivan Orlov authored
Add new control element 'PCM Slave Access Mode' which shows the access mode (interleaved/non-interleaved) for the PCM playing device. Add corresponding control change notification calls. Signed-off-by:
Ivan Orlov <ivan.orlov0322@gmail.com> Reviewed-by:
Jaroslav Kysela <perex@perex.cz> Link: https://lore.kernel.org/r/20230927113555.14877-2-ivan.orlov0322@gmail.comSigned-off-by:
-
Ivan Orlov authored
The current version of the loopback driver supports interleaved access mode only. This patch introduces support for the non-interleaved access mode. When in the interleaved mode, the 'copy_play_buf' function copies data from the playback to the capture buffer using one memcpy call. This call copies samples for multiple, interleaved channels. In the non-interleaved mode we have multiple channel buffers, so we have to perform multiple memcpy calls to copy samples channel after channel. Add new function called 'copy_play_buf_part_n', which copies a part of each channel buffer from playback to capture. Modify the 'copy_play_buf' to use the corresponding memory copy function(just memcpy / copy_play_buf_part_n) depending on the access mode. Signed-off-by:
-
Thomas Perl authored
While grep'ing for SIS7012, I noticed that there is only one reference to it in sound/pci/intel8x0m.c, while most of its code lives in sound/pci/intel8x0.c. This probably was a simple copy'n'paste mistake, as the sound/pci/intel8x0m.c driver implements support for SIS7013 (see DEVICE_SIS), and the two devices seem to share the same behavior / registers. Signed-off-by:
Thomas Perl <m@thp.io> Link: https://lore.kernel.org/r/5E4B5CA1-1109-4C82-A581-838ACF19A15D@thp.ioSigned-off-by:
-
Geoffrey D. Bennett authored
The Focusrite Clarett+ series uses the same protocol as the Scarlett Gen 2 and Gen 3 series. This patch adds support for the Clarett+ 2Pre and Clarett+ 4Pre similarly to the existing 8Pre support by adding appropriate entries to the scarlett2 driver. The Clarett 2Pre USB and 4Pre USB presumably use the same protocol as well, so support for them can easily be added if someone can test. Signed-off-by:
Geoffrey D. Bennett <g@b4.vu> Link: https://lore.kernel.org/r/ZRL7qjC3tYQllT3H@m.b4.vuSigned-off-by:
-
Cezary Rojewski authored
snd_hdac_ext_host_stream_setup() abstracts the procedure details away. Simplify the code by using it. Acked-by:
Mark Brown <broonie@kernel.org> Signed-off-by:
-
Cezary Rojewski authored
snd_hdac_ext_host_stream_setup() abstracts the procedure details away. Simplify the code by using it. Acked-by:
-
Cezary Rojewski authored
HDAudio stream setup procedure differs between revisions of the controller device. Currently the differences are handled directly within AudioDSP platform drivers with if-statements. Implement a more generic approach and expose a function that a platform driver may use to ensure the correct procedure is followed each time. Signed-off-by:
-
Cezary Rojewski authored
Software shall read SDxFIFOS calculated by the hardware and notify if invalid value is programmed before continuing the stream preparation. Signed-off-by:
-
- 30 Sep, 2023 1 commit
-
-
Dan Carpenter authored
The "ret" variable is zero but we should return -EINVAL. Fixes: 2144833e ("ALSA: hda: cirrus_scodec: Add KUnit test") Signed-off-by:
Dan Carpenter <dan.carpenter@linaro.org> Reviewed-by:
Richard Fitzgerald <rf@opensource.cirrus.com> Link: https://lore.kernel.org/r/5eea7fd5-67c8-4ed4-b5b3-b85dfb7572cc@moroto.mountainSigned-off-by:
-
- 23 Sep, 2023 2 commits
-
-
Kees Cook authored
Prepare for the coming implementation by GCC and Clang of the __counted_by attribute. Flexible array members annotated with __counted_by can have their accesses bounds-checked at run-time checking via CONFIG_UBSAN_BOUNDS (for array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family functions). As found with Coccinelle[1], add __counted_by for struct snd_usx2y_urb_seq. Additionally, since the element count member must be set before accessing the annotated flexible array member, move its initialization earlier. [1] https://github.com/kees/kernel-tools/blob/trunk/coccinelle/examples/counted_by.cocci Cc: Jaroslav Kysela <perex@perex.cz> Cc: Takashi Iwai <tiwai@suse.com> Cc: alsa-devel@alsa-project.org Signed-off-by:
Kees Cook <keescook@chromium.org> Link: https://lore.kernel.org/r/20230922175046.work.766-kees@kernel.orgSigned-off-by:
-
Kees Cook authored
Prepare for the coming implementation by GCC and Clang of the __counted_by attribute. Flexible array members annotated with __counted_by can have their accesses bounds-checked at run-time checking via CONFIG_UBSAN_BOUNDS (for array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family functions). As found with Coccinelle[1], add __counted_by for struct hda_conn_list. [1] https://github.com/kees/kernel-tools/blob/trunk/coccinelle/examples/counted_by.cocci Cc: Jaroslav Kysela <perex@perex.cz> Cc: Takashi Iwai <tiwai@suse.com> Cc: Cezary Rojewski <cezary.rojewski@intel.com> Cc: alsa-devel@alsa-project.org Signed-off-by:
-
- 22 Sep, 2023 4 commits
-
-
Stefan Binding authored
When the CS35L41 amp is requested to mute using the ACPI notification mechanism, userspace is not notified that the amp is muted. To allow userspace to know about the mute, add an ALSA control which tracks the forced mute override. This control does not track the overall mute state of the amp, since the amp is only unmuted during playback anyway, instead it tracks the mute override request from the ACPI notification. Signed-off-by:
Stefan Binding <sbinding@opensource.cirrus.com> Link: https://lore.kernel.org/r/20230921162849.1988124-5-sbinding@opensource.cirrus.comSigned-off-by:
-
Stefan Binding authored
Some laptops require a hardware based mute system, where when a hotkey is pressed, it forces the amp to be muted. For CS35L41, when the hotkey is pressed, an acpi notification is sent to the CS35L41 Device Node. The driver needs to handle this notification and call a _DSM function to retrieve the mute state. Since the amp is only muted during playback, the driver will only mute or unmute if playback is occurring, otherwise it will save the mute state for when playback starts. This uses the ACPI Notification mechanism, where a handler has been registered in the component master, which notifies each amp through the component binding. Signed-off-by:
-
Stefan Binding authored
For systems which have support for ACPI notifications, add a mechanism to register a handler for ACPI notifications and then call the acpi_notify api on the bound components. Registering a handler in the Realtek HDA driver, allows a single handler to be registered, which then calls into all the components, rather than attempting to register the same handler multiple times, once for each component. Signed-off-by:
-
Stefan Binding authored
Some systems support a notification from ACPI, which can be used for different things. Only one handler can be registered for the acpi notification, but all amps need to receive that notification, we can register a single handler inside the component master, so that it can then notify through the component framework. This is required to support mute notifications from ACPI. Signed-off-by:
-
- 20 Sep, 2023 2 commits
-
-
Nícolas F. R. A. Prado authored
When parsing the configs, keep track of card configurations that match the current system but haven't matched any card, and report those as test failures as they represent that a card which was expected to be present on the system is missing. This allows the configuration files to not only be used to detect missing PCM devices (which is currently possible) but also that the soundcard hasn't been registered at all. Signed-off-by:
Nícolas F. R. A. Prado <nfraprado@collabora.com> Reviewed-by:
-
Richard Fitzgerald authored
The KUnit test for cirrus_scodec uses GPIO library functions so select GPIOLIB in Kconfig. This fixes the ld failures on builds that didn't already select GPIOLIB. ld: vmlinux.o: in function `cirrus_scodec_test_gpio_get': sound/pci/hda/cirrus_scodec_test.c:40: undefined reference to `gpiochip_get_data' ld: vmlinux.o: in function `cirrus_scodec_test_gpio_probe': sound/pci/hda/cirrus_scodec_test.c:94: undefined reference to `gpiochip_generic_request' ld: sound/pci/hda/cirrus_scodec_test.c:94: undefined reference to `gpiochip_generic_free' ld: sound/pci/hda/cirrus_scodec_test.c:95: undefined reference to `devm_gpiochip_add_data_with_key' Signed-off-by:
kernel test robot <lkp@intel.com> Closes: https://lore.kernel.org/oe-kbuild-all/202309201646.NnjfKPWk-lkp@intel.com/ Fixes: 2144833e ("ALSA: hda: cirrus_scodec: Add KUnit test") Link: https://lore.kernel.org/r/20230920090338.29345-1-rf@opensource.cirrus.comSigned-off-by:
-
- 18 Sep, 2023 6 commits
-
-
Richard Fitzgerald authored
Add a KUnit test for cirrus_scodec_get_speaker_id(). It is impractical to have enough hardware with every possible permutation of speaker id. So use a test harness to test all theoretically supported options. The test harness consists of: - a mock GPIO controller. - a mock struct device to represent the scodec driver - software nodes to provide the fwnode info that would normally come from ACPI. Signed-off-by:
-
Richard Fitzgerald authored
Add handling of the "spk-id-gpios" _DSD property. If present, the value indicated by the GPIOs is appended to the subsystem-id part of the firmware name to load the appropriate tunings for that speaker. Some manufacturers use multiple sources of speakers, which need different tunings for best performance. On these models the type of speaker fitted is indicated by the values of one or more GPIOs. The number formed by the GPIOs identifies the tuning required. The speaker ID is only used in combination with a _SUB identifier because the value is only meaningful if the exact model is known. The code to get the speaker ID value has been implemented as a new library so that the cs35l41_hda driver can be switched in future to share common code. This library can be extended for other common functionality shared by Cirrus Logic amp drivers. Signed-off-by:
-
Geoffrey D. Bennett authored
This driver was originally developed for the Focusrite Scarlett Gen 2 series, but now also supports the Scarlett Gen 3 series, the Clarett 8Pre USB, and the Clarett+ 8Pre. The messages output by the driver on initialisation and error include the identifying text "Scarlett Gen 2/3", but this is no longer accurate, and writing "Scarlett Gen 2/3/Clarett USB/Clarett+" would be unwieldy. Add series_name field to the scarlett2_device_entry struct so that concise and accurate messages can be output. Signed-off-by:
-
Geoffrey D. Bennett authored
The Clarett 8Pre USB works the same as the Clarett+ 8Pre, only the USB ID is different. Tested-by:
Philippe Perrot <philippe@perrot-net.fr> Signed-off-by:
-
Geoffrey D. Bennett authored
By moving the USB IDs from the device_info struct into scarlett2_devices[], that will allow for devices with different USB IDs to share the same device_info. Tested-by:
-
Geoffrey D. Bennett authored
Early versions of this mixer driver did not work on all hardware, so out of caution the driver was disabled by default and had to be explicitly enabled with device_setup=1. Since commit 764fa6e6 ("ALSA: usb-audio: scarlett2: Fix device hang with ehci-pci") no more problems of this nature have been reported. Therefore, enable the driver by default but provide a new device_setup option to disable the driver in case that is needed. - device_setup value of 0 now means "enable" rather than "disable". - device_setup value of 1 is now ignored. - device_setup value of 4 now means "disable". Signed-off-by:
-
- 16 Sep, 2023 1 commit
-
-
Takashi Iwai authored
At the implementation of the optional proc fs in rawmidi, I forgot that rmidi->ops itself is optional and can be NULL. Add the proper NULL check for avoiding the Oops. Fixes: fa030f66 ("ALSA: ump: Additional proc output") Reported-and-tested-by:
Mark Hills <mark@xwax.org> Closes: https://lore.kernel.org/r/ef9118c3-a2eb-d0ff-1efa-cc5fb6416bde@xwax.org Cc: <stable@vger.kernel.org> Link: https://lore.kernel.org/r/20230916060725.11726-1-tiwai@suse.deSigned-off-by:
-
- 15 Sep, 2023 14 commits
-
-
Takashi Iwai authored
The filling of card->longname can be gracefully truncated, as it's only informative. Use scnprintf() and suppress the superfluous compile warning with -Wformat-truncation. Link: https://lore.kernel.org/r/20230915091313.5988-4-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
The filling of card->longname can be gracefully truncated, as it's only informative. Use scnprintf() and suppress the superfluous compile warning with -Wformat-truncation. Link: https://lore.kernel.org/r/20230915091313.5988-3-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
The filling of card->longname can be gracefully truncated, as it's only informative. Use scnprintf() and suppress the superfluous compile warning with -Wformat-truncation. Link: https://lore.kernel.org/r/20230915091313.5988-2-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
add_control_with_pfx() constructs a mixer name element with the fixed size, and it got compile warnings with -Wformat-truncation. Although the size overflow is very unlikely, let's have a sanity check of the string size and returns the error if it really doesn't fit instead of silent truncation. Link: https://lore.kernel.org/r/20230915082802.28684-14-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
CMIPCI driver got compile warnings with -Wformat-truncation at a couple of plain sprintf() usages. Use scnprintf() for filling the longname string for avoiding the warnings. Link: https://lore.kernel.org/r/20230915082802.28684-13-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
The compile warnings at filling MIDI stream name strings are all false-positive; the number of streams can't go so high. For suppressing the warning, replace snprintf() with scnprintf(). As stated in the above, truncation doesn't matter. Link: https://lore.kernel.org/r/20230915082802.28684-12-tiwai@suse.deReviewed-by:
Takashi Sakamoto <o-takashi@sakamocchi.jp> Tested-by:
-
Takashi Iwai authored
The filling of card->longname can be gracefully truncated, as it's only informative. Use scnprintf() and suppress the superfluous compile warning with -Wformat-truncation. Link: https://lore.kernel.org/r/20230915082802.28684-11-tiwai@suse.deReviewed-by:
-
Takashi Iwai authored
The compile warning with -Wformat-truncation at xen_snd_front_cfg_card() is false-positive; the loop can be only for SNDRV_PCM_DEVICES which is at most 32. For suppressing the warning, replace snprintf() with scnprintf(). As stated in the above, truncation doesn't matter. Link: https://lore.kernel.org/r/20230915082802.28684-10-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
The filling of card->longname can be gracefully truncated, as it's only informative. Use scnprintf() and suppress the superfluous compile warning with -Wformat-truncation. Link: https://lore.kernel.org/r/20230915082802.28684-9-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
The filling of card->longname can be gracefully truncated, as it's only informative. Use scnprintf() and suppress the superfluous compile warning with -Wformat-truncation. Link: https://lore.kernel.org/r/20230915082802.28684-8-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
The filling of card->longname can be gracefully truncated, as it's only informative. Use scnprintf() and suppress the superfluous compile warning with -Wformat-truncation. Link: https://lore.kernel.org/r/20230915082802.28684-7-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
The warning with -Wformat-truncation at sscape_upload_microcode() is false-positive; the version number can be only a single digit, hence fitting with the given string size. For suppressing the warning, replace snprintf() with scnprintf(). As stated in the above, truncation doesn't matter. Link: https://lore.kernel.org/r/20230915082802.28684-6-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
The filling of card->longname can be gracefully truncated, as it's only informative. Use scnprintf() and suppress the superfluous compile warning with -Wformat-truncation. Link: https://lore.kernel.org/r/20230915082802.28684-5-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
The recent enablement of -Wformat-truncation leads to a false-positive warning for mixer_scarlett_gen2.c. For suppressing the warning, replace snprintf() with scnprintf(). As stated in the above, truncation doesn't matter. Link: https://lore.kernel.org/r/20230915082802.28684-4-tiwai@suse.deSigned-off-by:
-
