- 20 May, 2021 1 commit
-
-
Takashi Sakamoto authored
Some macros are added to refactor codes related to CIP header. Signed-off-by:
Takashi Sakamoto <o-takashi@sakamocchi.jp> Link: https://lore.kernel.org/r/20210520040154.80450-2-o-takashi@sakamocchi.jpSigned-off-by:
Takashi Iwai <tiwai@suse.de>
-
- 19 May, 2021 4 commits
-
-
Takashi Iwai authored
-
Takashi Iwai authored
Currently we bail out when the device returns an error or an invalid value for the current clock selector value via uac_clock_selector_get_val(). But it's possible that the device is really uninitialized and waits for the setup of the proper route at first. For handling such a case, this patch lets the driver dealing with the error or the invalid error more gracefully, choosing the clock source automatically instead. Link: https://lore.kernel.org/r/20210518152112.8016-3-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
This patch just does refactoring of the UAC2/3 clock setup code. There should be no functional changes. The major changes are: * Provide union objects for pointing both UAC2 and UAC3 objects * Unify clock source, selector and multiplier helper functions * Unify __uac_clock_find_source() to deal with both UAC2 and UAC3 equally Link: https://lore.kernel.org/r/20210518152112.8016-2-tiwai@suse.deSigned-off-by:
-
Colin Ian King authored
The array bounds check on clock_names is currently checking the size of the entire array rather than the number of elements in the array leading to a potential array bounds read error. Fix this by using the ARRAY_SIZE macro instead of sizeof. Addresses-Coverity: ("Out-of-bounds read") Fixes: 08fdced6 ("ALSA: rawmidi: Add framing mode") Signed-off-by:Colin Ian King <colin.king@canonical.com> Link: https://lore.kernel.org/r/20210519105424.55221-1-colin.king@canonical.comSigned-off-by:
-
- 18 May, 2021 22 commits
-
-
Takashi Sakamoto authored
This commit fulfils sequence descriptors for skipped cycle when it's one cycle. This is preparation for future integration. Signed-off-by:
-
Takashi Sakamoto authored
Within devices supported by drivers in ALSA firewire stack, OXFW-based devices and Fireface devices are known to skip isochronous cycle for packet transmission. The former is due to the jumbo payload quirk. The latter is due to vendor protocol in which empty packet is not transferred in blocking mode. Although nothing to do just for handling events of the packet, packet continuity is necessarily for media clock recovery. This commit checks whether any cycle is continue or not. Signed-off-by:
-
Takashi Sakamoto authored
The size of CIP payload is now passed to helper function to parse CIP header. Signed-off-by:
-
Takashi Sakamoto authored
This commit refactors regarding to function argument for the length of isochronous packet payload. Signed-off-by:
-
Takashi Sakamoto authored
Some macros and functions are renamed so that they compute isochronous cycle within maximum count of second in isochronous context of 1394 OHCI. Signed-off-by:
-
Takashi Sakamoto authored
A internal structure is used to gather parameters relevant to sequence descriptor. Signed-off-by:
-
Takashi Sakamoto authored
Two quadlets are at least included in isochronous packet payload for Common Isochronous Packet (CIP) format in IEC 61883-1. However, it's better to equip ALSA IEC 61883-1/6 packet streaming engine for contrary packet. This commit handles isochronous cycle to process such packet so that the cycle is skipped. Signed-off-by:
-
Takashi Sakamoto authored
The number of PCM frame per period is common between PCM substreams handled in AMDTP stream in AMDTP domain. Signed-off-by:
-
Takashi Iwai authored
-
Takashi Sakamoto authored
Stanton SCS.1m and Apogee Duet FireWire use blocking transmission method unlike the other models. Signed-off-by:
-
Takashi Sakamoto authored
A new entry is added to the quirk enumeration for wrong_dbs quirk to obsolete structure member. Signed-off-by:
-
Takashi Sakamoto authored
A new macro is added to describe the maximum number of cycles to accept cycle skip by jumbo payload quirk. Signed-off-by:
-
Takashi Sakamoto authored
This commit adds enumeration to describe quirks of OXFW ASICs. Signed-off-by:
-
Takashi Sakamoto authored
ALSA OXFW supports two types of ASICS; OXFW970 and OXFW971. The former is known to have a quirk we call 'jumbo payload' that some isochronous cycles are skipped to transfer isochronous packets during handling asynchronous transaction. The quirk seems to correspond to firmware initially delivered by Oxford Semiconductor since the quirk is not confirmed for Mackie Onyx Satellite in which the revised firmware is available. The quirk is not confirmed in the latter. This commit adds code comment to describe the quirk. Signed-off-by:
-
Takashi Sakamoto authored
Loud Technologies Mackie Onyx Satellite is identified as the model with OXFW970 ASIC. This commit adds explicit entry for the model. Signed-off-by:
-
Takashi Sakamoto authored
Loud Technologies Tapco Link.FireWire 4x6 is identified as the model with OXFW970 ASIC. This commit adds explicit entry for the model. Signed-off-by:
-
Takashi Sakamoto authored
This commit changes condition statement to call mackie models detection just for the device entry. Additionally, comment is added for Onyx 1640i. Signed-off-by:
-
Takashi Sakamoto authored
All of the devices known to be based on OXFW ASICs have the same layout of configuration ROM, in which unit directory includes vendor, model, specifier_id and version immediate values. Especially, the pair of specifier_id and version is fixed to represent AV/C general protocol. This commit refactors device entries to fulfil with these 4 elements. Signed-off-by:
-
Takashi Sakamoto authored
It brings some inconvenience in practice to use enumerated type for variable to which bitwise OR with enumerator constant is assigned. This commit replaces declarations of enumerated type with int type. Suggested-by:
-
Takashi Sakamoto authored
This reverts commit 0edabdfe. I've explained that optional FireWire card for d.2 is also built-in to d.2 Pro, however it's wrong. The optional card uses DM1000 ASIC and has 'Mackie DJ Mixer' in its model name of configuration ROM. On the other hand, built-in FireWire card for d.2 Pro and d.4 Pro uses OXFW971 ASIC and has 'd.Pro' in its model name according to manuals and user experiences. The former card is not the card for d.2 Pro. They are similar in appearance but different internally. Signed-off-by:
-
Takashi Sakamoto authored
At high sampling transfer frequency, TC Electronic Konnekt Live transfers/receives 6 audio data frames in multi bit linear audio data channel of data block in CIP payload. Current hard-coded stream format is wrong. Cc: <stable@vger.kernel.org> Fixes: f1f0f330 ("ALSA: dice: add parameters of stream formats for models produced by TC Electronic") Signed-off-by:
-
Takashi Sakamoto authored
ALSA dice driver detects jumbo payload at high sampling transfer frequency for below models: * Avid M-Box 3 Pro * M-Audio Profire 610 * M-Audio Profire 2626 Although many DICE-based devices have a quirk at high sampling transfer frequency to multiplex double number of PCM frames into data block than the number in IEC 61883-1/6, the above devices are just compliant to IEC 61883-1/6. This commit disables the mode of double_pcm_frames for the models. Signed-off-by:
-
- 17 May, 2021 12 commits
-
-
Takashi Iwai authored
It's nowhere actually used. Link: https://lore.kernel.org/r/20210517131545.27252-12-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
Minor code refactoring by merging the superfluous function calls. The functions were split in the past for covering pre-history USB driver code, but this is utterly useless. Link: https://lore.kernel.org/r/20210517131545.27252-11-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
Unlike some other functions, we can't pass NULL pointer to free_pages_exact(). Add a proper NULL check for avoiding possible Oops. Link: https://lore.kernel.org/r/20210517131545.27252-10-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
Currently us428ctls_shmem pages are allocated dynamically upon the mmap call, but this is quite racy. Since the shared memory itself is mandatory for the mmap, let's allocate it at the beginning of the card initialization. Also, fix the initialization of the wait queue, too. Link: https://lore.kernel.org/r/20210517131545.27252-9-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
The PCM shmem pages are allocated in snd_usx2y_usbpcm_prepare(). Theoretically the prepare callback may be called simultaneously for both playback and capture, hence this allocation can be racy. Make sure that the allocation is performed exclusively by extending the pcm_mutex lock to cover the allocation code, too. Link: https://lore.kernel.org/r/20210517131545.27252-8-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
Theoretically the initialization functions in usx2y drivers may be called multiple times as the driver gets initialized via hwpdep ioctl. Meanwhile, those functions including memory allocations don't check whether they are called twice, and they forget the old resources, which would lead to memory leaks. This patch adds the sanity checks about the doubly initializations to give kernel WARNING, and returns an error in such a case. Also, each allocation assures to release the resources at its error path properly. Link: https://lore.kernel.org/r/20210517131545.27252-7-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
The initialization os usx2y driver is multi-staged, and the PCM and other device creations are done after the DSP is loaded and initialized. Upon the initialization, when an error happens, the driver tries to call snd_card_free(). But this is dangerous, and in general, the driver cannot kill itself during its operation. Hence better to drop the snd_card_free() call from there. Link: https://lore.kernel.org/r/20210517131545.27252-6-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
usx2y drivers may expose the allocated pages via mmap, but it performs zero-clear only for the struct size, not aligned with the page size. This leaves out some uninitialized trailing bytes. This patch fixes the clearance to cover all memory that are exposed to user-space. Link: https://lore.kernel.org/r/20210517131545.27252-5-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
This patch fixes various trivial coding-style issues in usx2y code, such as: * the assginments in if condition * comparison order with constants * NULL / zero checks * unsigned -> unsigned int * addition of braces in control blocks * debug print with function names * move local variables in block into function head * reduction of too nested indentations No functional changes. Link: https://lore.kernel.org/r/20210517131545.27252-4-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
This patch corrects merely the spaces in the usx2y code, including the superfluous trailing space in the debug prints and a slight reformat of some comment lines. Nothing really touches about the code itself. Link: https://lore.kernel.org/r/20210517131545.27252-3-tiwai@suse.deSigned-off-by:
-
Takashi Iwai authored
For improving readability, convert camelCase fields, variables and functions to the plain names with underscore. Also align the macros to be capital letters. All done via sed, no functional changes. Note that you'll still see many coding style issues even after this patch; the fixes will follow. Link: https://lore.kernel.org/r/20210517131545.27252-2-tiwai@suse.deSigned-off-by:
-
David Henningsson authored
This commit adds a new framing mode that frames all MIDI data into 32-byte frames with a timestamp. The main benefit is that we can get accurate timestamps even if userspace wakeup and processing is not immediate. Testing on a Celeron N3150 with this mode has a max jitter of 2.8 ms, compared to the in-kernel seq implementation which has a max jitter of 5 ms during idle and much worse when running scheduler stress tests in parallel. Signed-off-by:
David Henningsson <coding@diwic.se> Reviewed-by:
Jaroslav Kysela <perex@perex.cz> Link: https://lore.kernel.org/r/20210515071533.55332-1-coding@diwic.seSigned-off-by:
-
- 16 May, 2021 1 commit
-
-
Takashi Iwai authored
The interrupt handler of intel8x0 calls snd_intel8x0_update() whenever the hardware sets the corresponding status bit for each stream. This works fine for most cases as long as the hardware behaves properly. But when the hardware gives a wrong bit set, this leads to a zero- division Oops, and reportedly, this seems what happened on a VM. For fixing the crash, this patch adds a internal flag indicating that the stream is ready to be updated, and check it (as well as the flag being in suspended) to ignore such spurious update. Cc: <stable@vger.kernel.org> Reported-and-tested-by:
Sergey Senozhatsky <senozhatsky@chromium.org> Link: https://lore.kernel.org/r/s5h5yzi7uh0.wl-tiwai@suse.deSigned-off-by:
-
