linux/sound/firewire/dice/dice-stream.c
Takashi Sakamoto 9b84f0f74d ALSA: dice: fix regression for Lexicon I-ONIX FW810S
For Lexicon I-ONIX FW810S, the call of ioctl(2) with
SNDRV_PCM_IOCTL_HW_PARAMS can returns -ETIMEDOUT. This is a regression due
to the commit 41319eb56e ("ALSA: dice: wait just for
NOTIFY_CLOCK_ACCEPTED after GLOBAL_CLOCK_SELECT operation"). The device
does not emit NOTIFY_CLOCK_ACCEPTED notification when accepting
GLOBAL_CLOCK_SELECT operation with the same parameters as current ones.

This commit fixes the regression. When receiving no notification, return
-ETIMEDOUT as long as operating for any change.

Fixes: 41319eb56e ("ALSA: dice: wait just for NOTIFY_CLOCK_ACCEPTED after GLOBAL_CLOCK_SELECT operation")
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Link: https://lore.kernel.org/r/20221130130604.29774-1-o-takashi@sakamocchi.jp
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2022-11-30 16:15:02 +01:00

709 lines
17 KiB
C

// SPDX-License-Identifier: GPL-2.0-only
/*
* dice_stream.c - a part of driver for DICE based devices
*
* Copyright (c) Clemens Ladisch <clemens@ladisch.de>
* Copyright (c) 2014 Takashi Sakamoto <o-takashi@sakamocchi.jp>
*/
#include "dice.h"
#define READY_TIMEOUT_MS 200
#define NOTIFICATION_TIMEOUT_MS 100
struct reg_params {
unsigned int count;
unsigned int size;
};
const unsigned int snd_dice_rates[SND_DICE_RATES_COUNT] = {
/* mode 0 */
[0] = 32000,
[1] = 44100,
[2] = 48000,
/* mode 1 */
[3] = 88200,
[4] = 96000,
/* mode 2 */
[5] = 176400,
[6] = 192000,
};
int snd_dice_stream_get_rate_mode(struct snd_dice *dice, unsigned int rate,
enum snd_dice_rate_mode *mode)
{
/* Corresponding to each entry in snd_dice_rates. */
static const enum snd_dice_rate_mode modes[] = {
[0] = SND_DICE_RATE_MODE_LOW,
[1] = SND_DICE_RATE_MODE_LOW,
[2] = SND_DICE_RATE_MODE_LOW,
[3] = SND_DICE_RATE_MODE_MIDDLE,
[4] = SND_DICE_RATE_MODE_MIDDLE,
[5] = SND_DICE_RATE_MODE_HIGH,
[6] = SND_DICE_RATE_MODE_HIGH,
};
int i;
for (i = 0; i < ARRAY_SIZE(snd_dice_rates); i++) {
if (!(dice->clock_caps & BIT(i)))
continue;
if (snd_dice_rates[i] != rate)
continue;
*mode = modes[i];
return 0;
}
return -EINVAL;
}
static int select_clock(struct snd_dice *dice, unsigned int rate)
{
__be32 reg, new;
u32 data;
int i;
int err;
err = snd_dice_transaction_read_global(dice, GLOBAL_CLOCK_SELECT,
&reg, sizeof(reg));
if (err < 0)
return err;
data = be32_to_cpu(reg);
data &= ~CLOCK_RATE_MASK;
for (i = 0; i < ARRAY_SIZE(snd_dice_rates); ++i) {
if (snd_dice_rates[i] == rate)
break;
}
if (i == ARRAY_SIZE(snd_dice_rates))
return -EINVAL;
data |= i << CLOCK_RATE_SHIFT;
if (completion_done(&dice->clock_accepted))
reinit_completion(&dice->clock_accepted);
new = cpu_to_be32(data);
err = snd_dice_transaction_write_global(dice, GLOBAL_CLOCK_SELECT,
&new, sizeof(new));
if (err < 0)
return err;
if (wait_for_completion_timeout(&dice->clock_accepted,
msecs_to_jiffies(NOTIFICATION_TIMEOUT_MS)) == 0) {
if (reg != new)
return -ETIMEDOUT;
}
return 0;
}
static int get_register_params(struct snd_dice *dice,
struct reg_params *tx_params,
struct reg_params *rx_params)
{
__be32 reg[2];
int err;
err = snd_dice_transaction_read_tx(dice, TX_NUMBER, reg, sizeof(reg));
if (err < 0)
return err;
tx_params->count =
min_t(unsigned int, be32_to_cpu(reg[0]), MAX_STREAMS);
tx_params->size = be32_to_cpu(reg[1]) * 4;
err = snd_dice_transaction_read_rx(dice, RX_NUMBER, reg, sizeof(reg));
if (err < 0)
return err;
rx_params->count =
min_t(unsigned int, be32_to_cpu(reg[0]), MAX_STREAMS);
rx_params->size = be32_to_cpu(reg[1]) * 4;
return 0;
}
static void release_resources(struct snd_dice *dice)
{
int i;
for (i = 0; i < MAX_STREAMS; ++i) {
fw_iso_resources_free(&dice->tx_resources[i]);
fw_iso_resources_free(&dice->rx_resources[i]);
}
}
static void stop_streams(struct snd_dice *dice, enum amdtp_stream_direction dir,
struct reg_params *params)
{
__be32 reg;
unsigned int i;
for (i = 0; i < params->count; i++) {
reg = cpu_to_be32((u32)-1);
if (dir == AMDTP_IN_STREAM) {
snd_dice_transaction_write_tx(dice,
params->size * i + TX_ISOCHRONOUS,
&reg, sizeof(reg));
} else {
snd_dice_transaction_write_rx(dice,
params->size * i + RX_ISOCHRONOUS,
&reg, sizeof(reg));
}
}
}
static int keep_resources(struct snd_dice *dice, struct amdtp_stream *stream,
struct fw_iso_resources *resources, unsigned int rate,
unsigned int pcm_chs, unsigned int midi_ports)
{
bool double_pcm_frames;
unsigned int i;
int err;
// At 176.4/192.0 kHz, Dice has a quirk to transfer two PCM frames in
// one data block of AMDTP packet. Thus sampling transfer frequency is
// a half of PCM sampling frequency, i.e. PCM frames at 192.0 kHz are
// transferred on AMDTP packets at 96 kHz. Two successive samples of a
// channel are stored consecutively in the packet. This quirk is called
// as 'Dual Wire'.
// For this quirk, blocking mode is required and PCM buffer size should
// be aligned to SYT_INTERVAL.
double_pcm_frames = (rate > 96000 && !dice->disable_double_pcm_frames);
if (double_pcm_frames) {
rate /= 2;
pcm_chs *= 2;
}
err = amdtp_am824_set_parameters(stream, rate, pcm_chs, midi_ports,
double_pcm_frames);
if (err < 0)
return err;
if (double_pcm_frames) {
pcm_chs /= 2;
for (i = 0; i < pcm_chs; i++) {
amdtp_am824_set_pcm_position(stream, i, i * 2);
amdtp_am824_set_pcm_position(stream, i + pcm_chs,
i * 2 + 1);
}
}
return fw_iso_resources_allocate(resources,
amdtp_stream_get_max_payload(stream),
fw_parent_device(dice->unit)->max_speed);
}
static int keep_dual_resources(struct snd_dice *dice, unsigned int rate,
enum amdtp_stream_direction dir,
struct reg_params *params)
{
enum snd_dice_rate_mode mode;
int i;
int err;
err = snd_dice_stream_get_rate_mode(dice, rate, &mode);
if (err < 0)
return err;
for (i = 0; i < params->count; ++i) {
__be32 reg[2];
struct amdtp_stream *stream;
struct fw_iso_resources *resources;
unsigned int pcm_cache;
unsigned int pcm_chs;
unsigned int midi_ports;
if (dir == AMDTP_IN_STREAM) {
stream = &dice->tx_stream[i];
resources = &dice->tx_resources[i];
pcm_cache = dice->tx_pcm_chs[i][mode];
err = snd_dice_transaction_read_tx(dice,
params->size * i + TX_NUMBER_AUDIO,
reg, sizeof(reg));
} else {
stream = &dice->rx_stream[i];
resources = &dice->rx_resources[i];
pcm_cache = dice->rx_pcm_chs[i][mode];
err = snd_dice_transaction_read_rx(dice,
params->size * i + RX_NUMBER_AUDIO,
reg, sizeof(reg));
}
if (err < 0)
return err;
pcm_chs = be32_to_cpu(reg[0]);
midi_ports = be32_to_cpu(reg[1]);
// These are important for developer of this driver.
if (pcm_chs != pcm_cache) {
dev_info(&dice->unit->device,
"cache mismatch: pcm: %u:%u, midi: %u\n",
pcm_chs, pcm_cache, midi_ports);
return -EPROTO;
}
err = keep_resources(dice, stream, resources, rate, pcm_chs,
midi_ports);
if (err < 0)
return err;
}
return 0;
}
static void finish_session(struct snd_dice *dice, struct reg_params *tx_params,
struct reg_params *rx_params)
{
stop_streams(dice, AMDTP_IN_STREAM, tx_params);
stop_streams(dice, AMDTP_OUT_STREAM, rx_params);
snd_dice_transaction_clear_enable(dice);
}
int snd_dice_stream_reserve_duplex(struct snd_dice *dice, unsigned int rate,
unsigned int events_per_period,
unsigned int events_per_buffer)
{
unsigned int curr_rate;
int err;
// Check sampling transmission frequency.
err = snd_dice_transaction_get_rate(dice, &curr_rate);
if (err < 0)
return err;
if (rate == 0)
rate = curr_rate;
if (dice->substreams_counter == 0 || curr_rate != rate) {
struct reg_params tx_params, rx_params;
amdtp_domain_stop(&dice->domain);
err = get_register_params(dice, &tx_params, &rx_params);
if (err < 0)
return err;
finish_session(dice, &tx_params, &rx_params);
release_resources(dice);
// Just after owning the unit (GLOBAL_OWNER), the unit can
// return invalid stream formats. Selecting clock parameters
// have an effect for the unit to refine it.
err = select_clock(dice, rate);
if (err < 0)
return err;
// After changing sampling transfer frequency, the value of
// register can be changed.
err = get_register_params(dice, &tx_params, &rx_params);
if (err < 0)
return err;
err = keep_dual_resources(dice, rate, AMDTP_IN_STREAM,
&tx_params);
if (err < 0)
goto error;
err = keep_dual_resources(dice, rate, AMDTP_OUT_STREAM,
&rx_params);
if (err < 0)
goto error;
err = amdtp_domain_set_events_per_period(&dice->domain,
events_per_period, events_per_buffer);
if (err < 0)
goto error;
}
return 0;
error:
release_resources(dice);
return err;
}
static int start_streams(struct snd_dice *dice, enum amdtp_stream_direction dir,
unsigned int rate, struct reg_params *params)
{
unsigned int max_speed = fw_parent_device(dice->unit)->max_speed;
int i;
int err;
for (i = 0; i < params->count; i++) {
struct amdtp_stream *stream;
struct fw_iso_resources *resources;
__be32 reg;
if (dir == AMDTP_IN_STREAM) {
stream = dice->tx_stream + i;
resources = dice->tx_resources + i;
} else {
stream = dice->rx_stream + i;
resources = dice->rx_resources + i;
}
reg = cpu_to_be32(resources->channel);
if (dir == AMDTP_IN_STREAM) {
err = snd_dice_transaction_write_tx(dice,
params->size * i + TX_ISOCHRONOUS,
&reg, sizeof(reg));
} else {
err = snd_dice_transaction_write_rx(dice,
params->size * i + RX_ISOCHRONOUS,
&reg, sizeof(reg));
}
if (err < 0)
return err;
if (dir == AMDTP_IN_STREAM) {
reg = cpu_to_be32(max_speed);
err = snd_dice_transaction_write_tx(dice,
params->size * i + TX_SPEED,
&reg, sizeof(reg));
if (err < 0)
return err;
}
err = amdtp_domain_add_stream(&dice->domain, stream,
resources->channel, max_speed);
if (err < 0)
return err;
}
return 0;
}
/*
* MEMO: After this function, there're two states of streams:
* - None streams are running.
* - All streams are running.
*/
int snd_dice_stream_start_duplex(struct snd_dice *dice)
{
unsigned int generation = dice->rx_resources[0].generation;
struct reg_params tx_params, rx_params;
unsigned int i;
unsigned int rate;
enum snd_dice_rate_mode mode;
int err;
if (dice->substreams_counter == 0)
return -EIO;
err = get_register_params(dice, &tx_params, &rx_params);
if (err < 0)
return err;
// Check error of packet streaming.
for (i = 0; i < MAX_STREAMS; ++i) {
if (amdtp_streaming_error(&dice->tx_stream[i]) ||
amdtp_streaming_error(&dice->rx_stream[i])) {
amdtp_domain_stop(&dice->domain);
finish_session(dice, &tx_params, &rx_params);
break;
}
}
if (generation != fw_parent_device(dice->unit)->card->generation) {
for (i = 0; i < MAX_STREAMS; ++i) {
if (i < tx_params.count)
fw_iso_resources_update(dice->tx_resources + i);
if (i < rx_params.count)
fw_iso_resources_update(dice->rx_resources + i);
}
}
// Check required streams are running or not.
err = snd_dice_transaction_get_rate(dice, &rate);
if (err < 0)
return err;
err = snd_dice_stream_get_rate_mode(dice, rate, &mode);
if (err < 0)
return err;
for (i = 0; i < MAX_STREAMS; ++i) {
if (dice->tx_pcm_chs[i][mode] > 0 &&
!amdtp_stream_running(&dice->tx_stream[i]))
break;
if (dice->rx_pcm_chs[i][mode] > 0 &&
!amdtp_stream_running(&dice->rx_stream[i]))
break;
}
if (i < MAX_STREAMS) {
// Start both streams.
err = start_streams(dice, AMDTP_IN_STREAM, rate, &tx_params);
if (err < 0)
goto error;
err = start_streams(dice, AMDTP_OUT_STREAM, rate, &rx_params);
if (err < 0)
goto error;
err = snd_dice_transaction_set_enable(dice);
if (err < 0) {
dev_err(&dice->unit->device,
"fail to enable interface\n");
goto error;
}
// MEMO: The device immediately starts packet transmission when enabled. Some
// devices are strictly to generate any discontinuity in the sequence of tx packet
// when they receives invalid sequence of presentation time in CIP header. The
// sequence replay for media clock recovery can suppress the behaviour.
err = amdtp_domain_start(&dice->domain, 0, true, false);
if (err < 0)
goto error;
if (!amdtp_domain_wait_ready(&dice->domain, READY_TIMEOUT_MS)) {
err = -ETIMEDOUT;
goto error;
}
}
return 0;
error:
amdtp_domain_stop(&dice->domain);
finish_session(dice, &tx_params, &rx_params);
return err;
}
/*
* MEMO: After this function, there're two states of streams:
* - None streams are running.
* - All streams are running.
*/
void snd_dice_stream_stop_duplex(struct snd_dice *dice)
{
struct reg_params tx_params, rx_params;
if (dice->substreams_counter == 0) {
if (get_register_params(dice, &tx_params, &rx_params) >= 0)
finish_session(dice, &tx_params, &rx_params);
amdtp_domain_stop(&dice->domain);
release_resources(dice);
}
}
static int init_stream(struct snd_dice *dice, enum amdtp_stream_direction dir,
unsigned int index)
{
struct amdtp_stream *stream;
struct fw_iso_resources *resources;
int err;
if (dir == AMDTP_IN_STREAM) {
stream = &dice->tx_stream[index];
resources = &dice->tx_resources[index];
} else {
stream = &dice->rx_stream[index];
resources = &dice->rx_resources[index];
}
err = fw_iso_resources_init(resources, dice->unit);
if (err < 0)
goto end;
resources->channels_mask = 0x00000000ffffffffuLL;
err = amdtp_am824_init(stream, dice->unit, dir, CIP_BLOCKING);
if (err < 0) {
amdtp_stream_destroy(stream);
fw_iso_resources_destroy(resources);
}
end:
return err;
}
/*
* This function should be called before starting streams or after stopping
* streams.
*/
static void destroy_stream(struct snd_dice *dice,
enum amdtp_stream_direction dir,
unsigned int index)
{
struct amdtp_stream *stream;
struct fw_iso_resources *resources;
if (dir == AMDTP_IN_STREAM) {
stream = &dice->tx_stream[index];
resources = &dice->tx_resources[index];
} else {
stream = &dice->rx_stream[index];
resources = &dice->rx_resources[index];
}
amdtp_stream_destroy(stream);
fw_iso_resources_destroy(resources);
}
int snd_dice_stream_init_duplex(struct snd_dice *dice)
{
int i, err;
for (i = 0; i < MAX_STREAMS; i++) {
err = init_stream(dice, AMDTP_IN_STREAM, i);
if (err < 0) {
for (; i >= 0; i--)
destroy_stream(dice, AMDTP_IN_STREAM, i);
goto end;
}
}
for (i = 0; i < MAX_STREAMS; i++) {
err = init_stream(dice, AMDTP_OUT_STREAM, i);
if (err < 0) {
for (; i >= 0; i--)
destroy_stream(dice, AMDTP_OUT_STREAM, i);
for (i = 0; i < MAX_STREAMS; i++)
destroy_stream(dice, AMDTP_IN_STREAM, i);
goto end;
}
}
err = amdtp_domain_init(&dice->domain);
if (err < 0) {
for (i = 0; i < MAX_STREAMS; ++i) {
destroy_stream(dice, AMDTP_OUT_STREAM, i);
destroy_stream(dice, AMDTP_IN_STREAM, i);
}
}
end:
return err;
}
void snd_dice_stream_destroy_duplex(struct snd_dice *dice)
{
unsigned int i;
for (i = 0; i < MAX_STREAMS; i++) {
destroy_stream(dice, AMDTP_IN_STREAM, i);
destroy_stream(dice, AMDTP_OUT_STREAM, i);
}
amdtp_domain_destroy(&dice->domain);
}
void snd_dice_stream_update_duplex(struct snd_dice *dice)
{
struct reg_params tx_params, rx_params;
/*
* On a bus reset, the DICE firmware disables streaming and then goes
* off contemplating its own navel for hundreds of milliseconds before
* it can react to any of our attempts to reenable streaming. This
* means that we lose synchronization anyway, so we force our streams
* to stop so that the application can restart them in an orderly
* manner.
*/
dice->global_enabled = false;
if (get_register_params(dice, &tx_params, &rx_params) == 0) {
amdtp_domain_stop(&dice->domain);
stop_streams(dice, AMDTP_IN_STREAM, &tx_params);
stop_streams(dice, AMDTP_OUT_STREAM, &rx_params);
}
}
int snd_dice_stream_detect_current_formats(struct snd_dice *dice)
{
unsigned int rate;
enum snd_dice_rate_mode mode;
__be32 reg[2];
struct reg_params tx_params, rx_params;
int i;
int err;
/* If extended protocol is available, detect detail spec. */
err = snd_dice_detect_extension_formats(dice);
if (err >= 0)
return err;
/*
* Available stream format is restricted at current mode of sampling
* clock.
*/
err = snd_dice_transaction_get_rate(dice, &rate);
if (err < 0)
return err;
err = snd_dice_stream_get_rate_mode(dice, rate, &mode);
if (err < 0)
return err;
/*
* Just after owning the unit (GLOBAL_OWNER), the unit can return
* invalid stream formats. Selecting clock parameters have an effect
* for the unit to refine it.
*/
err = select_clock(dice, rate);
if (err < 0)
return err;
err = get_register_params(dice, &tx_params, &rx_params);
if (err < 0)
return err;
for (i = 0; i < tx_params.count; ++i) {
err = snd_dice_transaction_read_tx(dice,
tx_params.size * i + TX_NUMBER_AUDIO,
reg, sizeof(reg));
if (err < 0)
return err;
dice->tx_pcm_chs[i][mode] = be32_to_cpu(reg[0]);
dice->tx_midi_ports[i] = max_t(unsigned int,
be32_to_cpu(reg[1]), dice->tx_midi_ports[i]);
}
for (i = 0; i < rx_params.count; ++i) {
err = snd_dice_transaction_read_rx(dice,
rx_params.size * i + RX_NUMBER_AUDIO,
reg, sizeof(reg));
if (err < 0)
return err;
dice->rx_pcm_chs[i][mode] = be32_to_cpu(reg[0]);
dice->rx_midi_ports[i] = max_t(unsigned int,
be32_to_cpu(reg[1]), dice->rx_midi_ports[i]);
}
return 0;
}
static void dice_lock_changed(struct snd_dice *dice)
{
dice->dev_lock_changed = true;
wake_up(&dice->hwdep_wait);
}
int snd_dice_stream_lock_try(struct snd_dice *dice)
{
int err;
spin_lock_irq(&dice->lock);
if (dice->dev_lock_count < 0) {
err = -EBUSY;
goto out;
}
if (dice->dev_lock_count++ == 0)
dice_lock_changed(dice);
err = 0;
out:
spin_unlock_irq(&dice->lock);
return err;
}
void snd_dice_stream_lock_release(struct snd_dice *dice)
{
spin_lock_irq(&dice->lock);
if (WARN_ON(dice->dev_lock_count <= 0))
goto out;
if (--dice->dev_lock_count == 0)
dice_lock_changed(dice);
out:
spin_unlock_irq(&dice->lock);
}