OpenCloudOS-Kernel/drivers/soundwire/intel.c

2042 lines
51 KiB
C
Raw Normal View History

// SPDX-License-Identifier: (GPL-2.0 OR BSD-3-Clause)
// Copyright(c) 2015-17 Intel Corporation.
/*
* Soundwire Intel Master Driver
*/
#include <linux/acpi.h>
#include <linux/debugfs.h>
#include <linux/delay.h>
#include <linux/module.h>
#include <linux/interrupt.h>
#include <linux/io.h>
#include <linux/auxiliary_bus.h>
#include <sound/pcm_params.h>
#include <linux/pm_runtime.h>
#include <sound/soc.h>
#include <linux/soundwire/sdw_registers.h>
#include <linux/soundwire/sdw.h>
#include <linux/soundwire/sdw_intel.h>
#include "cadence_master.h"
#include "bus.h"
#include "intel.h"
/* IDA min selected to avoid conflicts with HDaudio/iDISP SDI values */
#define INTEL_DEV_NUM_IDA_MIN 4
#define INTEL_MASTER_SUSPEND_DELAY_MS 3000
soundwire: cadence: add paranoid check on self-clearing bits The Cadence IP exposes a small number of self-clearing bits in the MCP_CONTROL and MCP_CONFIG_UPDATE registers. We currently do not check that those bits are indeed cleared, e.g. during resume operations. That could lead to resuming peripheral devices too early. In addition, if we happen to read these registers, update one of the fields and write the register back, we may be writing stale data that might have been cleared in hardware. These sort of race conditions could lead to e.g. doing a hw_reset twice or stopping a clock that just restarted. There is no clear way of avoiding these potential race conditions other than making sure that these registers fields are cleared before any read-modify-write sequence. If we detect this sort of errors, we only log them since there is no clear recovery possible. The only way out is likely to restart the IP with a suspend/resume cycle. Note that the checks are performed before updating the registers, as well as after the Intel 'sync go' sequence in multi-link mode. That should cover both the start and end of suspend/resume hardware configurations. The Multi-Master mode gates the configuration updates until the 'sync go' signal is asserted, so we only check on init and after the end of the 'sync go' sequence. The duration of the usleep_range() was defined by the GSYNC frequency used in multi-master mode. With a 4kHz frequency, any configuration change might be deferred by up to 250us. Extending the range to 1000-1500us should guarantee that the configuration change is completed without any significant impact on the overall resume time. Suggested-by: Bard Liao <yung-chuan.liao@linux.intel.com> Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Link: https://lore.kernel.org/r/20210714051349.13064-1-yung-chuan.liao@linux.intel.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2021-07-14 13:13:49 +08:00
#define INTEL_MASTER_RESET_ITERATIONS 10
/*
* debug/config flags for the Intel SoundWire Master.
*
* Since we may have multiple masters active, we can have up to 8
* flags reused in each byte, with master0 using the ls-byte, etc.
*/
#define SDW_INTEL_MASTER_DISABLE_PM_RUNTIME BIT(0)
#define SDW_INTEL_MASTER_DISABLE_CLOCK_STOP BIT(1)
#define SDW_INTEL_MASTER_DISABLE_PM_RUNTIME_IDLE BIT(2)
#define SDW_INTEL_MASTER_DISABLE_MULTI_LINK BIT(3)
static int md_flags;
module_param_named(sdw_md_flags, md_flags, int, 0444);
MODULE_PARM_DESC(sdw_md_flags, "SoundWire Intel Master device flags (0x0 all off)");
enum intel_pdi_type {
INTEL_PDI_IN = 0,
INTEL_PDI_OUT = 1,
INTEL_PDI_BD = 2,
};
#define cdns_to_intel(_cdns) container_of(_cdns, struct sdw_intel, cdns)
/*
* Read, write helpers for HW registers
*/
static inline int intel_readl(void __iomem *base, int offset)
{
return readl(base + offset);
}
static inline void intel_writel(void __iomem *base, int offset, int value)
{
writel(value, base + offset);
}
static inline u16 intel_readw(void __iomem *base, int offset)
{
return readw(base + offset);
}
static inline void intel_writew(void __iomem *base, int offset, u16 value)
{
writew(value, base + offset);
}
static int intel_wait_bit(void __iomem *base, int offset, u32 mask, u32 target)
{
int timeout = 10;
u32 reg_read;
do {
reg_read = readl(base + offset);
if ((reg_read & mask) == target)
return 0;
timeout--;
usleep_range(50, 100);
} while (timeout != 0);
return -EAGAIN;
}
static int intel_clear_bit(void __iomem *base, int offset, u32 value, u32 mask)
{
writel(value, base + offset);
return intel_wait_bit(base, offset, mask, 0);
}
static int intel_set_bit(void __iomem *base, int offset, u32 value, u32 mask)
{
writel(value, base + offset);
return intel_wait_bit(base, offset, mask, mask);
}
/*
* debugfs
*/
#ifdef CONFIG_DEBUG_FS
#define RD_BUF (2 * PAGE_SIZE)
static ssize_t intel_sprintf(void __iomem *mem, bool l,
char *buf, size_t pos, unsigned int reg)
{
int value;
if (l)
value = intel_readl(mem, reg);
else
value = intel_readw(mem, reg);
return scnprintf(buf + pos, RD_BUF - pos, "%4x\t%4x\n", reg, value);
}
static int intel_reg_show(struct seq_file *s_file, void *data)
{
struct sdw_intel *sdw = s_file->private;
void __iomem *s = sdw->link_res->shim;
void __iomem *a = sdw->link_res->alh;
char *buf;
ssize_t ret;
int i, j;
unsigned int links, reg;
buf = kzalloc(RD_BUF, GFP_KERNEL);
if (!buf)
return -ENOMEM;
links = intel_readl(s, SDW_SHIM_LCAP) & SDW_SHIM_LCAP_LCOUNT_MASK;
ret = scnprintf(buf, RD_BUF, "Register Value\n");
ret += scnprintf(buf + ret, RD_BUF - ret, "\nShim\n");
for (i = 0; i < links; i++) {
reg = SDW_SHIM_LCAP + i * 4;
ret += intel_sprintf(s, true, buf, ret, reg);
}
for (i = 0; i < links; i++) {
ret += scnprintf(buf + ret, RD_BUF - ret, "\nLink%d\n", i);
ret += intel_sprintf(s, false, buf, ret, SDW_SHIM_CTLSCAP(i));
ret += intel_sprintf(s, false, buf, ret, SDW_SHIM_CTLS0CM(i));
ret += intel_sprintf(s, false, buf, ret, SDW_SHIM_CTLS1CM(i));
ret += intel_sprintf(s, false, buf, ret, SDW_SHIM_CTLS2CM(i));
ret += intel_sprintf(s, false, buf, ret, SDW_SHIM_CTLS3CM(i));
ret += intel_sprintf(s, false, buf, ret, SDW_SHIM_PCMSCAP(i));
ret += scnprintf(buf + ret, RD_BUF - ret, "\n PCMSyCH registers\n");
/*
* the value 10 is the number of PDIs. We will need a
* cleanup to remove hard-coded Intel configurations
* from cadence_master.c
*/
for (j = 0; j < 10; j++) {
ret += intel_sprintf(s, false, buf, ret,
SDW_SHIM_PCMSYCHM(i, j));
ret += intel_sprintf(s, false, buf, ret,
SDW_SHIM_PCMSYCHC(i, j));
}
ret += scnprintf(buf + ret, RD_BUF - ret, "\n IOCTL, CTMCTL\n");
ret += intel_sprintf(s, false, buf, ret, SDW_SHIM_IOCTL(i));
ret += intel_sprintf(s, false, buf, ret, SDW_SHIM_CTMCTL(i));
}
ret += scnprintf(buf + ret, RD_BUF - ret, "\nWake registers\n");
ret += intel_sprintf(s, false, buf, ret, SDW_SHIM_WAKEEN);
ret += intel_sprintf(s, false, buf, ret, SDW_SHIM_WAKESTS);
ret += scnprintf(buf + ret, RD_BUF - ret, "\nALH STRMzCFG\n");
for (i = 0; i < SDW_ALH_NUM_STREAMS; i++)
ret += intel_sprintf(a, true, buf, ret, SDW_ALH_STRMZCFG(i));
seq_printf(s_file, "%s", buf);
kfree(buf);
return 0;
}
DEFINE_SHOW_ATTRIBUTE(intel_reg);
static int intel_set_m_datamode(void *data, u64 value)
{
struct sdw_intel *sdw = data;
struct sdw_bus *bus = &sdw->cdns.bus;
if (value > SDW_PORT_DATA_MODE_STATIC_1)
return -EINVAL;
/* Userspace changed the hardware state behind the kernel's back */
add_taint(TAINT_USER, LOCKDEP_STILL_OK);
bus->params.m_data_mode = value;
return 0;
}
DEFINE_DEBUGFS_ATTRIBUTE(intel_set_m_datamode_fops, NULL,
intel_set_m_datamode, "%llu\n");
static int intel_set_s_datamode(void *data, u64 value)
{
struct sdw_intel *sdw = data;
struct sdw_bus *bus = &sdw->cdns.bus;
if (value > SDW_PORT_DATA_MODE_STATIC_1)
return -EINVAL;
/* Userspace changed the hardware state behind the kernel's back */
add_taint(TAINT_USER, LOCKDEP_STILL_OK);
bus->params.s_data_mode = value;
return 0;
}
DEFINE_DEBUGFS_ATTRIBUTE(intel_set_s_datamode_fops, NULL,
intel_set_s_datamode, "%llu\n");
static void intel_debugfs_init(struct sdw_intel *sdw)
{
struct dentry *root = sdw->cdns.bus.debugfs;
if (!root)
return;
sdw->debugfs = debugfs_create_dir("intel-sdw", root);
debugfs_create_file("intel-registers", 0400, sdw->debugfs, sdw,
&intel_reg_fops);
debugfs_create_file("intel-m-datamode", 0200, sdw->debugfs, sdw,
&intel_set_m_datamode_fops);
debugfs_create_file("intel-s-datamode", 0200, sdw->debugfs, sdw,
&intel_set_s_datamode_fops);
sdw_cdns_debugfs_init(&sdw->cdns, sdw->debugfs);
}
static void intel_debugfs_exit(struct sdw_intel *sdw)
{
debugfs_remove_recursive(sdw->debugfs);
}
#else
static void intel_debugfs_init(struct sdw_intel *sdw) {}
static void intel_debugfs_exit(struct sdw_intel *sdw) {}
#endif /* CONFIG_DEBUG_FS */
/*
* shim ops
*/
/* this needs to be called with shim_lock */
static void intel_shim_glue_to_master_ip(struct sdw_intel *sdw)
{
void __iomem *shim = sdw->link_res->shim;
unsigned int link_id = sdw->instance;
u16 ioctl;
/* Switch to MIP from Glue logic */
ioctl = intel_readw(shim, SDW_SHIM_IOCTL(link_id));
ioctl &= ~(SDW_SHIM_IOCTL_DOE);
intel_writew(shim, SDW_SHIM_IOCTL(link_id), ioctl);
usleep_range(10, 15);
ioctl &= ~(SDW_SHIM_IOCTL_DO);
intel_writew(shim, SDW_SHIM_IOCTL(link_id), ioctl);
usleep_range(10, 15);
ioctl |= (SDW_SHIM_IOCTL_MIF);
intel_writew(shim, SDW_SHIM_IOCTL(link_id), ioctl);
usleep_range(10, 15);
ioctl &= ~(SDW_SHIM_IOCTL_BKE);
ioctl &= ~(SDW_SHIM_IOCTL_COE);
intel_writew(shim, SDW_SHIM_IOCTL(link_id), ioctl);
usleep_range(10, 15);
/* at this point Master IP has full control of the I/Os */
}
/* this needs to be called with shim_lock */
static void intel_shim_master_ip_to_glue(struct sdw_intel *sdw)
{
unsigned int link_id = sdw->instance;
void __iomem *shim = sdw->link_res->shim;
u16 ioctl;
/* Glue logic */
ioctl = intel_readw(shim, SDW_SHIM_IOCTL(link_id));
ioctl |= SDW_SHIM_IOCTL_BKE;
ioctl |= SDW_SHIM_IOCTL_COE;
intel_writew(shim, SDW_SHIM_IOCTL(link_id), ioctl);
usleep_range(10, 15);
ioctl &= ~(SDW_SHIM_IOCTL_MIF);
intel_writew(shim, SDW_SHIM_IOCTL(link_id), ioctl);
usleep_range(10, 15);
/* at this point Integration Glue has full control of the I/Os */
}
/* this needs to be called with shim_lock */
static void intel_shim_init(struct sdw_intel *sdw)
{
void __iomem *shim = sdw->link_res->shim;
unsigned int link_id = sdw->instance;
u16 ioctl = 0, act = 0;
/* Initialize Shim */
ioctl |= SDW_SHIM_IOCTL_BKE;
intel_writew(shim, SDW_SHIM_IOCTL(link_id), ioctl);
usleep_range(10, 15);
ioctl |= SDW_SHIM_IOCTL_WPDD;
intel_writew(shim, SDW_SHIM_IOCTL(link_id), ioctl);
usleep_range(10, 15);
ioctl |= SDW_SHIM_IOCTL_DO;
intel_writew(shim, SDW_SHIM_IOCTL(link_id), ioctl);
usleep_range(10, 15);
ioctl |= SDW_SHIM_IOCTL_DOE;
intel_writew(shim, SDW_SHIM_IOCTL(link_id), ioctl);
usleep_range(10, 15);
intel_shim_glue_to_master_ip(sdw);
u16p_replace_bits(&act, 0x1, SDW_SHIM_CTMCTL_DOAIS);
act |= SDW_SHIM_CTMCTL_DACTQE;
act |= SDW_SHIM_CTMCTL_DODS;
intel_writew(shim, SDW_SHIM_CTMCTL(link_id), act);
usleep_range(10, 15);
}
static int intel_shim_check_wake(struct sdw_intel *sdw)
{
void __iomem *shim;
u16 wake_sts;
shim = sdw->link_res->shim;
wake_sts = intel_readw(shim, SDW_SHIM_WAKESTS);
return wake_sts & BIT(sdw->instance);
}
static void intel_shim_wake(struct sdw_intel *sdw, bool wake_enable)
{
void __iomem *shim = sdw->link_res->shim;
unsigned int link_id = sdw->instance;
u16 wake_en, wake_sts;
mutex_lock(sdw->link_res->shim_lock);
wake_en = intel_readw(shim, SDW_SHIM_WAKEEN);
if (wake_enable) {
/* Enable the wakeup */
wake_en |= (SDW_SHIM_WAKEEN_ENABLE << link_id);
intel_writew(shim, SDW_SHIM_WAKEEN, wake_en);
} else {
/* Disable the wake up interrupt */
wake_en &= ~(SDW_SHIM_WAKEEN_ENABLE << link_id);
intel_writew(shim, SDW_SHIM_WAKEEN, wake_en);
/* Clear wake status */
wake_sts = intel_readw(shim, SDW_SHIM_WAKESTS);
wake_sts |= (SDW_SHIM_WAKESTS_STATUS << link_id);
intel_writew(shim, SDW_SHIM_WAKESTS, wake_sts);
}
mutex_unlock(sdw->link_res->shim_lock);
}
static int intel_link_power_up(struct sdw_intel *sdw)
{
unsigned int link_id = sdw->instance;
void __iomem *shim = sdw->link_res->shim;
u32 *shim_mask = sdw->link_res->shim_mask;
struct sdw_bus *bus = &sdw->cdns.bus;
struct sdw_master_prop *prop = &bus->prop;
u32 spa_mask, cpa_mask;
u32 link_control;
int ret = 0;
u32 syncprd;
u32 sync_reg;
mutex_lock(sdw->link_res->shim_lock);
/*
* The hardware relies on an internal counter, typically 4kHz,
* to generate the SoundWire SSP - which defines a 'safe'
* synchronization point between commands and audio transport
* and allows for multi link synchronization. The SYNCPRD value
* is only dependent on the oscillator clock provided to
* the IP, so adjust based on _DSD properties reported in DSDT
* tables. The values reported are based on either 24MHz
* (CNL/CML) or 38.4 MHz (ICL/TGL+).
*/
if (prop->mclk_freq % 6000000)
syncprd = SDW_SHIM_SYNC_SYNCPRD_VAL_38_4;
else
syncprd = SDW_SHIM_SYNC_SYNCPRD_VAL_24;
if (!*shim_mask) {
dev_dbg(sdw->cdns.dev, "powering up all links\n");
/* we first need to program the SyncPRD/CPU registers */
dev_dbg(sdw->cdns.dev,
"first link up, programming SYNCPRD\n");
/* set SyncPRD period */
sync_reg = intel_readl(shim, SDW_SHIM_SYNC);
u32p_replace_bits(&sync_reg, syncprd, SDW_SHIM_SYNC_SYNCPRD);
/* Set SyncCPU bit */
sync_reg |= SDW_SHIM_SYNC_SYNCCPU;
intel_writel(shim, SDW_SHIM_SYNC, sync_reg);
/* Link power up sequence */
link_control = intel_readl(shim, SDW_SHIM_LCTL);
/* only power-up enabled links */
spa_mask = FIELD_PREP(SDW_SHIM_LCTL_SPA_MASK, sdw->link_res->link_mask);
cpa_mask = FIELD_PREP(SDW_SHIM_LCTL_CPA_MASK, sdw->link_res->link_mask);
link_control |= spa_mask;
ret = intel_set_bit(shim, SDW_SHIM_LCTL, link_control, cpa_mask);
if (ret < 0) {
dev_err(sdw->cdns.dev, "Failed to power up link: %d\n", ret);
goto out;
}
/* SyncCPU will change once link is active */
ret = intel_wait_bit(shim, SDW_SHIM_SYNC,
SDW_SHIM_SYNC_SYNCCPU, 0);
if (ret < 0) {
dev_err(sdw->cdns.dev,
"Failed to set SHIM_SYNC: %d\n", ret);
goto out;
}
}
*shim_mask |= BIT(link_id);
sdw->cdns.link_up = true;
intel_shim_init(sdw);
out:
mutex_unlock(sdw->link_res->shim_lock);
return ret;
}
static int intel_link_power_down(struct sdw_intel *sdw)
{
u32 link_control, spa_mask, cpa_mask;
unsigned int link_id = sdw->instance;
void __iomem *shim = sdw->link_res->shim;
u32 *shim_mask = sdw->link_res->shim_mask;
int ret = 0;
mutex_lock(sdw->link_res->shim_lock);
if (!(*shim_mask & BIT(link_id)))
dev_err(sdw->cdns.dev,
"%s: Unbalanced power-up/down calls\n", __func__);
sdw->cdns.link_up = false;
intel_shim_master_ip_to_glue(sdw);
*shim_mask &= ~BIT(link_id);
if (!*shim_mask) {
dev_dbg(sdw->cdns.dev, "powering down all links\n");
/* Link power down sequence */
link_control = intel_readl(shim, SDW_SHIM_LCTL);
/* only power-down enabled links */
spa_mask = FIELD_PREP(SDW_SHIM_LCTL_SPA_MASK, ~sdw->link_res->link_mask);
cpa_mask = FIELD_PREP(SDW_SHIM_LCTL_CPA_MASK, sdw->link_res->link_mask);
link_control &= spa_mask;
ret = intel_clear_bit(shim, SDW_SHIM_LCTL, link_control, cpa_mask);
if (ret < 0) {
dev_err(sdw->cdns.dev, "%s: could not power down link\n", __func__);
/*
* we leave the sdw->cdns.link_up flag as false since we've disabled
* the link at this point and cannot handle interrupts any longer.
*/
}
}
mutex_unlock(sdw->link_res->shim_lock);
return ret;
}
static void intel_shim_sync_arm(struct sdw_intel *sdw)
{
void __iomem *shim = sdw->link_res->shim;
u32 sync_reg;
mutex_lock(sdw->link_res->shim_lock);
/* update SYNC register */
sync_reg = intel_readl(shim, SDW_SHIM_SYNC);
sync_reg |= (SDW_SHIM_SYNC_CMDSYNC << sdw->instance);
intel_writel(shim, SDW_SHIM_SYNC, sync_reg);
mutex_unlock(sdw->link_res->shim_lock);
}
static int intel_shim_sync_go_unlocked(struct sdw_intel *sdw)
{
void __iomem *shim = sdw->link_res->shim;
u32 sync_reg;
int ret;
/* Read SYNC register */
sync_reg = intel_readl(shim, SDW_SHIM_SYNC);
/*
* Set SyncGO bit to synchronously trigger a bank switch for
* all the masters. A write to SYNCGO bit clears CMDSYNC bit for all
* the Masters.
*/
sync_reg |= SDW_SHIM_SYNC_SYNCGO;
ret = intel_clear_bit(shim, SDW_SHIM_SYNC, sync_reg,
SDW_SHIM_SYNC_SYNCGO);
if (ret < 0)
dev_err(sdw->cdns.dev, "SyncGO clear failed: %d\n", ret);
return ret;
}
static int intel_shim_sync_go(struct sdw_intel *sdw)
{
int ret;
mutex_lock(sdw->link_res->shim_lock);
ret = intel_shim_sync_go_unlocked(sdw);
mutex_unlock(sdw->link_res->shim_lock);
return ret;
}
/*
* PDI routines
*/
static void intel_pdi_init(struct sdw_intel *sdw,
struct sdw_cdns_stream_config *config)
{
void __iomem *shim = sdw->link_res->shim;
unsigned int link_id = sdw->instance;
int pcm_cap;
/* PCM Stream Capability */
pcm_cap = intel_readw(shim, SDW_SHIM_PCMSCAP(link_id));
config->pcm_bd = FIELD_GET(SDW_SHIM_PCMSCAP_BSS, pcm_cap);
config->pcm_in = FIELD_GET(SDW_SHIM_PCMSCAP_ISS, pcm_cap);
config->pcm_out = FIELD_GET(SDW_SHIM_PCMSCAP_OSS, pcm_cap);
dev_dbg(sdw->cdns.dev, "PCM cap bd:%d in:%d out:%d\n",
config->pcm_bd, config->pcm_in, config->pcm_out);
}
static int
intel_pdi_get_ch_cap(struct sdw_intel *sdw, unsigned int pdi_num)
{
void __iomem *shim = sdw->link_res->shim;
unsigned int link_id = sdw->instance;
int count;
count = intel_readw(shim, SDW_SHIM_PCMSYCHC(link_id, pdi_num));
/*
* WORKAROUND: on all existing Intel controllers, pdi
* number 2 reports channel count as 1 even though it
* supports 8 channels. Performing hardcoding for pdi
* number 2.
*/
if (pdi_num == 2)
count = 7;
/* zero based values for channel count in register */
count++;
return count;
}
static int intel_pdi_get_ch_update(struct sdw_intel *sdw,
struct sdw_cdns_pdi *pdi,
unsigned int num_pdi,
unsigned int *num_ch)
{
int i, ch_count = 0;
for (i = 0; i < num_pdi; i++) {
pdi->ch_count = intel_pdi_get_ch_cap(sdw, pdi->num);
ch_count += pdi->ch_count;
pdi++;
}
*num_ch = ch_count;
return 0;
}
static int intel_pdi_stream_ch_update(struct sdw_intel *sdw,
struct sdw_cdns_streams *stream)
{
intel_pdi_get_ch_update(sdw, stream->bd, stream->num_bd,
&stream->num_ch_bd);
intel_pdi_get_ch_update(sdw, stream->in, stream->num_in,
&stream->num_ch_in);
intel_pdi_get_ch_update(sdw, stream->out, stream->num_out,
&stream->num_ch_out);
return 0;
}
static int intel_pdi_ch_update(struct sdw_intel *sdw)
{
intel_pdi_stream_ch_update(sdw, &sdw->cdns.pcm);
return 0;
}
static void
intel_pdi_shim_configure(struct sdw_intel *sdw, struct sdw_cdns_pdi *pdi)
{
void __iomem *shim = sdw->link_res->shim;
unsigned int link_id = sdw->instance;
int pdi_conf = 0;
/* the Bulk and PCM streams are not contiguous */
pdi->intel_alh_id = (link_id * 16) + pdi->num + 3;
if (pdi->num >= 2)
pdi->intel_alh_id += 2;
/*
* Program stream parameters to stream SHIM register
* This is applicable for PCM stream only.
*/
if (pdi->type != SDW_STREAM_PCM)
return;
if (pdi->dir == SDW_DATA_DIR_RX)
pdi_conf |= SDW_SHIM_PCMSYCM_DIR;
else
pdi_conf &= ~(SDW_SHIM_PCMSYCM_DIR);
u32p_replace_bits(&pdi_conf, pdi->intel_alh_id, SDW_SHIM_PCMSYCM_STREAM);
u32p_replace_bits(&pdi_conf, pdi->l_ch_num, SDW_SHIM_PCMSYCM_LCHN);
u32p_replace_bits(&pdi_conf, pdi->h_ch_num, SDW_SHIM_PCMSYCM_HCHN);
intel_writew(shim, SDW_SHIM_PCMSYCHM(link_id, pdi->num), pdi_conf);
}
static void
intel_pdi_alh_configure(struct sdw_intel *sdw, struct sdw_cdns_pdi *pdi)
{
void __iomem *alh = sdw->link_res->alh;
unsigned int link_id = sdw->instance;
unsigned int conf;
/* the Bulk and PCM streams are not contiguous */
pdi->intel_alh_id = (link_id * 16) + pdi->num + 3;
if (pdi->num >= 2)
pdi->intel_alh_id += 2;
/* Program Stream config ALH register */
conf = intel_readl(alh, SDW_ALH_STRMZCFG(pdi->intel_alh_id));
u32p_replace_bits(&conf, SDW_ALH_STRMZCFG_DMAT_VAL, SDW_ALH_STRMZCFG_DMAT);
u32p_replace_bits(&conf, pdi->ch_count - 1, SDW_ALH_STRMZCFG_CHN);
intel_writel(alh, SDW_ALH_STRMZCFG(pdi->intel_alh_id), conf);
}
static int intel_params_stream(struct sdw_intel *sdw,
int stream,
struct snd_soc_dai *dai,
struct snd_pcm_hw_params *hw_params,
int link_id, int alh_stream_id)
{
struct sdw_intel_link_res *res = sdw->link_res;
struct sdw_intel_stream_params_data params_data;
params_data.stream = stream; /* direction */
params_data.dai = dai;
params_data.hw_params = hw_params;
params_data.link_id = link_id;
params_data.alh_stream_id = alh_stream_id;
if (res->ops && res->ops->params_stream && res->dev)
return res->ops->params_stream(res->dev,
&params_data);
return -EIO;
}
static int intel_free_stream(struct sdw_intel *sdw,
int stream,
struct snd_soc_dai *dai,
int link_id)
{
struct sdw_intel_link_res *res = sdw->link_res;
struct sdw_intel_stream_free_data free_data;
free_data.stream = stream; /* direction */
free_data.dai = dai;
free_data.link_id = link_id;
if (res->ops && res->ops->free_stream && res->dev)
return res->ops->free_stream(res->dev,
&free_data);
return 0;
}
/*
* bank switch routines
*/
static int intel_pre_bank_switch(struct sdw_bus *bus)
{
struct sdw_cdns *cdns = bus_to_cdns(bus);
struct sdw_intel *sdw = cdns_to_intel(cdns);
/* Write to register only for multi-link */
if (!bus->multi_link)
return 0;
intel_shim_sync_arm(sdw);
return 0;
}
static int intel_post_bank_switch(struct sdw_bus *bus)
{
struct sdw_cdns *cdns = bus_to_cdns(bus);
struct sdw_intel *sdw = cdns_to_intel(cdns);
void __iomem *shim = sdw->link_res->shim;
int sync_reg, ret;
/* Write to register only for multi-link */
if (!bus->multi_link)
return 0;
mutex_lock(sdw->link_res->shim_lock);
/* Read SYNC register */
sync_reg = intel_readl(shim, SDW_SHIM_SYNC);
/*
* post_bank_switch() ops is called from the bus in loop for
* all the Masters in the steam with the expectation that
* we trigger the bankswitch for the only first Master in the list
* and do nothing for the other Masters
*
* So, set the SYNCGO bit only if CMDSYNC bit is set for any Master.
*/
if (!(sync_reg & SDW_SHIM_SYNC_CMDSYNC_MASK)) {
ret = 0;
goto unlock;
}
ret = intel_shim_sync_go_unlocked(sdw);
unlock:
mutex_unlock(sdw->link_res->shim_lock);
if (ret < 0)
dev_err(sdw->cdns.dev, "Post bank switch failed: %d\n", ret);
return ret;
}
/*
* DAI routines
*/
static int intel_startup(struct snd_pcm_substream *substream,
struct snd_soc_dai *dai)
{
struct sdw_cdns *cdns = snd_soc_dai_get_drvdata(dai);
int ret;
ret = pm_runtime_resume_and_get(cdns->dev);
if (ret < 0 && ret != -EACCES) {
dev_err_ratelimited(cdns->dev,
"pm_runtime_resume_and_get failed in %s, ret %d\n",
__func__, ret);
return ret;
}
return 0;
}
static int intel_hw_params(struct snd_pcm_substream *substream,
struct snd_pcm_hw_params *params,
struct snd_soc_dai *dai)
{
struct sdw_cdns *cdns = snd_soc_dai_get_drvdata(dai);
struct sdw_intel *sdw = cdns_to_intel(cdns);
struct sdw_cdns_dma_data *dma;
struct sdw_cdns_pdi *pdi;
struct sdw_stream_config sconfig;
struct sdw_port_config *pconfig;
int ch, dir;
int ret;
dma = snd_soc_dai_get_dma_data(dai, substream);
if (!dma)
return -EIO;
ch = params_channels(params);
if (substream->stream == SNDRV_PCM_STREAM_CAPTURE)
dir = SDW_DATA_DIR_RX;
else
dir = SDW_DATA_DIR_TX;
pdi = sdw_cdns_alloc_pdi(cdns, &cdns->pcm, ch, dir, dai->id);
if (!pdi) {
ret = -EINVAL;
goto error;
}
/* do run-time configurations for SHIM, ALH and PDI/PORT */
intel_pdi_shim_configure(sdw, pdi);
intel_pdi_alh_configure(sdw, pdi);
sdw_cdns_config_stream(cdns, ch, dir, pdi);
/* store pdi and hw_params, may be needed in prepare step */
soundwire: intel: improve suspend flows This patch provides both a simplification of the suspend flows and a better balanced operation during suspend/resume transition, as part of the transition of Sound Open Firmware (SOF) to dynamic pipelines: the DSP resources are only enabled when required instead of enabled on startup. The exiting code relies on a convoluted way of dealing with suspend signals. Since there is no .suspend DAI callback, we used the component .suspend and marked all the component DAI dmas as 'suspended'. The information was used in the .prepare stage to differentiate resume operations from xrun handling, and only reinitialize SHIM registers and DMA in the former case. While this solution has been working reliably for about 2 years, there is a much better solution consisting in trapping the TRIGGER_SUSPEND in the .trigger DAI ops. The DMA is still marked in the same way for the .prepare op to run, but in addition the callbacks sent to DSP firmware are now balanced. Normal operation: hw_params -> intel_params_stream hw_free -> intel_free_stream suspend -> intel_free_stream prepare -> intel_params_stream This balanced operation was not required with existing SOF firmware relying on static pipelines instantiated at every boot. With the on-going transition to dynamic pipelines, it's however a requirement to keep the use count for the DAI widget balanced across all transitions. The component suspend is not removed but instead modified to deal with a corner case: when a substream is PAUSED, the ALSA core does not throw the TRIGGER_SUSPEND. This is problematic since the refcount for all pipelines and widgets is not balanced, leading to issues on resume. The trigger callback keeps track of the 'paused' state with a new flag, which is tested during the component suspend called later to release the remaining DSP resources. These resources will be re-enabled in the .prepare step. The IPC used in the TRIGGER_SUSPEND to release DSP resources is not a problem since the BE dailink is already marked as non-atomic. Co-developed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Acked-By: Vinod Koul <vkoul@kernel.org> Link: https://lore.kernel.org/r/20211224021034.26635-4-yung-chuan.liao@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
2021-12-24 10:10:30 +08:00
dma->paused = false;
dma->suspended = false;
dma->pdi = pdi;
dma->hw_params = params;
/* Inform DSP about PDI stream number */
ret = intel_params_stream(sdw, substream->stream, dai, params,
sdw->instance,
pdi->intel_alh_id);
if (ret)
goto error;
sconfig.direction = dir;
sconfig.ch_count = ch;
sconfig.frame_rate = params_rate(params);
sconfig.type = dma->stream_type;
sconfig.bps = snd_pcm_format_width(params_format(params));
/* Port configuration */
pconfig = kzalloc(sizeof(*pconfig), GFP_KERNEL);
if (!pconfig) {
ret = -ENOMEM;
goto error;
}
pconfig->num = pdi->num;
pconfig->ch_mask = (1 << ch) - 1;
ret = sdw_stream_add_master(&cdns->bus, &sconfig,
pconfig, 1, dma->stream);
if (ret)
dev_err(cdns->dev, "add master to stream failed:%d\n", ret);
kfree(pconfig);
error:
return ret;
}
static int intel_prepare(struct snd_pcm_substream *substream,
struct snd_soc_dai *dai)
{
struct sdw_cdns *cdns = snd_soc_dai_get_drvdata(dai);
struct sdw_intel *sdw = cdns_to_intel(cdns);
struct sdw_cdns_dma_data *dma;
int ch, dir;
int ret = 0;
dma = snd_soc_dai_get_dma_data(dai, substream);
if (!dma) {
dev_err(dai->dev, "failed to get dma data in %s\n",
__func__);
return -EIO;
}
if (dma->suspended) {
dma->suspended = false;
/*
* .prepare() is called after system resume, where we
* need to reinitialize the SHIM/ALH/Cadence IP.
* .prepare() is also called to deal with underflows,
* but in those cases we cannot touch ALH/SHIM
* registers
*/
/* configure stream */
ch = params_channels(dma->hw_params);
if (substream->stream == SNDRV_PCM_STREAM_CAPTURE)
dir = SDW_DATA_DIR_RX;
else
dir = SDW_DATA_DIR_TX;
intel_pdi_shim_configure(sdw, dma->pdi);
intel_pdi_alh_configure(sdw, dma->pdi);
sdw_cdns_config_stream(cdns, ch, dir, dma->pdi);
/* Inform DSP about PDI stream number */
ret = intel_params_stream(sdw, substream->stream, dai,
dma->hw_params,
sdw->instance,
dma->pdi->intel_alh_id);
}
return ret;
}
static int
intel_hw_free(struct snd_pcm_substream *substream, struct snd_soc_dai *dai)
{
struct sdw_cdns *cdns = snd_soc_dai_get_drvdata(dai);
struct sdw_intel *sdw = cdns_to_intel(cdns);
struct sdw_cdns_dma_data *dma;
int ret;
dma = snd_soc_dai_get_dma_data(dai, substream);
if (!dma)
return -EIO;
/*
* The sdw stream state will transition to RELEASED when stream->
* master_list is empty. So the stream state will transition to
* DEPREPARED for the first cpu-dai and to RELEASED for the last
* cpu-dai.
*/
ret = sdw_stream_remove_master(&cdns->bus, dma->stream);
if (ret < 0) {
dev_err(dai->dev, "remove master from stream %s failed: %d\n",
dma->stream->name, ret);
return ret;
}
ret = intel_free_stream(sdw, substream->stream, dai, sdw->instance);
if (ret < 0) {
dev_err(dai->dev, "intel_free_stream: failed %d\n", ret);
return ret;
}
dma->hw_params = NULL;
dma->pdi = NULL;
return 0;
}
static void intel_shutdown(struct snd_pcm_substream *substream,
struct snd_soc_dai *dai)
{
struct sdw_cdns *cdns = snd_soc_dai_get_drvdata(dai);
pm_runtime_mark_last_busy(cdns->dev);
pm_runtime_put_autosuspend(cdns->dev);
}
static int intel_pcm_set_sdw_stream(struct snd_soc_dai *dai,
void *stream, int direction)
{
return cdns_set_sdw_stream(dai, stream, direction);
}
static void *intel_get_sdw_stream(struct snd_soc_dai *dai,
int direction)
{
struct sdw_cdns_dma_data *dma;
if (direction == SNDRV_PCM_STREAM_PLAYBACK)
dma = dai->playback_dma_data;
else
dma = dai->capture_dma_data;
if (!dma)
return ERR_PTR(-EINVAL);
return dma->stream;
}
soundwire: intel: improve suspend flows This patch provides both a simplification of the suspend flows and a better balanced operation during suspend/resume transition, as part of the transition of Sound Open Firmware (SOF) to dynamic pipelines: the DSP resources are only enabled when required instead of enabled on startup. The exiting code relies on a convoluted way of dealing with suspend signals. Since there is no .suspend DAI callback, we used the component .suspend and marked all the component DAI dmas as 'suspended'. The information was used in the .prepare stage to differentiate resume operations from xrun handling, and only reinitialize SHIM registers and DMA in the former case. While this solution has been working reliably for about 2 years, there is a much better solution consisting in trapping the TRIGGER_SUSPEND in the .trigger DAI ops. The DMA is still marked in the same way for the .prepare op to run, but in addition the callbacks sent to DSP firmware are now balanced. Normal operation: hw_params -> intel_params_stream hw_free -> intel_free_stream suspend -> intel_free_stream prepare -> intel_params_stream This balanced operation was not required with existing SOF firmware relying on static pipelines instantiated at every boot. With the on-going transition to dynamic pipelines, it's however a requirement to keep the use count for the DAI widget balanced across all transitions. The component suspend is not removed but instead modified to deal with a corner case: when a substream is PAUSED, the ALSA core does not throw the TRIGGER_SUSPEND. This is problematic since the refcount for all pipelines and widgets is not balanced, leading to issues on resume. The trigger callback keeps track of the 'paused' state with a new flag, which is tested during the component suspend called later to release the remaining DSP resources. These resources will be re-enabled in the .prepare step. The IPC used in the TRIGGER_SUSPEND to release DSP resources is not a problem since the BE dailink is already marked as non-atomic. Co-developed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Acked-By: Vinod Koul <vkoul@kernel.org> Link: https://lore.kernel.org/r/20211224021034.26635-4-yung-chuan.liao@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
2021-12-24 10:10:30 +08:00
static int intel_trigger(struct snd_pcm_substream *substream, int cmd, struct snd_soc_dai *dai)
{
struct sdw_cdns *cdns = snd_soc_dai_get_drvdata(dai);
struct sdw_intel *sdw = cdns_to_intel(cdns);
struct sdw_intel_link_res *res = sdw->link_res;
soundwire: intel: improve suspend flows This patch provides both a simplification of the suspend flows and a better balanced operation during suspend/resume transition, as part of the transition of Sound Open Firmware (SOF) to dynamic pipelines: the DSP resources are only enabled when required instead of enabled on startup. The exiting code relies on a convoluted way of dealing with suspend signals. Since there is no .suspend DAI callback, we used the component .suspend and marked all the component DAI dmas as 'suspended'. The information was used in the .prepare stage to differentiate resume operations from xrun handling, and only reinitialize SHIM registers and DMA in the former case. While this solution has been working reliably for about 2 years, there is a much better solution consisting in trapping the TRIGGER_SUSPEND in the .trigger DAI ops. The DMA is still marked in the same way for the .prepare op to run, but in addition the callbacks sent to DSP firmware are now balanced. Normal operation: hw_params -> intel_params_stream hw_free -> intel_free_stream suspend -> intel_free_stream prepare -> intel_params_stream This balanced operation was not required with existing SOF firmware relying on static pipelines instantiated at every boot. With the on-going transition to dynamic pipelines, it's however a requirement to keep the use count for the DAI widget balanced across all transitions. The component suspend is not removed but instead modified to deal with a corner case: when a substream is PAUSED, the ALSA core does not throw the TRIGGER_SUSPEND. This is problematic since the refcount for all pipelines and widgets is not balanced, leading to issues on resume. The trigger callback keeps track of the 'paused' state with a new flag, which is tested during the component suspend called later to release the remaining DSP resources. These resources will be re-enabled in the .prepare step. The IPC used in the TRIGGER_SUSPEND to release DSP resources is not a problem since the BE dailink is already marked as non-atomic. Co-developed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Acked-By: Vinod Koul <vkoul@kernel.org> Link: https://lore.kernel.org/r/20211224021034.26635-4-yung-chuan.liao@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
2021-12-24 10:10:30 +08:00
struct sdw_cdns_dma_data *dma;
int ret = 0;
/*
* The .trigger callback is used to send required IPC to audio
* firmware. The .free_stream callback will still be called
* by intel_free_stream() in the TRIGGER_SUSPEND case.
*/
if (res->ops && res->ops->trigger)
res->ops->trigger(dai, cmd, substream->stream);
soundwire: intel: improve suspend flows This patch provides both a simplification of the suspend flows and a better balanced operation during suspend/resume transition, as part of the transition of Sound Open Firmware (SOF) to dynamic pipelines: the DSP resources are only enabled when required instead of enabled on startup. The exiting code relies on a convoluted way of dealing with suspend signals. Since there is no .suspend DAI callback, we used the component .suspend and marked all the component DAI dmas as 'suspended'. The information was used in the .prepare stage to differentiate resume operations from xrun handling, and only reinitialize SHIM registers and DMA in the former case. While this solution has been working reliably for about 2 years, there is a much better solution consisting in trapping the TRIGGER_SUSPEND in the .trigger DAI ops. The DMA is still marked in the same way for the .prepare op to run, but in addition the callbacks sent to DSP firmware are now balanced. Normal operation: hw_params -> intel_params_stream hw_free -> intel_free_stream suspend -> intel_free_stream prepare -> intel_params_stream This balanced operation was not required with existing SOF firmware relying on static pipelines instantiated at every boot. With the on-going transition to dynamic pipelines, it's however a requirement to keep the use count for the DAI widget balanced across all transitions. The component suspend is not removed but instead modified to deal with a corner case: when a substream is PAUSED, the ALSA core does not throw the TRIGGER_SUSPEND. This is problematic since the refcount for all pipelines and widgets is not balanced, leading to issues on resume. The trigger callback keeps track of the 'paused' state with a new flag, which is tested during the component suspend called later to release the remaining DSP resources. These resources will be re-enabled in the .prepare step. The IPC used in the TRIGGER_SUSPEND to release DSP resources is not a problem since the BE dailink is already marked as non-atomic. Co-developed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Acked-By: Vinod Koul <vkoul@kernel.org> Link: https://lore.kernel.org/r/20211224021034.26635-4-yung-chuan.liao@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
2021-12-24 10:10:30 +08:00
dma = snd_soc_dai_get_dma_data(dai, substream);
if (!dma) {
dev_err(dai->dev, "failed to get dma data in %s\n",
__func__);
return -EIO;
}
switch (cmd) {
case SNDRV_PCM_TRIGGER_SUSPEND:
/*
* The .prepare callback is used to deal with xruns and resume operations.
* In the case of xruns, the DMAs and SHIM registers cannot be touched,
* but for resume operations the DMAs and SHIM registers need to be initialized.
* the .trigger callback is used to track the suspend case only.
*/
dma->suspended = true;
ret = intel_free_stream(sdw, substream->stream, dai, sdw->instance);
break;
case SNDRV_PCM_TRIGGER_PAUSE_PUSH:
dma->paused = true;
break;
case SNDRV_PCM_TRIGGER_STOP:
case SNDRV_PCM_TRIGGER_PAUSE_RELEASE:
dma->paused = false;
break;
default:
break;
}
return ret;
}
soundwire: intel: use pm_runtime_resume() on component probe During the card registration, transactions on the SoundWire bus can be initiated. If the ALSA card is registered after the bus suspends, timeouts can be seen while reading/writing codec registers. This is extremely easy to reproduce in driver bind/unbind tests. In an initial experiment, the ASoC soc-component.c code was modified to initiate a pm_runtime resume on a component probe. The results showed this was too invasive. Instead this patch suggests resuming the SoundWire component only. Because of the parent-child hierarchy enforced by the pm_runtime framework, it can be argued that the codec component probe should be enough to resume all necessary devices, and indeed the same resume will be applied to SoundWire codecs used on Intel platforms. Calling pm_runtime_resume() on both the Intel and codec sides has the benefit of resuming the bus without assuming any order during the card registration. The first component on a dailink to be probed will resume the bus. In addition, if a codec driver did not implement this transition, the Intel component would still resume the bus and avoid timeouts on card registration. BugLink: https://github.com/thesofproject/linux/issues/3651 Reviewed-by: Rander Wang <rander.wang@intel.com> Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com> Reviewed-by: Péter Ujfalusi <peter.ujfalusi@linux.intel.com> Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Link: https://lore.kernel.org/r/20220621225641.221170-4-pierre-louis.bossart@linux.intel.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2022-06-22 06:56:40 +08:00
static int intel_component_probe(struct snd_soc_component *component)
{
int ret;
/*
* make sure the device is pm_runtime_active before initiating
* bus transactions during the card registration.
* We use pm_runtime_resume() here, without taking a reference
* and releasing it immediately.
*/
ret = pm_runtime_resume(component->dev);
if (ret < 0 && ret != -EACCES)
return ret;
return 0;
}
soundwire: intel: improve suspend flows This patch provides both a simplification of the suspend flows and a better balanced operation during suspend/resume transition, as part of the transition of Sound Open Firmware (SOF) to dynamic pipelines: the DSP resources are only enabled when required instead of enabled on startup. The exiting code relies on a convoluted way of dealing with suspend signals. Since there is no .suspend DAI callback, we used the component .suspend and marked all the component DAI dmas as 'suspended'. The information was used in the .prepare stage to differentiate resume operations from xrun handling, and only reinitialize SHIM registers and DMA in the former case. While this solution has been working reliably for about 2 years, there is a much better solution consisting in trapping the TRIGGER_SUSPEND in the .trigger DAI ops. The DMA is still marked in the same way for the .prepare op to run, but in addition the callbacks sent to DSP firmware are now balanced. Normal operation: hw_params -> intel_params_stream hw_free -> intel_free_stream suspend -> intel_free_stream prepare -> intel_params_stream This balanced operation was not required with existing SOF firmware relying on static pipelines instantiated at every boot. With the on-going transition to dynamic pipelines, it's however a requirement to keep the use count for the DAI widget balanced across all transitions. The component suspend is not removed but instead modified to deal with a corner case: when a substream is PAUSED, the ALSA core does not throw the TRIGGER_SUSPEND. This is problematic since the refcount for all pipelines and widgets is not balanced, leading to issues on resume. The trigger callback keeps track of the 'paused' state with a new flag, which is tested during the component suspend called later to release the remaining DSP resources. These resources will be re-enabled in the .prepare step. The IPC used in the TRIGGER_SUSPEND to release DSP resources is not a problem since the BE dailink is already marked as non-atomic. Co-developed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Acked-By: Vinod Koul <vkoul@kernel.org> Link: https://lore.kernel.org/r/20211224021034.26635-4-yung-chuan.liao@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
2021-12-24 10:10:30 +08:00
static int intel_component_dais_suspend(struct snd_soc_component *component)
{
struct snd_soc_dai *dai;
/*
* In the corner case where a SUSPEND happens during a PAUSE, the ALSA core
* does not throw the TRIGGER_SUSPEND. This leaves the DAIs in an unbalanced state.
* Since the component suspend is called last, we can trap this corner case
* and force the DAIs to release their resources.
*/
for_each_component_dais(component, dai) {
struct sdw_cdns *cdns = snd_soc_dai_get_drvdata(dai);
struct sdw_intel *sdw = cdns_to_intel(cdns);
struct sdw_cdns_dma_data *dma;
int stream;
int ret;
dma = dai->playback_dma_data;
stream = SNDRV_PCM_STREAM_PLAYBACK;
if (!dma) {
dma = dai->capture_dma_data;
stream = SNDRV_PCM_STREAM_CAPTURE;
}
if (!dma)
continue;
if (dma->suspended)
continue;
soundwire: intel: improve suspend flows This patch provides both a simplification of the suspend flows and a better balanced operation during suspend/resume transition, as part of the transition of Sound Open Firmware (SOF) to dynamic pipelines: the DSP resources are only enabled when required instead of enabled on startup. The exiting code relies on a convoluted way of dealing with suspend signals. Since there is no .suspend DAI callback, we used the component .suspend and marked all the component DAI dmas as 'suspended'. The information was used in the .prepare stage to differentiate resume operations from xrun handling, and only reinitialize SHIM registers and DMA in the former case. While this solution has been working reliably for about 2 years, there is a much better solution consisting in trapping the TRIGGER_SUSPEND in the .trigger DAI ops. The DMA is still marked in the same way for the .prepare op to run, but in addition the callbacks sent to DSP firmware are now balanced. Normal operation: hw_params -> intel_params_stream hw_free -> intel_free_stream suspend -> intel_free_stream prepare -> intel_params_stream This balanced operation was not required with existing SOF firmware relying on static pipelines instantiated at every boot. With the on-going transition to dynamic pipelines, it's however a requirement to keep the use count for the DAI widget balanced across all transitions. The component suspend is not removed but instead modified to deal with a corner case: when a substream is PAUSED, the ALSA core does not throw the TRIGGER_SUSPEND. This is problematic since the refcount for all pipelines and widgets is not balanced, leading to issues on resume. The trigger callback keeps track of the 'paused' state with a new flag, which is tested during the component suspend called later to release the remaining DSP resources. These resources will be re-enabled in the .prepare step. The IPC used in the TRIGGER_SUSPEND to release DSP resources is not a problem since the BE dailink is already marked as non-atomic. Co-developed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Acked-By: Vinod Koul <vkoul@kernel.org> Link: https://lore.kernel.org/r/20211224021034.26635-4-yung-chuan.liao@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
2021-12-24 10:10:30 +08:00
if (dma->paused) {
dma->suspended = true;
ret = intel_free_stream(sdw, stream, dai, sdw->instance);
if (ret < 0)
return ret;
}
}
return 0;
}
static const struct snd_soc_dai_ops intel_pcm_dai_ops = {
.startup = intel_startup,
.hw_params = intel_hw_params,
.prepare = intel_prepare,
.hw_free = intel_hw_free,
soundwire: intel: improve suspend flows This patch provides both a simplification of the suspend flows and a better balanced operation during suspend/resume transition, as part of the transition of Sound Open Firmware (SOF) to dynamic pipelines: the DSP resources are only enabled when required instead of enabled on startup. The exiting code relies on a convoluted way of dealing with suspend signals. Since there is no .suspend DAI callback, we used the component .suspend and marked all the component DAI dmas as 'suspended'. The information was used in the .prepare stage to differentiate resume operations from xrun handling, and only reinitialize SHIM registers and DMA in the former case. While this solution has been working reliably for about 2 years, there is a much better solution consisting in trapping the TRIGGER_SUSPEND in the .trigger DAI ops. The DMA is still marked in the same way for the .prepare op to run, but in addition the callbacks sent to DSP firmware are now balanced. Normal operation: hw_params -> intel_params_stream hw_free -> intel_free_stream suspend -> intel_free_stream prepare -> intel_params_stream This balanced operation was not required with existing SOF firmware relying on static pipelines instantiated at every boot. With the on-going transition to dynamic pipelines, it's however a requirement to keep the use count for the DAI widget balanced across all transitions. The component suspend is not removed but instead modified to deal with a corner case: when a substream is PAUSED, the ALSA core does not throw the TRIGGER_SUSPEND. This is problematic since the refcount for all pipelines and widgets is not balanced, leading to issues on resume. The trigger callback keeps track of the 'paused' state with a new flag, which is tested during the component suspend called later to release the remaining DSP resources. These resources will be re-enabled in the .prepare step. The IPC used in the TRIGGER_SUSPEND to release DSP resources is not a problem since the BE dailink is already marked as non-atomic. Co-developed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Acked-By: Vinod Koul <vkoul@kernel.org> Link: https://lore.kernel.org/r/20211224021034.26635-4-yung-chuan.liao@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
2021-12-24 10:10:30 +08:00
.trigger = intel_trigger,
.shutdown = intel_shutdown,
ASoC/SoundWire: dai: expand 'stream' concept beyond SoundWire The HDAudio ASoC support relies on the set_tdm_slots() helper to store the HDaudio stream tag in the tx_mask. This only works because of the pre-existing order in soc-pcm.c, where the hw_params() is handled for codec_dais *before* cpu_dais. When the order is reversed, the stream_tag is used as a mask in the codec fixup functions: /* fixup params based on TDM slot masks */ if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK && codec_dai->tx_mask) soc_pcm_codec_params_fixup(&codec_params, codec_dai->tx_mask); As a result of this confusion, the codec_params_fixup() ends-up generating bad channel masks, depending on what stream_tag was allocated. We could add a flag to state that the tx_mask is really not a mask, but it would be quite ugly to persist in overloading concepts. Instead, this patch suggests a more generic get/set 'stream' API based on the existing model for SoundWire. We can expand the concept to store 'stream' opaque information that is specific to different DAI types. In the case of HDAudio DAIs, we only need to store a stream tag as an unsigned char pointer. The TDM rx_ and tx_masks should really only be used to store masks. Rename get_sdw_stream/set_sdw_stream callbacks and helpers as get_stream/set_stream. No functionality change beyond the rename. Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Reviewed-by: Rander Wang <rander.wang@intel.com> Reviewed-by: Ranjani Sridharan <ranjani.sridharan@intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Acked-By: Vinod Koul <vkoul@kernel.org> Link: https://lore.kernel.org/r/20211224021034.26635-5-yung-chuan.liao@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
2021-12-24 10:10:31 +08:00
.set_stream = intel_pcm_set_sdw_stream,
.get_stream = intel_get_sdw_stream,
};
static const struct snd_soc_component_driver dai_component = {
.name = "soundwire",
sound updates for 6.0-rc1 As diffstat shows, we've had lots of developments in a wide range at this time; the majority of changes are about ASoC, including subsystem-wide cleanups, continued SOF / Intel updates and a bunch of new drivers (as usual), while there have been some significant (but almost invisible) improvements in ALSA core side, too. Below are some highlights: Core: - Faster lookups of control elements with Xarray; normal user won't notice, but on the devices with tons of control elements, it can be visibly faster - Support for input validation for controls; this will harden for badly written drivers in general with a slight overhead - Deferred async signal handling for working around the potential deadlocks - Cleanup / refactoring raw MIDI locking code ASoC: - Restructing of the set_fmt() callbacks for making things clearer in situations like CODEC to CODEC links - Clean up and modernizing the DAI naming scheme setups - Merge of more of the Intel AVS driver stack, including some board integrations - New version 4 mechanism for communication with SOF DSPs - Suppoort for dynamically selecting the PLL to use at runtime on i.MX platforms - Improvements for CODEC to CODEC support in the generic cards - Support for AMD Jadeite and various machines, AMD RPL, Intel MetorLake DSPs, Mediatek MT8186 DSPs and MT6366, nVidia Tegra MDDRC, OPE and PEQ, NXP TFA9890, Qualcomm SDM845, WCD9335 and WAS883x, and Texas Instruments TAS2780 HD- and USB-audio: - Continued improvement for CS35L41 (sub)codec support - More quirks for various devices (HP, Lenovo, Dell, Clevo) -----BEGIN PGP SIGNATURE----- iQJCBAABCAAsFiEEIXTw5fNLNI7mMiVaLtJE4w1nLE8FAmLr5bcOHHRpd2FpQHN1 c2UuZGUACgkQLtJE4w1nLE93IQ/+OleeiGv7C487QN5MrBCkdFnSAiXsXDArcMgo Gt6XLubH54et1tqi2ms4gRQOqr4HiBelERuqmaCLMZfEgVDc0VhJnf2jjhluYq9+ o9+kcYKul6qTZeNZLPjEX8pBvDe7HzOl7yep++ZnKeH6DAKNQQLDjVuOcQU/BXdY kL8vYrLI3zfqj/pCePb5xpkBx4XdCrE3TfiCr3tAHVg9MyvSGOJyWs02mEBqQRnc rlLmkjQVQyln/AGK4RAPMmrrFytktAvBjmIDyFL7kAzhdxe0ouNzTvdxESeojNJv CVo/p3hl/+0LYjpD2x9v2pQuifOfpjwSCy6f8jsaF366sMwR1D45h051prILsxAF 05D5AOwMCnnJdFQFxw3mIkoDva3/PRX8GFfHsXoz+efc5Ibp8ksNYVgAJ3D2TTtt 7nAYMn0dimVDtw2LHiKantgWAs/rOqD3hDzGxFj2sR662ahsHr8pT8csnJAGoBvW 7kgx7ZzFo/wSyZJqVqV7p4g6J79ScehRwhqoiwZau9Eo+PhuxZUKvm4RwGFh0Vvg GbiVRPfLV4xQd/pDin6qRX1M7cgPc62qGLkhQHAzrX6H5ipwIbQrpyDGLjwdSEp8 XcQmzCG1zGOvb9A8BY1VBOQXxZRTqN58XujbZ6hsms7Uw8sqagxpYLA/e1bvt1E1 RQoHQRw= =1n0/ -----END PGP SIGNATURE----- Merge tag 'sound-6.0-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound Pull sound updates from Takashi Iwai: "As the diffstat shows, we've had lots of developments in a wide range at this time; the majority of changes are about ASoC, including subsystem-wide cleanups, continued SOF / Intel updates and a bunch of new drivers (as usual), while there have been some significant (but almost invisible) improvements in ALSA core side, too. Below are some highlights: Core: - Faster lookups of control elements with Xarray; normal user won't notice, but on the devices with tons of control elements, it can be visibly faster - Support for input validation for controls; this will harden for badly written drivers in general with a slight overhead - Deferred async signal handling for working around the potential deadlocks - Cleanup / refactoring raw MIDI locking code ASoC: - Restructing of the set_fmt() callbacks for making things clearer in situations like CODEC to CODEC links - Clean up and modernizing the DAI naming scheme setups - Merge of more of the Intel AVS driver stack, including some board integrations - New version 4 mechanism for communication with SOF DSPs - Suppoort for dynamically selecting the PLL to use at runtime on i.MX platforms - Improvements for CODEC to CODEC support in the generic cards - Support for AMD Jadeite and various machines, AMD RPL, Intel MetorLake DSPs, Mediatek MT8186 DSPs and MT6366, nVidia Tegra MDDRC, OPE and PEQ, NXP TFA9890, Qualcomm SDM845, WCD9335 and WAS883x, and Texas Instruments TAS2780 HD- and USB-audio: - Continued improvement for CS35L41 (sub)codec support - More quirks for various devices (HP, Lenovo, Dell, Clevo)" * tag 'sound-6.0-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound: (778 commits) ALSA: hda/realtek: Add quirk for HP Spectre x360 15-eb0xxx ALSA: line6: Replace sprintf() with sysfs_emit() ALSA: hda: Replace sprintf() with sysfs_emit() ALSA: pcm: Replace sprintf() with sysfs_emit() ALSA: core: Replace scnprintf() with sysfs_emit() ALSA: control-led: Replace sprintf() with sysfs_emit() ALSA: aoa: Replace sprintf() with sysfs_emit() ALSA: ac97: Replace sprintf() with sysfs_emit() ALSA: hda/realtek: Add quirk for Clevo NV45PZ ALSA: hda/realtek: Add quirk for Lenovo Yoga9 14IAP7 ALSA: control: Use deferred fasync helper ALSA: pcm: Use deferred fasync helper ALSA: timer: Use deferred fasync helper ALSA: core: Add async signal helpers ASoC: q6asm: use kcalloc() instead of kzalloc() ACPI: scan: Add CLSA0101 Laptop Support ALSA: hda: cs35l41: Support CLSA0101 ALSA: hda: cs35l41: Use the CS35L41 HDA internal define ASoC: dt-bindings: use spi-peripheral-props.yaml ASoC: codecs: va-macro: use fsgen as clock ...
2022-08-07 01:19:51 +08:00
.probe = intel_component_probe,
.suspend = intel_component_dais_suspend,
.legacy_dai_naming = 1,
};
static int intel_create_dai(struct sdw_cdns *cdns,
struct snd_soc_dai_driver *dais,
enum intel_pdi_type type,
u32 num, u32 off, u32 max_ch)
{
int i;
if (num == 0)
return 0;
/* TODO: Read supported rates/formats from hardware */
for (i = off; i < (off + num); i++) {
dais[i].name = devm_kasprintf(cdns->dev, GFP_KERNEL,
"SDW%d Pin%d",
cdns->instance, i);
if (!dais[i].name)
return -ENOMEM;
if (type == INTEL_PDI_BD || type == INTEL_PDI_OUT) {
dais[i].playback.channels_min = 1;
dais[i].playback.channels_max = max_ch;
dais[i].playback.rates = SNDRV_PCM_RATE_48000;
dais[i].playback.formats = SNDRV_PCM_FMTBIT_S16_LE;
}
if (type == INTEL_PDI_BD || type == INTEL_PDI_IN) {
dais[i].capture.channels_min = 1;
dais[i].capture.channels_max = max_ch;
dais[i].capture.rates = SNDRV_PCM_RATE_48000;
dais[i].capture.formats = SNDRV_PCM_FMTBIT_S16_LE;
}
dais[i].ops = &intel_pcm_dai_ops;
}
return 0;
}
static int intel_register_dai(struct sdw_intel *sdw)
{
struct sdw_cdns_stream_config config;
struct sdw_cdns *cdns = &sdw->cdns;
struct sdw_cdns_streams *stream;
struct snd_soc_dai_driver *dais;
int num_dai, ret, off = 0;
/* Read the PDI config and initialize cadence PDI */
intel_pdi_init(sdw, &config);
ret = sdw_cdns_pdi_init(cdns, config);
if (ret)
return ret;
intel_pdi_ch_update(sdw);
/* DAIs are created based on total number of PDIs supported */
num_dai = cdns->pcm.num_pdi;
dais = devm_kcalloc(cdns->dev, num_dai, sizeof(*dais), GFP_KERNEL);
if (!dais)
return -ENOMEM;
/* Create PCM DAIs */
stream = &cdns->pcm;
ret = intel_create_dai(cdns, dais, INTEL_PDI_IN, cdns->pcm.num_in,
off, stream->num_ch_in);
if (ret)
return ret;
off += cdns->pcm.num_in;
ret = intel_create_dai(cdns, dais, INTEL_PDI_OUT, cdns->pcm.num_out,
off, stream->num_ch_out);
if (ret)
return ret;
off += cdns->pcm.num_out;
ret = intel_create_dai(cdns, dais, INTEL_PDI_BD, cdns->pcm.num_bd,
off, stream->num_ch_bd);
if (ret)
return ret;
return devm_snd_soc_register_component(cdns->dev, &dai_component,
dais, num_dai);
}
static int intel_start_bus(struct sdw_intel *sdw)
{
struct device *dev = sdw->cdns.dev;
struct sdw_cdns *cdns = &sdw->cdns;
struct sdw_bus *bus = &cdns->bus;
int ret;
ret = sdw_cdns_enable_interrupt(cdns, true);
if (ret < 0) {
dev_err(dev, "%s: cannot enable interrupts: %d\n", __func__, ret);
return ret;
}
/*
* follow recommended programming flows to avoid timeouts when
* gsync is enabled
*/
if (bus->multi_link)
intel_shim_sync_arm(sdw);
ret = sdw_cdns_init(cdns);
if (ret < 0) {
dev_err(dev, "%s: unable to initialize Cadence IP: %d\n", __func__, ret);
goto err_interrupt;
}
ret = sdw_cdns_exit_reset(cdns);
if (ret < 0) {
dev_err(dev, "%s: unable to exit bus reset sequence: %d\n", __func__, ret);
goto err_interrupt;
}
if (bus->multi_link) {
ret = intel_shim_sync_go(sdw);
if (ret < 0) {
dev_err(dev, "%s: sync go failed: %d\n", __func__, ret);
goto err_interrupt;
}
}
sdw_cdns_check_self_clearing_bits(cdns, __func__,
true, INTEL_MASTER_RESET_ITERATIONS);
return 0;
err_interrupt:
sdw_cdns_enable_interrupt(cdns, false);
return ret;
}
static int intel_start_bus_after_reset(struct sdw_intel *sdw)
{
struct device *dev = sdw->cdns.dev;
struct sdw_cdns *cdns = &sdw->cdns;
struct sdw_bus *bus = &cdns->bus;
bool clock_stop0;
int status;
int ret;
/*
* An exception condition occurs for the CLK_STOP_BUS_RESET
* case if one or more masters remain active. In this condition,
* all the masters are powered on for they are in the same power
* domain. Master can preserve its context for clock stop0, so
* there is no need to clear slave status and reset bus.
*/
clock_stop0 = sdw_cdns_is_clock_stop(&sdw->cdns);
if (!clock_stop0) {
/*
* make sure all Slaves are tagged as UNATTACHED and
* provide reason for reinitialization
*/
status = SDW_UNATTACH_REQUEST_MASTER_RESET;
sdw_clear_slave_status(bus, status);
ret = sdw_cdns_enable_interrupt(cdns, true);
if (ret < 0) {
dev_err(dev, "cannot enable interrupts during resume\n");
return ret;
}
/*
* follow recommended programming flows to avoid
* timeouts when gsync is enabled
*/
if (bus->multi_link)
intel_shim_sync_arm(sdw);
/*
* Re-initialize the IP since it was powered-off
*/
sdw_cdns_init(&sdw->cdns);
} else {
ret = sdw_cdns_enable_interrupt(cdns, true);
if (ret < 0) {
dev_err(dev, "cannot enable interrupts during resume\n");
return ret;
}
}
ret = sdw_cdns_clock_restart(cdns, !clock_stop0);
if (ret < 0) {
dev_err(dev, "unable to restart clock during resume\n");
goto err_interrupt;
}
if (!clock_stop0) {
ret = sdw_cdns_exit_reset(cdns);
if (ret < 0) {
dev_err(dev, "unable to exit bus reset sequence during resume\n");
goto err_interrupt;
}
if (bus->multi_link) {
ret = intel_shim_sync_go(sdw);
if (ret < 0) {
dev_err(sdw->cdns.dev, "sync go failed during resume\n");
goto err_interrupt;
}
}
}
sdw_cdns_check_self_clearing_bits(cdns, __func__, true, INTEL_MASTER_RESET_ITERATIONS);
return 0;
err_interrupt:
sdw_cdns_enable_interrupt(cdns, false);
return ret;
}
static void intel_check_clock_stop(struct sdw_intel *sdw)
{
struct device *dev = sdw->cdns.dev;
bool clock_stop0;
clock_stop0 = sdw_cdns_is_clock_stop(&sdw->cdns);
if (!clock_stop0)
dev_err(dev, "%s: invalid configuration, clock was not stopped\n", __func__);
}
static int intel_start_bus_after_clock_stop(struct sdw_intel *sdw)
{
struct device *dev = sdw->cdns.dev;
struct sdw_cdns *cdns = &sdw->cdns;
int ret;
ret = sdw_cdns_enable_interrupt(cdns, true);
if (ret < 0) {
dev_err(dev, "%s: cannot enable interrupts: %d\n", __func__, ret);
return ret;
}
ret = sdw_cdns_clock_restart(cdns, false);
if (ret < 0) {
dev_err(dev, "%s: unable to restart clock: %d\n", __func__, ret);
sdw_cdns_enable_interrupt(cdns, false);
return ret;
}
sdw_cdns_check_self_clearing_bits(cdns, "intel_resume_runtime no_quirks",
true, INTEL_MASTER_RESET_ITERATIONS);
return 0;
}
static int intel_stop_bus(struct sdw_intel *sdw, bool clock_stop)
{
struct device *dev = sdw->cdns.dev;
struct sdw_cdns *cdns = &sdw->cdns;
bool wake_enable = false;
int ret;
if (clock_stop) {
ret = sdw_cdns_clock_stop(cdns, true);
if (ret < 0)
dev_err(dev, "%s: cannot stop clock: %d\n", __func__, ret);
else
wake_enable = true;
}
ret = sdw_cdns_enable_interrupt(cdns, false);
if (ret < 0) {
dev_err(dev, "%s: cannot disable interrupts: %d\n", __func__, ret);
return ret;
}
ret = intel_link_power_down(sdw);
if (ret) {
dev_err(dev, "%s: Link power down failed: %d\n", __func__, ret);
return ret;
}
intel_shim_wake(sdw, wake_enable);
return 0;
}
static int sdw_master_read_intel_prop(struct sdw_bus *bus)
{
struct sdw_master_prop *prop = &bus->prop;
struct fwnode_handle *link;
char name[32];
u32 quirk_mask;
/* Find master handle */
snprintf(name, sizeof(name),
"mipi-sdw-link-%d-subproperties", bus->link_id);
link = device_get_named_child_node(bus->dev, name);
if (!link) {
dev_err(bus->dev, "Master node %s not found\n", name);
return -EIO;
}
fwnode_property_read_u32(link,
"intel-sdw-ip-clock",
&prop->mclk_freq);
/* the values reported by BIOS are the 2x clock, not the bus clock */
prop->mclk_freq /= 2;
fwnode_property_read_u32(link,
"intel-quirk-mask",
&quirk_mask);
if (quirk_mask & SDW_INTEL_QUIRK_MASK_BUS_DISABLE)
prop->hw_disabled = true;
prop->quirks = SDW_MASTER_QUIRKS_CLEAR_INITIAL_CLASH |
SDW_MASTER_QUIRKS_CLEAR_INITIAL_PARITY;
return 0;
}
static int intel_prop_read(struct sdw_bus *bus)
{
/* Initialize with default handler to read all DisCo properties */
sdw_master_read_prop(bus);
/* read Intel-specific properties */
sdw_master_read_intel_prop(bus);
return 0;
}
static struct sdw_master_ops sdw_intel_ops = {
.read_prop = intel_prop_read,
.override_adr = sdw_dmi_override_adr,
.xfer_msg = cdns_xfer_msg,
.xfer_msg_defer = cdns_xfer_msg_defer,
.reset_page_addr = cdns_reset_page_addr,
.set_bus_conf = cdns_bus_conf,
.pre_bank_switch = intel_pre_bank_switch,
.post_bank_switch = intel_post_bank_switch,
.read_ping_status = cdns_read_ping_status,
};
/*
* probe and init (aux_dev_id argument is required by function prototype but not used)
*/
static int intel_link_probe(struct auxiliary_device *auxdev,
const struct auxiliary_device_id *aux_dev_id)
{
struct device *dev = &auxdev->dev;
struct sdw_intel_link_dev *ldev = auxiliary_dev_to_sdw_intel_link_dev(auxdev);
struct sdw_intel *sdw;
struct sdw_cdns *cdns;
struct sdw_bus *bus;
int ret;
sdw = devm_kzalloc(dev, sizeof(*sdw), GFP_KERNEL);
if (!sdw)
return -ENOMEM;
cdns = &sdw->cdns;
bus = &cdns->bus;
sdw->instance = auxdev->id;
sdw->link_res = &ldev->link_res;
cdns->dev = dev;
cdns->registers = sdw->link_res->registers;
cdns->instance = sdw->instance;
cdns->msg_count = 0;
bus->link_id = auxdev->id;
bus->dev_num_ida_min = INTEL_DEV_NUM_IDA_MIN;
sdw_cdns_probe(cdns);
/* Set ops */
bus->ops = &sdw_intel_ops;
/* set driver data, accessed by snd_soc_dai_get_drvdata() */
auxiliary_set_drvdata(auxdev, cdns);
/* use generic bandwidth allocation algorithm */
sdw->cdns.bus.compute_params = sdw_compute_params;
soundwire: intel: prevent pm_runtime resume prior to system suspend commit e38f9ff63e6d ("ACPI: scan: Do not add device IDs from _CID if _HID is not valid") exposes a race condition on a TGL RVP device leading to a timeout. The detailed analysis shows the RT711 codec driver scheduling a jack detection workqueue while attaching during a spurious pm_runtime resume, and the work function happens to be scheduled after the manager device is suspended. The direct link between this ACPI patch and a spurious pm_runtime resume is not obvious; the most likely explanation is that a change in the ACPI device linked list management modifies the order in which the pm_runtime device status is checked and exposes a race condition that was probably present for a very long time, but was not identified. We already have a check in the .prepare stage, where we will resume to full power from specific clock-stop modes. In all other cases, we don't need to resume to full power by default. Adding the SMART_SUSPEND flag prevents the spurious resume from happening. BugLink: https://github.com/thesofproject/linux/issues/3459 Fixes: 029bfd1cd53cd ("soundwire: intel: conditionally exit clock stop mode on system suspend") Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Reviewed-by: Rander Wang <rander.wang@intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Link: https://lore.kernel.org/r/20220420023241.14335-2-yung-chuan.liao@linux.intel.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2022-04-20 10:32:39 +08:00
/* avoid resuming from pm_runtime suspend if it's not required */
dev_pm_set_driver_flags(dev, DPM_FLAG_SMART_SUSPEND);
ret = sdw_bus_master_add(bus, dev, dev->fwnode);
if (ret) {
dev_err(dev, "sdw_bus_master_add fail: %d\n", ret);
return ret;
}
if (bus->prop.hw_disabled)
dev_info(dev,
"SoundWire master %d is disabled, will be ignored\n",
bus->link_id);
/*
* Ignore BIOS err_threshold, it's a really bad idea when dealing
* with multiple hardware synchronized links
*/
bus->prop.err_threshold = 0;
return 0;
}
int intel_link_startup(struct auxiliary_device *auxdev)
{
struct device *dev = &auxdev->dev;
struct sdw_cdns *cdns = auxiliary_get_drvdata(auxdev);
struct sdw_intel *sdw = cdns_to_intel(cdns);
struct sdw_bus *bus = &cdns->bus;
int link_flags;
bool multi_link;
u32 clock_stop_quirks;
int ret;
if (bus->prop.hw_disabled) {
dev_info(dev,
"SoundWire master %d is disabled, ignoring\n",
sdw->instance);
return 0;
}
link_flags = md_flags >> (bus->link_id * 8);
multi_link = !(link_flags & SDW_INTEL_MASTER_DISABLE_MULTI_LINK);
if (!multi_link) {
dev_dbg(dev, "Multi-link is disabled\n");
} else {
/*
* hardware-based synchronization is required regardless
* of the number of segments used by a stream: SSP-based
* synchronization is gated by gsync when the multi-master
* mode is set.
*/
bus->hw_sync_min_links = 1;
}
bus->multi_link = multi_link;
/* Initialize shim, controller */
ret = intel_link_power_up(sdw);
if (ret)
goto err_init;
/* Register DAIs */
ret = intel_register_dai(sdw);
if (ret) {
dev_err(dev, "DAI registration failed: %d\n", ret);
goto err_power_up;
}
intel_debugfs_init(sdw);
/* start bus */
ret = intel_start_bus(sdw);
if (ret) {
dev_err(dev, "bus start failed: %d\n", ret);
goto err_power_up;
}
/* Enable runtime PM */
if (!(link_flags & SDW_INTEL_MASTER_DISABLE_PM_RUNTIME)) {
pm_runtime_set_autosuspend_delay(dev,
INTEL_MASTER_SUSPEND_DELAY_MS);
pm_runtime_use_autosuspend(dev);
pm_runtime_mark_last_busy(dev);
pm_runtime_set_active(dev);
pm_runtime_enable(dev);
}
clock_stop_quirks = sdw->link_res->clock_stop_quirks;
if (clock_stop_quirks & SDW_INTEL_CLK_STOP_NOT_ALLOWED) {
/*
* To keep the clock running we need to prevent
* pm_runtime suspend from happening by increasing the
* reference count.
* This quirk is specified by the parent PCI device in
* case of specific latency requirements. It will have
* no effect if pm_runtime is disabled by the user via
* a module parameter for testing purposes.
*/
pm_runtime_get_noresume(dev);
}
/*
* The runtime PM status of Slave devices is "Unsupported"
* until they report as ATTACHED. If they don't, e.g. because
* there are no Slave devices populated or if the power-on is
* delayed or dependent on a power switch, the Master will
* remain active and prevent its parent from suspending.
*
* Conditionally force the pm_runtime core to re-evaluate the
* Master status in the absence of any Slave activity. A quirk
* is provided to e.g. deal with Slaves that may be powered on
* with a delay. A more complete solution would require the
* definition of Master properties.
*/
if (!(link_flags & SDW_INTEL_MASTER_DISABLE_PM_RUNTIME_IDLE))
pm_runtime_idle(dev);
soundwire: intel: skip suspend/resume/wake when link was not started The SoundWire Linux devices are created purely based on information provided by platform firmware (e.g. ACPI DSDT table). When the kernel finds a matching driver for the device address (_ADR), the probe will initialize required data structures and initialize pm ops. When the SoundWire link is started at a later point, the physical devices will synchronize on the SoundWire frames and report their attachment status, thereby triggering the enumeration and initialization of device registers. This two-step solution was a conscious design decision to allow e.g. a driver to use sideband mechanisms to turn power rails on. This can also allow OEMs to describe multiple platforms with the same DSDT table, the devices that are not physically present in hardware. The drawback of this approach is a bit of confusion, with more devices than are actually present in hardware. This results in 'ghost' devices, for which the driver successfully probes, but that will not generate any traffic on the bus. suspend-resume transitions are handled by drivers, and skipped when the devices are not physically present. This patch provides a work-around for a second-level of confusion in platform firmware: some platforms only use HDaudio links, but nevertheless expose SoundWire 'ghost' devices. This results in error messages in the Intel driver while trying to suspend/resume these links. The simplest solution is to add a boolean status flag to skip all suspend/resume/wake sequences if the link was never started. Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Link: https://lore.kernel.org/r/20210818024954.16873-3-yung-chuan.liao@linux.intel.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2021-08-18 10:49:53 +08:00
sdw->startup_done = true;
return 0;
err_power_up:
intel_link_power_down(sdw);
err_init:
return ret;
}
static void intel_link_remove(struct auxiliary_device *auxdev)
{
struct sdw_cdns *cdns = auxiliary_get_drvdata(auxdev);
struct sdw_intel *sdw = cdns_to_intel(cdns);
struct sdw_bus *bus = &cdns->bus;
/*
* Since pm_runtime is already disabled, we don't decrease
* the refcount when the clock_stop_quirk is
* SDW_INTEL_CLK_STOP_NOT_ALLOWED
*/
if (!bus->prop.hw_disabled) {
intel_debugfs_exit(sdw);
sdw_cdns_enable_interrupt(cdns, false);
}
sdw_bus_master_delete(bus);
}
int intel_link_process_wakeen_event(struct auxiliary_device *auxdev)
{
struct device *dev = &auxdev->dev;
struct sdw_intel *sdw;
struct sdw_bus *bus;
sdw = auxiliary_get_drvdata(auxdev);
bus = &sdw->cdns.bus;
soundwire: intel: skip suspend/resume/wake when link was not started The SoundWire Linux devices are created purely based on information provided by platform firmware (e.g. ACPI DSDT table). When the kernel finds a matching driver for the device address (_ADR), the probe will initialize required data structures and initialize pm ops. When the SoundWire link is started at a later point, the physical devices will synchronize on the SoundWire frames and report their attachment status, thereby triggering the enumeration and initialization of device registers. This two-step solution was a conscious design decision to allow e.g. a driver to use sideband mechanisms to turn power rails on. This can also allow OEMs to describe multiple platforms with the same DSDT table, the devices that are not physically present in hardware. The drawback of this approach is a bit of confusion, with more devices than are actually present in hardware. This results in 'ghost' devices, for which the driver successfully probes, but that will not generate any traffic on the bus. suspend-resume transitions are handled by drivers, and skipped when the devices are not physically present. This patch provides a work-around for a second-level of confusion in platform firmware: some platforms only use HDaudio links, but nevertheless expose SoundWire 'ghost' devices. This results in error messages in the Intel driver while trying to suspend/resume these links. The simplest solution is to add a boolean status flag to skip all suspend/resume/wake sequences if the link was never started. Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Link: https://lore.kernel.org/r/20210818024954.16873-3-yung-chuan.liao@linux.intel.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2021-08-18 10:49:53 +08:00
if (bus->prop.hw_disabled || !sdw->startup_done) {
dev_dbg(dev, "SoundWire master %d is disabled or not-started, ignoring\n",
bus->link_id);
return 0;
}
if (!intel_shim_check_wake(sdw))
return 0;
/* disable WAKEEN interrupt ASAP to prevent interrupt flood */
intel_shim_wake(sdw, false);
/*
* resume the Master, which will generate a bus reset and result in
* Slaves re-attaching and be re-enumerated. The SoundWire physical
* device which generated the wake will trigger an interrupt, which
* will in turn cause the corresponding Linux Slave device to be
* resumed and the Slave codec driver to check the status.
*/
pm_request_resume(dev);
return 0;
}
/*
* PM calls
*/
soundwire: intel: conditionally exit clock stop mode on system suspend Intel stress tests reported issues with the clock stop mode, specifically when trying to do a system suspend while the link is already pm_runtime suspended. In this case, we need to disable the shim wake, but when the PCI parent device is also pm_runtime suspended the SHIM registers are not accessible. Since this is an invalid corner case, this patch suggests a pm_runtime resume of the entire bus to full power (parent+child devices) before the system suspend so that the shim wake can be disabled. Unlike the suspend operation, the .prepare callbacks are propagated from root device to leaf devices. By adding a .prepare callback at the SoundWire link level, we can double-check the pm_runtime status of the device as well as its parent PCI device. When the problematic configuration is detected, the device is pm_runtime resumed - which by construction also resume its parent. An additional loop is added to resume all child devices. In theory we only need to restart the link, but doing so will also cause the physical devices to synchronize and re-initialize, while their Linux devices remain pm_runtime suspended. It's simpler to make sure the codec devices are fully resumed so that we don't have to deal with zombie states. This additional loop could have been avoided by adding a .prepare callback in SoundWire codec drivers. Functionally this would have been equivalent. The rationale for implementing a loop at the link level is only to reduce the amount of code required to deal at the codec level with an Intel corner case - in other words keep codec drivers independent from Intel platform-specific programming sequences. BugLink: https://github.com/thesofproject/linux/issues/2606 Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Link: https://lore.kernel.org/r/20210818024954.16873-4-yung-chuan.liao@linux.intel.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2021-08-18 10:49:54 +08:00
static int intel_resume_child_device(struct device *dev, void *data)
{
int ret;
struct sdw_slave *slave = dev_to_sdw_dev(dev);
if (!slave->probed) {
dev_dbg(dev, "skipping device, no probed driver\n");
soundwire: intel: conditionally exit clock stop mode on system suspend Intel stress tests reported issues with the clock stop mode, specifically when trying to do a system suspend while the link is already pm_runtime suspended. In this case, we need to disable the shim wake, but when the PCI parent device is also pm_runtime suspended the SHIM registers are not accessible. Since this is an invalid corner case, this patch suggests a pm_runtime resume of the entire bus to full power (parent+child devices) before the system suspend so that the shim wake can be disabled. Unlike the suspend operation, the .prepare callbacks are propagated from root device to leaf devices. By adding a .prepare callback at the SoundWire link level, we can double-check the pm_runtime status of the device as well as its parent PCI device. When the problematic configuration is detected, the device is pm_runtime resumed - which by construction also resume its parent. An additional loop is added to resume all child devices. In theory we only need to restart the link, but doing so will also cause the physical devices to synchronize and re-initialize, while their Linux devices remain pm_runtime suspended. It's simpler to make sure the codec devices are fully resumed so that we don't have to deal with zombie states. This additional loop could have been avoided by adding a .prepare callback in SoundWire codec drivers. Functionally this would have been equivalent. The rationale for implementing a loop at the link level is only to reduce the amount of code required to deal at the codec level with an Intel corner case - in other words keep codec drivers independent from Intel platform-specific programming sequences. BugLink: https://github.com/thesofproject/linux/issues/2606 Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Link: https://lore.kernel.org/r/20210818024954.16873-4-yung-chuan.liao@linux.intel.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2021-08-18 10:49:54 +08:00
return 0;
}
if (!slave->dev_num_sticky) {
dev_dbg(dev, "skipping device, never detected on bus\n");
soundwire: intel: conditionally exit clock stop mode on system suspend Intel stress tests reported issues with the clock stop mode, specifically when trying to do a system suspend while the link is already pm_runtime suspended. In this case, we need to disable the shim wake, but when the PCI parent device is also pm_runtime suspended the SHIM registers are not accessible. Since this is an invalid corner case, this patch suggests a pm_runtime resume of the entire bus to full power (parent+child devices) before the system suspend so that the shim wake can be disabled. Unlike the suspend operation, the .prepare callbacks are propagated from root device to leaf devices. By adding a .prepare callback at the SoundWire link level, we can double-check the pm_runtime status of the device as well as its parent PCI device. When the problematic configuration is detected, the device is pm_runtime resumed - which by construction also resume its parent. An additional loop is added to resume all child devices. In theory we only need to restart the link, but doing so will also cause the physical devices to synchronize and re-initialize, while their Linux devices remain pm_runtime suspended. It's simpler to make sure the codec devices are fully resumed so that we don't have to deal with zombie states. This additional loop could have been avoided by adding a .prepare callback in SoundWire codec drivers. Functionally this would have been equivalent. The rationale for implementing a loop at the link level is only to reduce the amount of code required to deal at the codec level with an Intel corner case - in other words keep codec drivers independent from Intel platform-specific programming sequences. BugLink: https://github.com/thesofproject/linux/issues/2606 Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Link: https://lore.kernel.org/r/20210818024954.16873-4-yung-chuan.liao@linux.intel.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2021-08-18 10:49:54 +08:00
return 0;
}
ret = pm_request_resume(dev);
if (ret < 0)
dev_err(dev, "%s: pm_request_resume failed: %d\n", __func__, ret);
return ret;
}
static int __maybe_unused intel_pm_prepare(struct device *dev)
{
struct sdw_cdns *cdns = dev_get_drvdata(dev);
struct sdw_intel *sdw = cdns_to_intel(cdns);
struct sdw_bus *bus = &cdns->bus;
u32 clock_stop_quirks;
int ret;
soundwire: intel: conditionally exit clock stop mode on system suspend Intel stress tests reported issues with the clock stop mode, specifically when trying to do a system suspend while the link is already pm_runtime suspended. In this case, we need to disable the shim wake, but when the PCI parent device is also pm_runtime suspended the SHIM registers are not accessible. Since this is an invalid corner case, this patch suggests a pm_runtime resume of the entire bus to full power (parent+child devices) before the system suspend so that the shim wake can be disabled. Unlike the suspend operation, the .prepare callbacks are propagated from root device to leaf devices. By adding a .prepare callback at the SoundWire link level, we can double-check the pm_runtime status of the device as well as its parent PCI device. When the problematic configuration is detected, the device is pm_runtime resumed - which by construction also resume its parent. An additional loop is added to resume all child devices. In theory we only need to restart the link, but doing so will also cause the physical devices to synchronize and re-initialize, while their Linux devices remain pm_runtime suspended. It's simpler to make sure the codec devices are fully resumed so that we don't have to deal with zombie states. This additional loop could have been avoided by adding a .prepare callback in SoundWire codec drivers. Functionally this would have been equivalent. The rationale for implementing a loop at the link level is only to reduce the amount of code required to deal at the codec level with an Intel corner case - in other words keep codec drivers independent from Intel platform-specific programming sequences. BugLink: https://github.com/thesofproject/linux/issues/2606 Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Link: https://lore.kernel.org/r/20210818024954.16873-4-yung-chuan.liao@linux.intel.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2021-08-18 10:49:54 +08:00
if (bus->prop.hw_disabled || !sdw->startup_done) {
dev_dbg(dev, "SoundWire master %d is disabled or not-started, ignoring\n",
bus->link_id);
return 0;
}
clock_stop_quirks = sdw->link_res->clock_stop_quirks;
if (pm_runtime_suspended(dev) &&
pm_runtime_suspended(dev->parent) &&
((clock_stop_quirks & SDW_INTEL_CLK_STOP_BUS_RESET) ||
!clock_stop_quirks)) {
/*
* if we've enabled clock stop, and the parent is suspended, the SHIM registers
* are not accessible and the shim wake cannot be disabled.
* The only solution is to resume the entire bus to full power
*/
/*
* If any operation in this block fails, we keep going since we don't want
* to prevent system suspend from happening and errors should be recoverable
* on resume.
*/
/*
* first resume the device for this link. This will also by construction
* resume the PCI parent device.
*/
ret = pm_request_resume(dev);
if (ret < 0) {
dev_err(dev, "%s: pm_request_resume failed: %d\n", __func__, ret);
return 0;
}
/*
* Continue resuming the entire bus (parent + child devices) to exit
* the clock stop mode. If there are no devices connected on this link
* this is a no-op.
* The resume to full power could have been implemented with a .prepare
* step in SoundWire codec drivers. This would however require a lot
* of code to handle an Intel-specific corner case. It is simpler in
* practice to add a loop at the link level.
*/
ret = device_for_each_child(bus->dev, NULL, intel_resume_child_device);
if (ret < 0)
dev_err(dev, "%s: intel_resume_child_device failed: %d\n", __func__, ret);
}
return 0;
}
static int __maybe_unused intel_suspend(struct device *dev)
{
struct sdw_cdns *cdns = dev_get_drvdata(dev);
struct sdw_intel *sdw = cdns_to_intel(cdns);
struct sdw_bus *bus = &cdns->bus;
u32 clock_stop_quirks;
int ret;
soundwire: intel: skip suspend/resume/wake when link was not started The SoundWire Linux devices are created purely based on information provided by platform firmware (e.g. ACPI DSDT table). When the kernel finds a matching driver for the device address (_ADR), the probe will initialize required data structures and initialize pm ops. When the SoundWire link is started at a later point, the physical devices will synchronize on the SoundWire frames and report their attachment status, thereby triggering the enumeration and initialization of device registers. This two-step solution was a conscious design decision to allow e.g. a driver to use sideband mechanisms to turn power rails on. This can also allow OEMs to describe multiple platforms with the same DSDT table, the devices that are not physically present in hardware. The drawback of this approach is a bit of confusion, with more devices than are actually present in hardware. This results in 'ghost' devices, for which the driver successfully probes, but that will not generate any traffic on the bus. suspend-resume transitions are handled by drivers, and skipped when the devices are not physically present. This patch provides a work-around for a second-level of confusion in platform firmware: some platforms only use HDaudio links, but nevertheless expose SoundWire 'ghost' devices. This results in error messages in the Intel driver while trying to suspend/resume these links. The simplest solution is to add a boolean status flag to skip all suspend/resume/wake sequences if the link was never started. Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Link: https://lore.kernel.org/r/20210818024954.16873-3-yung-chuan.liao@linux.intel.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2021-08-18 10:49:53 +08:00
if (bus->prop.hw_disabled || !sdw->startup_done) {
dev_dbg(dev, "SoundWire master %d is disabled or not-started, ignoring\n",
bus->link_id);
return 0;
}
if (pm_runtime_suspended(dev)) {
dev_dbg(dev, "pm_runtime status: suspended\n");
clock_stop_quirks = sdw->link_res->clock_stop_quirks;
soundwire: intel: conditionally exit clock stop mode on system suspend Intel stress tests reported issues with the clock stop mode, specifically when trying to do a system suspend while the link is already pm_runtime suspended. In this case, we need to disable the shim wake, but when the PCI parent device is also pm_runtime suspended the SHIM registers are not accessible. Since this is an invalid corner case, this patch suggests a pm_runtime resume of the entire bus to full power (parent+child devices) before the system suspend so that the shim wake can be disabled. Unlike the suspend operation, the .prepare callbacks are propagated from root device to leaf devices. By adding a .prepare callback at the SoundWire link level, we can double-check the pm_runtime status of the device as well as its parent PCI device. When the problematic configuration is detected, the device is pm_runtime resumed - which by construction also resume its parent. An additional loop is added to resume all child devices. In theory we only need to restart the link, but doing so will also cause the physical devices to synchronize and re-initialize, while their Linux devices remain pm_runtime suspended. It's simpler to make sure the codec devices are fully resumed so that we don't have to deal with zombie states. This additional loop could have been avoided by adding a .prepare callback in SoundWire codec drivers. Functionally this would have been equivalent. The rationale for implementing a loop at the link level is only to reduce the amount of code required to deal at the codec level with an Intel corner case - in other words keep codec drivers independent from Intel platform-specific programming sequences. BugLink: https://github.com/thesofproject/linux/issues/2606 Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Link: https://lore.kernel.org/r/20210818024954.16873-4-yung-chuan.liao@linux.intel.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2021-08-18 10:49:54 +08:00
if ((clock_stop_quirks & SDW_INTEL_CLK_STOP_BUS_RESET) ||
!clock_stop_quirks) {
if (pm_runtime_suspended(dev->parent)) {
/*
* paranoia check: this should not happen with the .prepare
* resume to full power
*/
dev_err(dev, "%s: invalid config: parent is suspended\n", __func__);
} else {
intel_shim_wake(sdw, false);
}
}
return 0;
}
ret = intel_stop_bus(sdw, false);
if (ret < 0) {
dev_err(dev, "%s: cannot stop bus: %d\n", __func__, ret);
return ret;
}
return 0;
}
static int __maybe_unused intel_suspend_runtime(struct device *dev)
{
struct sdw_cdns *cdns = dev_get_drvdata(dev);
struct sdw_intel *sdw = cdns_to_intel(cdns);
struct sdw_bus *bus = &cdns->bus;
u32 clock_stop_quirks;
int ret;
soundwire: intel: skip suspend/resume/wake when link was not started The SoundWire Linux devices are created purely based on information provided by platform firmware (e.g. ACPI DSDT table). When the kernel finds a matching driver for the device address (_ADR), the probe will initialize required data structures and initialize pm ops. When the SoundWire link is started at a later point, the physical devices will synchronize on the SoundWire frames and report their attachment status, thereby triggering the enumeration and initialization of device registers. This two-step solution was a conscious design decision to allow e.g. a driver to use sideband mechanisms to turn power rails on. This can also allow OEMs to describe multiple platforms with the same DSDT table, the devices that are not physically present in hardware. The drawback of this approach is a bit of confusion, with more devices than are actually present in hardware. This results in 'ghost' devices, for which the driver successfully probes, but that will not generate any traffic on the bus. suspend-resume transitions are handled by drivers, and skipped when the devices are not physically present. This patch provides a work-around for a second-level of confusion in platform firmware: some platforms only use HDaudio links, but nevertheless expose SoundWire 'ghost' devices. This results in error messages in the Intel driver while trying to suspend/resume these links. The simplest solution is to add a boolean status flag to skip all suspend/resume/wake sequences if the link was never started. Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Link: https://lore.kernel.org/r/20210818024954.16873-3-yung-chuan.liao@linux.intel.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2021-08-18 10:49:53 +08:00
if (bus->prop.hw_disabled || !sdw->startup_done) {
dev_dbg(dev, "SoundWire master %d is disabled or not-started, ignoring\n",
bus->link_id);
return 0;
}
clock_stop_quirks = sdw->link_res->clock_stop_quirks;
if (clock_stop_quirks & SDW_INTEL_CLK_STOP_TEARDOWN) {
ret = intel_stop_bus(sdw, false);
if (ret < 0) {
dev_err(dev, "%s: cannot stop bus during teardown: %d\n",
__func__, ret);
return ret;
}
} else if (clock_stop_quirks & SDW_INTEL_CLK_STOP_BUS_RESET || !clock_stop_quirks) {
ret = intel_stop_bus(sdw, true);
if (ret < 0) {
dev_err(dev, "%s: cannot stop bus during clock_stop: %d\n",
__func__, ret);
return ret;
}
} else {
dev_err(dev, "%s clock_stop_quirks %x unsupported\n",
__func__, clock_stop_quirks);
ret = -EINVAL;
}
return ret;
}
static int __maybe_unused intel_resume(struct device *dev)
{
struct sdw_cdns *cdns = dev_get_drvdata(dev);
struct sdw_intel *sdw = cdns_to_intel(cdns);
struct sdw_bus *bus = &cdns->bus;
int link_flags;
int ret;
soundwire: intel: skip suspend/resume/wake when link was not started The SoundWire Linux devices are created purely based on information provided by platform firmware (e.g. ACPI DSDT table). When the kernel finds a matching driver for the device address (_ADR), the probe will initialize required data structures and initialize pm ops. When the SoundWire link is started at a later point, the physical devices will synchronize on the SoundWire frames and report their attachment status, thereby triggering the enumeration and initialization of device registers. This two-step solution was a conscious design decision to allow e.g. a driver to use sideband mechanisms to turn power rails on. This can also allow OEMs to describe multiple platforms with the same DSDT table, the devices that are not physically present in hardware. The drawback of this approach is a bit of confusion, with more devices than are actually present in hardware. This results in 'ghost' devices, for which the driver successfully probes, but that will not generate any traffic on the bus. suspend-resume transitions are handled by drivers, and skipped when the devices are not physically present. This patch provides a work-around for a second-level of confusion in platform firmware: some platforms only use HDaudio links, but nevertheless expose SoundWire 'ghost' devices. This results in error messages in the Intel driver while trying to suspend/resume these links. The simplest solution is to add a boolean status flag to skip all suspend/resume/wake sequences if the link was never started. Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Link: https://lore.kernel.org/r/20210818024954.16873-3-yung-chuan.liao@linux.intel.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2021-08-18 10:49:53 +08:00
if (bus->prop.hw_disabled || !sdw->startup_done) {
dev_dbg(dev, "SoundWire master %d is disabled or not-started, ignoring\n",
bus->link_id);
return 0;
}
link_flags = md_flags >> (bus->link_id * 8);
if (pm_runtime_suspended(dev)) {
dev_dbg(dev, "pm_runtime status was suspended, forcing active\n");
/* follow required sequence from runtime_pm.rst */
pm_runtime_disable(dev);
pm_runtime_set_active(dev);
pm_runtime_mark_last_busy(dev);
pm_runtime_enable(dev);
link_flags = md_flags >> (bus->link_id * 8);
if (!(link_flags & SDW_INTEL_MASTER_DISABLE_PM_RUNTIME_IDLE))
pm_runtime_idle(dev);
}
ret = intel_link_power_up(sdw);
if (ret) {
dev_err(dev, "%s failed: %d\n", __func__, ret);
return ret;
}
/*
* make sure all Slaves are tagged as UNATTACHED and provide
* reason for reinitialization
*/
sdw_clear_slave_status(bus, SDW_UNATTACH_REQUEST_MASTER_RESET);
ret = intel_start_bus(sdw);
if (ret < 0) {
dev_err(dev, "cannot start bus during resume\n");
intel_link_power_down(sdw);
return ret;
}
/*
* after system resume, the pm_runtime suspend() may kick in
* during the enumeration, before any children device force the
* master device to remain active. Using pm_runtime_get()
* routines is not really possible, since it'd prevent the
* master from suspending.
* A reasonable compromise is to update the pm_runtime
* counters and delay the pm_runtime suspend by several
* seconds, by when all enumeration should be complete.
*/
pm_runtime_mark_last_busy(dev);
return 0;
}
static int __maybe_unused intel_resume_runtime(struct device *dev)
{
struct sdw_cdns *cdns = dev_get_drvdata(dev);
struct sdw_intel *sdw = cdns_to_intel(cdns);
struct sdw_bus *bus = &cdns->bus;
u32 clock_stop_quirks;
int ret;
soundwire: intel: skip suspend/resume/wake when link was not started The SoundWire Linux devices are created purely based on information provided by platform firmware (e.g. ACPI DSDT table). When the kernel finds a matching driver for the device address (_ADR), the probe will initialize required data structures and initialize pm ops. When the SoundWire link is started at a later point, the physical devices will synchronize on the SoundWire frames and report their attachment status, thereby triggering the enumeration and initialization of device registers. This two-step solution was a conscious design decision to allow e.g. a driver to use sideband mechanisms to turn power rails on. This can also allow OEMs to describe multiple platforms with the same DSDT table, the devices that are not physically present in hardware. The drawback of this approach is a bit of confusion, with more devices than are actually present in hardware. This results in 'ghost' devices, for which the driver successfully probes, but that will not generate any traffic on the bus. suspend-resume transitions are handled by drivers, and skipped when the devices are not physically present. This patch provides a work-around for a second-level of confusion in platform firmware: some platforms only use HDaudio links, but nevertheless expose SoundWire 'ghost' devices. This results in error messages in the Intel driver while trying to suspend/resume these links. The simplest solution is to add a boolean status flag to skip all suspend/resume/wake sequences if the link was never started. Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Link: https://lore.kernel.org/r/20210818024954.16873-3-yung-chuan.liao@linux.intel.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2021-08-18 10:49:53 +08:00
if (bus->prop.hw_disabled || !sdw->startup_done) {
dev_dbg(dev, "SoundWire master %d is disabled or not-started, ignoring\n",
bus->link_id);
return 0;
}
soundwire: intel: disable WAKEEN in pm_runtime resume When the manager device is pm_runtime resumed, we see a series of spurious wakes and attempts to resume the same device: soundwire_intel.link.0: intel_resume_runtime: start soundwire_intel.link.0: intel_link_power_up: powering up all links soundwire_intel.link.0: intel_link_power_up: first link up, programming SYNCPRD soundwire_intel.link.0: intel_shim_wake: WAKEEN disabled for link 0 soundwire_intel.link.0: intel_link_process_wakeen_event: pm_request_resume start soundwire_intel.link.0: intel_link_process_wakeen_event: pm_request_resume done soundwire_intel.link.0: intel_shim_wake: WAKEEN disabled for link 0 soundwire_intel.link.0: intel_link_process_wakeen_event: pm_request_resume start soundwire_intel.link.0: intel_link_process_wakeen_event: pm_request_resume done This sequence does not break anything but is totally unnecessary. Currently the wakes are only disabled after the peripheral generates a wake, e.g. for jack detection. If the resume is initiated by the host drivers as a result of userspace actions (play/record typically), we need to disable wake detection as well. Doing so prevents the spurious wakes and calls to pm_request_resume(). Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Reviewed-by: Rander Wang <rander.wang@intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Link: https://lore.kernel.org/r/20220420023241.14335-3-yung-chuan.liao@linux.intel.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2022-04-20 10:32:40 +08:00
/* unconditionally disable WAKEEN interrupt */
intel_shim_wake(sdw, false);
clock_stop_quirks = sdw->link_res->clock_stop_quirks;
if (clock_stop_quirks & SDW_INTEL_CLK_STOP_TEARDOWN) {
ret = intel_link_power_up(sdw);
if (ret) {
dev_err(dev, "%s: power_up failed after teardown: %d\n", __func__, ret);
return ret;
}
/*
* make sure all Slaves are tagged as UNATTACHED and provide
* reason for reinitialization
*/
sdw_clear_slave_status(bus, SDW_UNATTACH_REQUEST_MASTER_RESET);
ret = intel_start_bus(sdw);
if (ret < 0) {
dev_err(dev, "%s: cannot start bus after teardown: %d\n", __func__, ret);
intel_link_power_down(sdw);
return ret;
}
soundwire: cadence: add paranoid check on self-clearing bits The Cadence IP exposes a small number of self-clearing bits in the MCP_CONTROL and MCP_CONFIG_UPDATE registers. We currently do not check that those bits are indeed cleared, e.g. during resume operations. That could lead to resuming peripheral devices too early. In addition, if we happen to read these registers, update one of the fields and write the register back, we may be writing stale data that might have been cleared in hardware. These sort of race conditions could lead to e.g. doing a hw_reset twice or stopping a clock that just restarted. There is no clear way of avoiding these potential race conditions other than making sure that these registers fields are cleared before any read-modify-write sequence. If we detect this sort of errors, we only log them since there is no clear recovery possible. The only way out is likely to restart the IP with a suspend/resume cycle. Note that the checks are performed before updating the registers, as well as after the Intel 'sync go' sequence in multi-link mode. That should cover both the start and end of suspend/resume hardware configurations. The Multi-Master mode gates the configuration updates until the 'sync go' signal is asserted, so we only check on init and after the end of the 'sync go' sequence. The duration of the usleep_range() was defined by the GSYNC frequency used in multi-master mode. With a 4kHz frequency, any configuration change might be deferred by up to 250us. Extending the range to 1000-1500us should guarantee that the configuration change is completed without any significant impact on the overall resume time. Suggested-by: Bard Liao <yung-chuan.liao@linux.intel.com> Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Link: https://lore.kernel.org/r/20210714051349.13064-1-yung-chuan.liao@linux.intel.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2021-07-14 13:13:49 +08:00
} else if (clock_stop_quirks & SDW_INTEL_CLK_STOP_BUS_RESET) {
ret = intel_link_power_up(sdw);
if (ret) {
dev_err(dev, "%s: power_up failed after bus reset: %d\n", __func__, ret);
return ret;
}
ret = intel_start_bus_after_reset(sdw);
if (ret < 0) {
dev_err(dev, "%s: cannot start bus after reset: %d\n", __func__, ret);
intel_link_power_down(sdw);
return ret;
}
} else if (!clock_stop_quirks) {
intel_check_clock_stop(sdw);
ret = intel_link_power_up(sdw);
if (ret) {
dev_err(dev, "%s: power_up failed: %d\n", __func__, ret);
return ret;
}
ret = intel_start_bus_after_clock_stop(sdw);
if (ret < 0) {
dev_err(dev, "%s: cannot start bus after clock stop: %d\n", __func__, ret);
intel_link_power_down(sdw);
return ret;
}
} else {
dev_err(dev, "%s: clock_stop_quirks %x unsupported\n",
__func__, clock_stop_quirks);
ret = -EINVAL;
}
return ret;
}
static const struct dev_pm_ops intel_pm = {
soundwire: intel: conditionally exit clock stop mode on system suspend Intel stress tests reported issues with the clock stop mode, specifically when trying to do a system suspend while the link is already pm_runtime suspended. In this case, we need to disable the shim wake, but when the PCI parent device is also pm_runtime suspended the SHIM registers are not accessible. Since this is an invalid corner case, this patch suggests a pm_runtime resume of the entire bus to full power (parent+child devices) before the system suspend so that the shim wake can be disabled. Unlike the suspend operation, the .prepare callbacks are propagated from root device to leaf devices. By adding a .prepare callback at the SoundWire link level, we can double-check the pm_runtime status of the device as well as its parent PCI device. When the problematic configuration is detected, the device is pm_runtime resumed - which by construction also resume its parent. An additional loop is added to resume all child devices. In theory we only need to restart the link, but doing so will also cause the physical devices to synchronize and re-initialize, while their Linux devices remain pm_runtime suspended. It's simpler to make sure the codec devices are fully resumed so that we don't have to deal with zombie states. This additional loop could have been avoided by adding a .prepare callback in SoundWire codec drivers. Functionally this would have been equivalent. The rationale for implementing a loop at the link level is only to reduce the amount of code required to deal at the codec level with an Intel corner case - in other words keep codec drivers independent from Intel platform-specific programming sequences. BugLink: https://github.com/thesofproject/linux/issues/2606 Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com> Link: https://lore.kernel.org/r/20210818024954.16873-4-yung-chuan.liao@linux.intel.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2021-08-18 10:49:54 +08:00
.prepare = intel_pm_prepare,
SET_SYSTEM_SLEEP_PM_OPS(intel_suspend, intel_resume)
SET_RUNTIME_PM_OPS(intel_suspend_runtime, intel_resume_runtime, NULL)
};
static const struct auxiliary_device_id intel_link_id_table[] = {
{ .name = "soundwire_intel.link" },
{},
};
MODULE_DEVICE_TABLE(auxiliary, intel_link_id_table);
static struct auxiliary_driver sdw_intel_drv = {
.probe = intel_link_probe,
.remove = intel_link_remove,
.driver = {
/* auxiliary_driver_register() sets .name to be the modname */
.pm = &intel_pm,
},
.id_table = intel_link_id_table
};
module_auxiliary_driver(sdw_intel_drv);
MODULE_LICENSE("Dual BSD/GPL");
MODULE_DESCRIPTION("Intel Soundwire Link Driver");