linux/drivers/soundwire/cadence_master.c

1899 lines
49 KiB
C
Raw Normal View History

// SPDX-License-Identifier: (GPL-2.0 OR BSD-3-Clause)
// Copyright(c) 2015-17 Intel Corporation.
/*
* Cadence SoundWire Master module
* Used by Master driver
*/
#include <linux/delay.h>
#include <linux/device.h>
#include <linux/debugfs.h>
#include <linux/interrupt.h>
#include <linux/io.h>
#include <linux/module.h>
#include <linux/mod_devicetable.h>
#include <linux/pm_runtime.h>
#include <linux/soundwire/sdw_registers.h>
#include <linux/soundwire/sdw.h>
#include <sound/pcm_params.h>
#include <sound/soc.h>
#include <linux/workqueue.h>
#include "bus.h"
#include "cadence_master.h"
static int interrupt_mask;
module_param_named(cnds_mcp_int_mask, interrupt_mask, int, 0444);
MODULE_PARM_DESC(cdns_mcp_int_mask, "Cadence MCP IntMask");
#define CDNS_MCP_CONFIG 0x0
#define CDNS_MCP_CONFIG_BUS_REL BIT(6)
#define CDNS_IP_MCP_CONFIG 0x0 /* IP offset added at run-time */
#define CDNS_IP_MCP_CONFIG_MCMD_RETRY GENMASK(27, 24)
#define CDNS_IP_MCP_CONFIG_MPREQ_DELAY GENMASK(20, 16)
#define CDNS_IP_MCP_CONFIG_MMASTER BIT(7)
#define CDNS_IP_MCP_CONFIG_SNIFFER BIT(5)
#define CDNS_IP_MCP_CONFIG_CMD BIT(3)
#define CDNS_IP_MCP_CONFIG_OP GENMASK(2, 0)
#define CDNS_IP_MCP_CONFIG_OP_NORMAL 0
#define CDNS_MCP_CONTROL 0x4
#define CDNS_MCP_CONTROL_CMD_RST BIT(7)
#define CDNS_MCP_CONTROL_SOFT_RST BIT(6)
#define CDNS_MCP_CONTROL_HW_RST BIT(4)
#define CDNS_MCP_CONTROL_CLK_STOP_CLR BIT(2)
#define CDNS_IP_MCP_CONTROL 0x4 /* IP offset added at run-time */
#define CDNS_IP_MCP_CONTROL_RST_DELAY GENMASK(10, 8)
#define CDNS_IP_MCP_CONTROL_SW_RST BIT(5)
#define CDNS_IP_MCP_CONTROL_CLK_PAUSE BIT(3)
#define CDNS_IP_MCP_CONTROL_CMD_ACCEPT BIT(1)
#define CDNS_IP_MCP_CONTROL_BLOCK_WAKEUP BIT(0)
#define CDNS_IP_MCP_CMDCTRL 0x8 /* IP offset added at run-time */
#define CDNS_IP_MCP_CMDCTRL_INSERT_PARITY_ERR BIT(2)
#define CDNS_MCP_SSPSTAT 0xC
#define CDNS_MCP_FRAME_SHAPE 0x10
#define CDNS_MCP_FRAME_SHAPE_INIT 0x14
#define CDNS_MCP_FRAME_SHAPE_COL_MASK GENMASK(2, 0)
#define CDNS_MCP_FRAME_SHAPE_ROW_MASK GENMASK(7, 3)
#define CDNS_MCP_CONFIG_UPDATE 0x18
#define CDNS_MCP_CONFIG_UPDATE_BIT BIT(0)
#define CDNS_MCP_PHYCTRL 0x1C
#define CDNS_MCP_SSP_CTRL0 0x20
#define CDNS_MCP_SSP_CTRL1 0x28
#define CDNS_MCP_CLK_CTRL0 0x30
#define CDNS_MCP_CLK_CTRL1 0x38
#define CDNS_MCP_CLK_MCLKD_MASK GENMASK(7, 0)
#define CDNS_MCP_STAT 0x40
#define CDNS_MCP_STAT_ACTIVE_BANK BIT(20)
#define CDNS_MCP_STAT_CLK_STOP BIT(16)
#define CDNS_MCP_INTSTAT 0x44
#define CDNS_MCP_INTMASK 0x48
#define CDNS_MCP_INT_IRQ BIT(31)
#define CDNS_MCP_INT_RESERVED1 GENMASK(30, 17)
#define CDNS_MCP_INT_WAKEUP BIT(16)
#define CDNS_MCP_INT_SLAVE_RSVD BIT(15)
#define CDNS_MCP_INT_SLAVE_ALERT BIT(14)
#define CDNS_MCP_INT_SLAVE_ATTACH BIT(13)
#define CDNS_MCP_INT_SLAVE_NATTACH BIT(12)
#define CDNS_MCP_INT_SLAVE_MASK GENMASK(15, 12)
#define CDNS_MCP_INT_DPINT BIT(11)
#define CDNS_MCP_INT_CTRL_CLASH BIT(10)
#define CDNS_MCP_INT_DATA_CLASH BIT(9)
#define CDNS_MCP_INT_PARITY BIT(8)
#define CDNS_MCP_INT_CMD_ERR BIT(7)
#define CDNS_MCP_INT_RESERVED2 GENMASK(6, 4)
#define CDNS_MCP_INT_RX_NE BIT(3)
#define CDNS_MCP_INT_RX_WL BIT(2)
#define CDNS_MCP_INT_TXE BIT(1)
#define CDNS_MCP_INT_TXF BIT(0)
#define CDNS_MCP_INT_RESERVED (CDNS_MCP_INT_RESERVED1 | CDNS_MCP_INT_RESERVED2)
#define CDNS_MCP_INTSET 0x4C
#define CDNS_MCP_SLAVE_STAT 0x50
#define CDNS_MCP_SLAVE_STAT_MASK GENMASK(1, 0)
#define CDNS_MCP_SLAVE_INTSTAT0 0x54
#define CDNS_MCP_SLAVE_INTSTAT1 0x58
#define CDNS_MCP_SLAVE_INTSTAT_NPRESENT BIT(0)
#define CDNS_MCP_SLAVE_INTSTAT_ATTACHED BIT(1)
#define CDNS_MCP_SLAVE_INTSTAT_ALERT BIT(2)
#define CDNS_MCP_SLAVE_INTSTAT_RESERVED BIT(3)
#define CDNS_MCP_SLAVE_STATUS_BITS GENMASK(3, 0)
#define CDNS_MCP_SLAVE_STATUS_NUM 4
#define CDNS_MCP_SLAVE_INTMASK0 0x5C
#define CDNS_MCP_SLAVE_INTMASK1 0x60
#define CDNS_MCP_SLAVE_INTMASK0_MASK GENMASK(31, 0)
#define CDNS_MCP_SLAVE_INTMASK1_MASK GENMASK(15, 0)
#define CDNS_MCP_PORT_INTSTAT 0x64
#define CDNS_MCP_PDI_STAT 0x6C
#define CDNS_MCP_FIFOLEVEL 0x78
#define CDNS_MCP_FIFOSTAT 0x7C
#define CDNS_MCP_RX_FIFO_AVAIL GENMASK(5, 0)
#define CDNS_IP_MCP_CMD_BASE 0x80 /* IP offset added at run-time */
#define CDNS_IP_MCP_RESP_BASE 0x80 /* IP offset added at run-time */
/* FIFO can hold 8 commands */
#define CDNS_MCP_CMD_LEN 8
#define CDNS_MCP_CMD_WORD_LEN 0x4
#define CDNS_MCP_CMD_SSP_TAG BIT(31)
#define CDNS_MCP_CMD_COMMAND GENMASK(30, 28)
#define CDNS_MCP_CMD_DEV_ADDR GENMASK(27, 24)
#define CDNS_MCP_CMD_REG_ADDR GENMASK(23, 8)
#define CDNS_MCP_CMD_REG_DATA GENMASK(7, 0)
#define CDNS_MCP_CMD_READ 2
#define CDNS_MCP_CMD_WRITE 3
#define CDNS_MCP_RESP_RDATA GENMASK(15, 8)
#define CDNS_MCP_RESP_ACK BIT(0)
#define CDNS_MCP_RESP_NACK BIT(1)
#define CDNS_DP_SIZE 128
#define CDNS_DPN_B0_CONFIG(n) (0x100 + CDNS_DP_SIZE * (n))
#define CDNS_DPN_B0_CH_EN(n) (0x104 + CDNS_DP_SIZE * (n))
#define CDNS_DPN_B0_SAMPLE_CTRL(n) (0x108 + CDNS_DP_SIZE * (n))
#define CDNS_DPN_B0_OFFSET_CTRL(n) (0x10C + CDNS_DP_SIZE * (n))
#define CDNS_DPN_B0_HCTRL(n) (0x110 + CDNS_DP_SIZE * (n))
#define CDNS_DPN_B0_ASYNC_CTRL(n) (0x114 + CDNS_DP_SIZE * (n))
#define CDNS_DPN_B1_CONFIG(n) (0x118 + CDNS_DP_SIZE * (n))
#define CDNS_DPN_B1_CH_EN(n) (0x11C + CDNS_DP_SIZE * (n))
#define CDNS_DPN_B1_SAMPLE_CTRL(n) (0x120 + CDNS_DP_SIZE * (n))
#define CDNS_DPN_B1_OFFSET_CTRL(n) (0x124 + CDNS_DP_SIZE * (n))
#define CDNS_DPN_B1_HCTRL(n) (0x128 + CDNS_DP_SIZE * (n))
#define CDNS_DPN_B1_ASYNC_CTRL(n) (0x12C + CDNS_DP_SIZE * (n))
#define CDNS_DPN_CONFIG_BPM BIT(18)
#define CDNS_DPN_CONFIG_BGC GENMASK(17, 16)
#define CDNS_DPN_CONFIG_WL GENMASK(12, 8)
#define CDNS_DPN_CONFIG_PORT_DAT GENMASK(3, 2)
#define CDNS_DPN_CONFIG_PORT_FLOW GENMASK(1, 0)
#define CDNS_DPN_SAMPLE_CTRL_SI GENMASK(15, 0)
#define CDNS_DPN_OFFSET_CTRL_1 GENMASK(7, 0)
#define CDNS_DPN_OFFSET_CTRL_2 GENMASK(15, 8)
#define CDNS_DPN_HCTRL_HSTOP GENMASK(3, 0)
#define CDNS_DPN_HCTRL_HSTART GENMASK(7, 4)
#define CDNS_DPN_HCTRL_LCTRL GENMASK(10, 8)
#define CDNS_PORTCTRL 0x130
#define CDNS_PORTCTRL_TEST_FAILED BIT(1)
#define CDNS_PORTCTRL_DIRN BIT(7)
#define CDNS_PORTCTRL_BANK_INVERT BIT(8)
#define CDNS_PORT_OFFSET 0x80
#define CDNS_PDI_CONFIG(n) (0x1100 + (n) * 16)
#define CDNS_PDI_CONFIG_SOFT_RESET BIT(24)
#define CDNS_PDI_CONFIG_CHANNEL GENMASK(15, 8)
#define CDNS_PDI_CONFIG_PORT GENMASK(4, 0)
/* Driver defaults */
#define CDNS_TX_TIMEOUT 500
#define CDNS_SCP_RX_FIFOLEVEL 0x2
/*
* register accessor helpers
*/
static inline u32 cdns_readl(struct sdw_cdns *cdns, int offset)
{
return readl(cdns->registers + offset);
}
static inline void cdns_writel(struct sdw_cdns *cdns, int offset, u32 value)
{
writel(value, cdns->registers + offset);
}
static inline u32 cdns_ip_readl(struct sdw_cdns *cdns, int offset)
{
return cdns_readl(cdns, cdns->ip_offset + offset);
}
static inline void cdns_ip_writel(struct sdw_cdns *cdns, int offset, u32 value)
{
return cdns_writel(cdns, cdns->ip_offset + offset, value);
}
static inline void cdns_updatel(struct sdw_cdns *cdns,
int offset, u32 mask, u32 val)
{
u32 tmp;
tmp = cdns_readl(cdns, offset);
tmp = (tmp & ~mask) | val;
cdns_writel(cdns, offset, tmp);
}
static inline void cdns_ip_updatel(struct sdw_cdns *cdns,
int offset, u32 mask, u32 val)
{
cdns_updatel(cdns, cdns->ip_offset + offset, mask, val);
}
static int cdns_set_wait(struct sdw_cdns *cdns, int offset, u32 mask, u32 value)
{
int timeout = 10;
u32 reg_read;
/* Wait for bit to be set */
do {
reg_read = readl(cdns->registers + offset);
if ((reg_read & mask) == value)
return 0;
timeout--;
usleep_range(50, 100);
} while (timeout != 0);
return -ETIMEDOUT;
}
static int cdns_clear_bit(struct sdw_cdns *cdns, int offset, u32 value)
{
writel(value, cdns->registers + offset);
/* Wait for bit to be self cleared */
return cdns_set_wait(cdns, offset, value, 0);
}
/*
* all changes to the MCP_CONFIG, MCP_CONTROL, MCP_CMDCTRL and MCP_PHYCTRL
* need to be confirmed with a write to MCP_CONFIG_UPDATE
*/
static int cdns_config_update(struct sdw_cdns *cdns)
{
int ret;
if (sdw_cdns_is_clock_stop(cdns)) {
dev_err(cdns->dev, "Cannot program MCP_CONFIG_UPDATE in ClockStopMode\n");
return -EINVAL;
}
ret = cdns_clear_bit(cdns, CDNS_MCP_CONFIG_UPDATE,
CDNS_MCP_CONFIG_UPDATE_BIT);
if (ret < 0)
dev_err(cdns->dev, "Config update timedout\n");
return ret;
}
/**
* sdw_cdns_config_update() - Update configurations
* @cdns: Cadence instance
*/
void sdw_cdns_config_update(struct sdw_cdns *cdns)
{
/* commit changes */
cdns_writel(cdns, CDNS_MCP_CONFIG_UPDATE, CDNS_MCP_CONFIG_UPDATE_BIT);
}
EXPORT_SYMBOL(sdw_cdns_config_update);
/**
* sdw_cdns_config_update_set_wait() - wait until configuration update bit is self-cleared
* @cdns: Cadence instance
*/
int sdw_cdns_config_update_set_wait(struct sdw_cdns *cdns)
{
/* the hardware recommendation is to wait at least 300us */
return cdns_set_wait(cdns, CDNS_MCP_CONFIG_UPDATE,
CDNS_MCP_CONFIG_UPDATE_BIT, 0);
}
EXPORT_SYMBOL(sdw_cdns_config_update_set_wait);
/*
* debugfs
*/
#ifdef CONFIG_DEBUG_FS
#define RD_BUF (2 * PAGE_SIZE)
static ssize_t cdns_sprintf(struct sdw_cdns *cdns,
char *buf, size_t pos, unsigned int reg)
{
return scnprintf(buf + pos, RD_BUF - pos,
"%4x\t%8x\n", reg, cdns_readl(cdns, reg));
}
static int cdns_reg_show(struct seq_file *s, void *data)
{
struct sdw_cdns *cdns = s->private;
char *buf;
ssize_t ret;
int num_ports;
int i, j;
buf = kzalloc(RD_BUF, GFP_KERNEL);
if (!buf)
return -ENOMEM;
ret = scnprintf(buf, RD_BUF, "Register Value\n");
ret += scnprintf(buf + ret, RD_BUF - ret, "\nMCP Registers\n");
/* 8 MCP registers */
for (i = CDNS_MCP_CONFIG; i <= CDNS_MCP_PHYCTRL; i += sizeof(u32))
ret += cdns_sprintf(cdns, buf, ret, i);
ret += scnprintf(buf + ret, RD_BUF - ret,
"\nStatus & Intr Registers\n");
/* 13 Status & Intr registers (offsets 0x70 and 0x74 not defined) */
for (i = CDNS_MCP_STAT; i <= CDNS_MCP_FIFOSTAT; i += sizeof(u32))
ret += cdns_sprintf(cdns, buf, ret, i);
ret += scnprintf(buf + ret, RD_BUF - ret,
"\nSSP & Clk ctrl Registers\n");
ret += cdns_sprintf(cdns, buf, ret, CDNS_MCP_SSP_CTRL0);
ret += cdns_sprintf(cdns, buf, ret, CDNS_MCP_SSP_CTRL1);
ret += cdns_sprintf(cdns, buf, ret, CDNS_MCP_CLK_CTRL0);
ret += cdns_sprintf(cdns, buf, ret, CDNS_MCP_CLK_CTRL1);
ret += scnprintf(buf + ret, RD_BUF - ret,
"\nDPn B0 Registers\n");
num_ports = cdns->num_ports;
for (i = 0; i < num_ports; i++) {
ret += scnprintf(buf + ret, RD_BUF - ret,
"\nDP-%d\n", i);
for (j = CDNS_DPN_B0_CONFIG(i);
j < CDNS_DPN_B0_ASYNC_CTRL(i); j += sizeof(u32))
ret += cdns_sprintf(cdns, buf, ret, j);
}
ret += scnprintf(buf + ret, RD_BUF - ret,
"\nDPn B1 Registers\n");
for (i = 0; i < num_ports; i++) {
ret += scnprintf(buf + ret, RD_BUF - ret,
"\nDP-%d\n", i);
for (j = CDNS_DPN_B1_CONFIG(i);
j < CDNS_DPN_B1_ASYNC_CTRL(i); j += sizeof(u32))
ret += cdns_sprintf(cdns, buf, ret, j);
}
ret += scnprintf(buf + ret, RD_BUF - ret,
"\nDPn Control Registers\n");
for (i = 0; i < num_ports; i++)
ret += cdns_sprintf(cdns, buf, ret,
CDNS_PORTCTRL + i * CDNS_PORT_OFFSET);
ret += scnprintf(buf + ret, RD_BUF - ret,
"\nPDIn Config Registers\n");
/* number of PDI and ports is interchangeable */
for (i = 0; i < num_ports; i++)
ret += cdns_sprintf(cdns, buf, ret, CDNS_PDI_CONFIG(i));
seq_printf(s, "%s", buf);
kfree(buf);
return 0;
}
DEFINE_SHOW_ATTRIBUTE(cdns_reg);
static int cdns_hw_reset(void *data, u64 value)
{
struct sdw_cdns *cdns = data;
int ret;
if (value != 1)
return -EINVAL;
/* Userspace changed the hardware state behind the kernel's back */
add_taint(TAINT_USER, LOCKDEP_STILL_OK);
ret = sdw_cdns_exit_reset(cdns);
dev_dbg(cdns->dev, "link hw_reset done: %d\n", ret);
return ret;
}
DEFINE_DEBUGFS_ATTRIBUTE(cdns_hw_reset_fops, NULL, cdns_hw_reset, "%llu\n");
static int cdns_parity_error_injection(void *data, u64 value)
{
struct sdw_cdns *cdns = data;
struct sdw_bus *bus;
int ret;
if (value != 1)
return -EINVAL;
bus = &cdns->bus;
/*
* Resume Master device. If this results in a bus reset, the
* Slave devices will re-attach and be re-enumerated.
*/
ret = pm_runtime_resume_and_get(bus->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;
}
/*
* wait long enough for Slave(s) to be in steady state. This
* does not need to be super precise.
*/
msleep(200);
/*
* Take the bus lock here to make sure that any bus transactions
* will be queued while we inject a parity error on a dummy read
*/
mutex_lock(&bus->bus_lock);
/* program hardware to inject parity error */
cdns_ip_updatel(cdns, CDNS_IP_MCP_CMDCTRL,
CDNS_IP_MCP_CMDCTRL_INSERT_PARITY_ERR,
CDNS_IP_MCP_CMDCTRL_INSERT_PARITY_ERR);
/* commit changes */
ret = cdns_clear_bit(cdns, CDNS_MCP_CONFIG_UPDATE, CDNS_MCP_CONFIG_UPDATE_BIT);
if (ret < 0)
goto unlock;
/* do a broadcast dummy read to avoid bus clashes */
ret = sdw_bread_no_pm_unlocked(&cdns->bus, 0xf, SDW_SCP_DEVID_0);
dev_info(cdns->dev, "parity error injection, read: %d\n", ret);
/* program hardware to disable parity error */
cdns_ip_updatel(cdns, CDNS_IP_MCP_CMDCTRL,
CDNS_IP_MCP_CMDCTRL_INSERT_PARITY_ERR,
0);
/* commit changes */
ret = cdns_clear_bit(cdns, CDNS_MCP_CONFIG_UPDATE, CDNS_MCP_CONFIG_UPDATE_BIT);
if (ret < 0)
goto unlock;
/* Userspace changed the hardware state behind the kernel's back */
add_taint(TAINT_USER, LOCKDEP_STILL_OK);
unlock:
/* Continue bus operation with parity error injection disabled */
mutex_unlock(&bus->bus_lock);
/*
* allow Master device to enter pm_runtime suspend. This may
* also result in Slave devices suspending.
*/
pm_runtime_mark_last_busy(bus->dev);
pm_runtime_put_autosuspend(bus->dev);
return 0;
}
DEFINE_DEBUGFS_ATTRIBUTE(cdns_parity_error_fops, NULL,
cdns_parity_error_injection, "%llu\n");
static int cdns_set_pdi_loopback_source(void *data, u64 value)
{
struct sdw_cdns *cdns = data;
unsigned int pdi_out_num = cdns->pcm.num_bd + cdns->pcm.num_out;
if (value > pdi_out_num)
return -EINVAL;
/* Userspace changed the hardware state behind the kernel's back */
add_taint(TAINT_USER, LOCKDEP_STILL_OK);
cdns->pdi_loopback_source = value;
return 0;
}
DEFINE_DEBUGFS_ATTRIBUTE(cdns_pdi_loopback_source_fops, NULL, cdns_set_pdi_loopback_source, "%llu\n");
static int cdns_set_pdi_loopback_target(void *data, u64 value)
{
struct sdw_cdns *cdns = data;
unsigned int pdi_in_num = cdns->pcm.num_bd + cdns->pcm.num_in;
if (value > pdi_in_num)
return -EINVAL;
/* Userspace changed the hardware state behind the kernel's back */
add_taint(TAINT_USER, LOCKDEP_STILL_OK);
cdns->pdi_loopback_target = value;
return 0;
}
DEFINE_DEBUGFS_ATTRIBUTE(cdns_pdi_loopback_target_fops, NULL, cdns_set_pdi_loopback_target, "%llu\n");
/**
* sdw_cdns_debugfs_init() - Cadence debugfs init
* @cdns: Cadence instance
* @root: debugfs root
*/
void sdw_cdns_debugfs_init(struct sdw_cdns *cdns, struct dentry *root)
{
debugfs_create_file("cdns-registers", 0400, root, cdns, &cdns_reg_fops);
debugfs_create_file("cdns-hw-reset", 0200, root, cdns,
&cdns_hw_reset_fops);
debugfs_create_file("cdns-parity-error-injection", 0200, root, cdns,
&cdns_parity_error_fops);
cdns->pdi_loopback_source = -1;
cdns->pdi_loopback_target = -1;
debugfs_create_file("cdns-pdi-loopback-source", 0200, root, cdns,
&cdns_pdi_loopback_source_fops);
debugfs_create_file("cdns-pdi-loopback-target", 0200, root, cdns,
&cdns_pdi_loopback_target_fops);
}
EXPORT_SYMBOL_GPL(sdw_cdns_debugfs_init);
#endif /* CONFIG_DEBUG_FS */
/*
* IO Calls
*/
static enum sdw_command_response
cdns_fill_msg_resp(struct sdw_cdns *cdns,
struct sdw_msg *msg, int count, int offset)
{
int nack = 0, no_ack = 0;
int i;
/* check message response */
for (i = 0; i < count; i++) {
if (!(cdns->response_buf[i] & CDNS_MCP_RESP_ACK)) {
no_ack = 1;
dev_vdbg(cdns->dev, "Msg Ack not received, cmd %d\n", i);
}
if (cdns->response_buf[i] & CDNS_MCP_RESP_NACK) {
nack = 1;
dev_err_ratelimited(cdns->dev, "Msg NACK received, cmd %d\n", i);
}
}
if (nack) {
dev_err_ratelimited(cdns->dev, "Msg NACKed for Slave %d\n", msg->dev_num);
return SDW_CMD_FAIL;
}
if (no_ack) {
dev_dbg_ratelimited(cdns->dev, "Msg ignored for Slave %d\n", msg->dev_num);
return SDW_CMD_IGNORED;
}
if (msg->flags == SDW_MSG_FLAG_READ) {
/* fill response */
for (i = 0; i < count; i++)
msg->buf[i + offset] = FIELD_GET(CDNS_MCP_RESP_RDATA,
cdns->response_buf[i]);
}
return SDW_CMD_OK;
}
static void cdns_read_response(struct sdw_cdns *cdns)
{
u32 num_resp, cmd_base;
int i;
/* RX_FIFO_AVAIL can be 2 entries more than the FIFO size */
BUILD_BUG_ON(ARRAY_SIZE(cdns->response_buf) < CDNS_MCP_CMD_LEN + 2);
num_resp = cdns_readl(cdns, CDNS_MCP_FIFOSTAT);
num_resp &= CDNS_MCP_RX_FIFO_AVAIL;
if (num_resp > ARRAY_SIZE(cdns->response_buf)) {
dev_warn(cdns->dev, "RX AVAIL %d too long\n", num_resp);
num_resp = ARRAY_SIZE(cdns->response_buf);
}
cmd_base = CDNS_IP_MCP_CMD_BASE;
for (i = 0; i < num_resp; i++) {
cdns->response_buf[i] = cdns_ip_readl(cdns, cmd_base);
cmd_base += CDNS_MCP_CMD_WORD_LEN;
}
}
static enum sdw_command_response
_cdns_xfer_msg(struct sdw_cdns *cdns, struct sdw_msg *msg, int cmd,
int offset, int count, bool defer)
{
unsigned long time;
u32 base, i, data;
u16 addr;
/* Program the watermark level for RX FIFO */
if (cdns->msg_count != count) {
cdns_writel(cdns, CDNS_MCP_FIFOLEVEL, count);
cdns->msg_count = count;
}
base = CDNS_IP_MCP_CMD_BASE;
addr = msg->addr + offset;
for (i = 0; i < count; i++) {
data = FIELD_PREP(CDNS_MCP_CMD_DEV_ADDR, msg->dev_num);
data |= FIELD_PREP(CDNS_MCP_CMD_COMMAND, cmd);
data |= FIELD_PREP(CDNS_MCP_CMD_REG_ADDR, addr);
addr++;
if (msg->flags == SDW_MSG_FLAG_WRITE)
data |= msg->buf[i + offset];
data |= FIELD_PREP(CDNS_MCP_CMD_SSP_TAG, msg->ssp_sync);
cdns_ip_writel(cdns, base, data);
base += CDNS_MCP_CMD_WORD_LEN;
}
if (defer)
return SDW_CMD_OK;
/* wait for timeout or response */
time = wait_for_completion_timeout(&cdns->tx_complete,
msecs_to_jiffies(CDNS_TX_TIMEOUT));
if (!time) {
dev_err(cdns->dev, "IO transfer timed out, cmd %d device %d addr %x len %d\n",
cmd, msg->dev_num, msg->addr, msg->len);
msg->len = 0;
/* Drain anything in the RX_FIFO */
cdns_read_response(cdns);
return SDW_CMD_TIMEOUT;
}
return cdns_fill_msg_resp(cdns, msg, count, offset);
}
static enum sdw_command_response
cdns_program_scp_addr(struct sdw_cdns *cdns, struct sdw_msg *msg)
{
int nack = 0, no_ack = 0;
unsigned long time;
u32 data[2], base;
int i;
/* Program the watermark level for RX FIFO */
if (cdns->msg_count != CDNS_SCP_RX_FIFOLEVEL) {
cdns_writel(cdns, CDNS_MCP_FIFOLEVEL, CDNS_SCP_RX_FIFOLEVEL);
cdns->msg_count = CDNS_SCP_RX_FIFOLEVEL;
}
data[0] = FIELD_PREP(CDNS_MCP_CMD_DEV_ADDR, msg->dev_num);
data[0] |= FIELD_PREP(CDNS_MCP_CMD_COMMAND, 0x3);
data[1] = data[0];
data[0] |= FIELD_PREP(CDNS_MCP_CMD_REG_ADDR, SDW_SCP_ADDRPAGE1);
data[1] |= FIELD_PREP(CDNS_MCP_CMD_REG_ADDR, SDW_SCP_ADDRPAGE2);
data[0] |= msg->addr_page1;
data[1] |= msg->addr_page2;
base = CDNS_IP_MCP_CMD_BASE;
cdns_ip_writel(cdns, base, data[0]);
base += CDNS_MCP_CMD_WORD_LEN;
cdns_ip_writel(cdns, base, data[1]);
time = wait_for_completion_timeout(&cdns->tx_complete,
msecs_to_jiffies(CDNS_TX_TIMEOUT));
if (!time) {
dev_err(cdns->dev, "SCP Msg trf timed out\n");
msg->len = 0;
return SDW_CMD_TIMEOUT;
}
/* check response the writes */
for (i = 0; i < 2; i++) {
if (!(cdns->response_buf[i] & CDNS_MCP_RESP_ACK)) {
no_ack = 1;
dev_err(cdns->dev, "Program SCP Ack not received\n");
if (cdns->response_buf[i] & CDNS_MCP_RESP_NACK) {
nack = 1;
dev_err(cdns->dev, "Program SCP NACK received\n");
}
}
}
/* For NACK, NO ack, don't return err if we are in Broadcast mode */
if (nack) {
dev_err_ratelimited(cdns->dev,
"SCP_addrpage NACKed for Slave %d\n", msg->dev_num);
return SDW_CMD_FAIL;
}
if (no_ack) {
dev_dbg_ratelimited(cdns->dev,
"SCP_addrpage ignored for Slave %d\n", msg->dev_num);
return SDW_CMD_IGNORED;
}
return SDW_CMD_OK;
}
static int cdns_prep_msg(struct sdw_cdns *cdns, struct sdw_msg *msg, int *cmd)
{
int ret;
if (msg->page) {
ret = cdns_program_scp_addr(cdns, msg);
if (ret) {
msg->len = 0;
return ret;
}
}
switch (msg->flags) {
case SDW_MSG_FLAG_READ:
*cmd = CDNS_MCP_CMD_READ;
break;
case SDW_MSG_FLAG_WRITE:
*cmd = CDNS_MCP_CMD_WRITE;
break;
default:
dev_err(cdns->dev, "Invalid msg cmd: %d\n", msg->flags);
return -EINVAL;
}
return 0;
}
enum sdw_command_response
cdns_xfer_msg(struct sdw_bus *bus, struct sdw_msg *msg)
{
struct sdw_cdns *cdns = bus_to_cdns(bus);
int cmd = 0, ret, i;
ret = cdns_prep_msg(cdns, msg, &cmd);
if (ret)
return SDW_CMD_FAIL_OTHER;
for (i = 0; i < msg->len / CDNS_MCP_CMD_LEN; i++) {
ret = _cdns_xfer_msg(cdns, msg, cmd, i * CDNS_MCP_CMD_LEN,
CDNS_MCP_CMD_LEN, false);
if (ret != SDW_CMD_OK)
return ret;
}
if (!(msg->len % CDNS_MCP_CMD_LEN))
return SDW_CMD_OK;
return _cdns_xfer_msg(cdns, msg, cmd, i * CDNS_MCP_CMD_LEN,
msg->len % CDNS_MCP_CMD_LEN, false);
}
EXPORT_SYMBOL(cdns_xfer_msg);
enum sdw_command_response
cdns_xfer_msg_defer(struct sdw_bus *bus)
{
struct sdw_cdns *cdns = bus_to_cdns(bus);
struct sdw_defer *defer = &bus->defer_msg;
struct sdw_msg *msg = defer->msg;
int cmd = 0, ret;
/* for defer only 1 message is supported */
if (msg->len > 1)
return -ENOTSUPP;
ret = cdns_prep_msg(cdns, msg, &cmd);
if (ret)
return SDW_CMD_FAIL_OTHER;
return _cdns_xfer_msg(cdns, msg, cmd, 0, msg->len, true);
}
EXPORT_SYMBOL(cdns_xfer_msg_defer);
u32 cdns_read_ping_status(struct sdw_bus *bus)
{
struct sdw_cdns *cdns = bus_to_cdns(bus);
return cdns_readl(cdns, CDNS_MCP_SLAVE_STAT);
}
EXPORT_SYMBOL(cdns_read_ping_status);
/*
* IRQ handling
*/
static int cdns_update_slave_status(struct sdw_cdns *cdns,
u64 slave_intstat)
{
enum sdw_slave_status status[SDW_MAX_DEVICES + 1];
bool is_slave = false;
u32 mask;
u32 val;
int i, set_status;
memset(status, 0, sizeof(status));
for (i = 0; i <= SDW_MAX_DEVICES; i++) {
mask = (slave_intstat >> (i * CDNS_MCP_SLAVE_STATUS_NUM)) &
CDNS_MCP_SLAVE_STATUS_BITS;
set_status = 0;
if (mask) {
is_slave = true;
if (mask & CDNS_MCP_SLAVE_INTSTAT_RESERVED) {
status[i] = SDW_SLAVE_RESERVED;
set_status++;
}
if (mask & CDNS_MCP_SLAVE_INTSTAT_ATTACHED) {
status[i] = SDW_SLAVE_ATTACHED;
set_status++;
}
if (mask & CDNS_MCP_SLAVE_INTSTAT_ALERT) {
status[i] = SDW_SLAVE_ALERT;
set_status++;
}
if (mask & CDNS_MCP_SLAVE_INTSTAT_NPRESENT) {
status[i] = SDW_SLAVE_UNATTACHED;
set_status++;
}
}
/*
* check that there was a single reported Slave status and when
* there is not use the latest status extracted from PING commands
*/
if (set_status != 1) {
val = cdns_readl(cdns, CDNS_MCP_SLAVE_STAT);
val >>= (i * 2);
switch (val & 0x3) {
case 0:
status[i] = SDW_SLAVE_UNATTACHED;
break;
case 1:
status[i] = SDW_SLAVE_ATTACHED;
break;
case 2:
status[i] = SDW_SLAVE_ALERT;
break;
case 3:
default:
status[i] = SDW_SLAVE_RESERVED;
break;
}
}
}
if (is_slave)
return sdw_handle_slave_status(&cdns->bus, status);
return 0;
}
/**
* sdw_cdns_irq() - Cadence interrupt handler
* @irq: irq number
* @dev_id: irq context
*/
irqreturn_t sdw_cdns_irq(int irq, void *dev_id)
{
struct sdw_cdns *cdns = dev_id;
u32 int_status;
/* Check if the link is up */
if (!cdns->link_up)
return IRQ_NONE;
int_status = cdns_readl(cdns, CDNS_MCP_INTSTAT);
/* check for reserved values read as zero */
if (int_status & CDNS_MCP_INT_RESERVED)
return IRQ_NONE;
if (!(int_status & CDNS_MCP_INT_IRQ))
return IRQ_NONE;
if (int_status & CDNS_MCP_INT_RX_WL) {
struct sdw_bus *bus = &cdns->bus;
struct sdw_defer *defer = &bus->defer_msg;
cdns_read_response(cdns);
if (defer && defer->msg) {
cdns_fill_msg_resp(cdns, defer->msg,
defer->length, 0);
complete(&defer->complete);
} else {
complete(&cdns->tx_complete);
}
}
if (int_status & CDNS_MCP_INT_PARITY) {
/* Parity error detected by Master */
dev_err_ratelimited(cdns->dev, "Parity error\n");
}
if (int_status & CDNS_MCP_INT_CTRL_CLASH) {
/* Slave is driving bit slot during control word */
dev_err_ratelimited(cdns->dev, "Bus clash for control word\n");
}
if (int_status & CDNS_MCP_INT_DATA_CLASH) {
/*
* Multiple slaves trying to drive bit slot, or issue with
* ownership of data bits or Slave gone bonkers
*/
dev_err_ratelimited(cdns->dev, "Bus clash for data word\n");
}
if (cdns->bus.params.m_data_mode != SDW_PORT_DATA_MODE_NORMAL &&
int_status & CDNS_MCP_INT_DPINT) {
u32 port_intstat;
/* just log which ports report an error */
port_intstat = cdns_readl(cdns, CDNS_MCP_PORT_INTSTAT);
dev_err_ratelimited(cdns->dev, "DP interrupt: PortIntStat %8x\n",
port_intstat);
/* clear status w/ write1 */
cdns_writel(cdns, CDNS_MCP_PORT_INTSTAT, port_intstat);
}
if (int_status & CDNS_MCP_INT_SLAVE_MASK) {
/* Mask the Slave interrupt and wake thread */
cdns_updatel(cdns, CDNS_MCP_INTMASK,
CDNS_MCP_INT_SLAVE_MASK, 0);
int_status &= ~CDNS_MCP_INT_SLAVE_MASK;
/*
* Deal with possible race condition between interrupt
* handling and disabling interrupts on suspend.
*
* If the master is in the process of disabling
* interrupts, don't schedule a workqueue
*/
if (cdns->interrupt_enabled)
schedule_work(&cdns->work);
}
cdns_writel(cdns, CDNS_MCP_INTSTAT, int_status);
return IRQ_HANDLED;
}
EXPORT_SYMBOL(sdw_cdns_irq);
/**
* cdns_update_slave_status_work - update slave status in a work since we will need to handle
* other interrupts eg. CDNS_MCP_INT_RX_WL during the update slave
* process.
* @work: cdns worker thread
*/
static void cdns_update_slave_status_work(struct work_struct *work)
{
struct sdw_cdns *cdns =
container_of(work, struct sdw_cdns, work);
u32 slave0, slave1;
u64 slave_intstat;
u32 device0_status;
int retry_count = 0;
soundwire: cadence: Fix lost ATTACHED interrupts when enumerating The correct way to handle interrupts is to clear the bits we are about to handle _before_ handling them. Thus if the condition then re-asserts during the handling we won't lose it. This patch changes cdns_update_slave_status_work() to do this. The previous code cleared the interrupts after handling them. The problem with this is that when handling enumeration of devices the ATTACH statuses can be accidentally cleared and so some or all of the devices never complete their enumeration. Thus we can have a situation like this: - one or more devices are reverting to ID #0 - accumulated status bits indicate some devices attached and some on ID #0. (Remember: status bits are sticky until they are handled) - Because of device on #0 sdw_handle_slave_status() programs the device ID and exits without handling the other status, expecting to get an ATTACHED from this reprogrammed device. - The device immediately starts reporting ATTACHED in PINGs, which will assert its CDNS_MCP_SLAVE_INTSTAT_ATTACHED bit. - cdns_update_slave_status_work() clears INTSTAT0/1. If the initial status had CDNS_MCP_SLAVE_INTSTAT_ATTACHED bit set it will be cleared. - The ATTACHED change for the device has now been lost. - cdns_update_slave_status_work() clears CDNS_MCP_INT_SLAVE_MASK so if the new ATTACHED state had set it, it will be cleared without ever having been handled. Unless there is some other state change from another device to cause a new interrupt, the ATTACHED state of the reprogrammed device will never cause an interrupt so its enumeration will not be completed. Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Link: https://lore.kernel.org/r/20220914160248.1047627-5-rf@opensource.cirrus.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2022-09-14 19:02:47 +03:00
/*
* Clear main interrupt first so we don't lose any assertions
* that happen during this function.
*/
cdns_writel(cdns, CDNS_MCP_INTSTAT, CDNS_MCP_INT_SLAVE_MASK);
slave0 = cdns_readl(cdns, CDNS_MCP_SLAVE_INTSTAT0);
slave1 = cdns_readl(cdns, CDNS_MCP_SLAVE_INTSTAT1);
soundwire: cadence: Fix lost ATTACHED interrupts when enumerating The correct way to handle interrupts is to clear the bits we are about to handle _before_ handling them. Thus if the condition then re-asserts during the handling we won't lose it. This patch changes cdns_update_slave_status_work() to do this. The previous code cleared the interrupts after handling them. The problem with this is that when handling enumeration of devices the ATTACH statuses can be accidentally cleared and so some or all of the devices never complete their enumeration. Thus we can have a situation like this: - one or more devices are reverting to ID #0 - accumulated status bits indicate some devices attached and some on ID #0. (Remember: status bits are sticky until they are handled) - Because of device on #0 sdw_handle_slave_status() programs the device ID and exits without handling the other status, expecting to get an ATTACHED from this reprogrammed device. - The device immediately starts reporting ATTACHED in PINGs, which will assert its CDNS_MCP_SLAVE_INTSTAT_ATTACHED bit. - cdns_update_slave_status_work() clears INTSTAT0/1. If the initial status had CDNS_MCP_SLAVE_INTSTAT_ATTACHED bit set it will be cleared. - The ATTACHED change for the device has now been lost. - cdns_update_slave_status_work() clears CDNS_MCP_INT_SLAVE_MASK so if the new ATTACHED state had set it, it will be cleared without ever having been handled. Unless there is some other state change from another device to cause a new interrupt, the ATTACHED state of the reprogrammed device will never cause an interrupt so its enumeration will not be completed. Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Link: https://lore.kernel.org/r/20220914160248.1047627-5-rf@opensource.cirrus.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2022-09-14 19:02:47 +03:00
/*
* Clear the bits before handling so we don't lose any
* bits that re-assert.
*/
cdns_writel(cdns, CDNS_MCP_SLAVE_INTSTAT0, slave0);
cdns_writel(cdns, CDNS_MCP_SLAVE_INTSTAT1, slave1);
/* combine the two status */
slave_intstat = ((u64)slave1 << 32) | slave0;
dev_dbg_ratelimited(cdns->dev, "Slave status change: 0x%llx\n", slave_intstat);
update_status:
cdns_update_slave_status(cdns, slave_intstat);
/*
* When there is more than one peripheral per link, it's
* possible that a deviceB becomes attached after we deal with
* the attachment of deviceA. Since the hardware does a
* logical AND, the attachment of the second device does not
* change the status seen by the driver.
*
* In that case, clearing the registers above would result in
* the deviceB never being detected - until a change of status
* is observed on the bus.
*
* To avoid this race condition, re-check if any device0 needs
* attention with PING commands. There is no need to check for
* ALERTS since they are not allowed until a non-zero
* device_number is assigned.
soundwire: cadence: Fix lost ATTACHED interrupts when enumerating The correct way to handle interrupts is to clear the bits we are about to handle _before_ handling them. Thus if the condition then re-asserts during the handling we won't lose it. This patch changes cdns_update_slave_status_work() to do this. The previous code cleared the interrupts after handling them. The problem with this is that when handling enumeration of devices the ATTACH statuses can be accidentally cleared and so some or all of the devices never complete their enumeration. Thus we can have a situation like this: - one or more devices are reverting to ID #0 - accumulated status bits indicate some devices attached and some on ID #0. (Remember: status bits are sticky until they are handled) - Because of device on #0 sdw_handle_slave_status() programs the device ID and exits without handling the other status, expecting to get an ATTACHED from this reprogrammed device. - The device immediately starts reporting ATTACHED in PINGs, which will assert its CDNS_MCP_SLAVE_INTSTAT_ATTACHED bit. - cdns_update_slave_status_work() clears INTSTAT0/1. If the initial status had CDNS_MCP_SLAVE_INTSTAT_ATTACHED bit set it will be cleared. - The ATTACHED change for the device has now been lost. - cdns_update_slave_status_work() clears CDNS_MCP_INT_SLAVE_MASK so if the new ATTACHED state had set it, it will be cleared without ever having been handled. Unless there is some other state change from another device to cause a new interrupt, the ATTACHED state of the reprogrammed device will never cause an interrupt so its enumeration will not be completed. Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Link: https://lore.kernel.org/r/20220914160248.1047627-5-rf@opensource.cirrus.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2022-09-14 19:02:47 +03:00
*
* Do not clear the INTSTAT0/1. While looping to enumerate devices on
* #0 there could be status changes on other devices - these must
* be kept in the INTSTAT so they can be handled when all #0 devices
* have been handled.
*/
device0_status = cdns_readl(cdns, CDNS_MCP_SLAVE_STAT);
device0_status &= 3;
if (device0_status == SDW_SLAVE_ATTACHED) {
if (retry_count++ < SDW_MAX_DEVICES) {
dev_dbg_ratelimited(cdns->dev,
"Device0 detected after clearing status, iteration %d\n",
retry_count);
slave_intstat = CDNS_MCP_SLAVE_INTSTAT_ATTACHED;
goto update_status;
} else {
dev_err_ratelimited(cdns->dev,
"Device0 detected after %d iterations\n",
retry_count);
}
}
soundwire: cadence: Fix lost ATTACHED interrupts when enumerating The correct way to handle interrupts is to clear the bits we are about to handle _before_ handling them. Thus if the condition then re-asserts during the handling we won't lose it. This patch changes cdns_update_slave_status_work() to do this. The previous code cleared the interrupts after handling them. The problem with this is that when handling enumeration of devices the ATTACH statuses can be accidentally cleared and so some or all of the devices never complete their enumeration. Thus we can have a situation like this: - one or more devices are reverting to ID #0 - accumulated status bits indicate some devices attached and some on ID #0. (Remember: status bits are sticky until they are handled) - Because of device on #0 sdw_handle_slave_status() programs the device ID and exits without handling the other status, expecting to get an ATTACHED from this reprogrammed device. - The device immediately starts reporting ATTACHED in PINGs, which will assert its CDNS_MCP_SLAVE_INTSTAT_ATTACHED bit. - cdns_update_slave_status_work() clears INTSTAT0/1. If the initial status had CDNS_MCP_SLAVE_INTSTAT_ATTACHED bit set it will be cleared. - The ATTACHED change for the device has now been lost. - cdns_update_slave_status_work() clears CDNS_MCP_INT_SLAVE_MASK so if the new ATTACHED state had set it, it will be cleared without ever having been handled. Unless there is some other state change from another device to cause a new interrupt, the ATTACHED state of the reprogrammed device will never cause an interrupt so its enumeration will not be completed. Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Link: https://lore.kernel.org/r/20220914160248.1047627-5-rf@opensource.cirrus.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
2022-09-14 19:02:47 +03:00
/* unmask Slave interrupt now */
cdns_updatel(cdns, CDNS_MCP_INTMASK,
CDNS_MCP_INT_SLAVE_MASK, CDNS_MCP_INT_SLAVE_MASK);
}
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 08:13:49 +03:00
/* paranoia check to make sure self-cleared bits are indeed cleared */
void sdw_cdns_check_self_clearing_bits(struct sdw_cdns *cdns, const char *string,
bool initial_delay, int reset_iterations)
{
u32 ip_mcp_control;
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 08:13:49 +03:00
u32 mcp_control;
u32 mcp_config_update;
int i;
if (initial_delay)
usleep_range(1000, 1500);
ip_mcp_control = cdns_ip_readl(cdns, CDNS_IP_MCP_CONTROL);
/* the following bits should be cleared immediately */
if (ip_mcp_control & CDNS_IP_MCP_CONTROL_SW_RST)
dev_err(cdns->dev, "%s failed: IP_MCP_CONTROL_SW_RST is not cleared\n", string);
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 08:13:49 +03:00
mcp_control = cdns_readl(cdns, CDNS_MCP_CONTROL);
/* the following bits should be cleared immediately */
if (mcp_control & CDNS_MCP_CONTROL_CMD_RST)
dev_err(cdns->dev, "%s failed: MCP_CONTROL_CMD_RST is not cleared\n", string);
if (mcp_control & CDNS_MCP_CONTROL_SOFT_RST)
dev_err(cdns->dev, "%s failed: MCP_CONTROL_SOFT_RST is not cleared\n", string);
if (mcp_control & CDNS_MCP_CONTROL_CLK_STOP_CLR)
dev_err(cdns->dev, "%s failed: MCP_CONTROL_CLK_STOP_CLR is not cleared\n", string);
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 08:13:49 +03:00
mcp_config_update = cdns_readl(cdns, CDNS_MCP_CONFIG_UPDATE);
if (mcp_config_update & CDNS_MCP_CONFIG_UPDATE_BIT)
dev_err(cdns->dev, "%s failed: MCP_CONFIG_UPDATE_BIT is not cleared\n", string);
i = 0;
while (mcp_control & CDNS_MCP_CONTROL_HW_RST) {
if (i == reset_iterations) {
dev_err(cdns->dev, "%s failed: MCP_CONTROL_HW_RST is not cleared\n", string);
break;
}
dev_dbg(cdns->dev, "%s: MCP_CONTROL_HW_RST is not cleared at iteration %d\n", string, i);
i++;
usleep_range(1000, 1500);
mcp_control = cdns_readl(cdns, CDNS_MCP_CONTROL);
}
}
EXPORT_SYMBOL(sdw_cdns_check_self_clearing_bits);
/*
* init routines
*/
/**
* sdw_cdns_exit_reset() - Program reset parameters and start bus operations
* @cdns: Cadence instance
*/
int sdw_cdns_exit_reset(struct sdw_cdns *cdns)
{
/* keep reset delay unchanged to 4096 cycles */
/* use hardware generated reset */
cdns_updatel(cdns, CDNS_MCP_CONTROL,
CDNS_MCP_CONTROL_HW_RST,
CDNS_MCP_CONTROL_HW_RST);
/* commit changes */
return cdns_config_update(cdns);
}
EXPORT_SYMBOL(sdw_cdns_exit_reset);
/**
* cdns_enable_slave_interrupts() - Enable SDW slave interrupts
* @cdns: Cadence instance
* @state: boolean for true/false
*/
static void cdns_enable_slave_interrupts(struct sdw_cdns *cdns, bool state)
{
u32 mask;
mask = cdns_readl(cdns, CDNS_MCP_INTMASK);
if (state)
mask |= CDNS_MCP_INT_SLAVE_MASK;
else
mask &= ~CDNS_MCP_INT_SLAVE_MASK;
cdns_writel(cdns, CDNS_MCP_INTMASK, mask);
}
/**
* sdw_cdns_enable_interrupt() - Enable SDW interrupts
* @cdns: Cadence instance
* @state: True if we are trying to enable interrupt.
*/
int sdw_cdns_enable_interrupt(struct sdw_cdns *cdns, bool state)
{
u32 slave_intmask0 = 0;
u32 slave_intmask1 = 0;
u32 mask = 0;
if (!state)
goto update_masks;
slave_intmask0 = CDNS_MCP_SLAVE_INTMASK0_MASK;
slave_intmask1 = CDNS_MCP_SLAVE_INTMASK1_MASK;
/* enable detection of all slave state changes */
mask = CDNS_MCP_INT_SLAVE_MASK;
/* enable detection of bus issues */
mask |= CDNS_MCP_INT_CTRL_CLASH | CDNS_MCP_INT_DATA_CLASH |
CDNS_MCP_INT_PARITY;
/* port interrupt limited to test modes for now */
if (cdns->bus.params.m_data_mode != SDW_PORT_DATA_MODE_NORMAL)
mask |= CDNS_MCP_INT_DPINT;
/* enable detection of RX fifo level */
mask |= CDNS_MCP_INT_RX_WL;
/*
* CDNS_MCP_INT_IRQ needs to be set otherwise all previous
* settings are irrelevant
*/
mask |= CDNS_MCP_INT_IRQ;
if (interrupt_mask) /* parameter override */
mask = interrupt_mask;
update_masks:
/* clear slave interrupt status before enabling interrupt */
if (state) {
u32 slave_state;
slave_state = cdns_readl(cdns, CDNS_MCP_SLAVE_INTSTAT0);
cdns_writel(cdns, CDNS_MCP_SLAVE_INTSTAT0, slave_state);
slave_state = cdns_readl(cdns, CDNS_MCP_SLAVE_INTSTAT1);
cdns_writel(cdns, CDNS_MCP_SLAVE_INTSTAT1, slave_state);
}
cdns->interrupt_enabled = state;
/*
* Complete any on-going status updates before updating masks,
* and cancel queued status updates.
*
* There could be a race with a new interrupt thrown before
* the 3 mask updates below are complete, so in the interrupt
* we use the 'interrupt_enabled' status to prevent new work
* from being queued.
*/
if (!state)
cancel_work_sync(&cdns->work);
cdns_writel(cdns, CDNS_MCP_SLAVE_INTMASK0, slave_intmask0);
cdns_writel(cdns, CDNS_MCP_SLAVE_INTMASK1, slave_intmask1);
cdns_writel(cdns, CDNS_MCP_INTMASK, mask);
return 0;
}
EXPORT_SYMBOL(sdw_cdns_enable_interrupt);
static int cdns_allocate_pdi(struct sdw_cdns *cdns,
struct sdw_cdns_pdi **stream,
u32 num, u32 pdi_offset)
{
struct sdw_cdns_pdi *pdi;
int i;
if (!num)
return 0;
pdi = devm_kcalloc(cdns->dev, num, sizeof(*pdi), GFP_KERNEL);
if (!pdi)
return -ENOMEM;
for (i = 0; i < num; i++) {
pdi[i].num = i + pdi_offset;
}
*stream = pdi;
return 0;
}
/**
* sdw_cdns_pdi_init() - PDI initialization routine
*
* @cdns: Cadence instance
* @config: Stream configurations
*/
int sdw_cdns_pdi_init(struct sdw_cdns *cdns,
struct sdw_cdns_stream_config config)
{
struct sdw_cdns_streams *stream;
int offset;
int ret;
cdns->pcm.num_bd = config.pcm_bd;
cdns->pcm.num_in = config.pcm_in;
cdns->pcm.num_out = config.pcm_out;
/* Allocate PDIs for PCMs */
stream = &cdns->pcm;
/* we allocate PDI0 and PDI1 which are used for Bulk */
offset = 0;
ret = cdns_allocate_pdi(cdns, &stream->bd,
stream->num_bd, offset);
if (ret)
return ret;
offset += stream->num_bd;
ret = cdns_allocate_pdi(cdns, &stream->in,
stream->num_in, offset);
if (ret)
return ret;
offset += stream->num_in;
ret = cdns_allocate_pdi(cdns, &stream->out,
stream->num_out, offset);
if (ret)
return ret;
/* Update total number of PCM PDIs */
stream->num_pdi = stream->num_bd + stream->num_in + stream->num_out;
cdns->num_ports = stream->num_pdi;
return 0;
}
EXPORT_SYMBOL(sdw_cdns_pdi_init);
static u32 cdns_set_initial_frame_shape(int n_rows, int n_cols)
{
u32 val;
int c;
int r;
r = sdw_find_row_index(n_rows);
c = sdw_find_col_index(n_cols);
val = FIELD_PREP(CDNS_MCP_FRAME_SHAPE_ROW_MASK, r);
val |= FIELD_PREP(CDNS_MCP_FRAME_SHAPE_COL_MASK, c);
return val;
}
static void cdns_init_clock_ctrl(struct sdw_cdns *cdns)
{
struct sdw_bus *bus = &cdns->bus;
struct sdw_master_prop *prop = &bus->prop;
u32 val;
u32 ssp_interval;
int divider;
/* Set clock divider */
divider = (prop->mclk_freq / prop->max_clk_freq) - 1;
cdns_updatel(cdns, CDNS_MCP_CLK_CTRL0,
CDNS_MCP_CLK_MCLKD_MASK, divider);
cdns_updatel(cdns, CDNS_MCP_CLK_CTRL1,
CDNS_MCP_CLK_MCLKD_MASK, divider);
/*
* Frame shape changes after initialization have to be done
* with the bank switch mechanism
*/
val = cdns_set_initial_frame_shape(prop->default_row,
prop->default_col);
cdns_writel(cdns, CDNS_MCP_FRAME_SHAPE_INIT, val);
/* Set SSP interval to default value */
ssp_interval = prop->default_frame_rate / SDW_CADENCE_GSYNC_HZ;
cdns_writel(cdns, CDNS_MCP_SSP_CTRL0, ssp_interval);
cdns_writel(cdns, CDNS_MCP_SSP_CTRL1, ssp_interval);
}
/**
* sdw_cdns_init() - Cadence initialization
* @cdns: Cadence instance
*/
int sdw_cdns_init(struct sdw_cdns *cdns)
{
u32 val;
cdns_init_clock_ctrl(cdns);
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 08:13:49 +03:00
sdw_cdns_check_self_clearing_bits(cdns, __func__, false, 0);
/* reset msg_count to default value of FIFOLEVEL */
cdns->msg_count = cdns_readl(cdns, CDNS_MCP_FIFOLEVEL);
/* flush command FIFOs */
cdns_updatel(cdns, CDNS_MCP_CONTROL, CDNS_MCP_CONTROL_CMD_RST,
CDNS_MCP_CONTROL_CMD_RST);
/* Set cmd accept mode */
cdns_ip_updatel(cdns, CDNS_IP_MCP_CONTROL, CDNS_IP_MCP_CONTROL_CMD_ACCEPT,
CDNS_IP_MCP_CONTROL_CMD_ACCEPT);
/* Configure mcp config */
val = cdns_readl(cdns, CDNS_MCP_CONFIG);
/* Disable auto bus release */
val &= ~CDNS_MCP_CONFIG_BUS_REL;
cdns_writel(cdns, CDNS_MCP_CONFIG, val);
/* Configure IP mcp config */
val = cdns_ip_readl(cdns, CDNS_IP_MCP_CONFIG);
/* enable bus operations with clock and data */
val &= ~CDNS_IP_MCP_CONFIG_OP;
val |= CDNS_IP_MCP_CONFIG_OP_NORMAL;
/* Set cmd mode for Tx and Rx cmds */
val &= ~CDNS_IP_MCP_CONFIG_CMD;
/* Disable sniffer mode */
val &= ~CDNS_IP_MCP_CONFIG_SNIFFER;
if (cdns->bus.multi_link)
/* Set Multi-master mode to take gsync into account */
val |= CDNS_IP_MCP_CONFIG_MMASTER;
/* leave frame delay to hardware default of 0x1F */
/* leave command retry to hardware default of 0 */
cdns_ip_writel(cdns, CDNS_IP_MCP_CONFIG, val);
/* changes will be committed later */
return 0;
}
EXPORT_SYMBOL(sdw_cdns_init);
int cdns_bus_conf(struct sdw_bus *bus, struct sdw_bus_params *params)
{
struct sdw_master_prop *prop = &bus->prop;
struct sdw_cdns *cdns = bus_to_cdns(bus);
int mcp_clkctrl_off;
int divider;
if (!params->curr_dr_freq) {
dev_err(cdns->dev, "NULL curr_dr_freq\n");
return -EINVAL;
}
divider = prop->mclk_freq * SDW_DOUBLE_RATE_FACTOR /
params->curr_dr_freq;
divider--; /* divider is 1/(N+1) */
if (params->next_bank)
mcp_clkctrl_off = CDNS_MCP_CLK_CTRL1;
else
mcp_clkctrl_off = CDNS_MCP_CLK_CTRL0;
cdns_updatel(cdns, mcp_clkctrl_off, CDNS_MCP_CLK_MCLKD_MASK, divider);
return 0;
}
EXPORT_SYMBOL(cdns_bus_conf);
static int cdns_port_params(struct sdw_bus *bus,
struct sdw_port_params *p_params, unsigned int bank)
{
struct sdw_cdns *cdns = bus_to_cdns(bus);
int dpn_config_off_source;
int dpn_config_off_target;
int target_num = p_params->num;
int source_num = p_params->num;
bool override = false;
int dpn_config;
if (target_num == cdns->pdi_loopback_target &&
cdns->pdi_loopback_source != -1) {
source_num = cdns->pdi_loopback_source;
override = true;
}
if (bank) {
dpn_config_off_source = CDNS_DPN_B1_CONFIG(source_num);
dpn_config_off_target = CDNS_DPN_B1_CONFIG(target_num);
} else {
dpn_config_off_source = CDNS_DPN_B0_CONFIG(source_num);
dpn_config_off_target = CDNS_DPN_B0_CONFIG(target_num);
}
dpn_config = cdns_readl(cdns, dpn_config_off_source);
/* use port params if there is no loopback, otherwise use source as is */
if (!override) {
u32p_replace_bits(&dpn_config, p_params->bps - 1, CDNS_DPN_CONFIG_WL);
u32p_replace_bits(&dpn_config, p_params->flow_mode, CDNS_DPN_CONFIG_PORT_FLOW);
u32p_replace_bits(&dpn_config, p_params->data_mode, CDNS_DPN_CONFIG_PORT_DAT);
}
cdns_writel(cdns, dpn_config_off_target, dpn_config);
return 0;
}
static int cdns_transport_params(struct sdw_bus *bus,
struct sdw_transport_params *t_params,
enum sdw_reg_bank bank)
{
struct sdw_cdns *cdns = bus_to_cdns(bus);
int dpn_config;
int dpn_config_off_source;
int dpn_config_off_target;
int dpn_hctrl;
int dpn_hctrl_off_source;
int dpn_hctrl_off_target;
int dpn_offsetctrl;
int dpn_offsetctrl_off_source;
int dpn_offsetctrl_off_target;
int dpn_samplectrl;
int dpn_samplectrl_off_source;
int dpn_samplectrl_off_target;
int source_num = t_params->port_num;
int target_num = t_params->port_num;
bool override = false;
if (target_num == cdns->pdi_loopback_target &&
cdns->pdi_loopback_source != -1) {
source_num = cdns->pdi_loopback_source;
override = true;
}
/*
* Note: Only full data port is supported on the Master side for
* both PCM and PDM ports.
*/
if (bank) {
dpn_config_off_source = CDNS_DPN_B1_CONFIG(source_num);
dpn_hctrl_off_source = CDNS_DPN_B1_HCTRL(source_num);
dpn_offsetctrl_off_source = CDNS_DPN_B1_OFFSET_CTRL(source_num);
dpn_samplectrl_off_source = CDNS_DPN_B1_SAMPLE_CTRL(source_num);
dpn_config_off_target = CDNS_DPN_B1_CONFIG(target_num);
dpn_hctrl_off_target = CDNS_DPN_B1_HCTRL(target_num);
dpn_offsetctrl_off_target = CDNS_DPN_B1_OFFSET_CTRL(target_num);
dpn_samplectrl_off_target = CDNS_DPN_B1_SAMPLE_CTRL(target_num);
} else {
dpn_config_off_source = CDNS_DPN_B0_CONFIG(source_num);
dpn_hctrl_off_source = CDNS_DPN_B0_HCTRL(source_num);
dpn_offsetctrl_off_source = CDNS_DPN_B0_OFFSET_CTRL(source_num);
dpn_samplectrl_off_source = CDNS_DPN_B0_SAMPLE_CTRL(source_num);
dpn_config_off_target = CDNS_DPN_B0_CONFIG(target_num);
dpn_hctrl_off_target = CDNS_DPN_B0_HCTRL(target_num);
dpn_offsetctrl_off_target = CDNS_DPN_B0_OFFSET_CTRL(target_num);
dpn_samplectrl_off_target = CDNS_DPN_B0_SAMPLE_CTRL(target_num);
}
dpn_config = cdns_readl(cdns, dpn_config_off_source);
if (!override) {
u32p_replace_bits(&dpn_config, t_params->blk_grp_ctrl, CDNS_DPN_CONFIG_BGC);
u32p_replace_bits(&dpn_config, t_params->blk_pkg_mode, CDNS_DPN_CONFIG_BPM);
}
cdns_writel(cdns, dpn_config_off_target, dpn_config);
if (!override) {
dpn_offsetctrl = 0;
u32p_replace_bits(&dpn_offsetctrl, t_params->offset1, CDNS_DPN_OFFSET_CTRL_1);
u32p_replace_bits(&dpn_offsetctrl, t_params->offset2, CDNS_DPN_OFFSET_CTRL_2);
} else {
dpn_offsetctrl = cdns_readl(cdns, dpn_offsetctrl_off_source);
}
cdns_writel(cdns, dpn_offsetctrl_off_target, dpn_offsetctrl);
if (!override) {
dpn_hctrl = 0;
u32p_replace_bits(&dpn_hctrl, t_params->hstart, CDNS_DPN_HCTRL_HSTART);
u32p_replace_bits(&dpn_hctrl, t_params->hstop, CDNS_DPN_HCTRL_HSTOP);
u32p_replace_bits(&dpn_hctrl, t_params->lane_ctrl, CDNS_DPN_HCTRL_LCTRL);
} else {
dpn_hctrl = cdns_readl(cdns, dpn_hctrl_off_source);
}
cdns_writel(cdns, dpn_hctrl_off_target, dpn_hctrl);
if (!override)
dpn_samplectrl = t_params->sample_interval - 1;
else
dpn_samplectrl = cdns_readl(cdns, dpn_samplectrl_off_source);
cdns_writel(cdns, dpn_samplectrl_off_target, dpn_samplectrl);
return 0;
}
static int cdns_port_enable(struct sdw_bus *bus,
struct sdw_enable_ch *enable_ch, unsigned int bank)
{
struct sdw_cdns *cdns = bus_to_cdns(bus);
int dpn_chnen_off, ch_mask;
if (bank)
dpn_chnen_off = CDNS_DPN_B1_CH_EN(enable_ch->port_num);
else
dpn_chnen_off = CDNS_DPN_B0_CH_EN(enable_ch->port_num);
ch_mask = enable_ch->ch_mask * enable_ch->enable;
cdns_writel(cdns, dpn_chnen_off, ch_mask);
return 0;
}
static const struct sdw_master_port_ops cdns_port_ops = {
.dpn_set_port_params = cdns_port_params,
.dpn_set_port_transport_params = cdns_transport_params,
.dpn_port_enable_ch = cdns_port_enable,
};
/**
* sdw_cdns_is_clock_stop: Check clock status
*
* @cdns: Cadence instance
*/
bool sdw_cdns_is_clock_stop(struct sdw_cdns *cdns)
{
return !!(cdns_readl(cdns, CDNS_MCP_STAT) & CDNS_MCP_STAT_CLK_STOP);
}
EXPORT_SYMBOL(sdw_cdns_is_clock_stop);
/**
* sdw_cdns_clock_stop: Cadence clock stop configuration routine
*
* @cdns: Cadence instance
* @block_wake: prevent wakes if required by the platform
*/
int sdw_cdns_clock_stop(struct sdw_cdns *cdns, bool block_wake)
{
bool slave_present = false;
struct sdw_slave *slave;
int 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 08:13:49 +03:00
sdw_cdns_check_self_clearing_bits(cdns, __func__, false, 0);
/* Check suspend status */
if (sdw_cdns_is_clock_stop(cdns)) {
dev_dbg(cdns->dev, "Clock is already stopped\n");
return 0;
}
/*
* Before entering clock stop we mask the Slave
* interrupts. This helps avoid having to deal with e.g. a
* Slave becoming UNATTACHED while the clock is being stopped
*/
cdns_enable_slave_interrupts(cdns, false);
/*
* For specific platforms, it is required to be able to put
* master into a state in which it ignores wake-up trials
* in clock stop state
*/
if (block_wake)
cdns_ip_updatel(cdns, CDNS_IP_MCP_CONTROL,
CDNS_IP_MCP_CONTROL_BLOCK_WAKEUP,
CDNS_IP_MCP_CONTROL_BLOCK_WAKEUP);
list_for_each_entry(slave, &cdns->bus.slaves, node) {
if (slave->status == SDW_SLAVE_ATTACHED ||
slave->status == SDW_SLAVE_ALERT) {
slave_present = true;
break;
}
}
/* commit changes */
ret = cdns_config_update(cdns);
if (ret < 0) {
dev_err(cdns->dev, "%s: config_update failed\n", __func__);
return ret;
}
/* Prepare slaves for clock stop */
if (slave_present) {
ret = sdw_bus_prep_clk_stop(&cdns->bus);
if (ret < 0 && ret != -ENODATA) {
dev_err(cdns->dev, "prepare clock stop failed %d\n", ret);
return ret;
}
}
/*
* Enter clock stop mode and only report errors if there are
* Slave devices present (ALERT or ATTACHED)
*/
ret = sdw_bus_clk_stop(&cdns->bus);
if (ret < 0 && slave_present && ret != -ENODATA) {
dev_err(cdns->dev, "bus clock stop failed %d\n", ret);
return ret;
}
ret = cdns_set_wait(cdns, CDNS_MCP_STAT,
CDNS_MCP_STAT_CLK_STOP,
CDNS_MCP_STAT_CLK_STOP);
if (ret < 0)
dev_err(cdns->dev, "Clock stop failed %d\n", ret);
return ret;
}
EXPORT_SYMBOL(sdw_cdns_clock_stop);
/**
* sdw_cdns_clock_restart: Cadence PM clock restart configuration routine
*
* @cdns: Cadence instance
* @bus_reset: context may be lost while in low power modes and the bus
* may require a Severe Reset and re-enumeration after a wake.
*/
int sdw_cdns_clock_restart(struct sdw_cdns *cdns, bool bus_reset)
{
int ret;
/* unmask Slave interrupts that were masked when stopping the clock */
cdns_enable_slave_interrupts(cdns, true);
ret = cdns_clear_bit(cdns, CDNS_MCP_CONTROL,
CDNS_MCP_CONTROL_CLK_STOP_CLR);
if (ret < 0) {
dev_err(cdns->dev, "Couldn't exit from clock stop\n");
return ret;
}
ret = cdns_set_wait(cdns, CDNS_MCP_STAT, CDNS_MCP_STAT_CLK_STOP, 0);
if (ret < 0) {
dev_err(cdns->dev, "clock stop exit failed %d\n", ret);
return ret;
}
cdns_ip_updatel(cdns, CDNS_IP_MCP_CONTROL,
CDNS_IP_MCP_CONTROL_BLOCK_WAKEUP, 0);
cdns_ip_updatel(cdns, CDNS_IP_MCP_CONTROL, CDNS_IP_MCP_CONTROL_CMD_ACCEPT,
CDNS_IP_MCP_CONTROL_CMD_ACCEPT);
if (!bus_reset) {
/* enable bus operations with clock and data */
cdns_ip_updatel(cdns, CDNS_IP_MCP_CONFIG,
CDNS_IP_MCP_CONFIG_OP,
CDNS_IP_MCP_CONFIG_OP_NORMAL);
ret = cdns_config_update(cdns);
if (ret < 0) {
dev_err(cdns->dev, "%s: config_update failed\n", __func__);
return ret;
}
ret = sdw_bus_exit_clk_stop(&cdns->bus);
if (ret < 0)
dev_err(cdns->dev, "bus failed to exit clock stop %d\n", ret);
}
return ret;
}
EXPORT_SYMBOL(sdw_cdns_clock_restart);
/**
* sdw_cdns_probe() - Cadence probe routine
* @cdns: Cadence instance
*/
int sdw_cdns_probe(struct sdw_cdns *cdns)
{
init_completion(&cdns->tx_complete);
cdns->bus.port_ops = &cdns_port_ops;
INIT_WORK(&cdns->work, cdns_update_slave_status_work);
return 0;
}
EXPORT_SYMBOL(sdw_cdns_probe);
int cdns_set_sdw_stream(struct snd_soc_dai *dai,
void *stream, int direction)
{
struct sdw_cdns *cdns = snd_soc_dai_get_drvdata(dai);
struct sdw_cdns_dai_runtime *dai_runtime;
dai_runtime = cdns->dai_runtime_array[dai->id];
if (stream) {
/* first paranoia check */
if (dai_runtime) {
dev_err(dai->dev,
"dai_runtime already allocated for dai %s\n",
dai->name);
return -EINVAL;
}
/* allocate and set dai_runtime info */
dai_runtime = kzalloc(sizeof(*dai_runtime), GFP_KERNEL);
if (!dai_runtime)
return -ENOMEM;
dai_runtime->stream_type = SDW_STREAM_PCM;
dai_runtime->bus = &cdns->bus;
dai_runtime->link_id = cdns->instance;
dai_runtime->stream = stream;
dai_runtime->direction = direction;
cdns->dai_runtime_array[dai->id] = dai_runtime;
} else {
/* second paranoia check */
if (!dai_runtime) {
dev_err(dai->dev,
"dai_runtime not allocated for dai %s\n",
dai->name);
return -EINVAL;
}
/* for NULL stream we release allocated dai_runtime */
kfree(dai_runtime);
cdns->dai_runtime_array[dai->id] = NULL;
}
return 0;
}
EXPORT_SYMBOL(cdns_set_sdw_stream);
/**
* cdns_find_pdi() - Find a free PDI
*
* @cdns: Cadence instance
* @offset: Starting offset
* @num: Number of PDIs
* @pdi: PDI instances
* @dai_id: DAI id
*
* Find a PDI for a given PDI array. The PDI num and dai_id are
* expected to match, return NULL otherwise.
*/
static struct sdw_cdns_pdi *cdns_find_pdi(struct sdw_cdns *cdns,
unsigned int offset,
unsigned int num,
struct sdw_cdns_pdi *pdi,
int dai_id)
{
int i;
for (i = offset; i < offset + num; i++)
if (pdi[i].num == dai_id)
return &pdi[i];
return NULL;
}
/**
* sdw_cdns_config_stream: Configure a stream
*
* @cdns: Cadence instance
* @ch: Channel count
* @dir: Data direction
* @pdi: PDI to be used
*/
void sdw_cdns_config_stream(struct sdw_cdns *cdns,
u32 ch, u32 dir, struct sdw_cdns_pdi *pdi)
{
u32 offset, val = 0;
if (dir == SDW_DATA_DIR_RX) {
val = CDNS_PORTCTRL_DIRN;
if (cdns->bus.params.m_data_mode != SDW_PORT_DATA_MODE_NORMAL)
val |= CDNS_PORTCTRL_TEST_FAILED;
}
offset = CDNS_PORTCTRL + pdi->num * CDNS_PORT_OFFSET;
cdns_updatel(cdns, offset,
CDNS_PORTCTRL_DIRN | CDNS_PORTCTRL_TEST_FAILED,
val);
val = pdi->num;
val |= CDNS_PDI_CONFIG_SOFT_RESET;
val |= FIELD_PREP(CDNS_PDI_CONFIG_CHANNEL, (1 << ch) - 1);
cdns_writel(cdns, CDNS_PDI_CONFIG(pdi->num), val);
}
EXPORT_SYMBOL(sdw_cdns_config_stream);
/**
* sdw_cdns_alloc_pdi() - Allocate a PDI
*
* @cdns: Cadence instance
* @stream: Stream to be allocated
* @ch: Channel count
* @dir: Data direction
* @dai_id: DAI id
*/
struct sdw_cdns_pdi *sdw_cdns_alloc_pdi(struct sdw_cdns *cdns,
struct sdw_cdns_streams *stream,
u32 ch, u32 dir, int dai_id)
{
struct sdw_cdns_pdi *pdi = NULL;
if (dir == SDW_DATA_DIR_RX)
pdi = cdns_find_pdi(cdns, 0, stream->num_in, stream->in,
dai_id);
else
pdi = cdns_find_pdi(cdns, 0, stream->num_out, stream->out,
dai_id);
/* check if we found a PDI, else find in bi-directional */
if (!pdi)
pdi = cdns_find_pdi(cdns, 2, stream->num_bd, stream->bd,
dai_id);
if (pdi) {
pdi->l_ch_num = 0;
pdi->h_ch_num = ch - 1;
pdi->dir = dir;
pdi->ch_count = ch;
}
return pdi;
}
EXPORT_SYMBOL(sdw_cdns_alloc_pdi);
MODULE_LICENSE("Dual BSD/GPL");
MODULE_DESCRIPTION("Cadence Soundwire Library");