OpenCloudOS-Kernel/drivers/w1/masters/omap_hdq.c

689 lines
17 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0-only
/*
* Copyright (C) 2007,2012 Texas Instruments, Inc.
*/
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/platform_device.h>
#include <linux/interrupt.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 16:04:11 +08:00
#include <linux/slab.h>
#include <linux/err.h>
#include <linux/io.h>
#include <linux/sched.h>
#include <linux/pm_runtime.h>
#include <linux/of.h>
#include <linux/w1.h>
#define MOD_NAME "OMAP_HDQ:"
#define OMAP_HDQ_REVISION 0x00
#define OMAP_HDQ_TX_DATA 0x04
#define OMAP_HDQ_RX_DATA 0x08
#define OMAP_HDQ_CTRL_STATUS 0x0c
#define OMAP_HDQ_CTRL_STATUS_SINGLE BIT(7)
#define OMAP_HDQ_CTRL_STATUS_INTERRUPTMASK BIT(6)
#define OMAP_HDQ_CTRL_STATUS_CLOCKENABLE BIT(5)
#define OMAP_HDQ_CTRL_STATUS_GO BIT(4)
#define OMAP_HDQ_CTRL_STATUS_PRESENCE BIT(3)
#define OMAP_HDQ_CTRL_STATUS_INITIALIZATION BIT(2)
#define OMAP_HDQ_CTRL_STATUS_DIR BIT(1)
#define OMAP_HDQ_INT_STATUS 0x10
#define OMAP_HDQ_INT_STATUS_TXCOMPLETE BIT(2)
#define OMAP_HDQ_INT_STATUS_RXCOMPLETE BIT(1)
#define OMAP_HDQ_INT_STATUS_TIMEOUT BIT(0)
#define OMAP_HDQ_FLAG_CLEAR 0
#define OMAP_HDQ_FLAG_SET 1
#define OMAP_HDQ_TIMEOUT (HZ/5)
#define OMAP_HDQ_MAX_USER 4
static DECLARE_WAIT_QUEUE_HEAD(hdq_wait_queue);
static int w1_id;
module_param(w1_id, int, 0400);
MODULE_PARM_DESC(w1_id, "1-wire id for the slave detection in HDQ mode");
struct hdq_data {
struct device *dev;
void __iomem *hdq_base;
w1: omap-hdq: fix interrupt handling which did show spurious timeouts Since commit 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") was applied, I did see timeouts and wrong values when reading a bq27000 connected to hdq of the omap3. This occurred mainly after boot but remained and only sometimes settled down after several reads. root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=0 POWER_SUPPLY_CURRENT_NOW=0 POWER_SUPPLY_CAPACITY=0 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=-2731 POWER_SUPPLY_TIME_TO_EMPTY_NOW=0 POWER_SUPPLY_TIME_TO_EMPTY_AVG=0 POWER_SUPPLY_TIME_TO_FULL_NOW=0 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=0 POWER_SUPPLY_CHARGE_NOW=0 POWER_SUPPLY_CHARGE_FULL_DESIGN=0 POWER_SUPPLY_CYCLE_COUNT=0 POWER_SUPPLY_ENERGY_NOW=0 POWER_SUPPLY_POWER_AVG=0 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real    0m15.761s user    0m0.001s sys     0m0.025s root@letux:~# Sometimes the effect did disappear after accessing the device multiple times, speed went up and results became correct. All this indicates that some interrupts from the hdq controller are lost by the driver. Enabling debugging revealed that there were spurious tx and rx timeouts, i.e. the driver does not always recognise interrupts. The main problem is that rx and tx interrupts share a single variable which was sometimes reset to 0 wiping out other interrupts. And it was overwritten by a second interrupt, independent of whether the previous interrupt was already processed or not. This patch improves interrupt handling to avoid such races and loss of interrupt flags. The ideas are: * only the hdq_isr() sets bits in hdq_status * it does not reset any bits * it does wake_up() if any interrupt is pending * bits are only reset by the read/write/break functions if they were waited for * this makes sure that no interrupts can be lost * rx/tx/timeout bits are completely decoupled from each other (and not reset all after waiting for any of them) * which bits to reset is now specified by a new parameter to hdq_reset_irqstatus() * hdq_reset_irqstatus() also returns the state before resetting so that we can encapsulate the spinlock * this should now handle the case that the write and read are both already finished quickly before the hdq_write_byte() ends. * Or that two interrupts occur in succession before they are processed by the driver. Old code may have reset all status bits making the next hdq_read_byte() timeout. * the spinlock now always protects changing of bits in function hdq_reset_irqstatus() which could become a read-write-modify problem if the interrupt handler tries to read-modify-write exactly at the same moment * we add mutex protection also for hdq_write_byte() just to be safe to not to disturb a hdq_read_byte() triggered by some other thread/process. This patch was tested on a GTA04 and results in no boot problems any more. And first read after boot is now ok: root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=3970000 POWER_SUPPLY_CURRENT_NOW=354144 POWER_SUPPLY_CAPACITY=82 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=266 POWER_SUPPLY_TIME_TO_EMPTY_NOW=7680 POWER_SUPPLY_TIME_TO_EMPTY_AVG=7380 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=934856 POWER_SUPPLY_CHARGE_NOW=763976 POWER_SUPPLY_CHARGE_FULL_DESIGN=1233792 POWER_SUPPLY_CYCLE_COUNT=82 POWER_SUPPLY_ENERGY_NOW=2852840 POWER_SUPPLY_POWER_AVG=1392840 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real 0m0.233s user 0m0.000s sys 0m0.025s root@letux:~# It was also tested with dev_dbg enabled and more printk that all activities behave correctly, especially hdq_write_byte(), hdq_read_byte(), omap_hdq_break(). Not tested is omap_w1_triplet(). Fixes: 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by: H. Nikolaus Schaller <hns@goldelico.com> Link: https://lore.kernel.org/r/68fc8623ae741878beef049273696d2377526165.1590255176.git.hns@goldelico.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-05-24 01:32:56 +08:00
/* lock read/write/break operations */
struct mutex hdq_mutex;
w1: omap-hdq: fix interrupt handling which did show spurious timeouts Since commit 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") was applied, I did see timeouts and wrong values when reading a bq27000 connected to hdq of the omap3. This occurred mainly after boot but remained and only sometimes settled down after several reads. root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=0 POWER_SUPPLY_CURRENT_NOW=0 POWER_SUPPLY_CAPACITY=0 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=-2731 POWER_SUPPLY_TIME_TO_EMPTY_NOW=0 POWER_SUPPLY_TIME_TO_EMPTY_AVG=0 POWER_SUPPLY_TIME_TO_FULL_NOW=0 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=0 POWER_SUPPLY_CHARGE_NOW=0 POWER_SUPPLY_CHARGE_FULL_DESIGN=0 POWER_SUPPLY_CYCLE_COUNT=0 POWER_SUPPLY_ENERGY_NOW=0 POWER_SUPPLY_POWER_AVG=0 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real    0m15.761s user    0m0.001s sys     0m0.025s root@letux:~# Sometimes the effect did disappear after accessing the device multiple times, speed went up and results became correct. All this indicates that some interrupts from the hdq controller are lost by the driver. Enabling debugging revealed that there were spurious tx and rx timeouts, i.e. the driver does not always recognise interrupts. The main problem is that rx and tx interrupts share a single variable which was sometimes reset to 0 wiping out other interrupts. And it was overwritten by a second interrupt, independent of whether the previous interrupt was already processed or not. This patch improves interrupt handling to avoid such races and loss of interrupt flags. The ideas are: * only the hdq_isr() sets bits in hdq_status * it does not reset any bits * it does wake_up() if any interrupt is pending * bits are only reset by the read/write/break functions if they were waited for * this makes sure that no interrupts can be lost * rx/tx/timeout bits are completely decoupled from each other (and not reset all after waiting for any of them) * which bits to reset is now specified by a new parameter to hdq_reset_irqstatus() * hdq_reset_irqstatus() also returns the state before resetting so that we can encapsulate the spinlock * this should now handle the case that the write and read are both already finished quickly before the hdq_write_byte() ends. * Or that two interrupts occur in succession before they are processed by the driver. Old code may have reset all status bits making the next hdq_read_byte() timeout. * the spinlock now always protects changing of bits in function hdq_reset_irqstatus() which could become a read-write-modify problem if the interrupt handler tries to read-modify-write exactly at the same moment * we add mutex protection also for hdq_write_byte() just to be safe to not to disturb a hdq_read_byte() triggered by some other thread/process. This patch was tested on a GTA04 and results in no boot problems any more. And first read after boot is now ok: root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=3970000 POWER_SUPPLY_CURRENT_NOW=354144 POWER_SUPPLY_CAPACITY=82 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=266 POWER_SUPPLY_TIME_TO_EMPTY_NOW=7680 POWER_SUPPLY_TIME_TO_EMPTY_AVG=7380 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=934856 POWER_SUPPLY_CHARGE_NOW=763976 POWER_SUPPLY_CHARGE_FULL_DESIGN=1233792 POWER_SUPPLY_CYCLE_COUNT=82 POWER_SUPPLY_ENERGY_NOW=2852840 POWER_SUPPLY_POWER_AVG=1392840 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real 0m0.233s user 0m0.000s sys 0m0.025s root@letux:~# It was also tested with dev_dbg enabled and more printk that all activities behave correctly, especially hdq_write_byte(), hdq_read_byte(), omap_hdq_break(). Not tested is omap_w1_triplet(). Fixes: 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by: H. Nikolaus Schaller <hns@goldelico.com> Link: https://lore.kernel.org/r/68fc8623ae741878beef049273696d2377526165.1590255176.git.hns@goldelico.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-05-24 01:32:56 +08:00
/* interrupt status and a lock for it */
u8 hdq_irqstatus;
spinlock_t hdq_spinlock;
/* mode: 0-HDQ 1-W1 */
int mode;
};
/* HDQ register I/O routines */
static inline u8 hdq_reg_in(struct hdq_data *hdq_data, u32 offset)
{
return __raw_readl(hdq_data->hdq_base + offset);
}
static inline void hdq_reg_out(struct hdq_data *hdq_data, u32 offset, u8 val)
{
__raw_writel(val, hdq_data->hdq_base + offset);
}
static inline u8 hdq_reg_merge(struct hdq_data *hdq_data, u32 offset,
u8 val, u8 mask)
{
u8 new_val = (__raw_readl(hdq_data->hdq_base + offset) & ~mask)
| (val & mask);
__raw_writel(new_val, hdq_data->hdq_base + offset);
return new_val;
}
/*
* Wait for one or more bits in flag change.
* HDQ_FLAG_SET: wait until any bit in the flag is set.
* HDQ_FLAG_CLEAR: wait until all bits in the flag are cleared.
* return 0 on success and -ETIMEDOUT in the case of timeout.
*/
static int hdq_wait_for_flag(struct hdq_data *hdq_data, u32 offset,
u8 flag, u8 flag_set, u8 *status)
{
int ret = 0;
unsigned long timeout = jiffies + OMAP_HDQ_TIMEOUT;
if (flag_set == OMAP_HDQ_FLAG_CLEAR) {
/* wait for the flag clear */
while (((*status = hdq_reg_in(hdq_data, offset)) & flag)
&& time_before(jiffies, timeout)) {
schedule_timeout_uninterruptible(1);
}
if (*status & flag)
ret = -ETIMEDOUT;
} else if (flag_set == OMAP_HDQ_FLAG_SET) {
/* wait for the flag set */
while (!((*status = hdq_reg_in(hdq_data, offset)) & flag)
&& time_before(jiffies, timeout)) {
schedule_timeout_uninterruptible(1);
}
if (!(*status & flag))
ret = -ETIMEDOUT;
} else
return -EINVAL;
return ret;
}
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
/* Clear saved irqstatus after using an interrupt */
w1: omap-hdq: fix interrupt handling which did show spurious timeouts Since commit 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") was applied, I did see timeouts and wrong values when reading a bq27000 connected to hdq of the omap3. This occurred mainly after boot but remained and only sometimes settled down after several reads. root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=0 POWER_SUPPLY_CURRENT_NOW=0 POWER_SUPPLY_CAPACITY=0 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=-2731 POWER_SUPPLY_TIME_TO_EMPTY_NOW=0 POWER_SUPPLY_TIME_TO_EMPTY_AVG=0 POWER_SUPPLY_TIME_TO_FULL_NOW=0 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=0 POWER_SUPPLY_CHARGE_NOW=0 POWER_SUPPLY_CHARGE_FULL_DESIGN=0 POWER_SUPPLY_CYCLE_COUNT=0 POWER_SUPPLY_ENERGY_NOW=0 POWER_SUPPLY_POWER_AVG=0 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real    0m15.761s user    0m0.001s sys     0m0.025s root@letux:~# Sometimes the effect did disappear after accessing the device multiple times, speed went up and results became correct. All this indicates that some interrupts from the hdq controller are lost by the driver. Enabling debugging revealed that there were spurious tx and rx timeouts, i.e. the driver does not always recognise interrupts. The main problem is that rx and tx interrupts share a single variable which was sometimes reset to 0 wiping out other interrupts. And it was overwritten by a second interrupt, independent of whether the previous interrupt was already processed or not. This patch improves interrupt handling to avoid such races and loss of interrupt flags. The ideas are: * only the hdq_isr() sets bits in hdq_status * it does not reset any bits * it does wake_up() if any interrupt is pending * bits are only reset by the read/write/break functions if they were waited for * this makes sure that no interrupts can be lost * rx/tx/timeout bits are completely decoupled from each other (and not reset all after waiting for any of them) * which bits to reset is now specified by a new parameter to hdq_reset_irqstatus() * hdq_reset_irqstatus() also returns the state before resetting so that we can encapsulate the spinlock * this should now handle the case that the write and read are both already finished quickly before the hdq_write_byte() ends. * Or that two interrupts occur in succession before they are processed by the driver. Old code may have reset all status bits making the next hdq_read_byte() timeout. * the spinlock now always protects changing of bits in function hdq_reset_irqstatus() which could become a read-write-modify problem if the interrupt handler tries to read-modify-write exactly at the same moment * we add mutex protection also for hdq_write_byte() just to be safe to not to disturb a hdq_read_byte() triggered by some other thread/process. This patch was tested on a GTA04 and results in no boot problems any more. And first read after boot is now ok: root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=3970000 POWER_SUPPLY_CURRENT_NOW=354144 POWER_SUPPLY_CAPACITY=82 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=266 POWER_SUPPLY_TIME_TO_EMPTY_NOW=7680 POWER_SUPPLY_TIME_TO_EMPTY_AVG=7380 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=934856 POWER_SUPPLY_CHARGE_NOW=763976 POWER_SUPPLY_CHARGE_FULL_DESIGN=1233792 POWER_SUPPLY_CYCLE_COUNT=82 POWER_SUPPLY_ENERGY_NOW=2852840 POWER_SUPPLY_POWER_AVG=1392840 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real 0m0.233s user 0m0.000s sys 0m0.025s root@letux:~# It was also tested with dev_dbg enabled and more printk that all activities behave correctly, especially hdq_write_byte(), hdq_read_byte(), omap_hdq_break(). Not tested is omap_w1_triplet(). Fixes: 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by: H. Nikolaus Schaller <hns@goldelico.com> Link: https://lore.kernel.org/r/68fc8623ae741878beef049273696d2377526165.1590255176.git.hns@goldelico.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-05-24 01:32:56 +08:00
static u8 hdq_reset_irqstatus(struct hdq_data *hdq_data, u8 bits)
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
{
unsigned long irqflags;
w1: omap-hdq: fix interrupt handling which did show spurious timeouts Since commit 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") was applied, I did see timeouts and wrong values when reading a bq27000 connected to hdq of the omap3. This occurred mainly after boot but remained and only sometimes settled down after several reads. root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=0 POWER_SUPPLY_CURRENT_NOW=0 POWER_SUPPLY_CAPACITY=0 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=-2731 POWER_SUPPLY_TIME_TO_EMPTY_NOW=0 POWER_SUPPLY_TIME_TO_EMPTY_AVG=0 POWER_SUPPLY_TIME_TO_FULL_NOW=0 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=0 POWER_SUPPLY_CHARGE_NOW=0 POWER_SUPPLY_CHARGE_FULL_DESIGN=0 POWER_SUPPLY_CYCLE_COUNT=0 POWER_SUPPLY_ENERGY_NOW=0 POWER_SUPPLY_POWER_AVG=0 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real    0m15.761s user    0m0.001s sys     0m0.025s root@letux:~# Sometimes the effect did disappear after accessing the device multiple times, speed went up and results became correct. All this indicates that some interrupts from the hdq controller are lost by the driver. Enabling debugging revealed that there were spurious tx and rx timeouts, i.e. the driver does not always recognise interrupts. The main problem is that rx and tx interrupts share a single variable which was sometimes reset to 0 wiping out other interrupts. And it was overwritten by a second interrupt, independent of whether the previous interrupt was already processed or not. This patch improves interrupt handling to avoid such races and loss of interrupt flags. The ideas are: * only the hdq_isr() sets bits in hdq_status * it does not reset any bits * it does wake_up() if any interrupt is pending * bits are only reset by the read/write/break functions if they were waited for * this makes sure that no interrupts can be lost * rx/tx/timeout bits are completely decoupled from each other (and not reset all after waiting for any of them) * which bits to reset is now specified by a new parameter to hdq_reset_irqstatus() * hdq_reset_irqstatus() also returns the state before resetting so that we can encapsulate the spinlock * this should now handle the case that the write and read are both already finished quickly before the hdq_write_byte() ends. * Or that two interrupts occur in succession before they are processed by the driver. Old code may have reset all status bits making the next hdq_read_byte() timeout. * the spinlock now always protects changing of bits in function hdq_reset_irqstatus() which could become a read-write-modify problem if the interrupt handler tries to read-modify-write exactly at the same moment * we add mutex protection also for hdq_write_byte() just to be safe to not to disturb a hdq_read_byte() triggered by some other thread/process. This patch was tested on a GTA04 and results in no boot problems any more. And first read after boot is now ok: root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=3970000 POWER_SUPPLY_CURRENT_NOW=354144 POWER_SUPPLY_CAPACITY=82 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=266 POWER_SUPPLY_TIME_TO_EMPTY_NOW=7680 POWER_SUPPLY_TIME_TO_EMPTY_AVG=7380 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=934856 POWER_SUPPLY_CHARGE_NOW=763976 POWER_SUPPLY_CHARGE_FULL_DESIGN=1233792 POWER_SUPPLY_CYCLE_COUNT=82 POWER_SUPPLY_ENERGY_NOW=2852840 POWER_SUPPLY_POWER_AVG=1392840 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real 0m0.233s user 0m0.000s sys 0m0.025s root@letux:~# It was also tested with dev_dbg enabled and more printk that all activities behave correctly, especially hdq_write_byte(), hdq_read_byte(), omap_hdq_break(). Not tested is omap_w1_triplet(). Fixes: 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by: H. Nikolaus Schaller <hns@goldelico.com> Link: https://lore.kernel.org/r/68fc8623ae741878beef049273696d2377526165.1590255176.git.hns@goldelico.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-05-24 01:32:56 +08:00
u8 status;
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
spin_lock_irqsave(&hdq_data->hdq_spinlock, irqflags);
w1: omap-hdq: fix interrupt handling which did show spurious timeouts Since commit 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") was applied, I did see timeouts and wrong values when reading a bq27000 connected to hdq of the omap3. This occurred mainly after boot but remained and only sometimes settled down after several reads. root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=0 POWER_SUPPLY_CURRENT_NOW=0 POWER_SUPPLY_CAPACITY=0 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=-2731 POWER_SUPPLY_TIME_TO_EMPTY_NOW=0 POWER_SUPPLY_TIME_TO_EMPTY_AVG=0 POWER_SUPPLY_TIME_TO_FULL_NOW=0 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=0 POWER_SUPPLY_CHARGE_NOW=0 POWER_SUPPLY_CHARGE_FULL_DESIGN=0 POWER_SUPPLY_CYCLE_COUNT=0 POWER_SUPPLY_ENERGY_NOW=0 POWER_SUPPLY_POWER_AVG=0 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real    0m15.761s user    0m0.001s sys     0m0.025s root@letux:~# Sometimes the effect did disappear after accessing the device multiple times, speed went up and results became correct. All this indicates that some interrupts from the hdq controller are lost by the driver. Enabling debugging revealed that there were spurious tx and rx timeouts, i.e. the driver does not always recognise interrupts. The main problem is that rx and tx interrupts share a single variable which was sometimes reset to 0 wiping out other interrupts. And it was overwritten by a second interrupt, independent of whether the previous interrupt was already processed or not. This patch improves interrupt handling to avoid such races and loss of interrupt flags. The ideas are: * only the hdq_isr() sets bits in hdq_status * it does not reset any bits * it does wake_up() if any interrupt is pending * bits are only reset by the read/write/break functions if they were waited for * this makes sure that no interrupts can be lost * rx/tx/timeout bits are completely decoupled from each other (and not reset all after waiting for any of them) * which bits to reset is now specified by a new parameter to hdq_reset_irqstatus() * hdq_reset_irqstatus() also returns the state before resetting so that we can encapsulate the spinlock * this should now handle the case that the write and read are both already finished quickly before the hdq_write_byte() ends. * Or that two interrupts occur in succession before they are processed by the driver. Old code may have reset all status bits making the next hdq_read_byte() timeout. * the spinlock now always protects changing of bits in function hdq_reset_irqstatus() which could become a read-write-modify problem if the interrupt handler tries to read-modify-write exactly at the same moment * we add mutex protection also for hdq_write_byte() just to be safe to not to disturb a hdq_read_byte() triggered by some other thread/process. This patch was tested on a GTA04 and results in no boot problems any more. And first read after boot is now ok: root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=3970000 POWER_SUPPLY_CURRENT_NOW=354144 POWER_SUPPLY_CAPACITY=82 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=266 POWER_SUPPLY_TIME_TO_EMPTY_NOW=7680 POWER_SUPPLY_TIME_TO_EMPTY_AVG=7380 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=934856 POWER_SUPPLY_CHARGE_NOW=763976 POWER_SUPPLY_CHARGE_FULL_DESIGN=1233792 POWER_SUPPLY_CYCLE_COUNT=82 POWER_SUPPLY_ENERGY_NOW=2852840 POWER_SUPPLY_POWER_AVG=1392840 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real 0m0.233s user 0m0.000s sys 0m0.025s root@letux:~# It was also tested with dev_dbg enabled and more printk that all activities behave correctly, especially hdq_write_byte(), hdq_read_byte(), omap_hdq_break(). Not tested is omap_w1_triplet(). Fixes: 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by: H. Nikolaus Schaller <hns@goldelico.com> Link: https://lore.kernel.org/r/68fc8623ae741878beef049273696d2377526165.1590255176.git.hns@goldelico.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-05-24 01:32:56 +08:00
status = hdq_data->hdq_irqstatus;
/* this is a read-modify-write */
hdq_data->hdq_irqstatus &= ~bits;
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
spin_unlock_irqrestore(&hdq_data->hdq_spinlock, irqflags);
w1: omap-hdq: fix interrupt handling which did show spurious timeouts Since commit 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") was applied, I did see timeouts and wrong values when reading a bq27000 connected to hdq of the omap3. This occurred mainly after boot but remained and only sometimes settled down after several reads. root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=0 POWER_SUPPLY_CURRENT_NOW=0 POWER_SUPPLY_CAPACITY=0 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=-2731 POWER_SUPPLY_TIME_TO_EMPTY_NOW=0 POWER_SUPPLY_TIME_TO_EMPTY_AVG=0 POWER_SUPPLY_TIME_TO_FULL_NOW=0 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=0 POWER_SUPPLY_CHARGE_NOW=0 POWER_SUPPLY_CHARGE_FULL_DESIGN=0 POWER_SUPPLY_CYCLE_COUNT=0 POWER_SUPPLY_ENERGY_NOW=0 POWER_SUPPLY_POWER_AVG=0 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real    0m15.761s user    0m0.001s sys     0m0.025s root@letux:~# Sometimes the effect did disappear after accessing the device multiple times, speed went up and results became correct. All this indicates that some interrupts from the hdq controller are lost by the driver. Enabling debugging revealed that there were spurious tx and rx timeouts, i.e. the driver does not always recognise interrupts. The main problem is that rx and tx interrupts share a single variable which was sometimes reset to 0 wiping out other interrupts. And it was overwritten by a second interrupt, independent of whether the previous interrupt was already processed or not. This patch improves interrupt handling to avoid such races and loss of interrupt flags. The ideas are: * only the hdq_isr() sets bits in hdq_status * it does not reset any bits * it does wake_up() if any interrupt is pending * bits are only reset by the read/write/break functions if they were waited for * this makes sure that no interrupts can be lost * rx/tx/timeout bits are completely decoupled from each other (and not reset all after waiting for any of them) * which bits to reset is now specified by a new parameter to hdq_reset_irqstatus() * hdq_reset_irqstatus() also returns the state before resetting so that we can encapsulate the spinlock * this should now handle the case that the write and read are both already finished quickly before the hdq_write_byte() ends. * Or that two interrupts occur in succession before they are processed by the driver. Old code may have reset all status bits making the next hdq_read_byte() timeout. * the spinlock now always protects changing of bits in function hdq_reset_irqstatus() which could become a read-write-modify problem if the interrupt handler tries to read-modify-write exactly at the same moment * we add mutex protection also for hdq_write_byte() just to be safe to not to disturb a hdq_read_byte() triggered by some other thread/process. This patch was tested on a GTA04 and results in no boot problems any more. And first read after boot is now ok: root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=3970000 POWER_SUPPLY_CURRENT_NOW=354144 POWER_SUPPLY_CAPACITY=82 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=266 POWER_SUPPLY_TIME_TO_EMPTY_NOW=7680 POWER_SUPPLY_TIME_TO_EMPTY_AVG=7380 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=934856 POWER_SUPPLY_CHARGE_NOW=763976 POWER_SUPPLY_CHARGE_FULL_DESIGN=1233792 POWER_SUPPLY_CYCLE_COUNT=82 POWER_SUPPLY_ENERGY_NOW=2852840 POWER_SUPPLY_POWER_AVG=1392840 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real 0m0.233s user 0m0.000s sys 0m0.025s root@letux:~# It was also tested with dev_dbg enabled and more printk that all activities behave correctly, especially hdq_write_byte(), hdq_read_byte(), omap_hdq_break(). Not tested is omap_w1_triplet(). Fixes: 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by: H. Nikolaus Schaller <hns@goldelico.com> Link: https://lore.kernel.org/r/68fc8623ae741878beef049273696d2377526165.1590255176.git.hns@goldelico.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-05-24 01:32:56 +08:00
return status;
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
}
/* write out a byte and fill *status with HDQ_INT_STATUS */
static int hdq_write_byte(struct hdq_data *hdq_data, u8 val, u8 *status)
{
int ret;
u8 tmp_status;
w1: omap-hdq: fix interrupt handling which did show spurious timeouts Since commit 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") was applied, I did see timeouts and wrong values when reading a bq27000 connected to hdq of the omap3. This occurred mainly after boot but remained and only sometimes settled down after several reads. root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=0 POWER_SUPPLY_CURRENT_NOW=0 POWER_SUPPLY_CAPACITY=0 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=-2731 POWER_SUPPLY_TIME_TO_EMPTY_NOW=0 POWER_SUPPLY_TIME_TO_EMPTY_AVG=0 POWER_SUPPLY_TIME_TO_FULL_NOW=0 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=0 POWER_SUPPLY_CHARGE_NOW=0 POWER_SUPPLY_CHARGE_FULL_DESIGN=0 POWER_SUPPLY_CYCLE_COUNT=0 POWER_SUPPLY_ENERGY_NOW=0 POWER_SUPPLY_POWER_AVG=0 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real    0m15.761s user    0m0.001s sys     0m0.025s root@letux:~# Sometimes the effect did disappear after accessing the device multiple times, speed went up and results became correct. All this indicates that some interrupts from the hdq controller are lost by the driver. Enabling debugging revealed that there were spurious tx and rx timeouts, i.e. the driver does not always recognise interrupts. The main problem is that rx and tx interrupts share a single variable which was sometimes reset to 0 wiping out other interrupts. And it was overwritten by a second interrupt, independent of whether the previous interrupt was already processed or not. This patch improves interrupt handling to avoid such races and loss of interrupt flags. The ideas are: * only the hdq_isr() sets bits in hdq_status * it does not reset any bits * it does wake_up() if any interrupt is pending * bits are only reset by the read/write/break functions if they were waited for * this makes sure that no interrupts can be lost * rx/tx/timeout bits are completely decoupled from each other (and not reset all after waiting for any of them) * which bits to reset is now specified by a new parameter to hdq_reset_irqstatus() * hdq_reset_irqstatus() also returns the state before resetting so that we can encapsulate the spinlock * this should now handle the case that the write and read are both already finished quickly before the hdq_write_byte() ends. * Or that two interrupts occur in succession before they are processed by the driver. Old code may have reset all status bits making the next hdq_read_byte() timeout. * the spinlock now always protects changing of bits in function hdq_reset_irqstatus() which could become a read-write-modify problem if the interrupt handler tries to read-modify-write exactly at the same moment * we add mutex protection also for hdq_write_byte() just to be safe to not to disturb a hdq_read_byte() triggered by some other thread/process. This patch was tested on a GTA04 and results in no boot problems any more. And first read after boot is now ok: root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=3970000 POWER_SUPPLY_CURRENT_NOW=354144 POWER_SUPPLY_CAPACITY=82 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=266 POWER_SUPPLY_TIME_TO_EMPTY_NOW=7680 POWER_SUPPLY_TIME_TO_EMPTY_AVG=7380 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=934856 POWER_SUPPLY_CHARGE_NOW=763976 POWER_SUPPLY_CHARGE_FULL_DESIGN=1233792 POWER_SUPPLY_CYCLE_COUNT=82 POWER_SUPPLY_ENERGY_NOW=2852840 POWER_SUPPLY_POWER_AVG=1392840 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real 0m0.233s user 0m0.000s sys 0m0.025s root@letux:~# It was also tested with dev_dbg enabled and more printk that all activities behave correctly, especially hdq_write_byte(), hdq_read_byte(), omap_hdq_break(). Not tested is omap_w1_triplet(). Fixes: 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by: H. Nikolaus Schaller <hns@goldelico.com> Link: https://lore.kernel.org/r/68fc8623ae741878beef049273696d2377526165.1590255176.git.hns@goldelico.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-05-24 01:32:56 +08:00
ret = mutex_lock_interruptible(&hdq_data->hdq_mutex);
if (ret < 0) {
ret = -EINTR;
goto rtn;
}
if (hdq_data->hdq_irqstatus)
dev_err(hdq_data->dev, "TX irqstatus not cleared (%02x)\n",
hdq_data->hdq_irqstatus);
*status = 0;
hdq_reg_out(hdq_data, OMAP_HDQ_TX_DATA, val);
/* set the GO bit */
hdq_reg_merge(hdq_data, OMAP_HDQ_CTRL_STATUS, OMAP_HDQ_CTRL_STATUS_GO,
OMAP_HDQ_CTRL_STATUS_DIR | OMAP_HDQ_CTRL_STATUS_GO);
/* wait for the TXCOMPLETE bit */
ret = wait_event_timeout(hdq_wait_queue,
w1: omap-hdq: fix interrupt handling which did show spurious timeouts Since commit 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") was applied, I did see timeouts and wrong values when reading a bq27000 connected to hdq of the omap3. This occurred mainly after boot but remained and only sometimes settled down after several reads. root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=0 POWER_SUPPLY_CURRENT_NOW=0 POWER_SUPPLY_CAPACITY=0 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=-2731 POWER_SUPPLY_TIME_TO_EMPTY_NOW=0 POWER_SUPPLY_TIME_TO_EMPTY_AVG=0 POWER_SUPPLY_TIME_TO_FULL_NOW=0 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=0 POWER_SUPPLY_CHARGE_NOW=0 POWER_SUPPLY_CHARGE_FULL_DESIGN=0 POWER_SUPPLY_CYCLE_COUNT=0 POWER_SUPPLY_ENERGY_NOW=0 POWER_SUPPLY_POWER_AVG=0 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real    0m15.761s user    0m0.001s sys     0m0.025s root@letux:~# Sometimes the effect did disappear after accessing the device multiple times, speed went up and results became correct. All this indicates that some interrupts from the hdq controller are lost by the driver. Enabling debugging revealed that there were spurious tx and rx timeouts, i.e. the driver does not always recognise interrupts. The main problem is that rx and tx interrupts share a single variable which was sometimes reset to 0 wiping out other interrupts. And it was overwritten by a second interrupt, independent of whether the previous interrupt was already processed or not. This patch improves interrupt handling to avoid such races and loss of interrupt flags. The ideas are: * only the hdq_isr() sets bits in hdq_status * it does not reset any bits * it does wake_up() if any interrupt is pending * bits are only reset by the read/write/break functions if they were waited for * this makes sure that no interrupts can be lost * rx/tx/timeout bits are completely decoupled from each other (and not reset all after waiting for any of them) * which bits to reset is now specified by a new parameter to hdq_reset_irqstatus() * hdq_reset_irqstatus() also returns the state before resetting so that we can encapsulate the spinlock * this should now handle the case that the write and read are both already finished quickly before the hdq_write_byte() ends. * Or that two interrupts occur in succession before they are processed by the driver. Old code may have reset all status bits making the next hdq_read_byte() timeout. * the spinlock now always protects changing of bits in function hdq_reset_irqstatus() which could become a read-write-modify problem if the interrupt handler tries to read-modify-write exactly at the same moment * we add mutex protection also for hdq_write_byte() just to be safe to not to disturb a hdq_read_byte() triggered by some other thread/process. This patch was tested on a GTA04 and results in no boot problems any more. And first read after boot is now ok: root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=3970000 POWER_SUPPLY_CURRENT_NOW=354144 POWER_SUPPLY_CAPACITY=82 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=266 POWER_SUPPLY_TIME_TO_EMPTY_NOW=7680 POWER_SUPPLY_TIME_TO_EMPTY_AVG=7380 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=934856 POWER_SUPPLY_CHARGE_NOW=763976 POWER_SUPPLY_CHARGE_FULL_DESIGN=1233792 POWER_SUPPLY_CYCLE_COUNT=82 POWER_SUPPLY_ENERGY_NOW=2852840 POWER_SUPPLY_POWER_AVG=1392840 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real 0m0.233s user 0m0.000s sys 0m0.025s root@letux:~# It was also tested with dev_dbg enabled and more printk that all activities behave correctly, especially hdq_write_byte(), hdq_read_byte(), omap_hdq_break(). Not tested is omap_w1_triplet(). Fixes: 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by: H. Nikolaus Schaller <hns@goldelico.com> Link: https://lore.kernel.org/r/68fc8623ae741878beef049273696d2377526165.1590255176.git.hns@goldelico.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-05-24 01:32:56 +08:00
(hdq_data->hdq_irqstatus & OMAP_HDQ_INT_STATUS_TXCOMPLETE),
OMAP_HDQ_TIMEOUT);
*status = hdq_reset_irqstatus(hdq_data, OMAP_HDQ_INT_STATUS_TXCOMPLETE);
if (ret == 0) {
dev_dbg(hdq_data->dev, "TX wait elapsed\n");
ret = -ETIMEDOUT;
goto out;
}
/* check irqstatus */
if (!(*status & OMAP_HDQ_INT_STATUS_TXCOMPLETE)) {
dev_dbg(hdq_data->dev, "timeout waiting for"
" TXCOMPLETE/RXCOMPLETE, %x\n", *status);
ret = -ETIMEDOUT;
goto out;
}
/* wait for the GO bit return to zero */
ret = hdq_wait_for_flag(hdq_data, OMAP_HDQ_CTRL_STATUS,
OMAP_HDQ_CTRL_STATUS_GO,
OMAP_HDQ_FLAG_CLEAR, &tmp_status);
if (ret) {
dev_dbg(hdq_data->dev, "timeout waiting GO bit"
" return to zero, %x\n", tmp_status);
}
out:
w1: omap-hdq: fix interrupt handling which did show spurious timeouts Since commit 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") was applied, I did see timeouts and wrong values when reading a bq27000 connected to hdq of the omap3. This occurred mainly after boot but remained and only sometimes settled down after several reads. root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=0 POWER_SUPPLY_CURRENT_NOW=0 POWER_SUPPLY_CAPACITY=0 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=-2731 POWER_SUPPLY_TIME_TO_EMPTY_NOW=0 POWER_SUPPLY_TIME_TO_EMPTY_AVG=0 POWER_SUPPLY_TIME_TO_FULL_NOW=0 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=0 POWER_SUPPLY_CHARGE_NOW=0 POWER_SUPPLY_CHARGE_FULL_DESIGN=0 POWER_SUPPLY_CYCLE_COUNT=0 POWER_SUPPLY_ENERGY_NOW=0 POWER_SUPPLY_POWER_AVG=0 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real    0m15.761s user    0m0.001s sys     0m0.025s root@letux:~# Sometimes the effect did disappear after accessing the device multiple times, speed went up and results became correct. All this indicates that some interrupts from the hdq controller are lost by the driver. Enabling debugging revealed that there were spurious tx and rx timeouts, i.e. the driver does not always recognise interrupts. The main problem is that rx and tx interrupts share a single variable which was sometimes reset to 0 wiping out other interrupts. And it was overwritten by a second interrupt, independent of whether the previous interrupt was already processed or not. This patch improves interrupt handling to avoid such races and loss of interrupt flags. The ideas are: * only the hdq_isr() sets bits in hdq_status * it does not reset any bits * it does wake_up() if any interrupt is pending * bits are only reset by the read/write/break functions if they were waited for * this makes sure that no interrupts can be lost * rx/tx/timeout bits are completely decoupled from each other (and not reset all after waiting for any of them) * which bits to reset is now specified by a new parameter to hdq_reset_irqstatus() * hdq_reset_irqstatus() also returns the state before resetting so that we can encapsulate the spinlock * this should now handle the case that the write and read are both already finished quickly before the hdq_write_byte() ends. * Or that two interrupts occur in succession before they are processed by the driver. Old code may have reset all status bits making the next hdq_read_byte() timeout. * the spinlock now always protects changing of bits in function hdq_reset_irqstatus() which could become a read-write-modify problem if the interrupt handler tries to read-modify-write exactly at the same moment * we add mutex protection also for hdq_write_byte() just to be safe to not to disturb a hdq_read_byte() triggered by some other thread/process. This patch was tested on a GTA04 and results in no boot problems any more. And first read after boot is now ok: root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=3970000 POWER_SUPPLY_CURRENT_NOW=354144 POWER_SUPPLY_CAPACITY=82 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=266 POWER_SUPPLY_TIME_TO_EMPTY_NOW=7680 POWER_SUPPLY_TIME_TO_EMPTY_AVG=7380 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=934856 POWER_SUPPLY_CHARGE_NOW=763976 POWER_SUPPLY_CHARGE_FULL_DESIGN=1233792 POWER_SUPPLY_CYCLE_COUNT=82 POWER_SUPPLY_ENERGY_NOW=2852840 POWER_SUPPLY_POWER_AVG=1392840 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real 0m0.233s user 0m0.000s sys 0m0.025s root@letux:~# It was also tested with dev_dbg enabled and more printk that all activities behave correctly, especially hdq_write_byte(), hdq_read_byte(), omap_hdq_break(). Not tested is omap_w1_triplet(). Fixes: 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by: H. Nikolaus Schaller <hns@goldelico.com> Link: https://lore.kernel.org/r/68fc8623ae741878beef049273696d2377526165.1590255176.git.hns@goldelico.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-05-24 01:32:56 +08:00
mutex_unlock(&hdq_data->hdq_mutex);
rtn:
return ret;
}
/* HDQ Interrupt service routine */
static irqreturn_t hdq_isr(int irq, void *_hdq)
{
struct hdq_data *hdq_data = _hdq;
unsigned long irqflags;
spin_lock_irqsave(&hdq_data->hdq_spinlock, irqflags);
w1: omap-hdq: fix interrupt handling which did show spurious timeouts Since commit 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") was applied, I did see timeouts and wrong values when reading a bq27000 connected to hdq of the omap3. This occurred mainly after boot but remained and only sometimes settled down after several reads. root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=0 POWER_SUPPLY_CURRENT_NOW=0 POWER_SUPPLY_CAPACITY=0 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=-2731 POWER_SUPPLY_TIME_TO_EMPTY_NOW=0 POWER_SUPPLY_TIME_TO_EMPTY_AVG=0 POWER_SUPPLY_TIME_TO_FULL_NOW=0 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=0 POWER_SUPPLY_CHARGE_NOW=0 POWER_SUPPLY_CHARGE_FULL_DESIGN=0 POWER_SUPPLY_CYCLE_COUNT=0 POWER_SUPPLY_ENERGY_NOW=0 POWER_SUPPLY_POWER_AVG=0 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real    0m15.761s user    0m0.001s sys     0m0.025s root@letux:~# Sometimes the effect did disappear after accessing the device multiple times, speed went up and results became correct. All this indicates that some interrupts from the hdq controller are lost by the driver. Enabling debugging revealed that there were spurious tx and rx timeouts, i.e. the driver does not always recognise interrupts. The main problem is that rx and tx interrupts share a single variable which was sometimes reset to 0 wiping out other interrupts. And it was overwritten by a second interrupt, independent of whether the previous interrupt was already processed or not. This patch improves interrupt handling to avoid such races and loss of interrupt flags. The ideas are: * only the hdq_isr() sets bits in hdq_status * it does not reset any bits * it does wake_up() if any interrupt is pending * bits are only reset by the read/write/break functions if they were waited for * this makes sure that no interrupts can be lost * rx/tx/timeout bits are completely decoupled from each other (and not reset all after waiting for any of them) * which bits to reset is now specified by a new parameter to hdq_reset_irqstatus() * hdq_reset_irqstatus() also returns the state before resetting so that we can encapsulate the spinlock * this should now handle the case that the write and read are both already finished quickly before the hdq_write_byte() ends. * Or that two interrupts occur in succession before they are processed by the driver. Old code may have reset all status bits making the next hdq_read_byte() timeout. * the spinlock now always protects changing of bits in function hdq_reset_irqstatus() which could become a read-write-modify problem if the interrupt handler tries to read-modify-write exactly at the same moment * we add mutex protection also for hdq_write_byte() just to be safe to not to disturb a hdq_read_byte() triggered by some other thread/process. This patch was tested on a GTA04 and results in no boot problems any more. And first read after boot is now ok: root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=3970000 POWER_SUPPLY_CURRENT_NOW=354144 POWER_SUPPLY_CAPACITY=82 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=266 POWER_SUPPLY_TIME_TO_EMPTY_NOW=7680 POWER_SUPPLY_TIME_TO_EMPTY_AVG=7380 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=934856 POWER_SUPPLY_CHARGE_NOW=763976 POWER_SUPPLY_CHARGE_FULL_DESIGN=1233792 POWER_SUPPLY_CYCLE_COUNT=82 POWER_SUPPLY_ENERGY_NOW=2852840 POWER_SUPPLY_POWER_AVG=1392840 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real 0m0.233s user 0m0.000s sys 0m0.025s root@letux:~# It was also tested with dev_dbg enabled and more printk that all activities behave correctly, especially hdq_write_byte(), hdq_read_byte(), omap_hdq_break(). Not tested is omap_w1_triplet(). Fixes: 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by: H. Nikolaus Schaller <hns@goldelico.com> Link: https://lore.kernel.org/r/68fc8623ae741878beef049273696d2377526165.1590255176.git.hns@goldelico.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-05-24 01:32:56 +08:00
hdq_data->hdq_irqstatus |= hdq_reg_in(hdq_data, OMAP_HDQ_INT_STATUS);
spin_unlock_irqrestore(&hdq_data->hdq_spinlock, irqflags);
dev_dbg(hdq_data->dev, "hdq_isr: %x\n", hdq_data->hdq_irqstatus);
if (hdq_data->hdq_irqstatus &
(OMAP_HDQ_INT_STATUS_TXCOMPLETE | OMAP_HDQ_INT_STATUS_RXCOMPLETE
| OMAP_HDQ_INT_STATUS_TIMEOUT)) {
/* wake up sleeping process */
wake_up(&hdq_wait_queue);
}
return IRQ_HANDLED;
}
/* W1 search callback function in HDQ mode */
static void omap_w1_search_bus(void *_hdq, struct w1_master *master_dev,
u8 search_type, w1_slave_found_callback slave_found)
{
u64 module_id, rn_le, cs, id;
if (w1_id)
module_id = w1_id;
else
module_id = 0x1;
rn_le = cpu_to_le64(module_id);
/*
* HDQ might not obey truly the 1-wire spec.
* So calculate CRC based on module parameter.
*/
cs = w1_calc_crc8((u8 *)&rn_le, 7);
id = (cs << 56) | module_id;
slave_found(master_dev, id);
}
/* Issue break pulse to the device */
static int omap_hdq_break(struct hdq_data *hdq_data)
{
int ret = 0;
u8 tmp_status;
ret = mutex_lock_interruptible(&hdq_data->hdq_mutex);
if (ret < 0) {
dev_dbg(hdq_data->dev, "Could not acquire mutex\n");
ret = -EINTR;
goto rtn;
}
if (hdq_data->hdq_irqstatus)
dev_err(hdq_data->dev, "break irqstatus not cleared (%02x)\n",
hdq_data->hdq_irqstatus);
/* set the INIT and GO bit */
hdq_reg_merge(hdq_data, OMAP_HDQ_CTRL_STATUS,
OMAP_HDQ_CTRL_STATUS_INITIALIZATION | OMAP_HDQ_CTRL_STATUS_GO,
OMAP_HDQ_CTRL_STATUS_DIR | OMAP_HDQ_CTRL_STATUS_INITIALIZATION |
OMAP_HDQ_CTRL_STATUS_GO);
/* wait for the TIMEOUT bit */
ret = wait_event_timeout(hdq_wait_queue,
w1: omap-hdq: fix interrupt handling which did show spurious timeouts Since commit 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") was applied, I did see timeouts and wrong values when reading a bq27000 connected to hdq of the omap3. This occurred mainly after boot but remained and only sometimes settled down after several reads. root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=0 POWER_SUPPLY_CURRENT_NOW=0 POWER_SUPPLY_CAPACITY=0 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=-2731 POWER_SUPPLY_TIME_TO_EMPTY_NOW=0 POWER_SUPPLY_TIME_TO_EMPTY_AVG=0 POWER_SUPPLY_TIME_TO_FULL_NOW=0 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=0 POWER_SUPPLY_CHARGE_NOW=0 POWER_SUPPLY_CHARGE_FULL_DESIGN=0 POWER_SUPPLY_CYCLE_COUNT=0 POWER_SUPPLY_ENERGY_NOW=0 POWER_SUPPLY_POWER_AVG=0 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real    0m15.761s user    0m0.001s sys     0m0.025s root@letux:~# Sometimes the effect did disappear after accessing the device multiple times, speed went up and results became correct. All this indicates that some interrupts from the hdq controller are lost by the driver. Enabling debugging revealed that there were spurious tx and rx timeouts, i.e. the driver does not always recognise interrupts. The main problem is that rx and tx interrupts share a single variable which was sometimes reset to 0 wiping out other interrupts. And it was overwritten by a second interrupt, independent of whether the previous interrupt was already processed or not. This patch improves interrupt handling to avoid such races and loss of interrupt flags. The ideas are: * only the hdq_isr() sets bits in hdq_status * it does not reset any bits * it does wake_up() if any interrupt is pending * bits are only reset by the read/write/break functions if they were waited for * this makes sure that no interrupts can be lost * rx/tx/timeout bits are completely decoupled from each other (and not reset all after waiting for any of them) * which bits to reset is now specified by a new parameter to hdq_reset_irqstatus() * hdq_reset_irqstatus() also returns the state before resetting so that we can encapsulate the spinlock * this should now handle the case that the write and read are both already finished quickly before the hdq_write_byte() ends. * Or that two interrupts occur in succession before they are processed by the driver. Old code may have reset all status bits making the next hdq_read_byte() timeout. * the spinlock now always protects changing of bits in function hdq_reset_irqstatus() which could become a read-write-modify problem if the interrupt handler tries to read-modify-write exactly at the same moment * we add mutex protection also for hdq_write_byte() just to be safe to not to disturb a hdq_read_byte() triggered by some other thread/process. This patch was tested on a GTA04 and results in no boot problems any more. And first read after boot is now ok: root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=3970000 POWER_SUPPLY_CURRENT_NOW=354144 POWER_SUPPLY_CAPACITY=82 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=266 POWER_SUPPLY_TIME_TO_EMPTY_NOW=7680 POWER_SUPPLY_TIME_TO_EMPTY_AVG=7380 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=934856 POWER_SUPPLY_CHARGE_NOW=763976 POWER_SUPPLY_CHARGE_FULL_DESIGN=1233792 POWER_SUPPLY_CYCLE_COUNT=82 POWER_SUPPLY_ENERGY_NOW=2852840 POWER_SUPPLY_POWER_AVG=1392840 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real 0m0.233s user 0m0.000s sys 0m0.025s root@letux:~# It was also tested with dev_dbg enabled and more printk that all activities behave correctly, especially hdq_write_byte(), hdq_read_byte(), omap_hdq_break(). Not tested is omap_w1_triplet(). Fixes: 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by: H. Nikolaus Schaller <hns@goldelico.com> Link: https://lore.kernel.org/r/68fc8623ae741878beef049273696d2377526165.1590255176.git.hns@goldelico.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-05-24 01:32:56 +08:00
(hdq_data->hdq_irqstatus & OMAP_HDQ_INT_STATUS_TIMEOUT),
OMAP_HDQ_TIMEOUT);
tmp_status = hdq_reset_irqstatus(hdq_data, OMAP_HDQ_INT_STATUS_TIMEOUT);
if (ret == 0) {
dev_dbg(hdq_data->dev, "break wait elapsed\n");
ret = -EINTR;
goto out;
}
/* check irqstatus */
if (!(tmp_status & OMAP_HDQ_INT_STATUS_TIMEOUT)) {
dev_dbg(hdq_data->dev, "timeout waiting for TIMEOUT, %x\n",
w1: omap-hdq: fix interrupt handling which did show spurious timeouts Since commit 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") was applied, I did see timeouts and wrong values when reading a bq27000 connected to hdq of the omap3. This occurred mainly after boot but remained and only sometimes settled down after several reads. root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=0 POWER_SUPPLY_CURRENT_NOW=0 POWER_SUPPLY_CAPACITY=0 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=-2731 POWER_SUPPLY_TIME_TO_EMPTY_NOW=0 POWER_SUPPLY_TIME_TO_EMPTY_AVG=0 POWER_SUPPLY_TIME_TO_FULL_NOW=0 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=0 POWER_SUPPLY_CHARGE_NOW=0 POWER_SUPPLY_CHARGE_FULL_DESIGN=0 POWER_SUPPLY_CYCLE_COUNT=0 POWER_SUPPLY_ENERGY_NOW=0 POWER_SUPPLY_POWER_AVG=0 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real    0m15.761s user    0m0.001s sys     0m0.025s root@letux:~# Sometimes the effect did disappear after accessing the device multiple times, speed went up and results became correct. All this indicates that some interrupts from the hdq controller are lost by the driver. Enabling debugging revealed that there were spurious tx and rx timeouts, i.e. the driver does not always recognise interrupts. The main problem is that rx and tx interrupts share a single variable which was sometimes reset to 0 wiping out other interrupts. And it was overwritten by a second interrupt, independent of whether the previous interrupt was already processed or not. This patch improves interrupt handling to avoid such races and loss of interrupt flags. The ideas are: * only the hdq_isr() sets bits in hdq_status * it does not reset any bits * it does wake_up() if any interrupt is pending * bits are only reset by the read/write/break functions if they were waited for * this makes sure that no interrupts can be lost * rx/tx/timeout bits are completely decoupled from each other (and not reset all after waiting for any of them) * which bits to reset is now specified by a new parameter to hdq_reset_irqstatus() * hdq_reset_irqstatus() also returns the state before resetting so that we can encapsulate the spinlock * this should now handle the case that the write and read are both already finished quickly before the hdq_write_byte() ends. * Or that two interrupts occur in succession before they are processed by the driver. Old code may have reset all status bits making the next hdq_read_byte() timeout. * the spinlock now always protects changing of bits in function hdq_reset_irqstatus() which could become a read-write-modify problem if the interrupt handler tries to read-modify-write exactly at the same moment * we add mutex protection also for hdq_write_byte() just to be safe to not to disturb a hdq_read_byte() triggered by some other thread/process. This patch was tested on a GTA04 and results in no boot problems any more. And first read after boot is now ok: root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=3970000 POWER_SUPPLY_CURRENT_NOW=354144 POWER_SUPPLY_CAPACITY=82 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=266 POWER_SUPPLY_TIME_TO_EMPTY_NOW=7680 POWER_SUPPLY_TIME_TO_EMPTY_AVG=7380 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=934856 POWER_SUPPLY_CHARGE_NOW=763976 POWER_SUPPLY_CHARGE_FULL_DESIGN=1233792 POWER_SUPPLY_CYCLE_COUNT=82 POWER_SUPPLY_ENERGY_NOW=2852840 POWER_SUPPLY_POWER_AVG=1392840 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real 0m0.233s user 0m0.000s sys 0m0.025s root@letux:~# It was also tested with dev_dbg enabled and more printk that all activities behave correctly, especially hdq_write_byte(), hdq_read_byte(), omap_hdq_break(). Not tested is omap_w1_triplet(). Fixes: 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by: H. Nikolaus Schaller <hns@goldelico.com> Link: https://lore.kernel.org/r/68fc8623ae741878beef049273696d2377526165.1590255176.git.hns@goldelico.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-05-24 01:32:56 +08:00
tmp_status);
ret = -ETIMEDOUT;
goto out;
}
/*
* check for the presence detect bit to get
* set to show that the slave is responding
*/
if (!(hdq_reg_in(hdq_data, OMAP_HDQ_CTRL_STATUS) &
OMAP_HDQ_CTRL_STATUS_PRESENCE)) {
dev_dbg(hdq_data->dev, "Presence bit not set\n");
ret = -ETIMEDOUT;
goto out;
}
/*
* wait for both INIT and GO bits rerurn to zero.
* zero wait time expected for interrupt mode.
*/
ret = hdq_wait_for_flag(hdq_data, OMAP_HDQ_CTRL_STATUS,
OMAP_HDQ_CTRL_STATUS_INITIALIZATION |
OMAP_HDQ_CTRL_STATUS_GO, OMAP_HDQ_FLAG_CLEAR,
&tmp_status);
if (ret)
dev_dbg(hdq_data->dev, "timeout waiting INIT&GO bits"
" return to zero, %x\n", tmp_status);
out:
mutex_unlock(&hdq_data->hdq_mutex);
rtn:
return ret;
}
static int hdq_read_byte(struct hdq_data *hdq_data, u8 *val)
{
int ret = 0;
u8 status;
ret = mutex_lock_interruptible(&hdq_data->hdq_mutex);
if (ret < 0) {
ret = -EINTR;
goto rtn;
}
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
if (pm_runtime_suspended(hdq_data->dev)) {
ret = -EINVAL;
goto out;
}
if (!(hdq_data->hdq_irqstatus & OMAP_HDQ_INT_STATUS_RXCOMPLETE)) {
hdq_reg_merge(hdq_data, OMAP_HDQ_CTRL_STATUS,
OMAP_HDQ_CTRL_STATUS_DIR | OMAP_HDQ_CTRL_STATUS_GO,
OMAP_HDQ_CTRL_STATUS_DIR | OMAP_HDQ_CTRL_STATUS_GO);
/*
* The RX comes immediately after TX.
*/
wait_event_timeout(hdq_wait_queue,
(hdq_data->hdq_irqstatus
w1: omap-hdq: fix interrupt handling which did show spurious timeouts Since commit 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") was applied, I did see timeouts and wrong values when reading a bq27000 connected to hdq of the omap3. This occurred mainly after boot but remained and only sometimes settled down after several reads. root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=0 POWER_SUPPLY_CURRENT_NOW=0 POWER_SUPPLY_CAPACITY=0 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=-2731 POWER_SUPPLY_TIME_TO_EMPTY_NOW=0 POWER_SUPPLY_TIME_TO_EMPTY_AVG=0 POWER_SUPPLY_TIME_TO_FULL_NOW=0 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=0 POWER_SUPPLY_CHARGE_NOW=0 POWER_SUPPLY_CHARGE_FULL_DESIGN=0 POWER_SUPPLY_CYCLE_COUNT=0 POWER_SUPPLY_ENERGY_NOW=0 POWER_SUPPLY_POWER_AVG=0 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real    0m15.761s user    0m0.001s sys     0m0.025s root@letux:~# Sometimes the effect did disappear after accessing the device multiple times, speed went up and results became correct. All this indicates that some interrupts from the hdq controller are lost by the driver. Enabling debugging revealed that there were spurious tx and rx timeouts, i.e. the driver does not always recognise interrupts. The main problem is that rx and tx interrupts share a single variable which was sometimes reset to 0 wiping out other interrupts. And it was overwritten by a second interrupt, independent of whether the previous interrupt was already processed or not. This patch improves interrupt handling to avoid such races and loss of interrupt flags. The ideas are: * only the hdq_isr() sets bits in hdq_status * it does not reset any bits * it does wake_up() if any interrupt is pending * bits are only reset by the read/write/break functions if they were waited for * this makes sure that no interrupts can be lost * rx/tx/timeout bits are completely decoupled from each other (and not reset all after waiting for any of them) * which bits to reset is now specified by a new parameter to hdq_reset_irqstatus() * hdq_reset_irqstatus() also returns the state before resetting so that we can encapsulate the spinlock * this should now handle the case that the write and read are both already finished quickly before the hdq_write_byte() ends. * Or that two interrupts occur in succession before they are processed by the driver. Old code may have reset all status bits making the next hdq_read_byte() timeout. * the spinlock now always protects changing of bits in function hdq_reset_irqstatus() which could become a read-write-modify problem if the interrupt handler tries to read-modify-write exactly at the same moment * we add mutex protection also for hdq_write_byte() just to be safe to not to disturb a hdq_read_byte() triggered by some other thread/process. This patch was tested on a GTA04 and results in no boot problems any more. And first read after boot is now ok: root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=3970000 POWER_SUPPLY_CURRENT_NOW=354144 POWER_SUPPLY_CAPACITY=82 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=266 POWER_SUPPLY_TIME_TO_EMPTY_NOW=7680 POWER_SUPPLY_TIME_TO_EMPTY_AVG=7380 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=934856 POWER_SUPPLY_CHARGE_NOW=763976 POWER_SUPPLY_CHARGE_FULL_DESIGN=1233792 POWER_SUPPLY_CYCLE_COUNT=82 POWER_SUPPLY_ENERGY_NOW=2852840 POWER_SUPPLY_POWER_AVG=1392840 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real 0m0.233s user 0m0.000s sys 0m0.025s root@letux:~# It was also tested with dev_dbg enabled and more printk that all activities behave correctly, especially hdq_write_byte(), hdq_read_byte(), omap_hdq_break(). Not tested is omap_w1_triplet(). Fixes: 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by: H. Nikolaus Schaller <hns@goldelico.com> Link: https://lore.kernel.org/r/68fc8623ae741878beef049273696d2377526165.1590255176.git.hns@goldelico.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-05-24 01:32:56 +08:00
& (OMAP_HDQ_INT_STATUS_RXCOMPLETE |
OMAP_HDQ_INT_STATUS_TIMEOUT)),
OMAP_HDQ_TIMEOUT);
w1: omap-hdq: fix interrupt handling which did show spurious timeouts Since commit 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") was applied, I did see timeouts and wrong values when reading a bq27000 connected to hdq of the omap3. This occurred mainly after boot but remained and only sometimes settled down after several reads. root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=0 POWER_SUPPLY_CURRENT_NOW=0 POWER_SUPPLY_CAPACITY=0 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=-2731 POWER_SUPPLY_TIME_TO_EMPTY_NOW=0 POWER_SUPPLY_TIME_TO_EMPTY_AVG=0 POWER_SUPPLY_TIME_TO_FULL_NOW=0 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=0 POWER_SUPPLY_CHARGE_NOW=0 POWER_SUPPLY_CHARGE_FULL_DESIGN=0 POWER_SUPPLY_CYCLE_COUNT=0 POWER_SUPPLY_ENERGY_NOW=0 POWER_SUPPLY_POWER_AVG=0 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real    0m15.761s user    0m0.001s sys     0m0.025s root@letux:~# Sometimes the effect did disappear after accessing the device multiple times, speed went up and results became correct. All this indicates that some interrupts from the hdq controller are lost by the driver. Enabling debugging revealed that there were spurious tx and rx timeouts, i.e. the driver does not always recognise interrupts. The main problem is that rx and tx interrupts share a single variable which was sometimes reset to 0 wiping out other interrupts. And it was overwritten by a second interrupt, independent of whether the previous interrupt was already processed or not. This patch improves interrupt handling to avoid such races and loss of interrupt flags. The ideas are: * only the hdq_isr() sets bits in hdq_status * it does not reset any bits * it does wake_up() if any interrupt is pending * bits are only reset by the read/write/break functions if they were waited for * this makes sure that no interrupts can be lost * rx/tx/timeout bits are completely decoupled from each other (and not reset all after waiting for any of them) * which bits to reset is now specified by a new parameter to hdq_reset_irqstatus() * hdq_reset_irqstatus() also returns the state before resetting so that we can encapsulate the spinlock * this should now handle the case that the write and read are both already finished quickly before the hdq_write_byte() ends. * Or that two interrupts occur in succession before they are processed by the driver. Old code may have reset all status bits making the next hdq_read_byte() timeout. * the spinlock now always protects changing of bits in function hdq_reset_irqstatus() which could become a read-write-modify problem if the interrupt handler tries to read-modify-write exactly at the same moment * we add mutex protection also for hdq_write_byte() just to be safe to not to disturb a hdq_read_byte() triggered by some other thread/process. This patch was tested on a GTA04 and results in no boot problems any more. And first read after boot is now ok: root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=3970000 POWER_SUPPLY_CURRENT_NOW=354144 POWER_SUPPLY_CAPACITY=82 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=266 POWER_SUPPLY_TIME_TO_EMPTY_NOW=7680 POWER_SUPPLY_TIME_TO_EMPTY_AVG=7380 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=934856 POWER_SUPPLY_CHARGE_NOW=763976 POWER_SUPPLY_CHARGE_FULL_DESIGN=1233792 POWER_SUPPLY_CYCLE_COUNT=82 POWER_SUPPLY_ENERGY_NOW=2852840 POWER_SUPPLY_POWER_AVG=1392840 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real 0m0.233s user 0m0.000s sys 0m0.025s root@letux:~# It was also tested with dev_dbg enabled and more printk that all activities behave correctly, especially hdq_write_byte(), hdq_read_byte(), omap_hdq_break(). Not tested is omap_w1_triplet(). Fixes: 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by: H. Nikolaus Schaller <hns@goldelico.com> Link: https://lore.kernel.org/r/68fc8623ae741878beef049273696d2377526165.1590255176.git.hns@goldelico.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-05-24 01:32:56 +08:00
status = hdq_reset_irqstatus(hdq_data,
OMAP_HDQ_INT_STATUS_RXCOMPLETE |
OMAP_HDQ_INT_STATUS_TIMEOUT);
hdq_reg_merge(hdq_data, OMAP_HDQ_CTRL_STATUS, 0,
OMAP_HDQ_CTRL_STATUS_DIR);
w1: omap-hdq: fix interrupt handling which did show spurious timeouts Since commit 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") was applied, I did see timeouts and wrong values when reading a bq27000 connected to hdq of the omap3. This occurred mainly after boot but remained and only sometimes settled down after several reads. root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=0 POWER_SUPPLY_CURRENT_NOW=0 POWER_SUPPLY_CAPACITY=0 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=-2731 POWER_SUPPLY_TIME_TO_EMPTY_NOW=0 POWER_SUPPLY_TIME_TO_EMPTY_AVG=0 POWER_SUPPLY_TIME_TO_FULL_NOW=0 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=0 POWER_SUPPLY_CHARGE_NOW=0 POWER_SUPPLY_CHARGE_FULL_DESIGN=0 POWER_SUPPLY_CYCLE_COUNT=0 POWER_SUPPLY_ENERGY_NOW=0 POWER_SUPPLY_POWER_AVG=0 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real    0m15.761s user    0m0.001s sys     0m0.025s root@letux:~# Sometimes the effect did disappear after accessing the device multiple times, speed went up and results became correct. All this indicates that some interrupts from the hdq controller are lost by the driver. Enabling debugging revealed that there were spurious tx and rx timeouts, i.e. the driver does not always recognise interrupts. The main problem is that rx and tx interrupts share a single variable which was sometimes reset to 0 wiping out other interrupts. And it was overwritten by a second interrupt, independent of whether the previous interrupt was already processed or not. This patch improves interrupt handling to avoid such races and loss of interrupt flags. The ideas are: * only the hdq_isr() sets bits in hdq_status * it does not reset any bits * it does wake_up() if any interrupt is pending * bits are only reset by the read/write/break functions if they were waited for * this makes sure that no interrupts can be lost * rx/tx/timeout bits are completely decoupled from each other (and not reset all after waiting for any of them) * which bits to reset is now specified by a new parameter to hdq_reset_irqstatus() * hdq_reset_irqstatus() also returns the state before resetting so that we can encapsulate the spinlock * this should now handle the case that the write and read are both already finished quickly before the hdq_write_byte() ends. * Or that two interrupts occur in succession before they are processed by the driver. Old code may have reset all status bits making the next hdq_read_byte() timeout. * the spinlock now always protects changing of bits in function hdq_reset_irqstatus() which could become a read-write-modify problem if the interrupt handler tries to read-modify-write exactly at the same moment * we add mutex protection also for hdq_write_byte() just to be safe to not to disturb a hdq_read_byte() triggered by some other thread/process. This patch was tested on a GTA04 and results in no boot problems any more. And first read after boot is now ok: root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=3970000 POWER_SUPPLY_CURRENT_NOW=354144 POWER_SUPPLY_CAPACITY=82 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=266 POWER_SUPPLY_TIME_TO_EMPTY_NOW=7680 POWER_SUPPLY_TIME_TO_EMPTY_AVG=7380 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=934856 POWER_SUPPLY_CHARGE_NOW=763976 POWER_SUPPLY_CHARGE_FULL_DESIGN=1233792 POWER_SUPPLY_CYCLE_COUNT=82 POWER_SUPPLY_ENERGY_NOW=2852840 POWER_SUPPLY_POWER_AVG=1392840 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real 0m0.233s user 0m0.000s sys 0m0.025s root@letux:~# It was also tested with dev_dbg enabled and more printk that all activities behave correctly, especially hdq_write_byte(), hdq_read_byte(), omap_hdq_break(). Not tested is omap_w1_triplet(). Fixes: 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by: H. Nikolaus Schaller <hns@goldelico.com> Link: https://lore.kernel.org/r/68fc8623ae741878beef049273696d2377526165.1590255176.git.hns@goldelico.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-05-24 01:32:56 +08:00
/* check irqstatus */
if (!(status & OMAP_HDQ_INT_STATUS_RXCOMPLETE)) {
dev_dbg(hdq_data->dev, "timeout waiting for"
" RXCOMPLETE, %x", status);
ret = -ETIMEDOUT;
goto out;
}
w1: omap-hdq: fix interrupt handling which did show spurious timeouts Since commit 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") was applied, I did see timeouts and wrong values when reading a bq27000 connected to hdq of the omap3. This occurred mainly after boot but remained and only sometimes settled down after several reads. root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=0 POWER_SUPPLY_CURRENT_NOW=0 POWER_SUPPLY_CAPACITY=0 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=-2731 POWER_SUPPLY_TIME_TO_EMPTY_NOW=0 POWER_SUPPLY_TIME_TO_EMPTY_AVG=0 POWER_SUPPLY_TIME_TO_FULL_NOW=0 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=0 POWER_SUPPLY_CHARGE_NOW=0 POWER_SUPPLY_CHARGE_FULL_DESIGN=0 POWER_SUPPLY_CYCLE_COUNT=0 POWER_SUPPLY_ENERGY_NOW=0 POWER_SUPPLY_POWER_AVG=0 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real    0m15.761s user    0m0.001s sys     0m0.025s root@letux:~# Sometimes the effect did disappear after accessing the device multiple times, speed went up and results became correct. All this indicates that some interrupts from the hdq controller are lost by the driver. Enabling debugging revealed that there were spurious tx and rx timeouts, i.e. the driver does not always recognise interrupts. The main problem is that rx and tx interrupts share a single variable which was sometimes reset to 0 wiping out other interrupts. And it was overwritten by a second interrupt, independent of whether the previous interrupt was already processed or not. This patch improves interrupt handling to avoid such races and loss of interrupt flags. The ideas are: * only the hdq_isr() sets bits in hdq_status * it does not reset any bits * it does wake_up() if any interrupt is pending * bits are only reset by the read/write/break functions if they were waited for * this makes sure that no interrupts can be lost * rx/tx/timeout bits are completely decoupled from each other (and not reset all after waiting for any of them) * which bits to reset is now specified by a new parameter to hdq_reset_irqstatus() * hdq_reset_irqstatus() also returns the state before resetting so that we can encapsulate the spinlock * this should now handle the case that the write and read are both already finished quickly before the hdq_write_byte() ends. * Or that two interrupts occur in succession before they are processed by the driver. Old code may have reset all status bits making the next hdq_read_byte() timeout. * the spinlock now always protects changing of bits in function hdq_reset_irqstatus() which could become a read-write-modify problem if the interrupt handler tries to read-modify-write exactly at the same moment * we add mutex protection also for hdq_write_byte() just to be safe to not to disturb a hdq_read_byte() triggered by some other thread/process. This patch was tested on a GTA04 and results in no boot problems any more. And first read after boot is now ok: root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=3970000 POWER_SUPPLY_CURRENT_NOW=354144 POWER_SUPPLY_CAPACITY=82 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=266 POWER_SUPPLY_TIME_TO_EMPTY_NOW=7680 POWER_SUPPLY_TIME_TO_EMPTY_AVG=7380 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=934856 POWER_SUPPLY_CHARGE_NOW=763976 POWER_SUPPLY_CHARGE_FULL_DESIGN=1233792 POWER_SUPPLY_CYCLE_COUNT=82 POWER_SUPPLY_ENERGY_NOW=2852840 POWER_SUPPLY_POWER_AVG=1392840 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real 0m0.233s user 0m0.000s sys 0m0.025s root@letux:~# It was also tested with dev_dbg enabled and more printk that all activities behave correctly, especially hdq_write_byte(), hdq_read_byte(), omap_hdq_break(). Not tested is omap_w1_triplet(). Fixes: 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by: H. Nikolaus Schaller <hns@goldelico.com> Link: https://lore.kernel.org/r/68fc8623ae741878beef049273696d2377526165.1590255176.git.hns@goldelico.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-05-24 01:32:56 +08:00
} else { /* interrupt had occurred before hdq_read_byte was called */
hdq_reset_irqstatus(hdq_data, OMAP_HDQ_INT_STATUS_RXCOMPLETE);
}
/* the data is ready. Read it in! */
*val = hdq_reg_in(hdq_data, OMAP_HDQ_RX_DATA);
out:
mutex_unlock(&hdq_data->hdq_mutex);
rtn:
return ret;
}
/*
* W1 triplet callback function - used for searching ROM addresses.
* Registered only when controller is in 1-wire mode.
*/
static u8 omap_w1_triplet(void *_hdq, u8 bdir)
{
u8 id_bit, comp_bit;
int err;
u8 ret = 0x3; /* no slaves responded */
struct hdq_data *hdq_data = _hdq;
u8 ctrl = OMAP_HDQ_CTRL_STATUS_SINGLE | OMAP_HDQ_CTRL_STATUS_GO |
OMAP_HDQ_CTRL_STATUS_INTERRUPTMASK;
u8 mask = ctrl | OMAP_HDQ_CTRL_STATUS_DIR;
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
err = pm_runtime_get_sync(hdq_data->dev);
if (err < 0) {
pm_runtime_put_noidle(hdq_data->dev);
return err;
}
err = mutex_lock_interruptible(&hdq_data->hdq_mutex);
if (err < 0) {
dev_dbg(hdq_data->dev, "Could not acquire mutex\n");
goto rtn;
}
/* read id_bit */
hdq_reg_merge(_hdq, OMAP_HDQ_CTRL_STATUS,
ctrl | OMAP_HDQ_CTRL_STATUS_DIR, mask);
err = wait_event_timeout(hdq_wait_queue,
(hdq_data->hdq_irqstatus
& OMAP_HDQ_INT_STATUS_RXCOMPLETE),
OMAP_HDQ_TIMEOUT);
w1: omap-hdq: fix interrupt handling which did show spurious timeouts Since commit 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") was applied, I did see timeouts and wrong values when reading a bq27000 connected to hdq of the omap3. This occurred mainly after boot but remained and only sometimes settled down after several reads. root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=0 POWER_SUPPLY_CURRENT_NOW=0 POWER_SUPPLY_CAPACITY=0 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=-2731 POWER_SUPPLY_TIME_TO_EMPTY_NOW=0 POWER_SUPPLY_TIME_TO_EMPTY_AVG=0 POWER_SUPPLY_TIME_TO_FULL_NOW=0 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=0 POWER_SUPPLY_CHARGE_NOW=0 POWER_SUPPLY_CHARGE_FULL_DESIGN=0 POWER_SUPPLY_CYCLE_COUNT=0 POWER_SUPPLY_ENERGY_NOW=0 POWER_SUPPLY_POWER_AVG=0 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real    0m15.761s user    0m0.001s sys     0m0.025s root@letux:~# Sometimes the effect did disappear after accessing the device multiple times, speed went up and results became correct. All this indicates that some interrupts from the hdq controller are lost by the driver. Enabling debugging revealed that there were spurious tx and rx timeouts, i.e. the driver does not always recognise interrupts. The main problem is that rx and tx interrupts share a single variable which was sometimes reset to 0 wiping out other interrupts. And it was overwritten by a second interrupt, independent of whether the previous interrupt was already processed or not. This patch improves interrupt handling to avoid such races and loss of interrupt flags. The ideas are: * only the hdq_isr() sets bits in hdq_status * it does not reset any bits * it does wake_up() if any interrupt is pending * bits are only reset by the read/write/break functions if they were waited for * this makes sure that no interrupts can be lost * rx/tx/timeout bits are completely decoupled from each other (and not reset all after waiting for any of them) * which bits to reset is now specified by a new parameter to hdq_reset_irqstatus() * hdq_reset_irqstatus() also returns the state before resetting so that we can encapsulate the spinlock * this should now handle the case that the write and read are both already finished quickly before the hdq_write_byte() ends. * Or that two interrupts occur in succession before they are processed by the driver. Old code may have reset all status bits making the next hdq_read_byte() timeout. * the spinlock now always protects changing of bits in function hdq_reset_irqstatus() which could become a read-write-modify problem if the interrupt handler tries to read-modify-write exactly at the same moment * we add mutex protection also for hdq_write_byte() just to be safe to not to disturb a hdq_read_byte() triggered by some other thread/process. This patch was tested on a GTA04 and results in no boot problems any more. And first read after boot is now ok: root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=3970000 POWER_SUPPLY_CURRENT_NOW=354144 POWER_SUPPLY_CAPACITY=82 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=266 POWER_SUPPLY_TIME_TO_EMPTY_NOW=7680 POWER_SUPPLY_TIME_TO_EMPTY_AVG=7380 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=934856 POWER_SUPPLY_CHARGE_NOW=763976 POWER_SUPPLY_CHARGE_FULL_DESIGN=1233792 POWER_SUPPLY_CYCLE_COUNT=82 POWER_SUPPLY_ENERGY_NOW=2852840 POWER_SUPPLY_POWER_AVG=1392840 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real 0m0.233s user 0m0.000s sys 0m0.025s root@letux:~# It was also tested with dev_dbg enabled and more printk that all activities behave correctly, especially hdq_write_byte(), hdq_read_byte(), omap_hdq_break(). Not tested is omap_w1_triplet(). Fixes: 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by: H. Nikolaus Schaller <hns@goldelico.com> Link: https://lore.kernel.org/r/68fc8623ae741878beef049273696d2377526165.1590255176.git.hns@goldelico.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-05-24 01:32:56 +08:00
/* Must clear irqstatus for another RXCOMPLETE interrupt */
hdq_reset_irqstatus(hdq_data, OMAP_HDQ_INT_STATUS_RXCOMPLETE);
if (err == 0) {
dev_dbg(hdq_data->dev, "RX wait elapsed\n");
goto out;
}
id_bit = (hdq_reg_in(_hdq, OMAP_HDQ_RX_DATA) & 0x01);
/* read comp_bit */
hdq_reg_merge(_hdq, OMAP_HDQ_CTRL_STATUS,
ctrl | OMAP_HDQ_CTRL_STATUS_DIR, mask);
err = wait_event_timeout(hdq_wait_queue,
(hdq_data->hdq_irqstatus
& OMAP_HDQ_INT_STATUS_RXCOMPLETE),
OMAP_HDQ_TIMEOUT);
w1: omap-hdq: fix interrupt handling which did show spurious timeouts Since commit 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") was applied, I did see timeouts and wrong values when reading a bq27000 connected to hdq of the omap3. This occurred mainly after boot but remained and only sometimes settled down after several reads. root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=0 POWER_SUPPLY_CURRENT_NOW=0 POWER_SUPPLY_CAPACITY=0 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=-2731 POWER_SUPPLY_TIME_TO_EMPTY_NOW=0 POWER_SUPPLY_TIME_TO_EMPTY_AVG=0 POWER_SUPPLY_TIME_TO_FULL_NOW=0 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=0 POWER_SUPPLY_CHARGE_NOW=0 POWER_SUPPLY_CHARGE_FULL_DESIGN=0 POWER_SUPPLY_CYCLE_COUNT=0 POWER_SUPPLY_ENERGY_NOW=0 POWER_SUPPLY_POWER_AVG=0 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real    0m15.761s user    0m0.001s sys     0m0.025s root@letux:~# Sometimes the effect did disappear after accessing the device multiple times, speed went up and results became correct. All this indicates that some interrupts from the hdq controller are lost by the driver. Enabling debugging revealed that there were spurious tx and rx timeouts, i.e. the driver does not always recognise interrupts. The main problem is that rx and tx interrupts share a single variable which was sometimes reset to 0 wiping out other interrupts. And it was overwritten by a second interrupt, independent of whether the previous interrupt was already processed or not. This patch improves interrupt handling to avoid such races and loss of interrupt flags. The ideas are: * only the hdq_isr() sets bits in hdq_status * it does not reset any bits * it does wake_up() if any interrupt is pending * bits are only reset by the read/write/break functions if they were waited for * this makes sure that no interrupts can be lost * rx/tx/timeout bits are completely decoupled from each other (and not reset all after waiting for any of them) * which bits to reset is now specified by a new parameter to hdq_reset_irqstatus() * hdq_reset_irqstatus() also returns the state before resetting so that we can encapsulate the spinlock * this should now handle the case that the write and read are both already finished quickly before the hdq_write_byte() ends. * Or that two interrupts occur in succession before they are processed by the driver. Old code may have reset all status bits making the next hdq_read_byte() timeout. * the spinlock now always protects changing of bits in function hdq_reset_irqstatus() which could become a read-write-modify problem if the interrupt handler tries to read-modify-write exactly at the same moment * we add mutex protection also for hdq_write_byte() just to be safe to not to disturb a hdq_read_byte() triggered by some other thread/process. This patch was tested on a GTA04 and results in no boot problems any more. And first read after boot is now ok: root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=3970000 POWER_SUPPLY_CURRENT_NOW=354144 POWER_SUPPLY_CAPACITY=82 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=266 POWER_SUPPLY_TIME_TO_EMPTY_NOW=7680 POWER_SUPPLY_TIME_TO_EMPTY_AVG=7380 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=934856 POWER_SUPPLY_CHARGE_NOW=763976 POWER_SUPPLY_CHARGE_FULL_DESIGN=1233792 POWER_SUPPLY_CYCLE_COUNT=82 POWER_SUPPLY_ENERGY_NOW=2852840 POWER_SUPPLY_POWER_AVG=1392840 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real 0m0.233s user 0m0.000s sys 0m0.025s root@letux:~# It was also tested with dev_dbg enabled and more printk that all activities behave correctly, especially hdq_write_byte(), hdq_read_byte(), omap_hdq_break(). Not tested is omap_w1_triplet(). Fixes: 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by: H. Nikolaus Schaller <hns@goldelico.com> Link: https://lore.kernel.org/r/68fc8623ae741878beef049273696d2377526165.1590255176.git.hns@goldelico.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-05-24 01:32:56 +08:00
/* Must clear irqstatus for another RXCOMPLETE interrupt */
hdq_reset_irqstatus(hdq_data, OMAP_HDQ_INT_STATUS_RXCOMPLETE);
if (err == 0) {
dev_dbg(hdq_data->dev, "RX wait elapsed\n");
goto out;
}
comp_bit = (hdq_reg_in(_hdq, OMAP_HDQ_RX_DATA) & 0x01);
if (id_bit && comp_bit) {
ret = 0x03; /* no slaves responded */
goto out;
}
if (!id_bit && !comp_bit) {
/* Both bits are valid, take the direction given */
ret = bdir ? 0x04 : 0;
} else {
/* Only one bit is valid, take that direction */
bdir = id_bit;
ret = id_bit ? 0x05 : 0x02;
}
/* write bdir bit */
hdq_reg_out(_hdq, OMAP_HDQ_TX_DATA, bdir);
hdq_reg_merge(_hdq, OMAP_HDQ_CTRL_STATUS, ctrl, mask);
err = wait_event_timeout(hdq_wait_queue,
(hdq_data->hdq_irqstatus
& OMAP_HDQ_INT_STATUS_TXCOMPLETE),
OMAP_HDQ_TIMEOUT);
w1: omap-hdq: fix interrupt handling which did show spurious timeouts Since commit 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") was applied, I did see timeouts and wrong values when reading a bq27000 connected to hdq of the omap3. This occurred mainly after boot but remained and only sometimes settled down after several reads. root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=0 POWER_SUPPLY_CURRENT_NOW=0 POWER_SUPPLY_CAPACITY=0 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=-2731 POWER_SUPPLY_TIME_TO_EMPTY_NOW=0 POWER_SUPPLY_TIME_TO_EMPTY_AVG=0 POWER_SUPPLY_TIME_TO_FULL_NOW=0 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=0 POWER_SUPPLY_CHARGE_NOW=0 POWER_SUPPLY_CHARGE_FULL_DESIGN=0 POWER_SUPPLY_CYCLE_COUNT=0 POWER_SUPPLY_ENERGY_NOW=0 POWER_SUPPLY_POWER_AVG=0 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real    0m15.761s user    0m0.001s sys     0m0.025s root@letux:~# Sometimes the effect did disappear after accessing the device multiple times, speed went up and results became correct. All this indicates that some interrupts from the hdq controller are lost by the driver. Enabling debugging revealed that there were spurious tx and rx timeouts, i.e. the driver does not always recognise interrupts. The main problem is that rx and tx interrupts share a single variable which was sometimes reset to 0 wiping out other interrupts. And it was overwritten by a second interrupt, independent of whether the previous interrupt was already processed or not. This patch improves interrupt handling to avoid such races and loss of interrupt flags. The ideas are: * only the hdq_isr() sets bits in hdq_status * it does not reset any bits * it does wake_up() if any interrupt is pending * bits are only reset by the read/write/break functions if they were waited for * this makes sure that no interrupts can be lost * rx/tx/timeout bits are completely decoupled from each other (and not reset all after waiting for any of them) * which bits to reset is now specified by a new parameter to hdq_reset_irqstatus() * hdq_reset_irqstatus() also returns the state before resetting so that we can encapsulate the spinlock * this should now handle the case that the write and read are both already finished quickly before the hdq_write_byte() ends. * Or that two interrupts occur in succession before they are processed by the driver. Old code may have reset all status bits making the next hdq_read_byte() timeout. * the spinlock now always protects changing of bits in function hdq_reset_irqstatus() which could become a read-write-modify problem if the interrupt handler tries to read-modify-write exactly at the same moment * we add mutex protection also for hdq_write_byte() just to be safe to not to disturb a hdq_read_byte() triggered by some other thread/process. This patch was tested on a GTA04 and results in no boot problems any more. And first read after boot is now ok: root@letux:~# time cat /sys/class/power_supply/bq27000-battery/uevent POWER_SUPPLY_NAME=bq27000-battery POWER_SUPPLY_STATUS=Discharging POWER_SUPPLY_PRESENT=1 POWER_SUPPLY_VOLTAGE_NOW=3970000 POWER_SUPPLY_CURRENT_NOW=354144 POWER_SUPPLY_CAPACITY=82 POWER_SUPPLY_CAPACITY_LEVEL=Normal POWER_SUPPLY_TEMP=266 POWER_SUPPLY_TIME_TO_EMPTY_NOW=7680 POWER_SUPPLY_TIME_TO_EMPTY_AVG=7380 POWER_SUPPLY_TECHNOLOGY=Li-ion POWER_SUPPLY_CHARGE_FULL=934856 POWER_SUPPLY_CHARGE_NOW=763976 POWER_SUPPLY_CHARGE_FULL_DESIGN=1233792 POWER_SUPPLY_CYCLE_COUNT=82 POWER_SUPPLY_ENERGY_NOW=2852840 POWER_SUPPLY_POWER_AVG=1392840 POWER_SUPPLY_HEALTH=Good POWER_SUPPLY_MANUFACTURER=Texas Instruments real 0m0.233s user 0m0.000s sys 0m0.025s root@letux:~# It was also tested with dev_dbg enabled and more printk that all activities behave correctly, especially hdq_write_byte(), hdq_read_byte(), omap_hdq_break(). Not tested is omap_w1_triplet(). Fixes: 27d13da8782a ("w1: omap-hdq: Simplify driver with PM runtime autosuspend") Cc: stable@vger.kernel.org # v5.6+ Signed-off-by: H. Nikolaus Schaller <hns@goldelico.com> Link: https://lore.kernel.org/r/68fc8623ae741878beef049273696d2377526165.1590255176.git.hns@goldelico.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2020-05-24 01:32:56 +08:00
/* Must clear irqstatus for another TXCOMPLETE interrupt */
hdq_reset_irqstatus(hdq_data, OMAP_HDQ_INT_STATUS_TXCOMPLETE);
if (err == 0) {
dev_dbg(hdq_data->dev, "TX wait elapsed\n");
goto out;
}
hdq_reg_merge(_hdq, OMAP_HDQ_CTRL_STATUS, 0,
OMAP_HDQ_CTRL_STATUS_SINGLE);
out:
mutex_unlock(&hdq_data->hdq_mutex);
rtn:
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
pm_runtime_mark_last_busy(hdq_data->dev);
pm_runtime_put_autosuspend(hdq_data->dev);
return ret;
}
/* reset callback */
static u8 omap_w1_reset_bus(void *_hdq)
{
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
struct hdq_data *hdq_data = _hdq;
int err;
err = pm_runtime_get_sync(hdq_data->dev);
if (err < 0) {
pm_runtime_put_noidle(hdq_data->dev);
return err;
}
omap_hdq_break(hdq_data);
pm_runtime_mark_last_busy(hdq_data->dev);
pm_runtime_put_autosuspend(hdq_data->dev);
return 0;
}
/* Read a byte of data from the device */
static u8 omap_w1_read_byte(void *_hdq)
{
struct hdq_data *hdq_data = _hdq;
u8 val = 0;
int ret;
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
ret = pm_runtime_get_sync(hdq_data->dev);
if (ret < 0) {
pm_runtime_put_noidle(hdq_data->dev);
return -1;
}
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
ret = hdq_read_byte(hdq_data, &val);
if (ret)
val = -1;
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
pm_runtime_mark_last_busy(hdq_data->dev);
pm_runtime_put_autosuspend(hdq_data->dev);
return val;
}
/* Write a byte of data to the device */
static void omap_w1_write_byte(void *_hdq, u8 byte)
{
struct hdq_data *hdq_data = _hdq;
int ret;
u8 status;
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
ret = pm_runtime_get_sync(hdq_data->dev);
if (ret < 0) {
pm_runtime_put_noidle(hdq_data->dev);
return;
}
/*
* We need to reset the slave before
* issuing the SKIP ROM command, else
* the slave will not work.
*/
if (byte == W1_SKIP_ROM)
omap_hdq_break(hdq_data);
ret = hdq_write_byte(hdq_data, byte, &status);
if (ret < 0) {
dev_dbg(hdq_data->dev, "TX failure:Ctrl status %x\n", status);
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
goto out_err;
}
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
out_err:
pm_runtime_mark_last_busy(hdq_data->dev);
pm_runtime_put_autosuspend(hdq_data->dev);
}
static struct w1_bus_master omap_w1_master = {
.read_byte = omap_w1_read_byte,
.write_byte = omap_w1_write_byte,
.reset_bus = omap_w1_reset_bus,
};
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
static int __maybe_unused omap_hdq_runtime_suspend(struct device *dev)
{
struct hdq_data *hdq_data = dev_get_drvdata(dev);
hdq_reg_out(hdq_data, 0, hdq_data->mode);
hdq_reg_in(hdq_data, OMAP_HDQ_INT_STATUS);
return 0;
}
static int __maybe_unused omap_hdq_runtime_resume(struct device *dev)
{
struct hdq_data *hdq_data = dev_get_drvdata(dev);
/* select HDQ/1W mode & enable clocks */
hdq_reg_out(hdq_data, OMAP_HDQ_CTRL_STATUS,
OMAP_HDQ_CTRL_STATUS_CLOCKENABLE |
OMAP_HDQ_CTRL_STATUS_INTERRUPTMASK |
hdq_data->mode);
hdq_reg_in(hdq_data, OMAP_HDQ_INT_STATUS);
return 0;
}
static const struct dev_pm_ops omap_hdq_pm_ops = {
SET_RUNTIME_PM_OPS(omap_hdq_runtime_suspend,
omap_hdq_runtime_resume, NULL)
};
static int omap_hdq_probe(struct platform_device *pdev)
{
struct device *dev = &pdev->dev;
struct hdq_data *hdq_data;
int ret, irq;
u8 rev;
const char *mode;
hdq_data = devm_kzalloc(dev, sizeof(*hdq_data), GFP_KERNEL);
if (!hdq_data)
return -ENOMEM;
hdq_data->dev = dev;
platform_set_drvdata(pdev, hdq_data);
hdq_data->hdq_base = devm_platform_ioremap_resource(pdev, 0);
if (IS_ERR(hdq_data->hdq_base))
return PTR_ERR(hdq_data->hdq_base);
mutex_init(&hdq_data->hdq_mutex);
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
ret = of_property_read_string(pdev->dev.of_node, "ti,mode", &mode);
if (ret < 0 || !strcmp(mode, "hdq")) {
hdq_data->mode = 0;
omap_w1_master.search = omap_w1_search_bus;
} else {
hdq_data->mode = 1;
omap_w1_master.triplet = omap_w1_triplet;
}
pm_runtime_enable(&pdev->dev);
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
pm_runtime_use_autosuspend(&pdev->dev);
pm_runtime_set_autosuspend_delay(&pdev->dev, 300);
ret = pm_runtime_get_sync(&pdev->dev);
if (ret < 0) {
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
pm_runtime_put_noidle(&pdev->dev);
dev_dbg(&pdev->dev, "pm_runtime_get_sync failed\n");
goto err_w1;
}
rev = hdq_reg_in(hdq_data, OMAP_HDQ_REVISION);
dev_info(&pdev->dev, "OMAP HDQ Hardware Rev %c.%c. Driver in %s mode\n",
(rev >> 4) + '0', (rev & 0x0f) + '0', "Interrupt");
spin_lock_init(&hdq_data->hdq_spinlock);
irq = platform_get_irq(pdev, 0);
if (irq < 0) {
dev_dbg(&pdev->dev, "Failed to get IRQ: %d\n", irq);
ret = irq;
goto err_irq;
}
ret = devm_request_irq(dev, irq, hdq_isr, 0, "omap_hdq", hdq_data);
if (ret < 0) {
dev_dbg(&pdev->dev, "could not request irq\n");
goto err_irq;
}
omap_hdq_break(hdq_data);
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
pm_runtime_mark_last_busy(&pdev->dev);
pm_runtime_put_autosuspend(&pdev->dev);
omap_w1_master.data = hdq_data;
ret = w1_add_master_device(&omap_w1_master);
if (ret) {
dev_dbg(&pdev->dev, "Failure in registering w1 master\n");
goto err_w1;
}
return 0;
err_irq:
pm_runtime_put_sync(&pdev->dev);
err_w1:
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
pm_runtime_dont_use_autosuspend(&pdev->dev);
pm_runtime_disable(&pdev->dev);
return ret;
}
static int omap_hdq_remove(struct platform_device *pdev)
{
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
int active;
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
active = pm_runtime_get_sync(&pdev->dev);
if (active < 0)
pm_runtime_put_noidle(&pdev->dev);
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
w1_remove_master_device(&omap_w1_master);
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
pm_runtime_dont_use_autosuspend(&pdev->dev);
if (active >= 0)
pm_runtime_put_sync(&pdev->dev);
pm_runtime_disable(&pdev->dev);
return 0;
}
static const struct of_device_id omap_hdq_dt_ids[] = {
{ .compatible = "ti,omap3-1w" },
{ .compatible = "ti,am4372-hdq" },
{}
};
MODULE_DEVICE_TABLE(of, omap_hdq_dt_ids);
static struct platform_driver omap_hdq_driver = {
.probe = omap_hdq_probe,
.remove = omap_hdq_remove,
.driver = {
.name = "omap_hdq",
.of_match_table = omap_hdq_dt_ids,
w1: omap-hdq: Simplify driver with PM runtime autosuspend We've had generic code handling module sysconfig and OCP reset registers for omap variants for many years now and all the drivers really needs to do is just call runtime PM functions. Looks like the omap-hdq driver got only partially updated over the years to use runtime PM, and still has lots of custom PM code left. We can replace all the custom code for sysconfig, OCP reset, and PM with just a few lines of runtime PM autosuspend code. In order to set the device mode properly when pm_runtime_get_sync() is called during probe, we need to also move parsing of "ti,mode" to happen earlier before we call pm_runtime_enable(). Since we now disable interrupts lazily in omap_hdq_runtime_suspend(), we must remove the call to hdq_disable_interrupt() in omap_w1_read_byte(). And we must clear irqstatus calling wait_event_timeout() on it, so let's add hdq_reset_irqstatus() for that. Note that the earlier driver specific usage count limit of four seems completely artificial and should not be an issue in normal use. Cc: Adam Ford <aford173@gmail.com> Cc: Andrew F. Davis <afd@ti.com> Cc: Andreas Kemnade <andreas@kemnade.info> Cc: H. Nikolaus Schaller <hns@goldelico.com> Cc: Vignesh R <vigneshr@ti.com> Tested-by: Andreas Kemnade <andreas@kemnade.info> # gta04 Tested-by: Adam Ford <aford173@gmail.com> #logicpd-torpedo-37xx-devkit Signed-off-by: Tony Lindgren <tony@atomide.com> Link: https://lore.kernel.org/r/20191217004048.46298-1-tony@atomide.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-12-17 08:40:48 +08:00
.pm = &omap_hdq_pm_ops,
},
};
module_platform_driver(omap_hdq_driver);
MODULE_AUTHOR("Texas Instruments");
MODULE_DESCRIPTION("HDQ-1W driver Library");
MODULE_LICENSE("GPL");