linux_dsm_epyc7002/drivers/sbus/char/bbc_envctrl.c

596 lines
16 KiB
C
Raw Normal View History

/* bbc_envctrl.c: UltraSPARC-III environment control driver.
*
* Copyright (C) 2001, 2008 David S. Miller (davem@davemloft.net)
*/
#include <linux/kthread.h>
#include <linux/delay.h>
#include <linux/kmod.h>
#include <linux/reboot.h>
#include <linux/of.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 15:04:11 +07:00
#include <linux/slab.h>
#include <linux/of_device.h>
#include <asm/oplib.h>
#include "bbc_i2c.h"
#include "max1617.h"
#undef ENVCTRL_TRACE
/* WARNING: Making changes to this driver is very dangerous.
* If you misprogram the sensor chips they can
* cut the power on you instantly.
*/
/* Two temperature sensors exist in the SunBLADE-1000 enclosure.
* Both are implemented using max1617 i2c devices. Each max1617
* monitors 2 temperatures, one for one of the cpu dies and the other
* for the ambient temperature.
*
* The max1617 is capable of being programmed with power-off
* temperature values, one low limit and one high limit. These
* can be controlled independently for the cpu or ambient temperature.
* If a limit is violated, the power is simply shut off. The frequency
* with which the max1617 does temperature sampling can be controlled
* as well.
*
* Three fans exist inside the machine, all three are controlled with
* an i2c digital to analog converter. There is a fan directed at the
* two processor slots, another for the rest of the enclosure, and the
* third is for the power supply. The first two fans may be speed
* controlled by changing the voltage fed to them. The third fan may
* only be completely off or on. The third fan is meant to only be
* disabled/enabled when entering/exiting the lowest power-saving
* mode of the machine.
*
* An environmental control kernel thread periodically monitors all
* temperature sensors. Based upon the samples it will adjust the
* fan speeds to try and keep the system within a certain temperature
* range (the goal being to make the fans as quiet as possible without
* allowing the system to get too hot).
*
* If the temperature begins to rise/fall outside of the acceptable
* operating range, a periodic warning will be sent to the kernel log.
* The fans will be put on full blast to attempt to deal with this
* situation. After exceeding the acceptable operating range by a
* certain threshold, the kernel thread will shut down the system.
* Here, the thread is attempting to shut the machine down cleanly
* before the hardware based power-off event is triggered.
*/
/* These settings are in Celsius. We use these defaults only
* if we cannot interrogate the cpu-fru SEEPROM.
*/
struct temp_limits {
s8 high_pwroff, high_shutdown, high_warn;
s8 low_warn, low_shutdown, low_pwroff;
};
static struct temp_limits cpu_temp_limits[2] = {
{ 100, 85, 80, 5, -5, -10 },
{ 100, 85, 80, 5, -5, -10 },
};
static struct temp_limits amb_temp_limits[2] = {
{ 65, 55, 40, 5, -5, -10 },
{ 65, 55, 40, 5, -5, -10 },
};
static LIST_HEAD(all_temps);
static LIST_HEAD(all_fans);
#define CPU_FAN_REG 0xf0
#define SYS_FAN_REG 0xf2
#define PSUPPLY_FAN_REG 0xf4
#define FAN_SPEED_MIN 0x0c
#define FAN_SPEED_MAX 0x3f
#define PSUPPLY_FAN_ON 0x1f
#define PSUPPLY_FAN_OFF 0x00
static void set_fan_speeds(struct bbc_fan_control *fp)
{
/* Put temperatures into range so we don't mis-program
* the hardware.
*/
if (fp->cpu_fan_speed < FAN_SPEED_MIN)
fp->cpu_fan_speed = FAN_SPEED_MIN;
if (fp->cpu_fan_speed > FAN_SPEED_MAX)
fp->cpu_fan_speed = FAN_SPEED_MAX;
if (fp->system_fan_speed < FAN_SPEED_MIN)
fp->system_fan_speed = FAN_SPEED_MIN;
if (fp->system_fan_speed > FAN_SPEED_MAX)
fp->system_fan_speed = FAN_SPEED_MAX;
#ifdef ENVCTRL_TRACE
printk("fan%d: Changed fan speed to cpu(%02x) sys(%02x)\n",
fp->index,
fp->cpu_fan_speed, fp->system_fan_speed);
#endif
bbc_i2c_writeb(fp->client, fp->cpu_fan_speed, CPU_FAN_REG);
bbc_i2c_writeb(fp->client, fp->system_fan_speed, SYS_FAN_REG);
bbc_i2c_writeb(fp->client,
(fp->psupply_fan_on ?
PSUPPLY_FAN_ON : PSUPPLY_FAN_OFF),
PSUPPLY_FAN_REG);
}
static void get_current_temps(struct bbc_cpu_temperature *tp)
{
tp->prev_amb_temp = tp->curr_amb_temp;
bbc_i2c_readb(tp->client,
(unsigned char *) &tp->curr_amb_temp,
MAX1617_AMB_TEMP);
tp->prev_cpu_temp = tp->curr_cpu_temp;
bbc_i2c_readb(tp->client,
(unsigned char *) &tp->curr_cpu_temp,
MAX1617_CPU_TEMP);
#ifdef ENVCTRL_TRACE
printk("temp%d: cpu(%d C) amb(%d C)\n",
tp->index,
(int) tp->curr_cpu_temp, (int) tp->curr_amb_temp);
#endif
}
static void do_envctrl_shutdown(struct bbc_cpu_temperature *tp)
{
static int shutting_down = 0;
char *type = "???";
s8 val = -1;
if (shutting_down != 0)
return;
if (tp->curr_amb_temp >= amb_temp_limits[tp->index].high_shutdown ||
tp->curr_amb_temp < amb_temp_limits[tp->index].low_shutdown) {
type = "ambient";
val = tp->curr_amb_temp;
} else if (tp->curr_cpu_temp >= cpu_temp_limits[tp->index].high_shutdown ||
tp->curr_cpu_temp < cpu_temp_limits[tp->index].low_shutdown) {
type = "CPU";
val = tp->curr_cpu_temp;
}
printk(KERN_CRIT "temp%d: Outside of safe %s "
"operating temperature, %d C.\n",
tp->index, type, val);
printk(KERN_CRIT "kenvctrld: Shutting down the system now.\n");
shutting_down = 1;
if (orderly_poweroff(true) < 0)
printk(KERN_CRIT "envctrl: shutdown execution failed\n");
}
#define WARN_INTERVAL (30 * HZ)
static void analyze_ambient_temp(struct bbc_cpu_temperature *tp, unsigned long *last_warn, int tick)
{
int ret = 0;
if (time_after(jiffies, (*last_warn + WARN_INTERVAL))) {
if (tp->curr_amb_temp >=
amb_temp_limits[tp->index].high_warn) {
printk(KERN_WARNING "temp%d: "
"Above safe ambient operating temperature, %d C.\n",
tp->index, (int) tp->curr_amb_temp);
ret = 1;
} else if (tp->curr_amb_temp <
amb_temp_limits[tp->index].low_warn) {
printk(KERN_WARNING "temp%d: "
"Below safe ambient operating temperature, %d C.\n",
tp->index, (int) tp->curr_amb_temp);
ret = 1;
}
if (ret)
*last_warn = jiffies;
} else if (tp->curr_amb_temp >= amb_temp_limits[tp->index].high_warn ||
tp->curr_amb_temp < amb_temp_limits[tp->index].low_warn)
ret = 1;
/* Now check the shutdown limits. */
if (tp->curr_amb_temp >= amb_temp_limits[tp->index].high_shutdown ||
tp->curr_amb_temp < amb_temp_limits[tp->index].low_shutdown) {
do_envctrl_shutdown(tp);
ret = 1;
}
if (ret) {
tp->fan_todo[FAN_AMBIENT] = FAN_FULLBLAST;
} else if ((tick & (8 - 1)) == 0) {
s8 amb_goal_hi = amb_temp_limits[tp->index].high_warn - 10;
s8 amb_goal_lo;
amb_goal_lo = amb_goal_hi - 3;
/* We do not try to avoid 'too cold' events. Basically we
* only try to deal with over-heating and fan noise reduction.
*/
if (tp->avg_amb_temp < amb_goal_hi) {
if (tp->avg_amb_temp >= amb_goal_lo)
tp->fan_todo[FAN_AMBIENT] = FAN_SAME;
else
tp->fan_todo[FAN_AMBIENT] = FAN_SLOWER;
} else {
tp->fan_todo[FAN_AMBIENT] = FAN_FASTER;
}
} else {
tp->fan_todo[FAN_AMBIENT] = FAN_SAME;
}
}
static void analyze_cpu_temp(struct bbc_cpu_temperature *tp, unsigned long *last_warn, int tick)
{
int ret = 0;
if (time_after(jiffies, (*last_warn + WARN_INTERVAL))) {
if (tp->curr_cpu_temp >=
cpu_temp_limits[tp->index].high_warn) {
printk(KERN_WARNING "temp%d: "
"Above safe CPU operating temperature, %d C.\n",
tp->index, (int) tp->curr_cpu_temp);
ret = 1;
} else if (tp->curr_cpu_temp <
cpu_temp_limits[tp->index].low_warn) {
printk(KERN_WARNING "temp%d: "
"Below safe CPU operating temperature, %d C.\n",
tp->index, (int) tp->curr_cpu_temp);
ret = 1;
}
if (ret)
*last_warn = jiffies;
} else if (tp->curr_cpu_temp >= cpu_temp_limits[tp->index].high_warn ||
tp->curr_cpu_temp < cpu_temp_limits[tp->index].low_warn)
ret = 1;
/* Now check the shutdown limits. */
if (tp->curr_cpu_temp >= cpu_temp_limits[tp->index].high_shutdown ||
tp->curr_cpu_temp < cpu_temp_limits[tp->index].low_shutdown) {
do_envctrl_shutdown(tp);
ret = 1;
}
if (ret) {
tp->fan_todo[FAN_CPU] = FAN_FULLBLAST;
} else if ((tick & (8 - 1)) == 0) {
s8 cpu_goal_hi = cpu_temp_limits[tp->index].high_warn - 10;
s8 cpu_goal_lo;
cpu_goal_lo = cpu_goal_hi - 3;
/* We do not try to avoid 'too cold' events. Basically we
* only try to deal with over-heating and fan noise reduction.
*/
if (tp->avg_cpu_temp < cpu_goal_hi) {
if (tp->avg_cpu_temp >= cpu_goal_lo)
tp->fan_todo[FAN_CPU] = FAN_SAME;
else
tp->fan_todo[FAN_CPU] = FAN_SLOWER;
} else {
tp->fan_todo[FAN_CPU] = FAN_FASTER;
}
} else {
tp->fan_todo[FAN_CPU] = FAN_SAME;
}
}
static void analyze_temps(struct bbc_cpu_temperature *tp, unsigned long *last_warn)
{
tp->avg_amb_temp = (s8)((int)((int)tp->avg_amb_temp + (int)tp->curr_amb_temp) / 2);
tp->avg_cpu_temp = (s8)((int)((int)tp->avg_cpu_temp + (int)tp->curr_cpu_temp) / 2);
analyze_ambient_temp(tp, last_warn, tp->sample_tick);
analyze_cpu_temp(tp, last_warn, tp->sample_tick);
tp->sample_tick++;
}
static enum fan_action prioritize_fan_action(int which_fan)
{
struct bbc_cpu_temperature *tp;
enum fan_action decision = FAN_STATE_MAX;
/* Basically, prioritize what the temperature sensors
* recommend we do, and perform that action on all the
* fans.
*/
list_for_each_entry(tp, &all_temps, glob_list) {
if (tp->fan_todo[which_fan] == FAN_FULLBLAST) {
decision = FAN_FULLBLAST;
break;
}
if (tp->fan_todo[which_fan] == FAN_SAME &&
decision != FAN_FASTER)
decision = FAN_SAME;
else if (tp->fan_todo[which_fan] == FAN_FASTER)
decision = FAN_FASTER;
else if (decision != FAN_FASTER &&
decision != FAN_SAME &&
tp->fan_todo[which_fan] == FAN_SLOWER)
decision = FAN_SLOWER;
}
if (decision == FAN_STATE_MAX)
decision = FAN_SAME;
return decision;
}
static int maybe_new_ambient_fan_speed(struct bbc_fan_control *fp)
{
enum fan_action decision = prioritize_fan_action(FAN_AMBIENT);
int ret;
if (decision == FAN_SAME)
return 0;
ret = 1;
if (decision == FAN_FULLBLAST) {
if (fp->system_fan_speed >= FAN_SPEED_MAX)
ret = 0;
else
fp->system_fan_speed = FAN_SPEED_MAX;
} else {
if (decision == FAN_FASTER) {
if (fp->system_fan_speed >= FAN_SPEED_MAX)
ret = 0;
else
fp->system_fan_speed += 2;
} else {
int orig_speed = fp->system_fan_speed;
if (orig_speed <= FAN_SPEED_MIN ||
orig_speed <= (fp->cpu_fan_speed - 3))
ret = 0;
else
fp->system_fan_speed -= 1;
}
}
return ret;
}
static int maybe_new_cpu_fan_speed(struct bbc_fan_control *fp)
{
enum fan_action decision = prioritize_fan_action(FAN_CPU);
int ret;
if (decision == FAN_SAME)
return 0;
ret = 1;
if (decision == FAN_FULLBLAST) {
if (fp->cpu_fan_speed >= FAN_SPEED_MAX)
ret = 0;
else
fp->cpu_fan_speed = FAN_SPEED_MAX;
} else {
if (decision == FAN_FASTER) {
if (fp->cpu_fan_speed >= FAN_SPEED_MAX)
ret = 0;
else {
fp->cpu_fan_speed += 2;
if (fp->system_fan_speed <
(fp->cpu_fan_speed - 3))
fp->system_fan_speed =
fp->cpu_fan_speed - 3;
}
} else {
if (fp->cpu_fan_speed <= FAN_SPEED_MIN)
ret = 0;
else
fp->cpu_fan_speed -= 1;
}
}
return ret;
}
static void maybe_new_fan_speeds(struct bbc_fan_control *fp)
{
int new;
new = maybe_new_ambient_fan_speed(fp);
new |= maybe_new_cpu_fan_speed(fp);
if (new)
set_fan_speeds(fp);
}
static void fans_full_blast(void)
{
struct bbc_fan_control *fp;
/* Since we will not be monitoring things anymore, put
* the fans on full blast.
*/
list_for_each_entry(fp, &all_fans, glob_list) {
fp->cpu_fan_speed = FAN_SPEED_MAX;
fp->system_fan_speed = FAN_SPEED_MAX;
fp->psupply_fan_on = 1;
set_fan_speeds(fp);
}
}
#define POLL_INTERVAL (5 * 1000)
static unsigned long last_warning_jiffies;
static struct task_struct *kenvctrld_task;
static int kenvctrld(void *__unused)
{
printk(KERN_INFO "bbc_envctrl: kenvctrld starting...\n");
last_warning_jiffies = jiffies - WARN_INTERVAL;
for (;;) {
struct bbc_cpu_temperature *tp;
struct bbc_fan_control *fp;
msleep_interruptible(POLL_INTERVAL);
if (kthread_should_stop())
break;
list_for_each_entry(tp, &all_temps, glob_list) {
get_current_temps(tp);
analyze_temps(tp, &last_warning_jiffies);
}
list_for_each_entry(fp, &all_fans, glob_list)
maybe_new_fan_speeds(fp);
}
printk(KERN_INFO "bbc_envctrl: kenvctrld exiting...\n");
fans_full_blast();
return 0;
}
static void attach_one_temp(struct bbc_i2c_bus *bp, struct of_device *op,
int temp_idx)
{
struct bbc_cpu_temperature *tp;
tp = kzalloc(sizeof(*tp), GFP_KERNEL);
if (!tp)
return;
tp->client = bbc_i2c_attach(bp, op);
if (!tp->client) {
kfree(tp);
return;
}
tp->index = temp_idx;
list_add(&tp->glob_list, &all_temps);
list_add(&tp->bp_list, &bp->temps);
/* Tell it to convert once every 5 seconds, clear all cfg
* bits.
*/
bbc_i2c_writeb(tp->client, 0x00, MAX1617_WR_CFG_BYTE);
bbc_i2c_writeb(tp->client, 0x02, MAX1617_WR_CVRATE_BYTE);
/* Program the hard temperature limits into the chip. */
bbc_i2c_writeb(tp->client, amb_temp_limits[tp->index].high_pwroff,
MAX1617_WR_AMB_HIGHLIM);
bbc_i2c_writeb(tp->client, amb_temp_limits[tp->index].low_pwroff,
MAX1617_WR_AMB_LOWLIM);
bbc_i2c_writeb(tp->client, cpu_temp_limits[tp->index].high_pwroff,
MAX1617_WR_CPU_HIGHLIM);
bbc_i2c_writeb(tp->client, cpu_temp_limits[tp->index].low_pwroff,
MAX1617_WR_CPU_LOWLIM);
get_current_temps(tp);
tp->prev_cpu_temp = tp->avg_cpu_temp = tp->curr_cpu_temp;
tp->prev_amb_temp = tp->avg_amb_temp = tp->curr_amb_temp;
tp->fan_todo[FAN_AMBIENT] = FAN_SAME;
tp->fan_todo[FAN_CPU] = FAN_SAME;
}
static void attach_one_fan(struct bbc_i2c_bus *bp, struct of_device *op,
int fan_idx)
{
struct bbc_fan_control *fp;
fp = kzalloc(sizeof(*fp), GFP_KERNEL);
if (!fp)
return;
fp->client = bbc_i2c_attach(bp, op);
if (!fp->client) {
kfree(fp);
return;
}
fp->index = fan_idx;
list_add(&fp->glob_list, &all_fans);
list_add(&fp->bp_list, &bp->fans);
/* The i2c device controlling the fans is write-only.
* So the only way to keep track of the current power
* level fed to the fans is via software. Choose half
* power for cpu/system and 'on' fo the powersupply fan
* and set it now.
*/
fp->psupply_fan_on = 1;
fp->cpu_fan_speed = (FAN_SPEED_MAX - FAN_SPEED_MIN) / 2;
fp->cpu_fan_speed += FAN_SPEED_MIN;
fp->system_fan_speed = (FAN_SPEED_MAX - FAN_SPEED_MIN) / 2;
fp->system_fan_speed += FAN_SPEED_MIN;
set_fan_speeds(fp);
}
static void destroy_one_temp(struct bbc_cpu_temperature *tp)
{
bbc_i2c_detach(tp->client);
kfree(tp);
}
static void destroy_all_temps(struct bbc_i2c_bus *bp)
{
struct bbc_cpu_temperature *tp, *tpos;
list_for_each_entry_safe(tp, tpos, &bp->temps, bp_list) {
list_del(&tp->bp_list);
list_del(&tp->glob_list);
destroy_one_temp(tp);
}
}
static void destroy_one_fan(struct bbc_fan_control *fp)
{
bbc_i2c_detach(fp->client);
kfree(fp);
}
static void destroy_all_fans(struct bbc_i2c_bus *bp)
{
struct bbc_fan_control *fp, *fpos;
list_for_each_entry_safe(fp, fpos, &bp->fans, bp_list) {
list_del(&fp->bp_list);
list_del(&fp->glob_list);
destroy_one_fan(fp);
}
}
int bbc_envctrl_init(struct bbc_i2c_bus *bp)
{
struct of_device *op;
int temp_index = 0;
int fan_index = 0;
int devidx = 0;
while ((op = bbc_i2c_getdev(bp, devidx++)) != NULL) {
if (!strcmp(op->dev.of_node->name, "temperature"))
attach_one_temp(bp, op, temp_index++);
if (!strcmp(op->dev.of_node->name, "fan-control"))
attach_one_fan(bp, op, fan_index++);
}
if (temp_index != 0 && fan_index != 0) {
kenvctrld_task = kthread_run(kenvctrld, NULL, "kenvctrld");
if (IS_ERR(kenvctrld_task)) {
int err = PTR_ERR(kenvctrld_task);
kenvctrld_task = NULL;
destroy_all_temps(bp);
destroy_all_fans(bp);
return err;
}
}
return 0;
}
void bbc_envctrl_cleanup(struct bbc_i2c_bus *bp)
{
if (kenvctrld_task)
kthread_stop(kenvctrld_task);
destroy_all_temps(bp);
destroy_all_fans(bp);
}