[RFC][PATCH] Ingo's smptimers patch experiment

[RFC][PATCH] Ingo's smptimers patch experiment

[Dipankar Sarma]

Hi Ingo,

I ported your smptimers patch to 2.5.3 and experimented with 
it a little bit. Basically I am curious about why we
we need to call run_all_timers() (which runs timers for all
CPUs) through the timer bh if locking fails in run_local_timers(). 
I experimented with a per-cpu tasklet to run only that CPU's 
timers and it seems to work. I tested UP and SMP kernels in a 
4way x86 box with LTP and chat benchmark.

Is there something that I am missing here ? A subtle race or
something that may result in losing firing of timers, perhaps ?

My experimental patch is included below.

Thanks
Dipankar
-- 
Dipankar Sarma  <dipankar@in.ibm.com> http://lse.sourceforge.net
Linux Technology Center, IBM Software Lab, Bangalore, India.

smptimers-2.5.3-2.patch
-----------------------

diff -urN linux-2.5.3/arch/i386/kernel/apic.c linux-2.5.3-smptimers/arch/i386/kernel/apic.c
--- linux-2.5.3/arch/i386/kernel/apic.c	Tue Jan 29 02:42:04 2002
+++ linux-2.5.3-smptimers/arch/i386/kernel/apic.c	Wed Feb  6 14:39:01 2002
@@ -1060,7 +1060,9 @@
 	irq_enter(cpu, 0);
 	smp_local_timer_interrupt(&regs);
 	irq_exit(cpu, 0);
-
+#if CONFIG_SMP
+	run_local_timers();
+#endif
 	if (softirq_pending(cpu))
 		do_softirq();
 }
diff -urN linux-2.5.3/arch/i386/mm/fault.c linux-2.5.3-smptimers/arch/i386/mm/fault.c
--- linux-2.5.3/arch/i386/mm/fault.c	Tue Jan  8 02:25:17 2002
+++ linux-2.5.3-smptimers/arch/i386/mm/fault.c	Wed Feb  6 14:39:01 2002
@@ -94,16 +94,12 @@
 	goto bad_area;
 }
 
-extern spinlock_t timerlist_lock;
-
 /*
  * Unlock any spinlocks which will prevent us from getting the
- * message out (timerlist_lock is acquired through the
- * console unblank code)
+ * message out
  */
 void bust_spinlocks(int yes)
 {
-	spin_lock_init(&timerlist_lock);
 	if (yes) {
 		oops_in_progress = 1;
 #ifdef CONFIG_SMP
diff -urN linux-2.5.3/arch/sparc/kernel/irq.c linux-2.5.3-smptimers/arch/sparc/kernel/irq.c
--- linux-2.5.3/arch/sparc/kernel/irq.c	Mon Jan 14 23:40:43 2002
+++ linux-2.5.3-smptimers/arch/sparc/kernel/irq.c	Wed Feb  6 14:39:01 2002
@@ -73,7 +73,7 @@
     prom_halt();
 }
 
-void (*init_timers)(void (*)(int, void *,struct pt_regs *)) =
+void (*sparc_init_timers)(void (*)(int, void *,struct pt_regs *)) =
     (void (*)(void (*)(int, void *,struct pt_regs *))) irq_panic;
 
 /*
diff -urN linux-2.5.3/arch/sparc/kernel/sun4c_irq.c linux-2.5.3-smptimers/arch/sparc/kernel/sun4c_irq.c
--- linux-2.5.3/arch/sparc/kernel/sun4c_irq.c	Fri Apr 27 10:47:26 2001
+++ linux-2.5.3-smptimers/arch/sparc/kernel/sun4c_irq.c	Wed Feb  6 14:39:01 2002
@@ -143,7 +143,7 @@
 	/* Errm.. not sure how to do this.. */
 }
 
-static void __init sun4c_init_timers(void (*counter_fn)(int, void *, struct pt_regs *))
+static void __init sun4c_sparc_init_timers(void (*counter_fn)(int, void *, struct pt_regs *))
 {
 	int irq;
 
@@ -221,7 +221,7 @@
 	BTFIXUPSET_CALL(clear_profile_irq, sun4c_clear_profile_irq, BTFIXUPCALL_NOP);
 	BTFIXUPSET_CALL(load_profile_irq, sun4c_load_profile_irq, BTFIXUPCALL_NOP);
 	BTFIXUPSET_CALL(__irq_itoa, sun4m_irq_itoa, BTFIXUPCALL_NORM);
-	init_timers = sun4c_init_timers;
+	sparc_init_timers = sun4c_sparc_init_timers;
 #ifdef CONFIG_SMP
 	BTFIXUPSET_CALL(set_cpu_int, sun4c_nop, BTFIXUPCALL_NOP);
 	BTFIXUPSET_CALL(clear_cpu_int, sun4c_nop, BTFIXUPCALL_NOP);
diff -urN linux-2.5.3/arch/sparc/kernel/sun4d_irq.c linux-2.5.3-smptimers/arch/sparc/kernel/sun4d_irq.c
--- linux-2.5.3/arch/sparc/kernel/sun4d_irq.c	Mon Jan 14 23:40:44 2002
+++ linux-2.5.3-smptimers/arch/sparc/kernel/sun4d_irq.c	Wed Feb  6 14:39:01 2002
@@ -436,7 +436,7 @@
 	bw_set_prof_limit(cpu, limit);
 }
 
-static void __init sun4d_init_timers(void (*counter_fn)(int, void *, struct pt_regs *))
+static void __init sun4d_sparc_init_timers(void (*counter_fn)(int, void *, struct pt_regs *))
 {
 	int irq;
 	extern struct prom_cpuinfo linux_cpus[NR_CPUS];
@@ -547,7 +547,7 @@
 	BTFIXUPSET_CALL(clear_profile_irq, sun4d_clear_profile_irq, BTFIXUPCALL_NORM);
 	BTFIXUPSET_CALL(load_profile_irq, sun4d_load_profile_irq, BTFIXUPCALL_NORM);
 	BTFIXUPSET_CALL(__irq_itoa, sun4d_irq_itoa, BTFIXUPCALL_NORM);
-	init_timers = sun4d_init_timers;
+	sparc_init_timers = sun4d_sparc_init_timers;
 #ifdef CONFIG_SMP
 	BTFIXUPSET_CALL(set_cpu_int, sun4d_set_cpu_int, BTFIXUPCALL_NORM);
 	BTFIXUPSET_CALL(clear_cpu_int, sun4d_clear_ipi, BTFIXUPCALL_NOP);
diff -urN linux-2.5.3/arch/sparc/kernel/sun4m_irq.c linux-2.5.3-smptimers/arch/sparc/kernel/sun4m_irq.c
--- linux-2.5.3/arch/sparc/kernel/sun4m_irq.c	Fri Apr 27 10:47:26 2001
+++ linux-2.5.3-smptimers/arch/sparc/kernel/sun4m_irq.c	Wed Feb  6 14:39:01 2002
@@ -223,7 +223,7 @@
 	return buff;
 }
 
-static void __init sun4m_init_timers(void (*counter_fn)(int, void *, struct pt_regs *))
+static void __init sun4m_sparc_init_timers(void (*counter_fn)(int, void *, struct pt_regs *))
 {
 	int reg_count, irq, cpu;
 	struct linux_prom_registers cnt_regs[PROMREG_MAX];
@@ -374,7 +374,7 @@
 	BTFIXUPSET_CALL(clear_profile_irq, sun4m_clear_profile_irq, BTFIXUPCALL_NORM);
 	BTFIXUPSET_CALL(load_profile_irq, sun4m_load_profile_irq, BTFIXUPCALL_NORM);
 	BTFIXUPSET_CALL(__irq_itoa, sun4m_irq_itoa, BTFIXUPCALL_NORM);
-	init_timers = sun4m_init_timers;
+	sparc_init_timers = sun4m_sparc_init_timers;
 #ifdef CONFIG_SMP
 	BTFIXUPSET_CALL(set_cpu_int, sun4m_send_ipi, BTFIXUPCALL_NORM);
 	BTFIXUPSET_CALL(clear_cpu_int, sun4m_clear_ipi, BTFIXUPCALL_NORM);
diff -urN linux-2.5.3/arch/sparc/kernel/time.c linux-2.5.3-smptimers/arch/sparc/kernel/time.c
--- linux-2.5.3/arch/sparc/kernel/time.c	Thu Jan 24 04:58:39 2002
+++ linux-2.5.3-smptimers/arch/sparc/kernel/time.c	Wed Feb  6 14:39:01 2002
@@ -379,7 +379,7 @@
 	else
 		clock_probe();
 
-	init_timers(timer_interrupt);
+	sparc_init_timers(timer_interrupt);
 	
 #ifdef CONFIG_SUN4
 	if(idprom->id_machtype == (SM_SUN4 | SM_4_330)) {
diff -urN linux-2.5.3/arch/sparc64/kernel/irq.c linux-2.5.3-smptimers/arch/sparc64/kernel/irq.c
--- linux-2.5.3/arch/sparc64/kernel/irq.c	Mon Jan 14 23:40:44 2002
+++ linux-2.5.3-smptimers/arch/sparc64/kernel/irq.c	Wed Feb  6 14:39:01 2002
@@ -1027,7 +1027,7 @@
 }
 
 /* This is gets the master TICK_INT timer going. */
-void init_timers(void (*cfunc)(int, void *, struct pt_regs *),
+void sparc_init_timers(void (*cfunc)(int, void *, struct pt_regs *),
 		 unsigned long *clock)
 {
 	unsigned long pstate;
diff -urN linux-2.5.3/arch/sparc64/kernel/time.c linux-2.5.3-smptimers/arch/sparc64/kernel/time.c
--- linux-2.5.3/arch/sparc64/kernel/time.c	Thu Jan 24 04:58:39 2002
+++ linux-2.5.3-smptimers/arch/sparc64/kernel/time.c	Wed Feb  6 20:50:27 2002
@@ -595,7 +595,7 @@
 	__restore_flags(flags);
 }
 
-extern void init_timers(void (*func)(int, void *, struct pt_regs *),
+extern void sparc_init_timers(void (*func)(int, void *, struct pt_regs *),
 			unsigned long *);
 
 void __init time_init(void)
@@ -606,7 +606,7 @@
 	 */
 	unsigned long clock;
 
-	init_timers(timer_interrupt, &clock);
+	sparc_init_timers(timer_interrupt, &clock);
 	timer_ticks_per_usec_quotient = ((1UL<<32) / (clock / 1000020));
 }
 
diff -urN linux-2.5.3/drivers/net/eepro100.c linux-2.5.3-smptimers/drivers/net/eepro100.c
--- linux-2.5.3/drivers/net/eepro100.c	Tue Jan  8 23:04:50 2002
+++ linux-2.5.3-smptimers/drivers/net/eepro100.c	Wed Feb  6 14:39:01 2002
@@ -1123,9 +1123,6 @@
 	/* We must continue to monitor the media. */
 	sp->timer.expires = RUN_AT(2*HZ); 			/* 2.0 sec. */
 	add_timer(&sp->timer);
-#if defined(timer_exit)
-	timer_exit(&sp->timer);
-#endif
 }
 
 static void speedo_show_state(struct net_device *dev)
diff -urN linux-2.5.3/include/asm-sparc/irq.h linux-2.5.3-smptimers/include/asm-sparc/irq.h
--- linux-2.5.3/include/asm-sparc/irq.h	Tue Aug 29 09:50:03 2000
+++ linux-2.5.3-smptimers/include/asm-sparc/irq.h	Wed Feb  6 14:39:01 2002
@@ -45,7 +45,7 @@
 #define clear_profile_irq(cpu) BTFIXUP_CALL(clear_profile_irq)(cpu)
 #define load_profile_irq(cpu,limit) BTFIXUP_CALL(load_profile_irq)(cpu,limit)
 
-extern void (*init_timers)(void (*lvl10_irq)(int, void *, struct pt_regs *));
+extern void (*sparc_init_timers)(void (*lvl10_irq)(int, void *, struct pt_regs *));
 extern void claim_ticker14(void (*irq_handler)(int, void *, struct pt_regs *),
 			   int irq,
 			   unsigned int timeout);
diff -urN linux-2.5.3/include/asm-sparc64/irq.h linux-2.5.3-smptimers/include/asm-sparc64/irq.h
--- linux-2.5.3/include/asm-sparc64/irq.h	Thu Jan 24 04:58:40 2002
+++ linux-2.5.3-smptimers/include/asm-sparc64/irq.h	Wed Feb  6 14:39:01 2002
@@ -113,7 +113,7 @@
 extern void disable_irq(unsigned int);
 #define disable_irq_nosync disable_irq
 extern void enable_irq(unsigned int);
-extern void init_timers(void (*lvl10_irq)(int, void *, struct pt_regs *),
+extern void sparc_init_timers(void (*lvl10_irq)(int, void *, struct pt_regs *),
 			unsigned long *);
 extern unsigned int build_irq(int pil, int inofixup, unsigned long iclr, unsigned long imap);
 extern unsigned int sbus_build_irq(void *sbus, unsigned int ino);
diff -urN linux-2.5.3/include/linux/smp.h linux-2.5.3-smptimers/include/linux/smp.h
--- linux-2.5.3/include/linux/smp.h	Wed Jan 30 11:11:12 2002
+++ linux-2.5.3-smptimers/include/linux/smp.h	Wed Feb  6 19:01:32 2002
@@ -76,7 +76,8 @@
 /*
  *	These macros fold the SMP functionality into a single CPU system
  */
- 
+
+#define NR_CPUS					1 
 #define smp_num_cpus				1
 #define smp_processor_id()			0
 #define hard_smp_processor_id()			0
diff -urN linux-2.5.3/include/linux/timer.h linux-2.5.3-smptimers/include/linux/timer.h
--- linux-2.5.3/include/linux/timer.h	Wed Jan 30 11:11:11 2002
+++ linux-2.5.3-smptimers/include/linux/timer.h	Wed Feb  6 19:01:32 2002
@@ -1,9 +1,6 @@
 #ifndef _LINUX_TIMER_H
 #define _LINUX_TIMER_H
 
-#include <linux/config.h>
-#include <linux/list.h>
-
 /*
  * In Linux 2.4, static timers have been removed from the kernel.
  * Timers may be dynamically created and destroyed, and should be initialized
@@ -13,22 +10,80 @@
  * timeouts. You can use this field to distinguish between the different
  * invocations.
  */
+
+#include <linux/config.h>
+#include <linux/smp.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/cache.h>
+
+/*
+ * Event timer code
+ */
+#define TVN_BITS 6
+#define TVR_BITS 8
+#define TVN_SIZE (1 << TVN_BITS)
+#define TVR_SIZE (1 << TVR_BITS)
+#define TVN_MASK (TVN_SIZE - 1)
+#define TVR_MASK (TVR_SIZE - 1)
+
+typedef struct tvec_s {
+	int index;
+	struct list_head vec[TVN_SIZE];
+} tvec_t;
+
+typedef struct tvec_root_s {
+	int index;
+	struct list_head vec[TVR_SIZE];
+} tvec_root_t;
+
+#define NOOF_TVECS 5
+
+typedef struct timer_list timer_t;
+
+struct tvec_t_base_s {
+	spinlock_t lock;
+	unsigned long timer_jiffies;
+	volatile timer_t * volatile running_timer;
+	tvec_root_t tv1;
+	tvec_t tv2;
+	tvec_t tv3;
+	tvec_t tv4;
+	tvec_t tv5;
+} ____cacheline_aligned_in_smp;
+
+typedef struct tvec_t_base_s tvec_base_t;
+
+/*
+ * This is the new and improved way of handling timers.
+ *
+ * The "data" field is in case you want to use the same
+ * timeout function for several timeouts. You can use this
+ * to distinguish between the different invocations.
+ */
 struct timer_list {
 	struct list_head list;
 	unsigned long expires;
 	unsigned long data;
 	void (*function)(unsigned long);
+	tvec_base_t *base;
 };
 
-extern void add_timer(struct timer_list * timer);
-extern int del_timer(struct timer_list * timer);
+extern void add_timer(timer_t * timer);
+extern int del_timer(timer_t * timer);
 
 #ifdef CONFIG_SMP
-extern int del_timer_sync(struct timer_list * timer);
+extern int del_timer_sync(timer_t * timer);
 extern void sync_timers(void);
+#define timer_enter(base, t) do { base->running_timer = t; mb(); } while (0)
+#define timer_exit(base) do { base->running_timer = NULL; } while (0)
+#define timer_is_running(base,t) (base->running_timer == t)
+#define timer_synchronize(base,t) while (timer_is_running(base,t)) barrier()
 #else
 #define del_timer_sync(t)	del_timer(t)
 #define sync_timers()		do { } while (0)
+#define timer_enter(base,t)          do { } while (0)
+#define timer_exit(base)            do { } while (0)
 #endif
 
 /*
@@ -38,17 +93,33 @@
  * If the timer is known to be not pending (ie, in the handler), mod_timer
  * is less efficient than a->expires = b; add_timer(a).
  */
-int mod_timer(struct timer_list *timer, unsigned long expires);
+int mod_timer(timer_t *timer, unsigned long expires);
 
 extern void it_real_fn(unsigned long);
 
-static inline void init_timer(struct timer_list * timer)
+extern void init_timers(void);
+extern void run_local_timers(void);
+
+extern tvec_base_t tvec_bases[NR_CPUS];
+
+static inline void init_timer(timer_t * timer)
 {
 	timer->list.next = timer->list.prev = NULL;
+	timer->base = tvec_bases + 0;
 }
 
-static inline int timer_pending (const struct timer_list * timer)
+#define TIMER_DEBUG 0
+#if TIMER_DEBUG
+# define CHECK_BASE(base) \
+	if (base && ((base < tvec_bases) || (base >= tvec_bases + NR_CPUS))) \
+		BUG()
+#else
+# define CHECK_BASE(base)
+#endif
+
+static inline int timer_pending(const timer_t * timer)
 {
+	CHECK_BASE(timer->base);
 	return timer->list.next != NULL;
 }
 
diff -urN linux-2.5.3/kernel/ksyms.c linux-2.5.3-smptimers/kernel/ksyms.c
--- linux-2.5.3/kernel/ksyms.c	Wed Jan 30 00:17:10 2002
+++ linux-2.5.3-smptimers/kernel/ksyms.c	Wed Feb  6 14:39:01 2002
@@ -383,6 +383,7 @@
 EXPORT_SYMBOL(del_timer_sync);
 #endif
 EXPORT_SYMBOL(mod_timer);
+EXPORT_SYMBOL(tvec_bases);
 EXPORT_SYMBOL(tq_timer);
 EXPORT_SYMBOL(tq_immediate);
 
diff -urN linux-2.5.3/kernel/sched.c linux-2.5.3-smptimers/kernel/sched.c
--- linux-2.5.3/kernel/sched.c	Tue Jan 29 04:42:47 2002
+++ linux-2.5.3-smptimers/kernel/sched.c	Wed Feb  6 14:39:01 2002
@@ -1322,7 +1322,7 @@
 	__restore_flags(flags);
 }
 
-extern void init_timervecs(void);
+extern void init_timers(void);
 extern void timer_bh(void);
 extern void tqueue_bh(void);
 extern void immediate_bh(void);
@@ -1361,8 +1361,7 @@
 	rq->idle = current;
 	wake_up_process(current);
 
-	init_timervecs();
-	init_bh(TIMER_BH, timer_bh);
+	init_timers();
 	init_bh(TQUEUE_BH, tqueue_bh);
 	init_bh(IMMEDIATE_BH, immediate_bh);
 
diff -urN linux-2.5.3/kernel/timer.c linux-2.5.3-smptimers/kernel/timer.c
--- linux-2.5.3/kernel/timer.c	Wed Jan 16 04:03:55 2002
+++ linux-2.5.3-smptimers/kernel/timer.c	Wed Feb  6 20:25:08 2002
@@ -13,10 +13,15 @@
  *              serialize accesses to xtime/lost_ticks).
  *                              Copyright (C) 1998  Andrea Arcangeli
  *  1999-03-10  Improved NTP compatibility by Ulrich Windl
+ *  2000-10-05  Implemented scalable SMP per-CPU timer handling.
+ *                              Copyright (C) 2000  Ingo Molnar
+ *              Designed by David S. Miller, Alexey Kuznetsov and Ingo Molnar
  */
 
 #include <linux/config.h>
+
 #include <linux/mm.h>
+#include <linux/init.h>
 #include <linux/timex.h>
 #include <linux/delay.h>
 #include <linux/smp_lock.h>
@@ -73,83 +78,56 @@
 unsigned long prof_len;
 unsigned long prof_shift;
 
-/*
- * Event timer code
- */
-#define TVN_BITS 6
-#define TVR_BITS 8
-#define TVN_SIZE (1 << TVN_BITS)
-#define TVR_SIZE (1 << TVR_BITS)
-#define TVN_MASK (TVN_SIZE - 1)
-#define TVR_MASK (TVR_SIZE - 1)
-
-struct timer_vec {
-	int index;
-	struct list_head vec[TVN_SIZE];
-};
-
-struct timer_vec_root {
-	int index;
-	struct list_head vec[TVR_SIZE];
-};
-
-static struct timer_vec tv5;
-static struct timer_vec tv4;
-static struct timer_vec tv3;
-static struct timer_vec tv2;
-static struct timer_vec_root tv1;
-
-static struct timer_vec * const tvecs[] = {
-	(struct timer_vec *)&tv1, &tv2, &tv3, &tv4, &tv5
-};
+tvec_base_t tvec_bases[NR_CPUS] __cacheline_aligned;
 
-#define NOOF_TVECS (sizeof(tvecs) / sizeof(tvecs[0]))
+/* To be changed to a per-CPU type with Rusty's patch */
+struct timer_tlet {
+	struct tasklet_struct tasklet;
+} ____cacheline_aligned_in_smp;
 
-void init_timervecs (void)
-{
-	int i;
+static struct timer_tlet timer_tlets[NR_CPUS] __cacheline_aligned;
 
-	for (i = 0; i < TVN_SIZE; i++) {
-		INIT_LIST_HEAD(tv5.vec + i);
-		INIT_LIST_HEAD(tv4.vec + i);
-		INIT_LIST_HEAD(tv3.vec + i);
-		INIT_LIST_HEAD(tv2.vec + i);
-	}
-	for (i = 0; i < TVR_SIZE; i++)
-		INIT_LIST_HEAD(tv1.vec + i);
-}
+/* jiffies at the most recent update of wall time */
+unsigned long wall_jiffies;
 
-static unsigned long timer_jiffies;
+/*
+ * This spinlock protect us from races in SMP while playing with xtime. -arca
+ */
+rwlock_t xtime_lock = RW_LOCK_UNLOCKED;
 
-static inline void internal_add_timer(struct timer_list *timer)
+/*
+ * This is the 'global' timer BH. This gets called only if one of
+ * the local timer interrupts couldnt run timers.
+ */
+static inline void internal_add_timer(tvec_base_t *base, timer_t *timer)
 {
 	/*
 	 * must be cli-ed when calling this
 	 */
 	unsigned long expires = timer->expires;
-	unsigned long idx = expires - timer_jiffies;
+	unsigned long idx = expires - base->timer_jiffies;
 	struct list_head * vec;
 
 	if (idx < TVR_SIZE) {
 		int i = expires & TVR_MASK;
-		vec = tv1.vec + i;
+		vec = base->tv1.vec + i;
 	} else if (idx < 1 << (TVR_BITS + TVN_BITS)) {
 		int i = (expires >> TVR_BITS) & TVN_MASK;
-		vec = tv2.vec + i;
+		vec = base->tv2.vec + i;
 	} else if (idx < 1 << (TVR_BITS + 2 * TVN_BITS)) {
 		int i = (expires >> (TVR_BITS + TVN_BITS)) & TVN_MASK;
-		vec =  tv3.vec + i;
+		vec = base->tv3.vec + i;
 	} else if (idx < 1 << (TVR_BITS + 3 * TVN_BITS)) {
 		int i = (expires >> (TVR_BITS + 2 * TVN_BITS)) & TVN_MASK;
-		vec = tv4.vec + i;
+		vec = base->tv4.vec + i;
 	} else if ((signed long) idx < 0) {
 		/* can happen if you add a timer with expires == jiffies,
 		 * or you set a timer to go off in the past
 		 */
-		vec = tv1.vec + tv1.index;
+		vec = base->tv1.vec + base->tv1.index;
 	} else if (idx <= 0xffffffffUL) {
 		int i = (expires >> (TVR_BITS + 3 * TVN_BITS)) & TVN_MASK;
-		vec = tv5.vec + i;
+		vec = base->tv5.vec + i;
 	} else {
 		/* Can only get here on architectures with 64-bit jiffies */
 		INIT_LIST_HEAD(&timer->list);
@@ -161,37 +139,27 @@
 	list_add(&timer->list, vec->prev);
 }
 
-/* Initialize both explicitly - let's try to have them in the same cache line */
-spinlock_t timerlist_lock = SPIN_LOCK_UNLOCKED;
-
-#ifdef CONFIG_SMP
-volatile struct timer_list * volatile running_timer;
-#define timer_enter(t) do { running_timer = t; mb(); } while (0)
-#define timer_exit() do { running_timer = NULL; } while (0)
-#define timer_is_running(t) (running_timer == t)
-#define timer_synchronize(t) while (timer_is_running(t)) barrier()
-#else
-#define timer_enter(t)		do { } while (0)
-#define timer_exit()		do { } while (0)
-#endif
-
-void add_timer(struct timer_list *timer)
+void add_timer(timer_t *timer)
 {
+	tvec_base_t * base = tvec_bases + smp_processor_id();
 	unsigned long flags;
 
-	spin_lock_irqsave(&timerlist_lock, flags);
+	CHECK_BASE(base);
+	CHECK_BASE(timer->base);
+	spin_lock_irqsave(&base->lock, flags);
 	if (timer_pending(timer))
 		goto bug;
-	internal_add_timer(timer);
-	spin_unlock_irqrestore(&timerlist_lock, flags);
+	internal_add_timer(base, timer);
+	timer->base = base;
+	spin_unlock_irqrestore(&base->lock, flags);
 	return;
 bug:
-	spin_unlock_irqrestore(&timerlist_lock, flags);
+	spin_unlock_irqrestore(&base->lock, flags);
 	printk("bug: kernel timer added twice at %p.\n",
 			__builtin_return_address(0));
 }
 
-static inline int detach_timer (struct timer_list *timer)
+static inline int detach_timer(timer_t *timer)
 {
 	if (!timer_pending(timer))
 		return 0;
@@ -199,28 +167,81 @@
 	return 1;
 }
 
-int mod_timer(struct timer_list *timer, unsigned long expires)
+/*
+ * mod_timer() has subtle locking semantics because parallel
+ * calls to it must happen serialized.
+ */
+int mod_timer(timer_t *timer, unsigned long expires)
 {
-	int ret;
+	tvec_base_t *old_base, *new_base;
 	unsigned long flags;
+	int ret;
+
+	new_base = tvec_bases + smp_processor_id();
+	CHECK_BASE(new_base);
+
+	__save_flags(flags);
+	__cli();
+repeat:
+	old_base = timer->base;
+	CHECK_BASE(old_base);
+
+	/*
+	 * Prevent deadlocks via ordering by old_base < new_base.
+	 */
+	if (old_base && (new_base != old_base)) {
+		if (old_base < new_base) {
+			spin_lock(&new_base->lock);
+			spin_lock(&old_base->lock);
+		} else {
+			spin_lock(&old_base->lock);
+			spin_lock(&new_base->lock);
+		}
+		/*
+		 * Subtle, we rely on timer->base being always
+		 * valid and being updated atomically.
+		 */
+		if (timer->base != old_base) {
+			spin_unlock(&new_base->lock);
+			spin_unlock(&old_base->lock);
+			goto repeat;
+		}
+	} else
+		spin_lock(&new_base->lock);
 
-	spin_lock_irqsave(&timerlist_lock, flags);
 	timer->expires = expires;
 	ret = detach_timer(timer);
-	internal_add_timer(timer);
-	spin_unlock_irqrestore(&timerlist_lock, flags);
+	internal_add_timer(new_base, timer);
+	timer->base = new_base;
+
+
+	if (old_base && (new_base != old_base))
+		spin_unlock(&old_base->lock);
+	spin_unlock_irqrestore(&new_base->lock, flags);
+
 	return ret;
 }
 
-int del_timer(struct timer_list * timer)
+int del_timer(timer_t * timer)
 {
-	int ret;
 	unsigned long flags;
+	tvec_base_t * base;
+	int ret;
 
-	spin_lock_irqsave(&timerlist_lock, flags);
+	CHECK_BASE(timer->base);
+	if (!timer->base)
+		return 0;
+repeat:
+ 	base = timer->base;
+	spin_lock_irqsave(&base->lock, flags);
+	if (base != timer->base) {
+		spin_unlock_irqrestore(&base->lock, flags);
+		goto repeat;
+	}
 	ret = detach_timer(timer);
 	timer->list.next = timer->list.prev = NULL;
-	spin_unlock_irqrestore(&timerlist_lock, flags);
+	spin_unlock_irqrestore(&base->lock, flags);
+
 	return ret;
 }
 
@@ -238,24 +259,34 @@
  * (for reference counting).
  */
 
-int del_timer_sync(struct timer_list * timer)
+int del_timer_sync(timer_t * timer)
 {
+	tvec_base_t * base;
 	int ret = 0;
 
+	CHECK_BASE(timer->base);
+	if (!timer->base)
+		return 0;
 	for (;;) {
 		unsigned long flags;
 		int running;
 
-		spin_lock_irqsave(&timerlist_lock, flags);
+repeat:
+	 	base = timer->base;
+		spin_lock_irqsave(&base->lock, flags);
+		if (base != timer->base) {
+			spin_unlock_irqrestore(&base->lock, flags);
+			goto repeat;
+		}
 		ret += detach_timer(timer);
 		timer->list.next = timer->list.prev = 0;
-		running = timer_is_running(timer);
-		spin_unlock_irqrestore(&timerlist_lock, flags);
+		running = timer_is_running(base, timer);
+		spin_unlock_irqrestore(&base->lock, flags);
 
 		if (!running)
 			break;
 
-		timer_synchronize(timer);
+		timer_synchronize(base, timer);
 	}
 
 	return ret;
@@ -263,7 +294,7 @@
 #endif
 
 
-static inline void cascade_timers(struct timer_vec *tv)
+static void cascade(tvec_base_t *base, tvec_t *tv)
 {
 	/* cascade all the timers from tv up one level */
 	struct list_head *head, *curr, *next;
@@ -275,54 +306,68 @@
 	 * detach them individually, just clear the list afterwards.
 	 */
 	while (curr != head) {
-		struct timer_list *tmp;
+		timer_t *tmp;
 
-		tmp = list_entry(curr, struct timer_list, list);
+		tmp = list_entry(curr, timer_t, list);
+		CHECK_BASE(tmp->base);
+		if (tmp->base != base)
+			BUG();
 		next = curr->next;
 		list_del(curr); // not needed
-		internal_add_timer(tmp);
+		internal_add_timer(base, tmp);
 		curr = next;
 	}
 	INIT_LIST_HEAD(head);
 	tv->index = (tv->index + 1) & TVN_MASK;
 }
 
-static inline void run_timer_list(void)
+static void __run_timers(tvec_base_t *base)
 {
-	spin_lock_irq(&timerlist_lock);
-	while ((long)(jiffies - timer_jiffies) >= 0) {
+	unsigned long flags;
+
+	spin_lock_irqsave(&base->lock, flags);
+	while ((long)(jiffies - base->timer_jiffies) >= 0) {
 		struct list_head *head, *curr;
-		if (!tv1.index) {
-			int n = 1;
-			do {
-				cascade_timers(tvecs[n]);
-			} while (tvecs[n]->index == 1 && ++n < NOOF_TVECS);
+
+		/*
+		 * Cascade timers:
+		 */
+		if (!base->tv1.index) {
+			cascade(base, &base->tv2);
+			if (base->tv2.index == 1) {
+				cascade(base, &base->tv3);
+				if (base->tv3.index == 1) {
+					cascade(base, &base->tv4);
+					if (base->tv4.index == 1)
+						cascade(base, &base->tv5);
+				}
+			}
 		}
 repeat:
-		head = tv1.vec + tv1.index;
+		head = base->tv1.vec + base->tv1.index;
 		curr = head->next;
 		if (curr != head) {
-			struct timer_list *timer;
 			void (*fn)(unsigned long);
 			unsigned long data;
+			timer_t *timer;
 
-			timer = list_entry(curr, struct timer_list, list);
+			timer = list_entry(curr, timer_t, list);
  			fn = timer->function;
- 			data= timer->data;
+ 			data = timer->data;
 
 			detach_timer(timer);
 			timer->list.next = timer->list.prev = NULL;
-			timer_enter(timer);
-			spin_unlock_irq(&timerlist_lock);
+			timer_enter(base, timer);
+			spin_unlock_irq(&base->lock);
 			fn(data);
-			spin_lock_irq(&timerlist_lock);
-			timer_exit();
+			spin_lock_irq(&base->lock);
+			timer_exit(base);
 			goto repeat;
 		}
-		++timer_jiffies; 
-		tv1.index = (tv1.index + 1) & TVR_MASK;
+		++base->timer_jiffies; 
+		base->tv1.index = (base->tv1.index + 1) & TVR_MASK;
 	}
-	spin_unlock_irq(&timerlist_lock);
+	spin_unlock_irqrestore(&base->lock, flags);
 }
 
 spinlock_t tqueue_lock = SPIN_LOCK_UNLOCKED;
@@ -638,40 +683,86 @@
 	}
 }
 
-/* jiffies at the most recent update of wall time */
-unsigned long wall_jiffies;
+static void run_timer_tasklet(unsigned long data)
+{
+	tvec_base_t *base = tvec_bases + smp_processor_id();
+	if ((long)(jiffies - base->timer_jiffies) >= 0)
+		__run_timers(base);
+}
+
+#ifndef CONFIG_X86
+/* Needed only for non-x86 for the time being *//
+static void run_all_timers(void)
+{
+	int i;
+
+	for (i = 0; i < smp_num_cpus; i++) {
+		tvec_base_t *base = tvec_bases + i;
+		if ((long)(jiffies - base->timer_jiffies) >= 0)
+			__run_timers(base);
+	}
+}
+#endif
 
 /*
- * This spinlock protect us from races in SMP while playing with xtime. -arca
+ * Called by the local, per-CPU timer interrupt on SMP.
+ *
+ * This function has to do all sorts of locking to make legacy
+ * cli()-users and BH-disablers work. If locking doesnt succeed
+ * now then we fall back to TIMER_BH.
  */
-rwlock_t xtime_lock = RW_LOCK_UNLOCKED;
+void run_local_timers(void)
+{
+	int cpu = smp_processor_id();
+	tvec_base_t *base = tvec_bases + cpu;
+
+	if (in_interrupt())
+		goto out_mark;
+
+	local_bh_disable();
+	local_irq_disable();
+	if (!spin_trylock(&global_bh_lock))
+		goto out_enable_mark;
+
+	if (!hardirq_trylock(cpu))
+		goto out_unlock_enable_mark;
+
+	if ((long)(jiffies - base->timer_jiffies) >= 0)
+		__run_timers(base);
+
+	hardirq_endlock(cpu);
+	spin_unlock(&global_bh_lock);
+	local_irq_enable();
+	local_bh_enable();
+	return;
+
+out_unlock_enable_mark:
+	spin_unlock(&global_bh_lock);
+
+out_enable_mark:
+	local_irq_enable();
+	local_bh_enable();
 
-static inline void update_times(void)
+out_mark:
+	tasklet_hi_schedule(&timer_tlets[cpu].tasklet);
+}
+
+/*
+ * Called by the timer interrupt. xtime_lock must already be taken
+ * by the timer IRQ!
+ */
+static void update_times(void)
 {
 	unsigned long ticks;
 
-	/*
-	 * update_times() is run from the raw timer_bh handler so we
-	 * just know that the irqs are locally enabled and so we don't
-	 * need to save/restore the flags of the local CPU here. -arca
-	 */
-	write_lock_irq(&xtime_lock);
-
 	ticks = jiffies - wall_jiffies;
 	if (ticks) {
 		wall_jiffies += ticks;
 		update_wall_time(ticks);
 	}
-	write_unlock_irq(&xtime_lock);
 	calc_load(ticks);
 }
 
-void timer_bh(void)
-{
-	update_times();
-	run_timer_list();
-}
-
 void do_timer(struct pt_regs *regs)
 {
 	(*(unsigned long *)&jiffies)++;
@@ -679,8 +770,18 @@
 	/* SMP process accounting uses the local APIC timer */
 
 	update_process_times(user_mode(regs));
+#ifdef CONFIG_X86
+	tasklet_hi_schedule(&timer_tlets[0].tasklet);
+#endif
 #endif
+	/*
+	 * Right now only x86-SMP calls run_local_timers() from a
+	 * per-CPU interrupt.
+	 */
+#ifndef CONFIG_X86
 	mark_bh(TIMER_BH);
+#endif
+	update_times();
 	if (TQ_ACTIVE(tq_timer))
 		mark_bh(TQUEUE_BH);
 }
@@ -923,5 +1024,28 @@
 		return -EINTR;
 	}
 	return 0;
+}
+
+void __init init_timers(void)
+{
+	int i, j;
+
+	for (i = 0; i < NR_CPUS; i++) {
+		tvec_base_t *base = tvec_bases + i;
+
+		spin_lock_init(&base->lock);
+		for (j = 0; j < TVN_SIZE; j++) {
+			INIT_LIST_HEAD(base->tv5.vec + j);
+			INIT_LIST_HEAD(base->tv4.vec + j);
+			INIT_LIST_HEAD(base->tv3.vec + j);
+			INIT_LIST_HEAD(base->tv2.vec + j);
+		}
+		for (j = 0; j < TVR_SIZE; j++)
+			INIT_LIST_HEAD(base->tv1.vec + j);
+		tasklet_init(&timer_tlets[i].tasklet, run_timer_tasklet, 0);
+	}
+#ifndef CONFIG_X86
+	init_bh(TIMER_BH, run_all_timers);
+#endif
 }
 

Re: [RFC][PATCH] Ingo's smptimers patch experiment

[Christoph Hellwig]

In article <20020206211925.A8720@in.ibm.com> you wrote:
> Hi Ingo,
>
> I ported your smptimers patch to 2.5.3 and experimented with 
> it a little bit. Basically I am curious about why we
> we need to call run_all_timers() (which runs timers for all
> CPUs) through the timer bh if locking fails in run_local_timers(). 

Some driver do ugly things with TIMER_BH, and Ingo's 2.4 patched
tried to stayed source compatible with 2.4 drivers.

For 2.5 I'd really like to see TIMER_BH (all BH's in fact) to gone.

	Christoph

-- 
Of course it doesn't work. We've performed a software upgrade.

Re: [RFC][PATCH] Ingo's smptimers patch experiment

[Dipankar Sarma]

On Wed, Feb 06, 2002 at 06:27:46PM +0100, Christoph Hellwig wrote:
> In article <20020206211925.A8720@in.ibm.com> you wrote:
> > Hi Ingo,
> >
> > I ported your smptimers patch to 2.5.3 and experimented with 
> > it a little bit. Basically I am curious about why we
> > we need to call run_all_timers() (which runs timers for all
> > CPUs) through the timer bh if locking fails in run_local_timers(). 
> 
> Some driver do ugly things with TIMER_BH, and Ingo's 2.4 patched
> tried to stayed source compatible with 2.4 drivers.
> 
> For 2.5 I'd really like to see TIMER_BH (all BH's in fact) to gone.

I can see that net driver relies on being able to disable all timers 
by doing in net/core/dev.c -

        tasklet_disable(bh_task_vec+TIMER_BH);

But this doesn't completely disable timers in Ingo's patch 
since timers can also be fired through run_local_timers() if 
locking succeeds, no TIMER_BH in that case.

There are only a few places where TIMER_BH is used. I will see if
I can make another smptimers patch that gets rid of them.

Thanks
Dipankar
-- 
Dipankar Sarma  <dipankar@in.ibm.com> http://lse.sourceforge.net
Linux Technology Center, IBM Software Lab, Bangalore, India.