* RE: [PATCH 0/3] fix ia64 clocksource
2007-07-18 14:58 [PATCH 0/3] fix ia64 clocksource Hidetoshi Seto
2007-07-20 10:44 ` Bob Picco
@ 2007-07-20 16:33 ` Luck, Tony
2007-07-20 16:55 ` Bob Picco
2007-07-23 0:57 ` Hidetoshi Seto
3 siblings, 0 replies; 5+ messages in thread
From: Luck, Tony @ 2007-07-20 16:33 UTC (permalink / raw)
To: linux-ia64
> I need to make the cleanups Tony identified. After that how would you like to
> proceed? Combining your patch 1/3 and my 1/2 seems logical.
I've done the cleanup, and merged Seto-san's parts 1,2,3 with your part 1, and
updated it all for the tip of Linus' tree. Result is below.
Part 2 stayed the same, except that I also deleted Documentation/time_interpolators.txt
I'm going to send both to Linus later today ... so read fast and shout if
I botched the merge.
Time taken for a single gettimeofday() call via the fsys.S path has increased
by about 8 cycles, but this would only be significant for applications doing
> 1,000,000 gettimeofday calls per second. The scalability problems are so
bad that I can't even get close to that rate on 4 cpus (peak cumulative rate
is about 600,000). The patched system still has problems, but scalabilty is
better (4 cpus can get to 1,100,000 cumulative calls per second).
-Tony
commit 89f8d0677c72742767f98b9f5fcfbd575e73d0ae
Author: Tony Luck <tony.luck@intel.com>
Date: Wed Jul 18 15:38:51 2007 -0700
[IA64] Convert to generic timekeeping/clocksource
This is a merge of Peter Keilty's initial patch for this with
Hidetoshi Seto's fixes and scaling improvements.
Signed-off-by: Tony Luck <tony.luck@intel.com>
diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt
index 9a54148..68115d7 100644
--- a/Documentation/kernel-parameters.txt
+++ b/Documentation/kernel-parameters.txt
@@ -1154,6 +1154,8 @@ and is between 256 and 4096 characters. It is defined in the file
nointroute [IA-64]
+ nojitter [IA64] Disables jitter checking for ITC timers.
+
nolapic [IA-32,APIC] Do not enable or use the local APIC.
nolapic_timer [IA-32,APIC] Do not use the local APIC timer.
diff --git a/arch/ia64/Kconfig b/arch/ia64/Kconfig
index 616c96e..36c7b96 100644
--- a/arch/ia64/Kconfig
+++ b/arch/ia64/Kconfig
@@ -62,7 +62,11 @@ config GENERIC_CALIBRATE_DELAY
bool
default y
-config TIME_INTERPOLATION
+config GENERIC_TIME
+ bool
+ default y
+
+config GENERIC_TIME_VSYSCALL
bool
default y
diff --git a/arch/ia64/configs/bigsur_defconfig b/arch/ia64/configs/bigsur_defconfig
index 90e9c2e..9eb48c0 100644
--- a/arch/ia64/configs/bigsur_defconfig
+++ b/arch/ia64/configs/bigsur_defconfig
@@ -85,7 +85,7 @@ CONFIG_MMU=y
CONFIG_SWIOTLB=y
CONFIG_RWSEM_XCHGADD_ALGORITHM=y
CONFIG_GENERIC_CALIBRATE_DELAY=y
-CONFIG_TIME_INTERPOLATION=y
+CONFIG_GENERIC_TIME=y
CONFIG_EFI=y
CONFIG_GENERIC_IOMAP=y
CONFIG_SCHED_NO_NO_OMIT_FRAME_POINTER=y
diff --git a/arch/ia64/configs/gensparse_defconfig b/arch/ia64/configs/gensparse_defconfig
index 0d29aa2..3a9ed95 100644
--- a/arch/ia64/configs/gensparse_defconfig
+++ b/arch/ia64/configs/gensparse_defconfig
@@ -86,7 +86,7 @@ CONFIG_MMU=y
CONFIG_SWIOTLB=y
CONFIG_RWSEM_XCHGADD_ALGORITHM=y
CONFIG_GENERIC_CALIBRATE_DELAY=y
-CONFIG_TIME_INTERPOLATION=y
+CONFIG_GENERIC_TIME=y
CONFIG_EFI=y
CONFIG_GENERIC_IOMAP=y
CONFIG_SCHED_NO_NO_OMIT_FRAME_POINTER=y
diff --git a/arch/ia64/configs/sim_defconfig b/arch/ia64/configs/sim_defconfig
index d9146c3..c420d9f 100644
--- a/arch/ia64/configs/sim_defconfig
+++ b/arch/ia64/configs/sim_defconfig
@@ -86,7 +86,7 @@ CONFIG_MMU=y
CONFIG_SWIOTLB=y
CONFIG_RWSEM_XCHGADD_ALGORITHM=y
CONFIG_GENERIC_CALIBRATE_DELAY=y
-CONFIG_TIME_INTERPOLATION=y
+CONFIG_GENERIC_TIME=y
CONFIG_EFI=y
CONFIG_GENERIC_IOMAP=y
CONFIG_SCHED_NO_NO_OMIT_FRAME_POINTER=y
diff --git a/arch/ia64/configs/sn2_defconfig b/arch/ia64/configs/sn2_defconfig
index 64e951d..4c9ffc4 100644
--- a/arch/ia64/configs/sn2_defconfig
+++ b/arch/ia64/configs/sn2_defconfig
@@ -93,7 +93,7 @@ CONFIG_SWIOTLB=y
CONFIG_RWSEM_XCHGADD_ALGORITHM=y
CONFIG_GENERIC_FIND_NEXT_BIT=y
CONFIG_GENERIC_CALIBRATE_DELAY=y
-CONFIG_TIME_INTERPOLATION=y
+CONFIG_GENERIC_TIME=y
CONFIG_DMI=y
CONFIG_EFI=y
CONFIG_GENERIC_IOMAP=y
diff --git a/arch/ia64/configs/tiger_defconfig b/arch/ia64/configs/tiger_defconfig
index a144693..e208747 100644
--- a/arch/ia64/configs/tiger_defconfig
+++ b/arch/ia64/configs/tiger_defconfig
@@ -98,7 +98,7 @@ CONFIG_RWSEM_XCHGADD_ALGORITHM=y
# CONFIG_ARCH_HAS_ILOG2_U64 is not set
CONFIG_GENERIC_FIND_NEXT_BIT=y
CONFIG_GENERIC_CALIBRATE_DELAY=y
-CONFIG_TIME_INTERPOLATION=y
+CONFIG_GENERIC_TIME=y
CONFIG_DMI=y
CONFIG_EFI=y
CONFIG_GENERIC_IOMAP=y
diff --git a/arch/ia64/configs/zx1_defconfig b/arch/ia64/configs/zx1_defconfig
index 1c7955c..4a060fc 100644
--- a/arch/ia64/configs/zx1_defconfig
+++ b/arch/ia64/configs/zx1_defconfig
@@ -96,7 +96,7 @@ CONFIG_RWSEM_XCHGADD_ALGORITHM=y
# CONFIG_ARCH_HAS_ILOG2_U64 is not set
CONFIG_GENERIC_FIND_NEXT_BIT=y
CONFIG_GENERIC_CALIBRATE_DELAY=y
-CONFIG_TIME_INTERPOLATION=y
+CONFIG_GENERIC_TIME=y
CONFIG_DMI=y
CONFIG_EFI=y
CONFIG_GENERIC_IOMAP=y
diff --git a/arch/ia64/defconfig b/arch/ia64/defconfig
index 90bd960..461f8ee 100644
--- a/arch/ia64/defconfig
+++ b/arch/ia64/defconfig
@@ -98,7 +98,7 @@ CONFIG_RWSEM_XCHGADD_ALGORITHM=y
# CONFIG_ARCH_HAS_ILOG2_U64 is not set
CONFIG_GENERIC_FIND_NEXT_BIT=y
CONFIG_GENERIC_CALIBRATE_DELAY=y
-CONFIG_TIME_INTERPOLATION=y
+CONFIG_GENERIC_TIME=y
CONFIG_DMI=y
CONFIG_EFI=y
CONFIG_GENERIC_IOMAP=y
diff --git a/arch/ia64/kernel/asm-offsets.c b/arch/ia64/kernel/asm-offsets.c
index 2236fab..0aebc6f 100644
--- a/arch/ia64/kernel/asm-offsets.c
+++ b/arch/ia64/kernel/asm-offsets.c
@@ -7,6 +7,7 @@
#define ASM_OFFSETS_C 1
#include <linux/sched.h>
+#include <linux/clocksource.h>
#include <asm-ia64/processor.h>
#include <asm-ia64/ptrace.h>
@@ -15,6 +16,7 @@
#include <asm-ia64/mca.h>
#include "../kernel/sigframe.h"
+#include "../kernel/fsyscall_gtod_data.h"
#define DEFINE(sym, val) \
asm volatile("\n->" #sym " %0 " #val : : "i" (val))
@@ -256,17 +258,24 @@ void foo(void)
BLANK();
/* used by fsys_gettimeofday in arch/ia64/kernel/fsys.S */
- DEFINE(IA64_TIME_INTERPOLATOR_ADDRESS_OFFSET, offsetof (struct time_interpolator, addr));
- DEFINE(IA64_TIME_INTERPOLATOR_SOURCE_OFFSET, offsetof (struct time_interpolator, source));
- DEFINE(IA64_TIME_INTERPOLATOR_SHIFT_OFFSET, offsetof (struct time_interpolator, shift));
- DEFINE(IA64_TIME_INTERPOLATOR_NSEC_OFFSET, offsetof (struct time_interpolator, nsec_per_cyc));
- DEFINE(IA64_TIME_INTERPOLATOR_OFFSET_OFFSET, offsetof (struct time_interpolator, offset));
- DEFINE(IA64_TIME_INTERPOLATOR_LAST_CYCLE_OFFSET, offsetof (struct time_interpolator, last_cycle));
- DEFINE(IA64_TIME_INTERPOLATOR_LAST_COUNTER_OFFSET, offsetof (struct time_interpolator, last_counter));
- DEFINE(IA64_TIME_INTERPOLATOR_JITTER_OFFSET, offsetof (struct time_interpolator, jitter));
- DEFINE(IA64_TIME_INTERPOLATOR_MASK_OFFSET, offsetof (struct time_interpolator, mask));
- DEFINE(IA64_TIME_SOURCE_CPU, TIME_SOURCE_CPU);
- DEFINE(IA64_TIME_SOURCE_MMIO64, TIME_SOURCE_MMIO64);
- DEFINE(IA64_TIME_SOURCE_MMIO32, TIME_SOURCE_MMIO32);
- DEFINE(IA64_TIMESPEC_TV_NSEC_OFFSET, offsetof (struct timespec, tv_nsec));
+ DEFINE(IA64_GTOD_LOCK_OFFSET,
+ offsetof (struct fsyscall_gtod_data_t, lock));
+ DEFINE(IA64_GTOD_WALL_TIME_OFFSET,
+ offsetof (struct fsyscall_gtod_data_t, wall_time));
+ DEFINE(IA64_GTOD_MONO_TIME_OFFSET,
+ offsetof (struct fsyscall_gtod_data_t, monotonic_time));
+ DEFINE(IA64_CLKSRC_MASK_OFFSET,
+ offsetof (struct fsyscall_gtod_data_t, clk_mask));
+ DEFINE(IA64_CLKSRC_MULT_OFFSET,
+ offsetof (struct fsyscall_gtod_data_t, clk_mult));
+ DEFINE(IA64_CLKSRC_SHIFT_OFFSET,
+ offsetof (struct fsyscall_gtod_data_t, clk_shift));
+ DEFINE(IA64_CLKSRC_MMIO_OFFSET,
+ offsetof (struct fsyscall_gtod_data_t, clk_fsys_mmio));
+ DEFINE(IA64_CLKSRC_CYCLE_LAST_OFFSET,
+ offsetof (struct fsyscall_gtod_data_t, clk_cycle_last));
+ DEFINE(IA64_ITC_JITTER_OFFSET,
+ offsetof (struct itc_jitter_data_t, itc_jitter));
+ DEFINE(IA64_ITC_LASTCYCLE_OFFSET,
+ offsetof (struct itc_jitter_data_t, itc_lastcycle));
}
diff --git a/arch/ia64/kernel/cyclone.c b/arch/ia64/kernel/cyclone.c
index e00b215..2fd96d9 100644
--- a/arch/ia64/kernel/cyclone.c
+++ b/arch/ia64/kernel/cyclone.c
@@ -3,6 +3,7 @@
#include <linux/time.h>
#include <linux/errno.h>
#include <linux/timex.h>
+#include <linux/clocksource.h>
#include <asm/io.h>
/* IBM Summit (EXA) Cyclone counter code*/
@@ -18,13 +19,21 @@ void __init cyclone_setup(void)
use_cyclone = 1;
}
+static void __iomem *cyclone_mc;
-struct time_interpolator cyclone_interpolator = {
- .source = TIME_SOURCE_MMIO64,
- .shift = 16,
- .frequency = CYCLONE_TIMER_FREQ,
- .drift = -100,
- .mask = (1LL << 40) - 1
+static cycle_t read_cyclone(void)
+{
+ return (cycle_t)readq((void __iomem *)cyclone_mc);
+}
+
+static struct clocksource clocksource_cyclone = {
+ .name = "cyclone",
+ .rating = 300,
+ .read = read_cyclone,
+ .mask = (1LL << 40) - 1,
+ .mult = 0, /*to be caluclated*/
+ .shift = 16,
+ .flags = CLOCK_SOURCE_IS_CONTINUOUS,
};
int __init init_cyclone_clock(void)
@@ -44,13 +53,15 @@ int __init init_cyclone_clock(void)
offset = (CYCLONE_CBAR_ADDR);
reg = (u64*)ioremap_nocache(offset, sizeof(u64));
if(!reg){
- printk(KERN_ERR "Summit chipset: Could not find valid CBAR register.\n");
+ printk(KERN_ERR "Summit chipset: Could not find valid CBAR"
+ " register.\n");
use_cyclone = 0;
return -ENODEV;
}
base = readq(reg);
if(!base){
- printk(KERN_ERR "Summit chipset: Could not find valid CBAR value.\n");
+ printk(KERN_ERR "Summit chipset: Could not find valid CBAR"
+ " value.\n");
use_cyclone = 0;
return -ENODEV;
}
@@ -60,7 +71,8 @@ int __init init_cyclone_clock(void)
offset = (base + CYCLONE_PMCC_OFFSET);
reg = (u64*)ioremap_nocache(offset, sizeof(u64));
if(!reg){
- printk(KERN_ERR "Summit chipset: Could not find valid PMCC register.\n");
+ printk(KERN_ERR "Summit chipset: Could not find valid PMCC"
+ " register.\n");
use_cyclone = 0;
return -ENODEV;
}
@@ -71,7 +83,8 @@ int __init init_cyclone_clock(void)
offset = (base + CYCLONE_MPCS_OFFSET);
reg = (u64*)ioremap_nocache(offset, sizeof(u64));
if(!reg){
- printk(KERN_ERR "Summit chipset: Could not find valid MPCS register.\n");
+ printk(KERN_ERR "Summit chipset: Could not find valid MPCS"
+ " register.\n");
use_cyclone = 0;
return -ENODEV;
}
@@ -82,7 +95,8 @@ int __init init_cyclone_clock(void)
offset = (base + CYCLONE_MPMC_OFFSET);
cyclone_timer = (u32*)ioremap_nocache(offset, sizeof(u32));
if(!cyclone_timer){
- printk(KERN_ERR "Summit chipset: Could not find valid MPMC register.\n");
+ printk(KERN_ERR "Summit chipset: Could not find valid MPMC"
+ " register.\n");
use_cyclone = 0;
return -ENODEV;
}
@@ -93,7 +107,8 @@ int __init init_cyclone_clock(void)
int stall = 100;
while(stall--) barrier();
if(readl(cyclone_timer) = old){
- printk(KERN_ERR "Summit chipset: Counter not counting! DISABLED\n");
+ printk(KERN_ERR "Summit chipset: Counter not counting!"
+ " DISABLED\n");
iounmap(cyclone_timer);
cyclone_timer = 0;
use_cyclone = 0;
@@ -101,8 +116,11 @@ int __init init_cyclone_clock(void)
}
}
/* initialize last tick */
- cyclone_interpolator.addr = cyclone_timer;
- register_time_interpolator(&cyclone_interpolator);
+ cyclone_mc = cyclone_timer;
+ clocksource_cyclone.fsys_mmio = cyclone_timer;
+ clocksource_cyclone.mult = clocksource_hz2mult(CYCLONE_TIMER_FREQ,
+ clocksource_cyclone.shift);
+ clocksource_register(&clocksource_cyclone);
return 0;
}
diff --git a/arch/ia64/kernel/fsys.S b/arch/ia64/kernel/fsys.S
index 3f926c2..4484197 100644
--- a/arch/ia64/kernel/fsys.S
+++ b/arch/ia64/kernel/fsys.S
@@ -147,12 +147,11 @@ ENTRY(fsys_set_tid_address)
FSYS_RETURN
END(fsys_set_tid_address)
-/*
- * Ensure that the time interpolator structure is compatible with the asm code
- */
-#if IA64_TIME_INTERPOLATOR_SOURCE_OFFSET !=0 || IA64_TIME_INTERPOLATOR_SHIFT_OFFSET != 2 \
- || IA64_TIME_INTERPOLATOR_JITTER_OFFSET != 3 || IA64_TIME_INTERPOLATOR_NSEC_OFFSET != 4
-#error fsys_gettimeofday incompatible with changes to struct time_interpolator
+#if IA64_GTOD_LOCK_OFFSET !=0
+#error fsys_gettimeofday incompatible with changes to struct fsyscall_gtod_data_t
+#endif
+#if IA64_ITC_JITTER_OFFSET !=0
+#error fsys_gettimeofday incompatible with changes to struct itc_jitter_data_t
#endif
#define CLOCK_REALTIME 0
#define CLOCK_MONOTONIC 1
@@ -179,126 +178,124 @@ ENTRY(fsys_gettimeofday)
// r11 = preserved: saved ar.pfs
// r12 = preserved: memory stack
// r13 = preserved: thread pointer
- // r14 = address of mask / mask
+ // r14 = address of mask / mask value
// r15 = preserved: system call number
// r16 = preserved: current task pointer
- // r17 = wall to monotonic use
- // r18 = time_interpolator->offset
- // r19 = address of wall_to_monotonic
- // r20 = pointer to struct time_interpolator / pointer to time_interpolator->address
- // r21 = shift factor
- // r22 = address of time interpolator->last_counter
- // r23 = address of time_interpolator->last_cycle
- // r24 = adress of time_interpolator->offset
- // r25 = last_cycle value
- // r26 = last_counter value
- // r27 = pointer to xtime
+ // r17 = (not used)
+ // r18 = (not used)
+ // r19 = address of itc_lastcycle
+ // r20 = struct fsyscall_gtod_data (= address of gtod_lock.sequence)
+ // r21 = address of mmio_ptr
+ // r22 = address of wall_time or monotonic_time
+ // r23 = address of shift / value
+ // r24 = address mult factor / cycle_last value
+ // r25 = itc_lastcycle value
+ // r26 = address clocksource cycle_last
+ // r27 = (not used)
// r28 = sequence number at the beginning of critcal section
- // r29 = address of seqlock
+ // r29 = address of itc_jitter
// r30 = time processing flags / memory address
// r31 = pointer to result
// Predicates
// p6,p7 short term use
// p8 = timesource ar.itc
// p9 = timesource mmio64
- // p10 = timesource mmio32
+ // p10 = timesource mmio32 - not used
// p11 = timesource not to be handled by asm code
- // p12 = memory time source ( = p9 | p10)
- // p13 = do cmpxchg with time_interpolator_last_cycle
+ // p12 = memory time source ( = p9 | p10) - not used
+ // p13 = do cmpxchg with itc_lastcycle
// p14 = Divide by 1000
// p15 = Add monotonic
//
- // Note that instructions are optimized for McKinley. McKinley can process two
- // bundles simultaneously and therefore we continuously try to feed the CPU
- // two bundles and then a stop.
- tnat.nz p6,p0 = r31 // branch deferred since it does not fit into bundle structure
+ // Note that instructions are optimized for McKinley. McKinley can
+ // process two bundles simultaneously and therefore we continuously
+ // try to feed the CPU two bundles and then a stop.
+ //
+ // Additional note that code has changed a lot. Optimization is TBD.
+ // Comments begin with "?" are maybe outdated.
+ tnat.nz p6,p0 = r31 // ? branch deferred to fit later bundle
mov pr = r30,0xc000 // Set predicates according to function
add r2 = TI_FLAGS+IA64_TASK_SIZE,r16
- movl r20 = time_interpolator
+ movl r20 = fsyscall_gtod_data // load fsyscall gettimeofday data address
;;
- ld8 r20 = [r20] // get pointer to time_interpolator structure
- movl r29 = xtime_lock
+ movl r29 = itc_jitter_data // itc_jitter
+ add r22 = IA64_GTOD_WALL_TIME_OFFSET,r20 // wall_time
ld4 r2 = [r2] // process work pending flags
- movl r27 = xtime
- ;; // only one bundle here
- ld8 r21 = [r20] // first quad with control information
+ ;;
+(p15) add r22 = IA64_GTOD_MONO_TIME_OFFSET,r20 // monotonic_time
+ add r21 = IA64_CLKSRC_MMIO_OFFSET,r20
+ add r19 = IA64_ITC_LASTCYCLE_OFFSET,r29
and r2 = TIF_ALLWORK_MASK,r2
-(p6) br.cond.spnt.few .fail_einval // deferred branch
+(p6) br.cond.spnt.few .fail_einval // ? deferred branch
;;
- add r10 = IA64_TIME_INTERPOLATOR_ADDRESS_OFFSET,r20
- extr r3 = r21,32,32 // time_interpolator->nsec_per_cyc
- extr r8 = r21,0,16 // time_interpolator->source
+ add r26 = IA64_CLKSRC_CYCLE_LAST_OFFSET,r20 // clksrc_cycle_last
cmp.ne p6, p0 = 0, r2 // Fallback if work is scheduled
(p6) br.cond.spnt.many fsys_fallback_syscall
;;
- cmp.eq p8,p12 = 0,r8 // Check for cpu timer
- cmp.eq p9,p0 = 1,r8 // MMIO64 ?
- extr r2 = r21,24,8 // time_interpolator->jitter
- cmp.eq p10,p0 = 2,r8 // MMIO32 ?
- cmp.ltu p11,p0 = 2,r8 // function or other clock
-(p11) br.cond.spnt.many fsys_fallback_syscall
+ // Begin critical section
+.time_redo:
+ ld4.acq r28 = [r20] // gtod_lock.sequence, Must take first
+ ;;
+ and r28 = ~1,r28 // And make sequence even to force retry if odd
;;
- setf.sig f7 = r3 // Setup for scaling of counter
-(p15) movl r19 = wall_to_monotonic
-(p12) ld8 r30 = [r10]
- cmp.ne p13,p0 = r2,r0 // need jitter compensation?
- extr r21 = r21,16,8 // shift factor
+ ld8 r30 = [r21] // clocksource->mmio_ptr
+ add r24 = IA64_CLKSRC_MULT_OFFSET,r20
+ ld4 r2 = [r29] // itc_jitter value
+ add r23 = IA64_CLKSRC_SHIFT_OFFSET,r20
+ add r14 = IA64_CLKSRC_MASK_OFFSET,r20
;;
-.time_redo:
- .pred.rel.mutex p8,p9,p10
- ld4.acq r28 = [r29] // xtime_lock.sequence. Must come first for locking purposes
+ ld4 r3 = [r24] // clocksource mult value
+ ld8 r14 = [r14] // clocksource mask value
+ cmp.eq p8,p9 = 0,r30 // use cpu timer if no mmio_ptr
;;
- and r28 = ~1,r28 // Make sequence even to force retry if odd
+ setf.sig f7 = r3 // Setup for mult scaling of counter
+(p8) cmp.ne p13,p0 = r2,r0 // need itc_jitter compensation, set p13
+ ld4 r23 = [r23] // clocksource shift value
+ ld8 r24 = [r26] // get clksrc_cycle_last value
+(p9) cmp.eq p13,p0 = 0,r30 // if mmio_ptr, clear p13 jitter control
;;
+ .pred.rel.mutex p8,p9
(p8) mov r2 = ar.itc // CPU_TIMER. 36 clocks latency!!!
- add r22 = IA64_TIME_INTERPOLATOR_LAST_COUNTER_OFFSET,r20
-(p9) ld8 r2 = [r30] // readq(ti->address). Could also have latency issues..
-(p10) ld4 r2 = [r30] // readw(ti->address)
-(p13) add r23 = IA64_TIME_INTERPOLATOR_LAST_CYCLE_OFFSET,r20
- ;; // could be removed by moving the last add upward
- ld8 r26 = [r22] // time_interpolator->last_counter
-(p13) ld8 r25 = [r23] // time interpolator->last_cycle
- add r24 = IA64_TIME_INTERPOLATOR_OFFSET_OFFSET,r20
-(p15) ld8 r17 = [r19],IA64_TIMESPEC_TV_NSEC_OFFSET
- ld8 r9 = [r27],IA64_TIMESPEC_TV_NSEC_OFFSET
- add r14 = IA64_TIME_INTERPOLATOR_MASK_OFFSET, r20
- ;;
- ld8 r18 = [r24] // time_interpolator->offset
- ld8 r8 = [r27],-IA64_TIMESPEC_TV_NSEC_OFFSET // xtime.tv_nsec
-(p13) sub r3 = r25,r2 // Diff needed before comparison (thanks davidm)
- ;;
- ld8 r14 = [r14] // time_interpolator->mask
-(p13) cmp.gt.unc p6,p7 = r3,r0 // check if it is less than last. p6,p7 cleared
- sub r10 = r2,r26 // current_counter - last_counter
- ;;
-(p6) sub r10 = r25,r26 // time we got was less than last_cycle
+(p9) ld8 r2 = [r30] // MMIO_TIMER. Could also have latency issues..
+(p13) ld8 r25 = [r19] // get itc_lastcycle value
+ ;; // ? could be removed by moving the last add upward
+ ld8 r9 = [r22],IA64_TIMESPEC_TV_NSEC_OFFSET // tv_sec
+ ;;
+ ld8 r8 = [r22],-IA64_TIMESPEC_TV_NSEC_OFFSET // tv_nsec
+(p13) sub r3 = r25,r2 // Diff needed before comparison (thanks davidm)
+ ;;
+(p13) cmp.gt.unc p6,p7 = r3,r0 // check if it is less than last. p6,p7 cleared
+ sub r10 = r2,r24 // current_cycle - last_cycle
+ ;;
+(p6) sub r10 = r25,r24 // time we got was less than last_cycle
(p7) mov ar.ccv = r25 // more than last_cycle. Prep for cmpxchg
;;
+(p7) cmpxchg8.rel r3 = [r19],r2,ar.ccv
+ ;;
+(p7) cmp.ne p7,p0 = r25,r3 // if cmpxchg not successful
+ ;;
+(p7) sub r10 = r3,r24 // then use new last_cycle instead
+ ;;
and r10 = r10,r14 // Apply mask
;;
setf.sig f8 = r10
nop.i 123
;;
-(p7) cmpxchg8.rel r3 = [r23],r2,ar.ccv
-EX(.fail_efault, probe.w.fault r31, 3) // This takes 5 cycles and we have spare time
+ // fault check takes 5 cycles and we have spare time
+EX(.fail_efault, probe.w.fault r31, 3)
xmpy.l f8 = f8,f7 // nsec_per_cyc*(counter-last_counter)
-(p15) add r9 = r9,r17 // Add wall to monotonic.secs to result secs
;;
-(p15) ld8 r17 = [r19],-IA64_TIMESPEC_TV_NSEC_OFFSET
-(p7) cmp.ne p7,p0 = r25,r3 // if cmpxchg not successful redo
- // simulate tbit.nz.or p7,p0 = r28,0
+ // ? simulate tbit.nz.or p7,p0 = r28,0
getf.sig r2 = f8
mf
- add r8 = r8,r18 // Add time interpolator offset
;;
- ld4 r10 = [r29] // xtime_lock.sequence
-(p15) add r8 = r8, r17 // Add monotonic.nsecs to nsecs
- shr.u r2 = r2,r21
- ;; // overloaded 3 bundles!
- // End critical section.
+ ld4 r10 = [r20] // gtod_lock.sequence
+ shr.u r2 = r2,r23 // shift by factor
+ ;; // ? overloaded 3 bundles!
add r8 = r8,r2 // Add xtime.nsecs
- cmp4.ne.or p7,p0 = r28,r10
-(p7) br.cond.dpnt.few .time_redo // sequence number changed ?
+ cmp4.ne p7,p0 = r28,r10
+(p7) br.cond.dpnt.few .time_redo // sequence number changed, redo
+ // End critical section.
// Now r8=tv->tv_nsec and r9=tv->tv_sec
mov r10 = r0
movl r2 = 1000000000
@@ -308,19 +305,19 @@ EX(.fail_efault, probe.w.fault r31, 3) // This takes 5 cycles and we have spare
.time_normalize:
mov r21 = r8
cmp.ge p6,p0 = r8,r2
-(p14) shr.u r20 = r8, 3 // We can repeat this if necessary just wasting some time
+(p14) shr.u r20 = r8, 3 // We can repeat this if necessary just wasting time
;;
(p14) setf.sig f8 = r20
(p6) sub r8 = r8,r2
-(p6) add r9 = 1,r9 // two nops before the branch.
-(p14) setf.sig f7 = r3 // Chances for repeats are 1 in 10000 for gettod
+(p6) add r9 = 1,r9 // two nops before the branch.
+(p14) setf.sig f7 = r3 // Chances for repeats are 1 in 10000 for gettod
(p6) br.cond.dpnt.few .time_normalize
;;
// Divided by 8 though shift. Now divide by 125
// The compiler was able to do that with a multiply
// and a shift and we do the same
-EX(.fail_efault, probe.w.fault r23, 3) // This also costs 5 cycles
-(p14) xmpy.hu f8 = f8, f7 // xmpy has 5 cycles latency so use it...
+EX(.fail_efault, probe.w.fault r23, 3) // This also costs 5 cycles
+(p14) xmpy.hu f8 = f8, f7 // xmpy has 5 cycles latency so use it
;;
mov r8 = r0
(p14) getf.sig r2 = f8
diff --git a/arch/ia64/kernel/fsyscall_gtod_data.h b/arch/ia64/kernel/fsyscall_gtod_data.h
new file mode 100644
index 0000000..490dab5
--- /dev/null
+++ b/arch/ia64/kernel/fsyscall_gtod_data.h
@@ -0,0 +1,23 @@
+/*
+ * (c) Copyright 2007 Hewlett-Packard Development Company, L.P.
+ * Contributed by Peter Keilty <peter.keilty@hp.com>
+ *
+ * fsyscall gettimeofday data
+ */
+
+struct fsyscall_gtod_data_t {
+ seqlock_t lock;
+ struct timespec wall_time;
+ struct timespec monotonic_time;
+ cycle_t clk_mask;
+ u32 clk_mult;
+ u32 clk_shift;
+ void *clk_fsys_mmio;
+ cycle_t clk_cycle_last;
+} __attribute__ ((aligned (L1_CACHE_BYTES)));
+
+struct itc_jitter_data_t {
+ int itc_jitter;
+ cycle_t itc_lastcycle;
+} __attribute__ ((aligned (L1_CACHE_BYTES)));
+
diff --git a/arch/ia64/kernel/time.c b/arch/ia64/kernel/time.c
index 3486fe7..627785c 100644
--- a/arch/ia64/kernel/time.c
+++ b/arch/ia64/kernel/time.c
@@ -19,6 +19,7 @@
#include <linux/interrupt.h>
#include <linux/efi.h>
#include <linux/timex.h>
+#include <linux/clocksource.h>
#include <asm/machvec.h>
#include <asm/delay.h>
@@ -28,6 +29,16 @@
#include <asm/sections.h>
#include <asm/system.h>
+#include "fsyscall_gtod_data.h"
+
+static cycle_t itc_get_cycles(void);
+
+struct fsyscall_gtod_data_t fsyscall_gtod_data = {
+ .lock = SEQLOCK_UNLOCKED,
+};
+
+struct itc_jitter_data_t itc_jitter_data;
+
volatile int time_keeper_id = 0; /* smp_processor_id() of time-keeper */
#ifdef CONFIG_IA64_DEBUG_IRQ
@@ -37,11 +48,16 @@ EXPORT_SYMBOL(last_cli_ip);
#endif
-static struct time_interpolator itc_interpolator = {
- .shift = 16,
- .mask = 0xffffffffffffffffLL,
- .source = TIME_SOURCE_CPU
+static struct clocksource clocksource_itc = {
+ .name = "itc",
+ .rating = 350,
+ .read = itc_get_cycles,
+ .mask = 0xffffffffffffffff,
+ .mult = 0, /*to be caluclated*/
+ .shift = 16,
+ .flags = CLOCK_SOURCE_IS_CONTINUOUS,
};
+static struct clocksource *itc_clocksource;
static irqreturn_t
timer_interrupt (int irq, void *dev_id)
@@ -210,8 +226,6 @@ ia64_init_itm (void)
+ itc_freq/2)/itc_freq;
if (!(sal_platform_features & IA64_SAL_PLATFORM_FEATURE_ITC_DRIFT)) {
- itc_interpolator.frequency = local_cpu_data->itc_freq;
- itc_interpolator.drift = itc_drift;
#ifdef CONFIG_SMP
/* On IA64 in an SMP configuration ITCs are never accurately synchronized.
* Jitter compensation requires a cmpxchg which may limit
@@ -223,15 +237,50 @@ ia64_init_itm (void)
* even going backward) if the ITC offsets between the individual CPUs
* are too large.
*/
- if (!nojitter) itc_interpolator.jitter = 1;
+ if (!nojitter)
+ itc_jitter_data.itc_jitter = 1;
#endif
- register_time_interpolator(&itc_interpolator);
}
/* Setup the CPU local timer tick */
ia64_cpu_local_tick();
+
+ if (!itc_clocksource) {
+ /* Sort out mult/shift values: */
+ clocksource_itc.mult + clocksource_hz2mult(local_cpu_data->itc_freq,
+ clocksource_itc.shift);
+ clocksource_register(&clocksource_itc);
+ itc_clocksource = &clocksource_itc;
+ }
}
+static cycle_t itc_get_cycles()
+{
+ u64 lcycle, now, ret;
+
+ if (!itc_jitter_data.itc_jitter)
+ return get_cycles();
+
+ lcycle = itc_jitter_data.itc_lastcycle;
+ now = get_cycles();
+ if (lcycle && time_after(lcycle, now))
+ return lcycle;
+
+ /*
+ * Keep track of the last timer value returned.
+ * In an SMP environment, you could lose out in contention of
+ * cmpxchg. If so, your cmpxchg returns new value which the
+ * winner of contention updated to. Use the new value instead.
+ */
+ ret = cmpxchg(&itc_jitter_data.itc_lastcycle, lcycle, now);
+ if (unlikely(ret != lcycle))
+ return ret;
+
+ return now;
+}
+
+
static struct irqaction timer_irqaction = {
.handler = timer_interrupt,
.flags = IRQF_DISABLED | IRQF_IRQPOLL,
@@ -307,3 +356,34 @@ ia64_setup_printk_clock(void)
if (!(sal_platform_features & IA64_SAL_PLATFORM_FEATURE_ITC_DRIFT))
ia64_printk_clock = ia64_itc_printk_clock;
}
+
+void update_vsyscall(struct timespec *wall, struct clocksource *c)
+{
+ unsigned long flags;
+
+ write_seqlock_irqsave(&fsyscall_gtod_data.lock, flags);
+
+ /* copy fsyscall clock data */
+ fsyscall_gtod_data.clk_mask = c->mask;
+ fsyscall_gtod_data.clk_mult = c->mult;
+ fsyscall_gtod_data.clk_shift = c->shift;
+ fsyscall_gtod_data.clk_fsys_mmio = c->fsys_mmio;
+ fsyscall_gtod_data.clk_cycle_last = c->cycle_last;
+
+ /* copy kernel time structures */
+ fsyscall_gtod_data.wall_time.tv_sec = wall->tv_sec;
+ fsyscall_gtod_data.wall_time.tv_nsec = wall->tv_nsec;
+ fsyscall_gtod_data.monotonic_time.tv_sec = wall_to_monotonic.tv_sec
+ + wall->tv_sec;
+ fsyscall_gtod_data.monotonic_time.tv_nsec = wall_to_monotonic.tv_nsec
+ + wall->tv_nsec;
+
+ /* normalize */
+ while (fsyscall_gtod_data.monotonic_time.tv_nsec >= NSEC_PER_SEC) {
+ fsyscall_gtod_data.monotonic_time.tv_nsec -= NSEC_PER_SEC;
+ fsyscall_gtod_data.monotonic_time.tv_sec++;
+ }
+
+ write_sequnlock_irqrestore(&fsyscall_gtod_data.lock, flags);
+}
+
diff --git a/arch/ia64/sn/kernel/sn2/timer.c b/arch/ia64/sn/kernel/sn2/timer.c
index 56a88b6..19e25d2 100644
--- a/arch/ia64/sn/kernel/sn2/timer.c
+++ b/arch/ia64/sn/kernel/sn2/timer.c
@@ -11,6 +11,7 @@
#include <linux/sched.h>
#include <linux/time.h>
#include <linux/interrupt.h>
+#include <linux/clocksource.h>
#include <asm/hw_irq.h>
#include <asm/system.h>
@@ -22,11 +23,21 @@
extern unsigned long sn_rtc_cycles_per_second;
-static struct time_interpolator sn2_interpolator = {
- .drift = -1,
- .shift = 10,
- .mask = (1LL << 55) - 1,
- .source = TIME_SOURCE_MMIO64
+static void __iomem *sn2_mc;
+
+static cycle_t read_sn2(void)
+{
+ return (cycle_t)readq(sn2_mc);
+}
+
+static struct clocksource clocksource_sn2 = {
+ .name = "sn2_rtc",
+ .rating = 300,
+ .read = read_sn2,
+ .mask = (1LL << 55) - 1,
+ .mult = 0,
+ .shift = 10,
+ .flags = CLOCK_SOURCE_IS_CONTINUOUS,
};
/*
@@ -47,9 +58,11 @@ ia64_sn_udelay (unsigned long usecs)
void __init sn_timer_init(void)
{
- sn2_interpolator.frequency = sn_rtc_cycles_per_second;
- sn2_interpolator.addr = RTC_COUNTER_ADDR;
- register_time_interpolator(&sn2_interpolator);
+ sn2_mc = RTC_COUNTER_ADDR;
+ clocksource_sn2.fsys_mmio = RTC_COUNTER_ADDR;
+ clocksource_sn2.mult = clocksource_hz2mult(sn_rtc_cycles_per_second,
+ clocksource_sn2.shift);
+ clocksource_register(&clocksource_sn2);
ia64_udelay = &ia64_sn_udelay;
}
diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c
index 80ffc78..bb5d23b 100644
--- a/drivers/acpi/processor_idle.c
+++ b/drivers/acpi/processor_idle.c
@@ -475,7 +475,7 @@ static void acpi_processor_idle(void)
/* Get end time (ticks) */
t2 = inl(acpi_gbl_FADT.xpm_timer_block.address);
-#ifdef CONFIG_GENERIC_TIME
+#if defined (CONFIG_GENERIC_TIME) && defined (CONFIG_X86_TSC)
/* TSC halts in C2, so notify users */
mark_tsc_unstable("possible TSC halt in C2");
#endif
@@ -517,7 +517,7 @@ static void acpi_processor_idle(void)
acpi_set_register(ACPI_BITREG_ARB_DISABLE, 0);
}
-#ifdef CONFIG_GENERIC_TIME
+#if defined (CONFIG_GENERIC_TIME) && defined (CONFIG_X86_TSC)
/* TSC halts in C3, so notify users */
mark_tsc_unstable("TSC halts in C3");
#endif
diff --git a/drivers/char/hpet.c b/drivers/char/hpet.c
index 0be700f..ba0e74a 100644
--- a/drivers/char/hpet.c
+++ b/drivers/char/hpet.c
@@ -29,6 +29,7 @@
#include <linux/bcd.h>
#include <linux/seq_file.h>
#include <linux/bitops.h>
+#include <linux/clocksource.h>
#include <asm/current.h>
#include <asm/uaccess.h>
@@ -51,8 +52,34 @@
#define HPET_RANGE_SIZE 1024 /* from HPET spec */
+#if BITS_PER_LONG = 64
+#define write_counter(V, MC) writeq(V, MC)
+#define read_counter(MC) readq(MC)
+#else
+#define write_counter(V, MC) writel(V, MC)
+#define read_counter(MC) readl(MC)
+#endif
+
static u32 hpet_nhpet, hpet_max_freq = HPET_USER_FREQ;
+static void __iomem *hpet_mctr;
+
+static cycle_t read_hpet(void)
+{
+ return (cycle_t)read_counter((void __iomem *)hpet_mctr);
+}
+
+static struct clocksource clocksource_hpet = {
+ .name = "hpet",
+ .rating = 250,
+ .read = read_hpet,
+ .mask = 0xffffffffffffffff,
+ .mult = 0, /*to be caluclated*/
+ .shift = 10,
+ .flags = CLOCK_SOURCE_IS_CONTINUOUS,
+};
+static struct clocksource *hpet_clocksource;
+
/* A lock for concurrent access by app and isr hpet activity. */
static DEFINE_SPINLOCK(hpet_lock);
/* A lock for concurrent intermodule access to hpet and isr hpet activity. */
@@ -79,7 +106,7 @@ struct hpets {
struct hpets *hp_next;
struct hpet __iomem *hp_hpet;
unsigned long hp_hpet_phys;
- struct time_interpolator *hp_interpolator;
+ struct clocksource *hp_clocksource;
unsigned long long hp_tick_freq;
unsigned long hp_delta;
unsigned int hp_ntimer;
@@ -94,13 +121,6 @@ static struct hpets *hpets;
#define HPET_PERIODIC 0x0004
#define HPET_SHARED_IRQ 0x0008
-#if BITS_PER_LONG = 64
-#define write_counter(V, MC) writeq(V, MC)
-#define read_counter(MC) readq(MC)
-#else
-#define write_counter(V, MC) writel(V, MC)
-#define read_counter(MC) readl(MC)
-#endif
#ifndef readq
static inline unsigned long long readq(void __iomem *addr)
@@ -737,27 +757,6 @@ static ctl_table dev_root[] = {
static struct ctl_table_header *sysctl_header;
-static void hpet_register_interpolator(struct hpets *hpetp)
-{
-#ifdef CONFIG_TIME_INTERPOLATION
- struct time_interpolator *ti;
-
- ti = kzalloc(sizeof(*ti), GFP_KERNEL);
- if (!ti)
- return;
-
- ti->source = TIME_SOURCE_MMIO64;
- ti->shift = 10;
- ti->addr = &hpetp->hp_hpet->hpet_mc;
- ti->frequency = hpetp->hp_tick_freq;
- ti->drift = HPET_DRIFT;
- ti->mask = -1;
-
- hpetp->hp_interpolator = ti;
- register_time_interpolator(ti);
-#endif
-}
-
/*
* Adjustment for when arming the timer with
* initial conditions. That is, main counter
@@ -909,7 +908,16 @@ int hpet_alloc(struct hpet_data *hdp)
}
hpetp->hp_delta = hpet_calibrate(hpetp);
- hpet_register_interpolator(hpetp);
+
+ if (!hpet_clocksource) {
+ hpet_mctr = (void __iomem *)&hpetp->hp_hpet->hpet_mc;
+ CLKSRC_FSYS_MMIO_SET(clocksource_hpet.fsys_mmio, hpet_mctr);
+ clocksource_hpet.mult = clocksource_hz2mult(hpetp->hp_tick_freq,
+ clocksource_hpet.shift);
+ clocksource_register(&clocksource_hpet);
+ hpetp->hp_clocksource = &clocksource_hpet;
+ hpet_clocksource = &clocksource_hpet;
+ }
return 0;
}
@@ -995,7 +1003,7 @@ static int hpet_acpi_add(struct acpi_device *device)
static int hpet_acpi_remove(struct acpi_device *device, int type)
{
- /* XXX need to unregister interpolator, dealloc mem, etc */
+ /* XXX need to unregister clocksource, dealloc mem, etc */
return -EINVAL;
}
diff --git a/include/linux/clocksource.h b/include/linux/clocksource.h
index bf297b0..562c86a 100644
--- a/include/linux/clocksource.h
+++ b/include/linux/clocksource.h
@@ -67,6 +67,12 @@ struct clocksource {
unsigned long flags;
cycle_t (*vread)(void);
void (*resume)(void);
+#ifdef CONFIG_IA64
+ void *fsys_mmio; /* used by fsyscall asm code */
+#define CLKSRC_FSYS_MMIO_SET(mmio, addr) ((mmio) = (addr))
+#else
+#define CLKSRC_FSYS_MMIO_SET(mmio, addr) (0)
+#endif
/* timekeeping specific data, ignore */
cycle_t cycle_interval;
^ permalink raw reply related [flat|nested] 5+ messages in thread