Commit 16761939 authored by Venkatesh Pallipadi's avatar Venkatesh Pallipadi Committed by Andi Kleen

[PATCH] x86-64: Handle 32 bit PerfMon Counter writes cleanly in x86_64 nmi_watchdog

P6 CPUs and Core/Core 2 CPUs which has 'architectural perf mon' feature,
only supports write of low 32 bits in Performance Monitoring Counters.
Bits 32..39 are sign extended based on bit 31 and bits 40..63 are reserved
and should be zero.

This patch:

Change x86_64 nmi handler to handle this case cleanly.
Signed-off-by: default avatarVenkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: default avatarAndi Kleen <ak@suse.de>
parent 4c3cbf75
...@@ -214,6 +214,23 @@ static __init void nmi_cpu_busy(void *data) ...@@ -214,6 +214,23 @@ static __init void nmi_cpu_busy(void *data)
} }
#endif #endif
static unsigned int adjust_for_32bit_ctr(unsigned int hz)
{
unsigned int retval = hz;
/*
* On Intel CPUs with ARCH_PERFMON only 32 bits in the counter
* are writable, with higher bits sign extending from bit 31.
* So, we can only program the counter with 31 bit values and
* 32nd bit should be 1, for 33.. to be 1.
* Find the appropriate nmi_hz
*/
if ((((u64)cpu_khz * 1000) / retval) > 0x7fffffffULL) {
retval = ((u64)cpu_khz * 1000) / 0x7fffffffUL + 1;
}
return retval;
}
int __init check_nmi_watchdog (void) int __init check_nmi_watchdog (void)
{ {
int *counts; int *counts;
...@@ -268,17 +285,8 @@ int __init check_nmi_watchdog (void) ...@@ -268,17 +285,8 @@ int __init check_nmi_watchdog (void)
struct nmi_watchdog_ctlblk *wd = &__get_cpu_var(nmi_watchdog_ctlblk); struct nmi_watchdog_ctlblk *wd = &__get_cpu_var(nmi_watchdog_ctlblk);
nmi_hz = 1; nmi_hz = 1;
/* if (wd->perfctr_msr == MSR_ARCH_PERFMON_PERFCTR0)
* On Intel CPUs with ARCH_PERFMON only 32 bits in the counter nmi_hz = adjust_for_32bit_ctr(nmi_hz);
* are writable, with higher bits sign extending from bit 31.
* So, we can only program the counter with 31 bit values and
* 32nd bit should be 1, for 33.. to be 1.
* Find the appropriate nmi_hz
*/
if (wd->perfctr_msr == MSR_ARCH_PERFMON_PERFCTR0 &&
((u64)cpu_khz * 1000) > 0x7fffffffULL) {
nmi_hz = ((u64)cpu_khz * 1000) / 0x7fffffffUL + 1;
}
} }
kfree(counts); kfree(counts);
...@@ -634,7 +642,9 @@ static int setup_intel_arch_watchdog(void) ...@@ -634,7 +642,9 @@ static int setup_intel_arch_watchdog(void)
/* setup the timer */ /* setup the timer */
wrmsr(evntsel_msr, evntsel, 0); wrmsr(evntsel_msr, evntsel, 0);
wrmsrl(perfctr_msr, -((u64)cpu_khz * 1000 / nmi_hz));
nmi_hz = adjust_for_32bit_ctr(nmi_hz);
wrmsr(perfctr_msr, (u32)(-((u64)cpu_khz * 1000 / nmi_hz)), 0);
apic_write(APIC_LVTPC, APIC_DM_NMI); apic_write(APIC_LVTPC, APIC_DM_NMI);
evntsel |= ARCH_PERFMON_EVENTSEL0_ENABLE; evntsel |= ARCH_PERFMON_EVENTSEL0_ENABLE;
...@@ -855,15 +865,23 @@ int __kprobes nmi_watchdog_tick(struct pt_regs * regs, unsigned reason) ...@@ -855,15 +865,23 @@ int __kprobes nmi_watchdog_tick(struct pt_regs * regs, unsigned reason)
dummy &= ~P4_CCCR_OVF; dummy &= ~P4_CCCR_OVF;
wrmsrl(wd->cccr_msr, dummy); wrmsrl(wd->cccr_msr, dummy);
apic_write(APIC_LVTPC, APIC_DM_NMI); apic_write(APIC_LVTPC, APIC_DM_NMI);
/* start the cycle over again */
wrmsrl(wd->perfctr_msr,
-((u64)cpu_khz * 1000 / nmi_hz));
} else if (wd->perfctr_msr == MSR_ARCH_PERFMON_PERFCTR0) { } else if (wd->perfctr_msr == MSR_ARCH_PERFMON_PERFCTR0) {
/* /*
* ArchPerfom/Core Duo needs to re-unmask * ArchPerfom/Core Duo needs to re-unmask
* the apic vector * the apic vector
*/ */
apic_write(APIC_LVTPC, APIC_DM_NMI); apic_write(APIC_LVTPC, APIC_DM_NMI);
/* ARCH_PERFMON has 32 bit counter writes */
wrmsr(wd->perfctr_msr,
(u32)(-((u64)cpu_khz * 1000 / nmi_hz)), 0);
} else {
/* start the cycle over again */
wrmsrl(wd->perfctr_msr,
-((u64)cpu_khz * 1000 / nmi_hz));
} }
/* start the cycle over again */
wrmsrl(wd->perfctr_msr, -((u64)cpu_khz * 1000 / nmi_hz));
rc = 1; rc = 1;
} else if (nmi_watchdog == NMI_IO_APIC) { } else if (nmi_watchdog == NMI_IO_APIC) {
/* don't know how to accurately check for this. /* don't know how to accurately check for this.
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment