diff options
| author | Martin Schwidefsky <schwidefsky@de.ibm.com> | 2013-10-17 12:38:17 +0200 |
|---|---|---|
| committer | Martin Schwidefsky <schwidefsky@de.ibm.com> | 2013-10-22 09:16:40 +0200 |
| commit | 8c071b0f19dfa230335d22ce56a8fab5bd20cedc (patch) | |
| tree | ffd0d6272cd4e3f33b3069b22cb52457030a7f7a /arch/s390/kvm | |
| parent | 9784bd4f1a6ea736ad9bf241f5a965e0a2913a5e (diff) | |
| download | linux-8c071b0f19dfa230335d22ce56a8fab5bd20cedc.tar.bz2 | |
s390/time: correct use of store clock fast
The result of the store-clock-fast (STCKF) instruction is a bit fuzzy.
It can happen that the value stored on one CPU is smaller than the value
stored on another CPU, although the order of the stores is the other
way around. This can cause deltas of get_tod_clock() values to become
negative when they should not be.
We need to be more careful with store-clock-fast, this patch partially
reverts git commit e4b7b4238e666682555461fa52eecd74652f36bb "time:
always use stckf instead of stck if available". The get_tod_clock()
function now uses the store-clock-extended (STCKE) instruction.
get_tod_clock_fast() can be used if the fuzziness of store-clock-fast
is acceptable e.g. for wait loops local to a CPU.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Diffstat (limited to 'arch/s390/kvm')
| -rw-r--r-- | arch/s390/kvm/interrupt.c | 6 |
1 files changed, 3 insertions, 3 deletions
diff --git a/arch/s390/kvm/interrupt.c b/arch/s390/kvm/interrupt.c index 7f35cb33e510..7f1f7ac5cf7f 100644 --- a/arch/s390/kvm/interrupt.c +++ b/arch/s390/kvm/interrupt.c @@ -385,7 +385,7 @@ static int kvm_cpu_has_interrupt(struct kvm_vcpu *vcpu) } if ((!rc) && (vcpu->arch.sie_block->ckc < - get_tod_clock() + vcpu->arch.sie_block->epoch)) { + get_tod_clock_fast() + vcpu->arch.sie_block->epoch)) { if ((!psw_extint_disabled(vcpu)) && (vcpu->arch.sie_block->gcr[0] & 0x800ul)) rc = 1; @@ -425,7 +425,7 @@ int kvm_s390_handle_wait(struct kvm_vcpu *vcpu) goto no_timer; } - now = get_tod_clock() + vcpu->arch.sie_block->epoch; + now = get_tod_clock_fast() + vcpu->arch.sie_block->epoch; if (vcpu->arch.sie_block->ckc < now) { __unset_cpu_idle(vcpu); return 0; @@ -515,7 +515,7 @@ void kvm_s390_deliver_pending_interrupts(struct kvm_vcpu *vcpu) } if ((vcpu->arch.sie_block->ckc < - get_tod_clock() + vcpu->arch.sie_block->epoch)) + get_tod_clock_fast() + vcpu->arch.sie_block->epoch)) __try_deliver_ckc_interrupt(vcpu); if (atomic_read(&fi->active)) { |