[PATCH v4] powerpc: kvm: fix rare but potential deadlock scene

[PATCH v4] powerpc: kvm: fix rare but potential deadlock scene

[Liu Ping Fan]

Since kvmppc_hv_find_lock_hpte() is called from both virtmode and
realmode, so it can trigger the deadlock.

Suppose the following scene:

Two physical cpuM, cpuN, two VM instances A, B, each VM has a group of
vcpus.

If on cpuM, vcpu_A_1 holds bitlock X (HPTE_V_HVLOCK), then is switched
out, and on cpuN, vcpu_A_2 try to lock X in realmode, then cpuN will be
caught in realmode for a long time.

What makes things even worse if the following happens,
  On cpuM, bitlockX is hold, on cpuN, Y is hold.
  vcpu_B_2 try to lock Y on cpuM in realmode
  vcpu_A_2 try to lock X on cpuN in realmode

Oops! deadlock happens

Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
---
v4: remove the over-engineered part and keep it simple, also add some notes.
  
---
 arch/powerpc/kvm/book3s_64_mmu_hv.c | 6 +++++-
 arch/powerpc/kvm/book3s_hv_rm_mmu.c | 4 ++++
 2 files changed, 9 insertions(+), 1 deletion(-)

diff --git a/arch/powerpc/kvm/book3s_64_mmu_hv.c b/arch/powerpc/kvm/book3s_64_mmu_hv.c
index 842f081..abf81fe 100644
--- a/arch/powerpc/kvm/book3s_64_mmu_hv.c
+++ b/arch/powerpc/kvm/book3s_64_mmu_hv.c
@@ -473,11 +473,14 @@ static int kvmppc_mmu_book3s_64_hv_xlate(struct kvm_vcpu *vcpu, gva_t eaddr,
 		slb_v = vcpu->kvm->arch.vrma_slb_v;
 	}
 
+	preempt_disable();
 	/* Find the HPTE in the hash table */
 	index = kvmppc_hv_find_lock_hpte(kvm, eaddr, slb_v,
 					 HPTE_V_VALID | HPTE_V_ABSENT);
-	if (index < 0)
+	if (index < 0) {
+		preempt_enable();
 		return -ENOENT;
+	}
 	hptep = (unsigned long *)(kvm->arch.hpt_virt + (index << 4));
 	v = hptep[0] & ~HPTE_V_HVLOCK;
 	gr = kvm->arch.revmap[index].guest_rpte;
@@ -485,6 +488,7 @@ static int kvmppc_mmu_book3s_64_hv_xlate(struct kvm_vcpu *vcpu, gva_t eaddr,
 	/* Unlock the HPTE */
 	asm volatile("lwsync" : : : "memory");
 	hptep[0] = v;
+	preempt_enable();
 
 	gpte->eaddr = eaddr;
 	gpte->vpage = ((v & HPTE_V_AVPN) << 4) | ((eaddr >> 12) & 0xfff);
diff --git a/arch/powerpc/kvm/book3s_hv_rm_mmu.c b/arch/powerpc/kvm/book3s_hv_rm_mmu.c
index 9c51544..ea17b30 100644
--- a/arch/powerpc/kvm/book3s_hv_rm_mmu.c
+++ b/arch/powerpc/kvm/book3s_hv_rm_mmu.c
@@ -749,6 +749,10 @@ static int slb_base_page_shift[4] = {
 	20,	/* 1M, unsupported */
 };
 
+/* When called from virtmode, this func should be protected by
+ * preempt_disable(), otherwise, the holding of HPTE_V_HVLOCK
+ * can trigger deadlock issue.
+ */
 long kvmppc_hv_find_lock_hpte(struct kvm *kvm, gva_t eaddr, unsigned long slb_v,
 			      unsigned long valid)
 {
-- 
1.8.1.4

[PATCH v2] powerpc: kvm: optimize "sc 1" as fast return

[Liu Ping Fan]

In some scene, e.g openstack CI, PR guest can trigger "sc 1" frequently,
this patch optimizes the path by directly delivering BOOK3S_INTERRUPT_SYSCALL
to HV guest, so powernv can return to HV guest without heavy exit, i.e,
no need to swap TLB, HTAB,.. etc

Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
---
 arch/powerpc/kvm/book3s_hv.c            |  6 ------
 arch/powerpc/kvm/book3s_hv_rmhandlers.S | 11 ++++++++++-
 2 files changed, 10 insertions(+), 7 deletions(-)

diff --git a/arch/powerpc/kvm/book3s_hv.c b/arch/powerpc/kvm/book3s_hv.c
index 62a2b5a..73dc852 100644
--- a/arch/powerpc/kvm/book3s_hv.c
+++ b/arch/powerpc/kvm/book3s_hv.c
@@ -628,12 +628,6 @@ static int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu,
 		/* hcall - punt to userspace */
 		int i;
 
-		if (vcpu->arch.shregs.msr & MSR_PR) {
-			/* sc 1 from userspace - reflect to guest syscall */
-			kvmppc_book3s_queue_irqprio(vcpu, BOOK3S_INTERRUPT_SYSCALL);
-			r = RESUME_GUEST;
-			break;
-		}
 		run->papr_hcall.nr = kvmppc_get_gpr(vcpu, 3);
 		for (i = 0; i < 9; ++i)
 			run->papr_hcall.args[i] = kvmppc_get_gpr(vcpu, 4 + i);
diff --git a/arch/powerpc/kvm/book3s_hv_rmhandlers.S b/arch/powerpc/kvm/book3s_hv_rmhandlers.S
index c71103b..a463f08 100644
--- a/arch/powerpc/kvm/book3s_hv_rmhandlers.S
+++ b/arch/powerpc/kvm/book3s_hv_rmhandlers.S
@@ -1388,7 +1388,8 @@ kvmppc_hisi:
 hcall_try_real_mode:
 	ld	r3,VCPU_GPR(R3)(r9)
 	andi.	r0,r11,MSR_PR
-	bne	guest_exit_cont
+	/* sc 1 from userspace - reflect to guest syscall */
+	bne	sc_1_fast_return
 	clrrdi	r3,r3,2
 	cmpldi	r3,hcall_real_table_end - hcall_real_table
 	bge	guest_exit_cont
@@ -1409,6 +1410,14 @@ hcall_try_real_mode:
 	ld	r11,VCPU_MSR(r4)
 	b	fast_guest_return
 
+sc_1_fast_return:
+	mtspr	SPRN_SRR0,r10
+	mtspr	SPRN_SRR1,r11
+	li	r10, BOOK3S_INTERRUPT_SYSCALL
+	li      r11, (MSR_ME << 1) | 1  /* synthesize MSR_SF | MSR_ME */
+	rotldi  r11, r11, 63
+	b	fast_guest_return
+
 	/* We've attempted a real mode hcall, but it's punted it back
 	 * to userspace.  We need to restore some clobbered volatiles
 	 * before resuming the pass-it-to-qemu path */
-- 
1.8.1.4

Re: [PATCH v2] powerpc: kvm: optimize "sc 1" as fast return

[Paul Mackerras]

On Fri, Nov 15, 2013 at 04:35:01PM +0800, Liu Ping Fan wrote:
>  
> +sc_1_fast_return:
> +	mtspr	SPRN_SRR0,r10
> +	mtspr	SPRN_SRR1,r11
> +	li	r10, BOOK3S_INTERRUPT_SYSCALL
> +	li      r11, (MSR_ME << 1) | 1  /* synthesize MSR_SF | MSR_ME */
> +	rotldi  r11, r11, 63

You need a "mr r4, r9" instruction here, because fast_guest_return
needs the vcpu pointer in r4.  Apart from that this looks fine.

> +	b	fast_guest_return

Paul.

Re: [PATCH v2] powerpc: kvm: optimize "sc 1" as fast return

[liu ping fan]

On Sat, Nov 16, 2013 at 3:00 PM, Paul Mackerras <paulus@samba.org> wrote:
> On Fri, Nov 15, 2013 at 04:35:01PM +0800, Liu Ping Fan wrote:
>>
>> +sc_1_fast_return:
>> +     mtspr   SPRN_SRR0,r10
>> +     mtspr   SPRN_SRR1,r11
>> +     li      r10, BOOK3S_INTERRUPT_SYSCALL
>> +     li      r11, (MSR_ME << 1) | 1  /* synthesize MSR_SF | MSR_ME */
>> +     rotldi  r11, r11, 63
>
> You need a "mr r4, r9" instruction here, because fast_guest_return
> needs the vcpu pointer in r4.  Apart from that this looks fine.
>
Will fix it.

Thanks and regards,
Pingfan
>> +     b       fast_guest_return
>
> Paul.

Re: [PATCH v4] powerpc: kvm: fix rare but potential deadlock scene

[Paul Mackerras]

On Fri, Nov 15, 2013 at 04:35:00PM +0800, Liu Ping Fan wrote:
> Since kvmppc_hv_find_lock_hpte() is called from both virtmode and
> realmode, so it can trigger the deadlock.
> 
> Suppose the following scene:
> 
> Two physical cpuM, cpuN, two VM instances A, B, each VM has a group of
> vcpus.
> 
> If on cpuM, vcpu_A_1 holds bitlock X (HPTE_V_HVLOCK), then is switched
> out, and on cpuN, vcpu_A_2 try to lock X in realmode, then cpuN will be
> caught in realmode for a long time.
> 
> What makes things even worse if the following happens,
>   On cpuM, bitlockX is hold, on cpuN, Y is hold.
>   vcpu_B_2 try to lock Y on cpuM in realmode
>   vcpu_A_2 try to lock X on cpuN in realmode
> 
> Oops! deadlock happens
> 
> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>

Reviewed-by: Paul Mackerras <paulus@samba.org>

Re: [PATCH v4] powerpc: kvm: fix rare but potential deadlock scene

[Alexander Graf]

On 16.11.2013, at 01:55, Paul Mackerras <paulus@samba.org> wrote:

> On Fri, Nov 15, 2013 at 04:35:00PM +0800, Liu Ping Fan wrote:
>> Since kvmppc_hv_find_lock_hpte() is called from both virtmode and
>> realmode, so it can trigger the deadlock.
>> 
>> Suppose the following scene:
>> 
>> Two physical cpuM, cpuN, two VM instances A, B, each VM has a group of
>> vcpus.
>> 
>> If on cpuM, vcpu_A_1 holds bitlock X (HPTE_V_HVLOCK), then is switched
>> out, and on cpuN, vcpu_A_2 try to lock X in realmode, then cpuN will be
>> caught in realmode for a long time.
>> 
>> What makes things even worse if the following happens,
>>  On cpuM, bitlockX is hold, on cpuN, Y is hold.
>>  vcpu_B_2 try to lock Y on cpuM in realmode
>>  vcpu_A_2 try to lock X on cpuN in realmode
>> 
>> Oops! deadlock happens
>> 
>> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
> 
> Reviewed-by: Paul Mackerras <paulus@samba.org>

Thanks, applied to kvm-ppc-queue.


Alex

Re: [PATCH v4] powerpc: kvm: fix rare but potential deadlock scene

[Alexander Graf]

On 18.11.2013, at 16:32, Alexander Graf <agraf@suse.de> wrote:

>=20
> On 16.11.2013, at 01:55, Paul Mackerras <paulus@samba.org> wrote:
>=20
>> On Fri, Nov 15, 2013 at 04:35:00PM +0800, Liu Ping Fan wrote:
>>> Since kvmppc_hv_find_lock_hpte() is called from both virtmode and
>>> realmode, so it can trigger the deadlock.
>>>=20
>>> Suppose the following scene:
>>>=20
>>> Two physical cpuM, cpuN, two VM instances A, B, each VM has a group =
of
>>> vcpus.
>>>=20
>>> If on cpuM, vcpu_A_1 holds bitlock X (HPTE_V_HVLOCK), then is =
switched
>>> out, and on cpuN, vcpu_A_2 try to lock X in realmode, then cpuN will =
be
>>> caught in realmode for a long time.
>>>=20
>>> What makes things even worse if the following happens,
>>> On cpuM, bitlockX is hold, on cpuN, Y is hold.
>>> vcpu_B_2 try to lock Y on cpuM in realmode
>>> vcpu_A_2 try to lock X on cpuN in realmode
>>>=20
>>> Oops! deadlock happens
>>>=20
>>> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
>>=20
>> Reviewed-by: Paul Mackerras <paulus@samba.org>
>=20
> Thanks, applied to kvm-ppc-queue.

Actually, I've changed my mind and moved the patch to the for-3.13 =
branch instead. Please make sure to CC kvm@vger on all patches you =
submit though.


Alex