Received-SPF: pass (google.com: domain of linux-kernel+bounces-22991-linux.lists.archive=gmail.com@vger.kernel.org designates 139.178.88.99 as permitted sender) client-ip=139.178.88.99;
Date: Thu, 11 Jan 2024 10:11:31 +0800
From: Yan Zhao <yan.y.zhao@intel.com>
To: Sean Christopherson <seanjc@google.com>
CC: Paolo Bonzini <pbonzini@redhat.com>, <kvm@vger.kernel.org>,
	<linux-kernel@vger.kernel.org>, Kai Huang <kai.huang@intel.com>
Subject: Re: [PATCH v2] KVM: x86/mmu: Retry fault before acquiring mmu_lock
 if mapping is changing
Message-ID: <ZZ9OUxnkX2Uwn3x2@yzhao56-desk.sh.intel.com>
Reply-To: Yan Zhao <yan.y.zhao@intel.com>
References: <20240110012045.505046-1-seanjc@google.com>
Content-Type: text/plain; charset="us-ascii"
Content-Disposition: inline
In-Reply-To: <20240110012045.505046-1-seanjc@google.com>
Precedence: bulk
MIME-Version: 1.0
X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1
X-MS-Exchange-AntiSpam-MessageData-0: =?us-ascii?Q?FqpTdgtIoy+Zr67oqs5MQb74Q/MCqHOObFkm9wyT0S9acCCpTkkipiYPLmpI?=
 =?us-ascii?Q?2mLa8ewAawj05pBcsGuICWkB6um/Q++Tq3gSwqMxf23htLqS+5B/MvC0ebXN?=
 =?us-ascii?Q?Hrj+ba43S84bveBdHfYb5rUK2Y7DKTKIRLH82FF57t8PlS7a6vzlf38/EzOo?=
 =?us-ascii?Q?GuMYFsrV5QMpB6+VLIjC/9F73nC2pOgYmpJ4cVuqV0kq/MFQCrke5x0a+RGt?=
 =?us-ascii?Q?AITfZm+RLwEQ6fpNPOkEukDXtuE+ZgXZl+pif+KL54zLFWWTWVsggfhOFwHt?=
 =?us-ascii?Q?OXTIyySofJ5nPTdk9hLB8QGePgHRo2dvVkj40EVik8verg8U9T4bG0XMhi66?=
 =?us-ascii?Q?0MDeYkKAIIlvj81zmZn5x/fKqZQEQG/W7XfMomvbz8eKgoE17YzrBVVZ2IHT?=
 =?us-ascii?Q?7i6uIT/H5zzbmcAuJb9VX2dGSs0JDto3GtTvGxp45r8gXudtseJx0hlaQZmn?=
 =?us-ascii?Q?AreKDpPCF/fDHpg8Sc0ZTJdMz2RT6IjM/MyUAoPGrxW/viXxySsU+it7f1tV?=
 =?us-ascii?Q?1WO+esBEGtwrepIKc3a+Xx6PD3kBxmD7WbbI2CGSPdjYgMmVAm8cA0miRsRS?=
 =?us-ascii?Q?V6U0YS6CfOmquv36W2gli2YE81sigXJUu/0fIP3Fz5q/nhgnVhGc6mwCr+6O?=
 =?us-ascii?Q?A8qXOV9pdC1UOfj9wXOjYpHxWQp4XjBWWvTSvhnaDkvJXRWZgeDiwLAIC01C?=
 =?us-ascii?Q?CP0ndCDoDOPQHMUWVDp0zvjR3TkplV/FhJAeuSnZ6iHmWkoB/84m+y1ERtJK?=
 =?us-ascii?Q?bXUhP8tgQGZgUeCDiQfQ7MKijWuIL50pkLgpFFOaeplPbN6QaaEMt8Mrld5m?=
 =?us-ascii?Q?wHapyTsQD8IEfegzTOJ3j5AIkgjav3GSyj/VCF4FNy1/bVrou6QBslj0rYRd?=
 =?us-ascii?Q?aKT3mguboYA/F7JFGuSEfv2SxBULboPm1tKPTmDjgjokgbBwzsN9Sq6Y2fWO?=
 =?us-ascii?Q?mWrMWZbjRoo9pSAivqUndexYZQt45RRByBuMy54CiSAm0r4xxM8SZJGvjyoC?=
 =?us-ascii?Q?p/vz9+FRR4kMLB7WecCQ7L/wg6Eao59/mlgyq7gi1P/bRLIrQuDL83r/I5eG?=
 =?us-ascii?Q?eD6CY6tricFHyhr4EjrEhtL0oKReKw/lIJnPbp4BCjSip2wKnFqmAW1IfsTH?=
 =?us-ascii?Q?oiOyrjFeI2J1mry4Cb9FLrNbnENmDhoIlzV0bs1ipo2UoHPJVpXDIN0EX64m?=
 =?us-ascii?Q?y9Dv8kvdvMP+TDV2k/Z1zTwurmo3RNPlnYmr+SNZO4oTlehllFewI0nBEKG1?=
 =?us-ascii?Q?X2yOjlhqnd7mW0I104bPJdWH9Ix87kjJnUDQ5x3yPbbrTBcXriPyF0hHvC/n?=
 =?us-ascii?Q?Zk7jvPn7nKfFwn1Dta23T+Z/0dKYvwFKTimktH/+gB0YsDH2KaofsWcaTV0w?=
 =?us-ascii?Q?A5SQijMiEmjEOGhkoajq/Wy0LS0qZl9FBnWBUpLdLcCJLwUibLHfVhLLcYyt?=
 =?us-ascii?Q?h6x3UNptAI1rFZDx0I2XF0tG9CS4NmIKp2j9ogs+sg7e5UJlJU7zYma2ruwu?=
 =?us-ascii?Q?F+nWrxoupLCa2pvGdG6MzjK+K9BDCB2Tv2aN9FcseZ2wLV9Nwx1I10pQL5/B?=
 =?us-ascii?Q?ChUMXaB++ajhZD0P2IZVpJG8tB5QQj/Cs7sjR35P?=
X-MS-Exchange-CrossTenant-Network-Message-Id: 7c3eaee7-927c-4d3a-e6d7-08dc124eb887
X-MS-Exchange-CrossTenant-AuthSource: DS7PR11MB5966.namprd11.prod.outlook.com
X-MS-Exchange-CrossTenant-AuthAs: Internal
X-MS-Exchange-CrossTenant-OriginalArrivalTime: 11 Jan 2024 02:40:49.3643
 (UTC)
X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted
X-MS-Exchange-CrossTenant-Id: 46c98d88-e344-4ed4-8496-4ed7712e255d
X-MS-Exchange-CrossTenant-MailboxType: HOSTED
X-MS-Exchange-CrossTenant-UserPrincipalName: HmwIdib94IMjJbt4i0iPLzW/f1MaE+zd6f7d67y8kXOkBgynetc7vtWPnmw80vLhhO+QshEIQYoWz1z8jwEmXA==
X-MS-Exchange-Transport-CrossTenantHeadersStamped: SN7PR11MB7113
X-OriginatorOrg: intel.com

Tested on my environemnt with below improvements observed:
For a VM with 8 vcpus + 16G memory + OVMF

Avg cycles of kvm_zap_gfn_range() is reduced from 1100k to 800k;
Avg cycles of kvm_unmap_gfn_range() is reduced rom 470 to 180.

Reviewed-by: Yan Zhao <yan.y.zhao@intel.com>

On Tue, Jan 09, 2024 at 05:20:45PM -0800, Sean Christopherson wrote:
> Retry page faults without acquiring mmu_lock if the resolved gfn is covered
> by an active invalidation.  Contending for mmu_lock is especially
> problematic on preemptible kernels as the mmu_notifier invalidation task
> will yield mmu_lock (see rwlock_needbreak()), delay the in-progress
> invalidation, and ultimately increase the latency of resolving the page
> fault.  And in the worst case scenario, yielding will be accompanied by a
> remote TLB flush, e.g. if the invalidation covers a large range of memory
> and vCPUs are accessing addresses that were already zapped.
> 
> Alternatively, the yielding issue could be mitigated by teaching KVM's MMU
> iterators to perform more work before yielding, but that wouldn't solve
> the lock contention and would negatively affect scenarios where a vCPU is
> trying to fault in an address that is NOT covered by the in-progress
> invalidation.
> 
> Add a dedicated lockess version of the range-based retry check to avoid
> false positives on the sanity check on start+end WARN, and so that it's
> super obvious that checking for a racing invalidation without holding
> mmu_lock is unsafe (though obviously useful).
> 
> Wrap mmu_invalidate_in_progress in READ_ONCE() to ensure that pre-checking
> invalidation in a loop won't put KVM into an infinite loop, e.g. due to
> caching the in-progress flag and never seeing it go to '0'.
> 
> Force a load of mmu_invalidate_seq as well, even though it isn't strictly
> necessary to avoid an infinite loop, as doing so improves the probability
> that KVM will detect an invalidation that already completed before
> acquiring mmu_lock and bailing anyways.
> 
> Do the pre-check even for non-preemptible kernels, as waiting to detect
> the invalidation until mmu_lock is held guarantees the vCPU will observe
> the worst case latency in terms of handling the fault, and can generate
> even more mmu_lock contention.  E.g. the vCPU will acquire mmu_lock,
> detect retry, drop mmu_lock, re-enter the guest, retake the fault, and
> eventually re-acquire mmu_lock.  This behavior is also why there are no
> new starvation issues due to losing the fairness guarantees provided by
> rwlocks: if the vCPU needs to retry, it _must_ drop mmu_lock, i.e. waiting
> on mmu_lock doesn't guarantee forward progress in the face of _another_
> mmu_notifier invalidation event.
> 
> Note, adding READ_ONCE() isn't entirely free, e.g. on x86, the READ_ONCE()
> may generate a load into a register instead of doing a direct comparison
> (MOV+TEST+Jcc instead of CMP+Jcc), but practically speaking the added cost
> is a few bytes of code and maaaaybe a cycle or three.
> 
> Reported-by: Yan Zhao <yan.y.zhao@intel.com>
> Closes: https://lore.kernel.org/all/ZNnPF4W26ZbAyGto@yzhao56-desk.sh.intel.com
> Acked-by: Kai Huang <kai.huang@intel.com>
> Signed-off-by: Sean Christopherson <seanjc@google.com>
> ---
> 
> Note, this version adds a dedicated helper, mmu_invalidate_retry_gfn_unsafe(),
> instead of making mmu_invalidate_retry_gfn() play nice with being called without
> mmu_lock held.  I was hesitant to drop the lockdep assertion before, and the
> recently introduced sanity check on the gfn start/end values pushed this past
> the threshold of being worth the duplicate code (preserving the start/end sanity
> check in lock-free code would comically difficult, and would add almost no value
> since it would have to be quite conservative to avoid false positives).
> 
> Kai, I kept your Ack even though the code is obviously a little different.
> Holler if you want me to drop it.
> 
> v2:
>  - Introduce a dedicated helper and collapse to a single patch (because
>    adding an unused helper would be quite silly).
>  - Add a comment to explain the "unsafe" check in kvm_faultin_pfn(). [Kai]
>  - Add Kai's Ack.
> 
> v1: https://lore.kernel.org/all/20230825020733.2849862-1-seanjc@google.com
> 
>  arch/x86/kvm/mmu/mmu.c   | 16 ++++++++++++++++
>  include/linux/kvm_host.h | 26 ++++++++++++++++++++++++++
>  2 files changed, 42 insertions(+)
> 
> diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
> index 3c844e428684..92f51540c4a7 100644
> --- a/arch/x86/kvm/mmu/mmu.c
> +++ b/arch/x86/kvm/mmu/mmu.c
> @@ -4415,6 +4415,22 @@ static int kvm_faultin_pfn(struct kvm_vcpu *vcpu, struct kvm_page_fault *fault,
>  	if (unlikely(!fault->slot))
>  		return kvm_handle_noslot_fault(vcpu, fault, access);
>  
> +	/*
> +	 * Pre-check for a relevant mmu_notifier invalidation event prior to
> +	 * acquiring mmu_lock.  If there is an in-progress invalidation and the
> +	 * kernel allows preemption, the invalidation task may drop mmu_lock
> +	 * and yield in response to mmu_lock being contended, which is *very*
> +	 * counter-productive as this vCPU can't actually make forward progress
> +	 * until the invalidation completes.  This "unsafe" check can get false
> +	 * negatives, i.e. KVM needs to re-check after acquiring mmu_lock.  Do
> +	 * the pre-check even for non-preemtible kernels, i.e. even if KVM will
> +	 * never yield mmu_lock in response to contention, as this vCPU ob
> +	 * *guaranteed* to need to retry, i.e. waiting until mmu_lock is held
> +	 * to detect retry guarantees the worst case latency for the vCPU.
> +	 */
> +	if (mmu_invalidate_retry_gfn_unsafe(vcpu->kvm, fault->mmu_seq, fault->gfn))
> +		return RET_PF_RETRY;
> +
>  	return RET_PF_CONTINUE;
>  }
>  
> diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
> index 7e7fd25b09b3..179df96b20f8 100644
> --- a/include/linux/kvm_host.h
> +++ b/include/linux/kvm_host.h
> @@ -2031,6 +2031,32 @@ static inline int mmu_invalidate_retry_gfn(struct kvm *kvm,
>  		return 1;
>  	return 0;
>  }
> +
> +/*
> + * This lockless version of the range-based retry check *must* be paired with a
> + * call to the locked version after acquiring mmu_lock, i.e. this is safe to
> + * use only as a pre-check to avoid contending mmu_lock.  This version *will*
> + * get false negatives and false positives.
> + */
> +static inline bool mmu_invalidate_retry_gfn_unsafe(struct kvm *kvm,
> +						   unsigned long mmu_seq,
> +						   gfn_t gfn)
> +{
> +	/*
> +	 * Use READ_ONCE() to ensure the in-progress flag and sequence counter
> +	 * are always read from memory, e.g. so that checking for retry in a
> +	 * loop won't result in an infinite retry loop.  Don't force loads for
> +	 * start+end, as the key to avoiding infinite retry loops is observing
> +	 * the 1=>0 transition of in-progress, i.e. getting false negatives
> +	 * due to stale start+end values is acceptable.
> +	 */
> +	if (unlikely(READ_ONCE(kvm->mmu_invalidate_in_progress)) &&
> +	    gfn >= kvm->mmu_invalidate_range_start &&
> +	    gfn < kvm->mmu_invalidate_range_end)
> +		return true;
> +
> +	return READ_ONCE(kvm->mmu_invalidate_seq) != mmu_seq;
> +}
>  #endif
>  
>  #ifdef CONFIG_HAVE_KVM_IRQ_ROUTING
> 
> base-commit: 1c6d984f523f67ecfad1083bb04c55d91977bb15
> -- 
> 2.43.0.472.g3155946c3a-goog
>