From: Sean Christopherson <seanjc@google.com>
To: isaku.yamahata@intel.com
Cc: kvm@vger.kernel.org, isaku.yamahata@gmail.com,
linux-kernel@vger.kernel.org,
Paolo Bonzini <pbonzini@redhat.com>,
Michael Roth <michael.roth@amd.com>,
David Matlack <dmatlack@google.com>,
Federico Parola <federico.parola@polito.it>
Subject: Re: [RFC PATCH 2/8] KVM: Add KVM_MAP_MEMORY vcpu ioctl to pre-populate guest memory
Date: Mon, 11 Mar 2024 10:23:28 -0700 [thread overview]
Message-ID: <Ze8-EFtsIONMyO3o@google.com> (raw)
In-Reply-To: <012b59708114ba121735769de94756fa5af3204d.1709288671.git.isaku.yamahata@intel.com>
On Fri, Mar 01, 2024, isaku.yamahata@intel.com wrote:
> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index d1fd9cb5d037..d77c9b79d76b 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -4419,6 +4419,69 @@ static int kvm_vcpu_ioctl_get_stats_fd(struct kvm_vcpu *vcpu)
> return fd;
> }
>
> +__weak int kvm_arch_vcpu_pre_map_memory(struct kvm_vcpu *vcpu)
> +{
> + return -EOPNOTSUPP;
> +}
> +
> +__weak int kvm_arch_vcpu_map_memory(struct kvm_vcpu *vcpu,
> + struct kvm_memory_mapping *mapping)
> +{
> + return -EOPNOTSUPP;
> +}
> +
> +static int kvm_vcpu_map_memory(struct kvm_vcpu *vcpu,
> + struct kvm_memory_mapping *mapping)
> +{
> + bool added = false;
> + int idx, r = 0;
Pointless initialization of 'r'.
> +
> + if (mapping->flags & ~(KVM_MEMORY_MAPPING_FLAG_WRITE |
> + KVM_MEMORY_MAPPING_FLAG_EXEC |
> + KVM_MEMORY_MAPPING_FLAG_USER |
> + KVM_MEMORY_MAPPING_FLAG_PRIVATE))
> + return -EINVAL;
> + if ((mapping->flags & KVM_MEMORY_MAPPING_FLAG_PRIVATE) &&
> + !kvm_arch_has_private_mem(vcpu->kvm))
> + return -EINVAL;
> +
> + /* Sanity check */
Pointless comment.
> + if (!IS_ALIGNED(mapping->source, PAGE_SIZE) ||
> + !mapping->nr_pages ||
> + mapping->base_gfn + mapping->nr_pages <= mapping->base_gfn)
> + return -EINVAL;
> +
> + vcpu_load(vcpu);
> + idx = srcu_read_lock(&vcpu->kvm->srcu);
> + r = kvm_arch_vcpu_pre_map_memory(vcpu);
This hooks is unnecessary, x86's kvm_mmu_reload() is optimized for the happy path
where the MMU is already loaded. Just make the call from kvm_arch_vcpu_map_memory().
> + if (r)
> + return r;
Which is a good thing, because this leaks the SRCU lock.
> +
> + while (mapping->nr_pages) {
> + if (signal_pending(current)) {
> + r = -ERESTARTSYS;
Why -ERESTARTSYS instead of -EINTR? The latter is KVM's typical response to a
pending signal.
> + break;
> + }
> +
> + if (need_resched())
No need to manually check need_resched(), the below is a _conditional_ resched.
The reason KVM explicitly checks need_resched() in MMU flows is because KVM needs
to drop mmu_lock before rescheduling, i.e. calling cond_resched() directly would
try to schedule() while holding a spinlock.
> + cond_resched();
> +
> + r = kvm_arch_vcpu_map_memory(vcpu, mapping);
> + if (r)
> + break;
> +
> + added = true;
> + }
> +
> + srcu_read_unlock(&vcpu->kvm->srcu, idx);
> + vcpu_put(vcpu);
> +
> + if (added && mapping->nr_pages > 0)
> + r = -EAGAIN;
No, this clobbers 'r', which might hold a fatal error code. I don't see any
reason for common code to ever force -EAGAIN, it can't possibly know if trying
again is reasonable.
> +
> + return r;
> +}
next prev parent reply other threads:[~2024-03-11 17:23 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-03-01 17:28 [RFC PATCH 0/8] KVM: Prepopulate guest memory API isaku.yamahata
2024-03-01 17:28 ` [RFC PATCH 1/8] KVM: Document KVM_MAP_MEMORY ioctl isaku.yamahata
2024-03-07 0:43 ` David Matlack
2024-03-07 1:29 ` Isaku Yamahata
2024-03-07 12:30 ` Huang, Kai
2024-03-07 20:33 ` Isaku Yamahata
2024-03-08 0:20 ` Huang, Kai
2024-03-08 0:56 ` David Matlack
2024-03-08 1:28 ` Sean Christopherson
2024-03-08 2:19 ` Isaku Yamahata
2024-03-10 23:12 ` Michael Roth
2024-03-11 1:05 ` Huang, Kai
2024-03-11 1:08 ` Huang, Kai
2024-03-12 1:34 ` Isaku Yamahata
2024-03-01 17:28 ` [RFC PATCH 2/8] KVM: Add KVM_MAP_MEMORY vcpu ioctl to pre-populate guest memory isaku.yamahata
2024-03-07 0:49 ` David Matlack
2024-03-07 2:52 ` Isaku Yamahata
2024-03-07 12:45 ` Huang, Kai
2024-03-07 20:41 ` Isaku Yamahata
2024-03-11 17:23 ` Sean Christopherson [this message]
2024-03-11 22:19 ` Isaku Yamahata
2024-03-01 17:28 ` [RFC PATCH 3/8] KVM: x86/mmu: Introduce initialier macro for struct kvm_page_fault isaku.yamahata
2024-03-11 17:24 ` Sean Christopherson
2024-03-11 22:56 ` Isaku Yamahata
2024-03-01 17:28 ` [RFC PATCH 4/8] KVM: x86/mmu: Factor out kvm_mmu_do_page_fault() isaku.yamahata
2024-03-01 17:28 ` [RFC PATCH 5/8] KVM: x86/mmu: Introduce kvm_mmu_map_page() for prepopulating guest memory isaku.yamahata
2024-03-07 0:38 ` David Matlack
2024-03-19 15:53 ` Isaku Yamahata
2024-03-11 17:29 ` Sean Christopherson
2024-03-11 22:57 ` Isaku Yamahata
2024-03-01 17:28 ` [RFC PATCH 6/8] KVM: x86: Implement kvm_arch_{, pre_}vcpu_map_memory() isaku.yamahata
2024-03-07 0:30 ` David Matlack
2024-03-07 0:36 ` David Matlack
2024-03-07 1:51 ` Isaku Yamahata
2024-03-19 16:26 ` Isaku Yamahata
2024-04-03 23:15 ` Sean Christopherson
2024-03-07 1:34 ` Isaku Yamahata
2024-03-11 23:26 ` Sean Christopherson
2024-03-12 12:38 ` Huang, Kai
2024-03-12 14:20 ` Sean Christopherson
2024-03-12 21:41 ` Huang, Kai
2024-03-12 21:46 ` Huang, Kai
2024-03-12 23:03 ` Sean Christopherson
2024-03-01 17:28 ` [RFC PATCH 7/8] KVM: x86: Add hooks in kvm_arch_vcpu_map_memory() isaku.yamahata
2024-03-01 17:28 ` [RFC PATCH 8/8] KVM: selftests: x86: Add test for KVM_MAP_MEMORY isaku.yamahata
2024-03-07 0:53 ` [RFC PATCH 0/8] KVM: Prepopulate guest memory API David Matlack
2024-03-07 2:09 ` Isaku Yamahata
2024-03-19 16:33 ` Isaku Yamahata
2024-04-03 18:30 ` Sean Christopherson
2024-04-03 22:00 ` Isaku Yamahata
2024-04-03 22:42 ` Sean Christopherson
2024-03-11 3:20 ` Michael Roth
2024-03-11 23:44 ` Sean Christopherson
2024-03-12 1:32 ` Isaku Yamahata
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=Ze8-EFtsIONMyO3o@google.com \
--to=seanjc@google.com \
--cc=dmatlack@google.com \
--cc=federico.parola@polito.it \
--cc=isaku.yamahata@gmail.com \
--cc=isaku.yamahata@intel.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=michael.roth@amd.com \
--cc=pbonzini@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).