2024-06-11 10:23:02

by Paolo Bonzini

Subject: [PATCH] KVM: interrupt kvm_gmem_populate() on signals

kvm_gmem_populate() is a potentially lengthy operation that can involve
multiple calls to the firmware. Interrupt it if a signal arrives.

Fixes: 1f6c06b177513 ("KVM: guest_memfd: Add interface for populating gmem pages with user data")
Cc: Isaku Yamahata <[email protected]>
Cc: Michael Roth <[email protected]>
Signed-off-by: Paolo Bonzini <[email protected]>
---
virt/kvm/guest_memfd.c | 5 +++++
1 file changed, 5 insertions(+)

diff --git a/virt/kvm/guest_memfd.c b/virt/kvm/guest_memfd.c
index 9714add38852..3bfe1824ec2d 100644
--- a/virt/kvm/guest_memfd.c
+++ b/virt/kvm/guest_memfd.c
@@ -629,6 +629,11 @@ long kvm_gmem_populate(struct kvm *kvm, gfn_t start_gfn, void __user *src, long
gfn_t gfn = start_gfn + i;
kvm_pfn_t pfn;

+ if (signal_pending(current)) {
+ ret = -EINTR;
+ break;
+ }
+
ret = __kvm_gmem_get_pfn(file, slot, gfn, &pfn, &max_order, false);
if (ret)
break;
--
2.43.0

2024-06-12 23:13:45

by Isaku Yamahata

Subject: Re: [PATCH] KVM: interrupt kvm_gmem_populate() on signals

On Tue, Jun 11, 2024 at 06:22:43AM -0400,
Paolo Bonzini <[email protected]> wrote:

> kvm_gmem_populate() is a potentially lengthy operation that can involve
> multiple calls to the firmware. Interrupt it if a signal arrives.

What about cond_resched() in the loop? kvm_gmem_allocate() has both.

The change itself looks good for TDX because KVM_TDX_INIT_MEMREGION checks the
signal. I can drop the duplicated check. Similar to cond_resched().
--
Isaku Yamahata <[email protected]>