Skip to contentSkip to conversation

KVM: s390: vsie: retry SIE when unable to get vsie_page

PATCH In review Eric Farman 4 participants 10 messages Dec 17, 2025 – Jan 20

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 KVM: s390: vsie: retry SIE when unable to get vsie_page

SIE may exit because of pending host work, such as handling an interrupt, in which case VSIE rewinds the guest PSW such that it is transparently resumed (see Fixes tag). There is still one scenario where those conditions are not present, but that the VSIE processor returns with effectively rc=0, resulting in additional (and unnecessary) guest work to be performed.

For this case, rewind the guest PSW as we do in the other non-error exits.

Fixes 33a729a1770b ("KVM: s390: vsie: retry SIE instruction on host intercepts")
---
 arch/s390/kvm/vsie.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
arch/s390/kvm/vsie.c +4 -2
index b526621d2a1b..b8064c6a4b57 100644
@@ -1498,11 +1498,13 @@ int kvm_s390_handle_vsie(struct kvm_vcpu *vcpu)
 	}
 
 	vsie_page = get_vsie_page(vcpu->kvm, scb_addr);
-	if (IS_ERR(vsie_page))
+	if (IS_ERR(vsie_page)) {
 		return PTR_ERR(vsie_page);
-	else if (!vsie_page)
+	} else if (!vsie_page) {
 		/* double use of sie control block - simply do nothing */
+		kvm_s390_rewind_psw(vcpu, 4);
 		return 0;
+	}
 
 	rc = pin_scb(vcpu, vsie_page, scb_addr);
 	if (rc)
Signature · 1 line
2.51.0
maintainer
#1 lore raw
Quoted 10 lines · Eric Farman

SIE may exit because of pending host work, such as handling an interrupt, in which case VSIE rewinds the guest PSW such that it is transparently resumed (see Fixes tag). There is still one scenario where those conditions are not present, but that the VSIE processor returns with effectively rc=0, resulting in additional (and unnecessary) guest work to be performed.

For this case, rewind the guest PSW as we do in the other non-error exits.

Fixes 33a729a1770b ("KVM: s390: vsie: retry SIE instruction on host intercepts")

This is purely cosmetic to have all instances look the same, right?

#2 lore raw
Quoted 7 lines · Janosch Frank
Quoted 6 lines · Eric Farman

SIE may exit because of pending host work, such as handling an interrupt, in which case VSIE rewinds the guest PSW such that it is transparently resumed (see Fixes tag). There is still one scenario where those conditions are not present, but that the VSIE processor returns with effectively rc=0, resulting in additional (and unnecessary) guest work to be performed.

Quoted 6 lines
Quoted 4 lines

For this case, rewind the guest PSW as we do in the other non-error exits.

Fixes 33a729a1770b ("KVM: s390: vsie: retry SIE instruction on host intercepts")

This is purely cosmetic to have all instances look the same, right?

Nope, I can take this path with particularly high I/O loads on the system, which ends up (incorrectly) sending the intercept to the guest.

maintainer
#3 lore raw
Quoted 5 lines · On Mon, 05 Jan 2026 10:46:53 -0500 Eric Farman <[email protected]> wrote:
Quoted 4 lines · Janosch Frank
Quoted 3 lines · Eric Farman

SIE may exit because of pending host work, such as handling an interrupt, in which case VSIE rewinds the guest PSW such that it is transparently resumed (see Fixes tag). There is still one scenario where those conditions

can you add a few words to (very briefly) explain what the scenario is?

Quoted 5 lines
Quoted 5 lines
Quoted 5 lines

are not present, but that the VSIE processor returns with effectively rc=0, resulting in additional (and unnecessary) guest work to be performed.

For this case, rewind the guest PSW as we do in the other non-error exits.

Quoted 7 lines
Quoted 4 lines
Quoted 2 lines
Fixes 33a729a1770b ("KVM: s390: vsie: retry SIE instruction on host intercepts")

This is purely cosmetic to have all instances look the same, right?

Nope, I can take this path with particularly high I/O loads on the system, which ends up (incorrectly) sending the intercept to the guest.

this is a good candidate for the explanation I mentioned above :)

(the patch itself looks fine)

#4 lore raw
Quoted 10 lines · Claudio Imbrenda
Quoted 5 lines · On Mon, 05 Jan 2026 10:46:53 -0500 Eric Farman <[email protected]> wrote:
Quoted 4 lines · Janosch Frank
Quoted 3 lines · Eric Farman

SIE may exit because of pending host work, such as handling an interrupt, in which case VSIE rewinds the guest PSW such that it is transparently resumed (see Fixes tag). There is still one scenario where those conditions

can you add a few words to (very briefly) explain what the scenario is?

Maybe if this paragraph were rewritten this way, instead?

--8<--
SIE may exit because of pending host work, such as handling an interrupt, in which case VSIE rewinds the guest PSW such that it is transparently resumed (see Fixes tag). Unlike those other places that return rc=0, this return leaves the guest PSW in place, requiring the guest to handle an intercept that was meant to be serviced by the host. This showed up when testing heavy I/O workloads, when multiple vcpus attempted to dispatch the same SIE block and incurred failures inserting them into the radix tree. -->8--

@Janosch, if that ends up being okay, can you update the patch or do you want me to send a v2?

Quoted 18 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines

are not present, but that the VSIE processor returns with effectively rc=0, resulting in additional (and unnecessary) guest work to be performed.

For this case, rewind the guest PSW as we do in the other non-error exits.

Quoted 7 lines
Quoted 4 lines
Quoted 2 lines
Fixes 33a729a1770b ("KVM: s390: vsie: retry SIE instruction on host intercepts")

This is purely cosmetic to have all instances look the same, right?

Nope, I can take this path with particularly high I/O loads on the system, which ends up (incorrectly) sending the intercept to the guest.

this is a good candidate for the explanation I mentioned above :)

(the patch itself looks fine)

#5 lore raw
Quoted 14 lines · Eric Farman
Quoted 10 lines · Claudio Imbrenda
Quoted 5 lines · On Mon, 05 Jan 2026 10:46:53 -0500 Eric Farman <[email protected]> wrote:
Quoted 4 lines · Janosch Frank
Quoted 3 lines · Eric Farman

SIE may exit because of pending host work, such as handling an interrupt, in which case VSIE rewinds the guest PSW such that it is transparently resumed (see Fixes tag). There is still one scenario where those conditions

can you add a few words to (very briefly) explain what the scenario is?

Maybe if this paragraph were rewritten this way, instead?

Quoted 9 lines

--8<--
SIE may exit because of pending host work, such as handling an interrupt, in which case VSIE rewinds the guest PSW such that it is transparently resumed (see Fixes tag). Unlike those other places that return rc=0, this return leaves the guest PSW in place, requiring the guest to handle an intercept that was meant to be serviced by the host. This showed up when testing heavy I/O workloads, when multiple vcpus attempted to dispatch the same SIE block and incurred failures inserting them into the radix tree. -->8--

Spoke to Claudio offline, and he suggested the following edit to the above:

--8<--
SIE may exit because of pending host work, such as handling an interrupt, in which case VSIE rewinds the guest PSW such that it is transparently resumed (see Fixes tag). Unlike those other places that return rc=0, this return leaves the guest PSW in place, requiring the guest to handle a spurious intercept. This showed up when testing heavy I/O workloads, when multiple vcpus attempted to dispatch the same SIE block and incurred failures inserting them into the radix tree.
-->8--

Quoted 21 lines

@Janosch, if that ends up being okay, can you update the patch or do you want me to send a v2?

Quoted 18 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines

are not present, but that the VSIE processor returns with effectively rc=0, resulting in additional (and unnecessary) guest work to be performed.

For this case, rewind the guest PSW as we do in the other non-error exits.

Quoted 7 lines
Quoted 4 lines
Quoted 2 lines
Fixes 33a729a1770b ("KVM: s390: vsie: retry SIE instruction on host intercepts")

This is purely cosmetic to have all instances look the same, right?

Nope, I can take this path with particularly high I/O loads on the system, which ends up (incorrectly) sending the intercept to the guest.

this is a good candidate for the explanation I mentioned above :)

(the patch itself looks fine)

maintainer Reviewed-by
#6 lore raw
Quoted 27 lines · On Fri, 16 Jan 2026 10:45:22 -0500 Eric Farman <[email protected]> wrote:
Quoted 14 lines · Eric Farman
Quoted 10 lines · Claudio Imbrenda
Quoted 5 lines · On Mon, 05 Jan 2026 10:46:53 -0500 Eric Farman <[email protected]> wrote:
Quoted 4 lines · Janosch Frank
Quoted 3 lines · Eric Farman

SIE may exit because of pending host work, such as handling an interrupt, in which case VSIE rewinds the guest PSW such that it is transparently resumed (see Fixes tag). There is still one scenario where those conditions

can you add a few words to (very briefly) explain what the scenario is?

Maybe if this paragraph were rewritten this way, instead?

Quoted 9 lines

--8<--
SIE may exit because of pending host work, such as handling an interrupt, in which case VSIE rewinds the guest PSW such that it is transparently resumed (see Fixes tag). Unlike those other places that return rc=0, this return leaves the guest PSW in place, requiring the guest to handle an intercept that was meant to be serviced by the host. This showed up when testing heavy I/O workloads, when multiple vcpus attempted to dispatch the same SIE block and incurred failures inserting them into the radix tree. -->8--

Spoke to Claudio offline, and he suggested the following edit to the above:

Quoted 9 lines

--8<--
SIE may exit because of pending host work, such as handling an interrupt, in which case VSIE rewinds the guest PSW such that it is transparently resumed (see Fixes tag). Unlike those other places that return rc=0, this return leaves the guest PSW in place, requiring the guest to handle a spurious intercept. This showed up when testing heavy I/O workloads, when multiple vcpus attempted to dispatch the same SIE block and incurred failures inserting them into the radix tree.
-->8--

with the above text:

Quoted 22 lines
Quoted 22 lines

@Janosch, if that ends up being okay, can you update the patch or do you want me to send a v2?

Quoted 18 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines

are not present, but that the VSIE processor returns with effectively rc=0, resulting in additional (and unnecessary) guest work to be performed.

For this case, rewind the guest PSW as we do in the other non-error exits.

Quoted 7 lines
Quoted 4 lines
Quoted 2 lines
Fixes 33a729a1770b ("KVM: s390: vsie: retry SIE instruction on host intercepts")

This is purely cosmetic to have all instances look the same, right?

Nope, I can take this path with particularly high I/O loads on the system, which ends up (incorrectly) sending the intercept to the guest.

this is a good candidate for the explanation I mentioned above :)

(the patch itself looks fine)

maintainer
#7 lore raw
In reply to Eric Farman
Quoted 27 lines · Eric Farman
Quoted 14 lines · Eric Farman
Quoted 10 lines · Claudio Imbrenda
Quoted 5 lines · On Mon, 05 Jan 2026 10:46:53 -0500 Eric Farman <[email protected]> wrote:
Quoted 4 lines · Janosch Frank
Quoted 3 lines · Eric Farman

SIE may exit because of pending host work, such as handling an interrupt, in which case VSIE rewinds the guest PSW such that it is transparently resumed (see Fixes tag). There is still one scenario where those conditions

can you add a few words to (very briefly) explain what the scenario is?

Maybe if this paragraph were rewritten this way, instead?

Quoted 9 lines

--8<--
SIE may exit because of pending host work, such as handling an interrupt, in which case VSIE rewinds the guest PSW such that it is transparently resumed (see Fixes tag). Unlike those other places that return rc=0, this return leaves the guest PSW in place, requiring the guest to handle an intercept that was meant to be serviced by the host. This showed up when testing heavy I/O workloads, when multiple vcpus attempted to dispatch the same SIE block and incurred failures inserting them into the radix tree. -->8--

Spoke to Claudio offline, and he suggested the following edit to the above:

Quoted 12 lines

--8<--
SIE may exit because of pending host work, such as handling an interrupt, in which case VSIE rewinds the guest PSW such that it is transparently resumed (see Fixes tag). Unlike those other places that return rc=0, this return leaves the guest PSW in place, requiring the guest to handle a spurious intercept. This showed up when testing heavy I/O workloads, when multiple vcpus attempted to dispatch the same SIE block and incurred failures inserting them into the radix tree.
-->8--

Quoted 2 lines

@Janosch, if that ends up being okay, can you update the patch or do you want me to send a v2?

I can fix that up.

#8 lore raw
Quoted 39 lines · Janosch Frank
Quoted 38 lines · Eric Farman
Quoted 14 lines · Eric Farman
Quoted 10 lines · Claudio Imbrenda
Quoted 5 lines · On Mon, 05 Jan 2026 10:46:53 -0500 Eric Farman <[email protected]> wrote:
Quoted 4 lines · Janosch Frank
Quoted 3 lines · Eric Farman

SIE may exit because of pending host work, such as handling an interrupt, in which case VSIE rewinds the guest PSW such that it is transparently resumed (see Fixes tag). There is still one scenario where those conditions

can you add a few words to (very briefly) explain what the scenario is?

Maybe if this paragraph were rewritten this way, instead?

Quoted 9 lines

--8<--
SIE may exit because of pending host work, such as handling an interrupt, in which case VSIE rewinds the guest PSW such that it is transparently resumed (see Fixes tag). Unlike those other places that return rc=0, this return leaves the guest PSW in place, requiring the guest to handle an intercept that was meant to be serviced by the host. This showed up when testing heavy I/O workloads, when multiple vcpus attempted to dispatch the same SIE block and incurred failures inserting them into the radix tree. -->8--

Spoke to Claudio offline, and he suggested the following edit to the above:

--8<--
SIE may exit because of pending host work, such as handling an interrupt, in which case VSIE rewinds the guest PSW such that it is transparently resumed (see Fixes tag). Unlike those other places that return rc=0, this return leaves the guest PSW in place, requiring the guest to handle a spurious intercept. This showed up when testing heavy I/O workloads, when multiple vcpus attempted to dispatch the same SIE block and incurred failures inserting them into the radix tree.
-->8--

Quoted 1 line
Quoted 3 lines
Quoted 1 line
Quoted 1 line

@Janosch, if that ends up being okay, can you update the patch or do you want me to send a v2?

I can fix that up.

Great, thank you!

Reviewed-by on v1 1/1
#9 lore raw
In reply to Eric Farman
Quoted 10 lines · Eric Farman

SIE may exit because of pending host work, such as handling an interrupt, in which case VSIE rewinds the guest PSW such that it is transparently resumed (see Fixes tag). There is still one scenario where those conditions are not present, but that the VSIE processor returns with effectively rc=0, resulting in additional (and unnecessary) guest work to be performed.

For this case, rewind the guest PSW as we do in the other non-error exits.

Fixes 33a729a1770b ("KVM: s390: vsie: retry SIE instruction on host intercepts")

LGTM