Skip to contentSkip to conversation

hw/s390x: Fix the function arguments in the pci stub file

PATCH In review Thomas Huth 3 participants 12 messages Jan 31, 2019 – Feb 11, 2019

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 hw/s390x: Fix the function arguments in the pci stub file

The arguments of the stub functions to not match the real implementation (and the prototypes in the header) anymore, so if you try to compile s390x without CONFIG_PCI, the build currently fails.

---
 hw/s390x/s390-pci-stub.c | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)
hw/s390x/s390-pci-stub.c +9 -7
index ad4c5a7..145759e 100644
@@ -29,38 +29,40 @@ void s390_pci_sclp_deconfigure(SCCB *sccb)
 }
 
 /* target/s390x/kvm.c */
-int clp_service_call(S390CPU *cpu, uint8_t r2)
+int clp_service_call(S390CPU *cpu, uint8_t r2, uintptr_t ra)
 {
     return -1;
 }
 
-int pcilg_service_call(S390CPU *cpu, uint8_t r1, uint8_t r2)
+int pcilg_service_call(S390CPU *cpu, uint8_t r1, uint8_t r2, uintptr_t ra)
 {
     return -1;
 }
 
-int pcistg_service_call(S390CPU *cpu, uint8_t r1, uint8_t r2)
+int pcistg_service_call(S390CPU *cpu, uint8_t r1, uint8_t r2, uintptr_t ra)
 {
     return -1;
 }
 
-int stpcifc_service_call(S390CPU *cpu, uint8_t r1, uint64_t fiba, uint8_t ar)
+int stpcifc_service_call(S390CPU *cpu, uint8_t r1, uint64_t fiba, uint8_t ar,
+                         uintptr_t ra)
 {
     return -1;
 }
 
-int rpcit_service_call(S390CPU *cpu, uint8_t r1, uint8_t r2)
+int rpcit_service_call(S390CPU *cpu, uint8_t r1, uint8_t r2, uintptr_t ra)
 {
     return -1;
 }
 
 int pcistb_service_call(S390CPU *cpu, uint8_t r1, uint8_t r3, uint64_t gaddr,
-                        uint8_t ar)
+                        uint8_t ar, uintptr_t ra)
 {
     return -1;
 }
 
-int mpcifc_service_call(S390CPU *cpu, uint8_t r1, uint64_t fiba, uint8_t ar)
+int mpcifc_service_call(S390CPU *cpu, uint8_t r1, uint64_t fiba, uint8_t ar,
+                        uintptr_t ra)
 {
     return -1;
 }
Signature · 1 line
1.8.3.1
#1 lore raw
Quoted 1 line · On Thu, 31 Jan 2019 18:47:08 +0100 Thomas Huth <[email protected]> wrote:

The arguments of the stub functions to not match the real implementation

s/to/do/

Quoted 3 lines

(and the prototypes in the header) anymore, so if you try to compile s390x without CONFIG_PCI, the build currently fails.

Fixes 468a93898a97 ("s390x/pci: pass the retaddr to all PCI instructions")
Quoted 4 lines
---
 hw/s390x/s390-pci-stub.c | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)

This file seems to be in danger of bitrot. Do you think it'll be easier to test rarely used configs like that after we switch to Kconfig?

#2 lore raw
Quoted 6 lines · Cornelia Huck
Quoted 1 line · On Thu, 31 Jan 2019 18:47:08 +0100 Thomas Huth <[email protected]> wrote:

The arguments of the stub functions to not match the real implementation

s/to/do/

D'oh!

Quoted 6 lines
Quoted 3 lines

(and the prototypes in the header) anymore, so if you try to compile s390x without CONFIG_PCI, the build currently fails.

Fixes 468a93898a97 ("s390x/pci: pass the retaddr to all PCI instructions")
Quoted 7 lines
Quoted 4 lines
---
 hw/s390x/s390-pci-stub.c | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)

This file seems to be in danger of bitrot. Do you think it'll be easier to test rarely used configs like that after we switch to Kconfig?

I hope so, yes. There will be a new --without-default-devices options for "configure" (which matches "make allnoconfig" from the kernel) - if we do it right in the Kconfig file for s390x, it should be possible to catch this problem with that option.

Thomas

#3 lore raw
Quoted 9 lines · Thomas Huth
Quoted 9 lines
Quoted 3 lines

(and the prototypes in the header) anymore, so if you try to compile s390x without CONFIG_PCI, the build currently fails.

Fixes 468a93898a97 ("s390x/pci: pass the retaddr to all PCI instructions")
Quoted 4 lines
---
 hw/s390x/s390-pci-stub.c | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)
Quoted 6 lines
Quoted 2 lines

This file seems to be in danger of bitrot. Do you think it'll be easier to test rarely used configs like that after we switch to Kconfig?

I hope so, yes. There will be a new --without-default-devices options for "configure" (which matches "make allnoconfig" from the kernel) - if we do it right in the Kconfig file for s390x, it should be possible to catch this problem with that option.

Yes, it will be in .travis.yml too.

Right now there is a "select PCI" in the hw/s390x/Kconfig file, but probably it's best to add a config S390_ZPCI with "default y if S390_CCW_VIRTIO" and "select PCI" in it. Not a blocker, but I can integrate it if you send me a fixup patch.

Paolo

#4 lore raw
Quoted 17 lines · Paolo Bonzini
Quoted 9 lines · Thomas Huth
Quoted 9 lines
Quoted 3 lines

(and the prototypes in the header) anymore, so if you try to compile s390x without CONFIG_PCI, the build currently fails.

Fixes 468a93898a97 ("s390x/pci: pass the retaddr to all PCI instructions")
Quoted 4 lines
---
 hw/s390x/s390-pci-stub.c | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)
Quoted 6 lines
Quoted 2 lines

This file seems to be in danger of bitrot. Do you think it'll be easier to test rarely used configs like that after we switch to Kconfig?

I hope so, yes. There will be a new --without-default-devices options for "configure" (which matches "make allnoconfig" from the kernel) - if we do it right in the Kconfig file for s390x, it should be possible to catch this problem with that option.

Quoted 6 lines

Yes, it will be in .travis.yml too.

Right now there is a "select PCI" in the hw/s390x/Kconfig file, but probably it's best to add a config S390_ZPCI with "default y if S390_CCW_VIRTIO" and "select PCI" in it. Not a blocker, but I can integrate it if you send me a fixup patch.

Yes, that's what I had in mind, too. I'll send a fixup patch...

Thomas

#5 lore raw
Quoted 18 lines · Thomas Huth
Quoted 17 lines · Paolo Bonzini
Quoted 15 lines · Thomas Huth
Quoted 11 lines
Quoted 3 lines

(and the prototypes in the header) anymore, so if you try to compile s390x without CONFIG_PCI, the build currently fails.

Fixes 468a93898a97 ("s390x/pci: pass the retaddr to all PCI instructions")
Quoted 4 lines
---
 hw/s390x/s390-pci-stub.c | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)

This file seems to be in danger of bitrot. Do you think it'll be easier to test rarely used configs like that after we switch to Kconfig?

I hope so, yes. There will be a new --without-default-devices options for "configure" (which matches "make allnoconfig" from the kernel) - if we do it right in the Kconfig file for s390x, it should be possible to catch this problem with that option.

Quoted 8 lines
Quoted 6 lines

Yes, it will be in .travis.yml too.

Right now there is a "select PCI" in the hw/s390x/Kconfig file, but probably it's best to add a config S390_ZPCI with "default y if S390_CCW_VIRTIO" and "select PCI" in it. Not a blocker, but I can integrate it if you send me a fixup patch.

Yes, that's what I had in mind, too. I'll send a fixup patch...

Actually, disabling the s390-pci code works from a compiling+linking point of view, but when you then try to start the machine, it fails:

$ s390x-softmmu/qemu-system-s390x -nographic
qemu-system-s390x: Unknown device 's390-pcihost' for default sysbus
Aborted (core dumped)

IIRC we originally wanted to make the "s390-pcihost" device really optional, but then decided not to do it since it would break migration. So the status of the "PCI disablement" got stuck somewhere inbetween, where we've created the switch for the Makefile and the stub file, but never really got it to a point where it could really really be disabled.

So I see two options now:

1) Finally really make the device optional, at least for new machine types, so we can really disable CONFIG_PCI and get a working executable.

2) Scratch the idea completely to make this optional, always link the s390-pci-bus.o and s390-pci-inst.o files unconditionally, and remove the s390-pci-stub.c file.

I assume options 2 is preferred, since we likely rather want to move into the PCI direction in the long run, instead of ignoring it...

Other opinions?

Thomas

#6 lore raw
Quoted 35 lines · On Fri, 1 Feb 2019 07:46:40 +0100 Thomas Huth <[email protected]> wrote:
Quoted 18 lines · Thomas Huth
Quoted 17 lines · Paolo Bonzini
Quoted 15 lines · Thomas Huth
Quoted 11 lines
Quoted 3 lines

(and the prototypes in the header) anymore, so if you try to compile s390x without CONFIG_PCI, the build currently fails.

Fixes 468a93898a97 ("s390x/pci: pass the retaddr to all PCI instructions")
Quoted 4 lines
---
 hw/s390x/s390-pci-stub.c | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)  

This file seems to be in danger of bitrot. Do you think it'll be easier to test rarely used configs like that after we switch to Kconfig?

I hope so, yes. There will be a new --without-default-devices options for "configure" (which matches "make allnoconfig" from the kernel) - if we do it right in the Kconfig file for s390x, it should be possible to catch this problem with that option.

Quoted 8 lines
Quoted 6 lines

Yes, it will be in .travis.yml too.

Right now there is a "select PCI" in the hw/s390x/Kconfig file, but probably it's best to add a config S390_ZPCI with "default y if S390_CCW_VIRTIO" and "select PCI" in it. Not a blocker, but I can integrate it if you send me a fixup patch.

Yes, that's what I had in mind, too. I'll send a fixup patch...

Actually, disabling the s390-pci code works from a compiling+linking point of view, but when you then try to start the machine, it fails:

$ s390x-softmmu/qemu-system-s390x -nographic
qemu-system-s390x: Unknown device 's390-pcihost' for default sysbus
Aborted (core dumped)
Quoted 5 lines

IIRC we originally wanted to make the "s390-pcihost" device really optional, but then decided not to do it since it would break migration. So the status of the "PCI disablement" got stuck somewhere inbetween, where we've created the switch for the Makefile and the stub file, but never really got it to a point where it could really really be disabled.

Well you can disable it, but it's not usable ;)

But yes, that annoying pcihost device needed for backwards compat got into the way.

Quoted 6 lines

So I see two options now:

1) Finally really make the device optional, at least for new machine types, so we can really disable CONFIG_PCI and get a working executable.

Quoted 6 lines

2) Scratch the idea completely to make this optional, always link the s390-pci-bus.o and s390-pci-inst.o files unconditionally, and remove the s390-pci-stub.c file.

I assume options 2 is preferred, since we likely rather want to move into the PCI direction in the long run, instead of ignoring it...

I think both options are viable, but option 1 is of course more work. The win there is that we could disable an entire subsystem.

I guess that the basic questions are: How important is it that subsystems can be compiled out, and do we see a use case for a pci-less s390 machine in the future? We really don't want to spend much time on something of dubious use...

#7 lore raw
Quoted 15 lines · On Fri, 1 Feb 2019 09:23:56 +0100 Cornelia Huck <[email protected]> wrote:
Quoted 11 lines · On Fri, 1 Feb 2019 07:46:40 +0100 Thomas Huth <[email protected]> wrote:

So I see two options now:

1) Finally really make the device optional, at least for new machine types, so we can really disable CONFIG_PCI and get a working executable.

2) Scratch the idea completely to make this optional, always link the s390-pci-bus.o and s390-pci-inst.o files unconditionally, and remove the s390-pci-stub.c file.

I assume options 2 is preferred, since we likely rather want to move into the PCI direction in the long run, instead of ignoring it...

Quoted 7 lines

I think both options are viable, but option 1 is of course more work. The win there is that we could disable an entire subsystem.

I guess that the basic questions are: How important is it that subsystems can be compiled out, and do we see a use case for a pci-less s390 machine in the future? We really don't want to spend much time on something of dubious use...

Any thoughts on this?

I'm currently tending towards option 2 (and can cook up a patch for that). Unless someone is already working on option 1 :)

Of course, I can also apply the original patch, but the end result is not very useful...

#8 lore raw
Quoted 26 lines · Cornelia Huck
Quoted 15 lines · On Fri, 1 Feb 2019 09:23:56 +0100 Cornelia Huck <[email protected]> wrote:
Quoted 11 lines · On Fri, 1 Feb 2019 07:46:40 +0100 Thomas Huth <[email protected]> wrote:

So I see two options now:

1) Finally really make the device optional, at least for new machine types, so we can really disable CONFIG_PCI and get a working executable.

2) Scratch the idea completely to make this optional, always link the s390-pci-bus.o and s390-pci-inst.o files unconditionally, and remove the s390-pci-stub.c file.

I assume options 2 is preferred, since we likely rather want to move into the PCI direction in the long run, instead of ignoring it...

Quoted 7 lines

I think both options are viable, but option 1 is of course more work. The win there is that we could disable an entire subsystem.

I guess that the basic questions are: How important is it that subsystems can be compiled out, and do we see a use case for a pci-less s390 machine in the future? We really don't want to spend much time on something of dubious use...

Quoted 4 lines

Any thoughts on this?

I'm currently tending towards option 2 (and can cook up a patch for that). Unless someone is already working on option 1 :)

Since nobody currently has a need to completely disable PCI, I think we should go with option 2.

Thomas

#9 lore raw
Quoted 27 lines · On Mon, 11 Feb 2019 11:54:45 +0100 Thomas Huth <[email protected]> wrote:
Quoted 26 lines · Cornelia Huck
Quoted 22 lines · On Fri, 1 Feb 2019 09:23:56 +0100 Cornelia Huck <[email protected]> wrote:
Quoted 11 lines · On Fri, 1 Feb 2019 07:46:40 +0100 Thomas Huth <[email protected]> wrote:

So I see two options now:

1) Finally really make the device optional, at least for new machine types, so we can really disable CONFIG_PCI and get a working executable.

2) Scratch the idea completely to make this optional, always link the s390-pci-bus.o and s390-pci-inst.o files unconditionally, and remove the s390-pci-stub.c file.

I assume options 2 is preferred, since we likely rather want to move into the PCI direction in the long run, instead of ignoring it...

I think both options are viable, but option 1 is of course more work. The win there is that we could disable an entire subsystem.

I guess that the basic questions are: How important is it that subsystems can be compiled out, and do we see a use case for a pci-less s390 machine in the future? We really don't want to spend much time on something of dubious use...

Quoted 7 lines
Quoted 4 lines

Any thoughts on this?

I'm currently tending towards option 2 (and can cook up a patch for that). Unless someone is already working on option 1 :)

Since nobody currently has a need to completely disable PCI, I think we should go with option 2.

Hm... I'm wondering if we also should move S390_FEAT_ZPCI from the max cpu model to the qemu cpu model (is there any reason not to turn it on by default in tcg?)

#10 lore raw
Quoted 35 lines · Cornelia Huck
Quoted 32 lines · On Mon, 11 Feb 2019 11:54:45 +0100 Thomas Huth <[email protected]> wrote:
Quoted 30 lines · Cornelia Huck
Quoted 15 lines · On Fri, 1 Feb 2019 09:23:56 +0100 Cornelia Huck <[email protected]> wrote:
Quoted 11 lines · On Fri, 1 Feb 2019 07:46:40 +0100 Thomas Huth <[email protected]> wrote:

So I see two options now:

1) Finally really make the device optional, at least for new machine types, so we can really disable CONFIG_PCI and get a working executable.

2) Scratch the idea completely to make this optional, always link the s390-pci-bus.o and s390-pci-inst.o files unconditionally, and remove the s390-pci-stub.c file.

I assume options 2 is preferred, since we likely rather want to move into the PCI direction in the long run, instead of ignoring it...

Quoted 7 lines

I think both options are viable, but option 1 is of course more work. The win there is that we could disable an entire subsystem.

I guess that the basic questions are: How important is it that subsystems can be compiled out, and do we see a use case for a pci-less s390 machine in the future? We really don't want to spend much time on something of dubious use...

Any thoughts on this?

I'm currently tending towards option 2 (and can cook up a patch for that). Unless someone is already working on option 1 :)

Quoted 6 lines
Quoted 2 lines

Since nobody currently has a need to completely disable PCI, I think we should go with option 2.

Hm... I'm wondering if we also should move S390_FEAT_ZPCI from the max cpu model to the qemu cpu model (is there any reason not to turn it on by default in tcg?)

Migration compatibility? Wouldn't that cause problems when migrating back to older versions of QEMU?

Thomas

#11 lore raw
Quoted 39 lines · On Mon, 11 Feb 2019 12:17:26 +0100 Thomas Huth <[email protected]> wrote:
Quoted 38 lines · Cornelia Huck
Quoted 34 lines · On Mon, 11 Feb 2019 11:54:45 +0100 Thomas Huth <[email protected]> wrote:
Quoted 26 lines · Cornelia Huck
Quoted 15 lines · On Fri, 1 Feb 2019 09:23:56 +0100 Cornelia Huck <[email protected]> wrote:
Quoted 11 lines · On Fri, 1 Feb 2019 07:46:40 +0100 Thomas Huth <[email protected]> wrote:

So I see two options now:

1) Finally really make the device optional, at least for new machine types, so we can really disable CONFIG_PCI and get a working executable.

2) Scratch the idea completely to make this optional, always link the s390-pci-bus.o and s390-pci-inst.o files unconditionally, and remove the s390-pci-stub.c file.

I assume options 2 is preferred, since we likely rather want to move into the PCI direction in the long run, instead of ignoring it...

Quoted 7 lines

I think both options are viable, but option 1 is of course more work. The win there is that we could disable an entire subsystem.

I guess that the basic questions are: How important is it that subsystems can be compiled out, and do we see a use case for a pci-less s390 machine in the future? We really don't want to spend much time on something of dubious use...

Quoted 4 lines

Any thoughts on this?

I'm currently tending towards option 2 (and can cook up a patch for that). Unless someone is already working on option 1 :)

Since nobody currently has a need to completely disable PCI, I think we should go with option 2.

Quoted 6 lines
Quoted 3 lines

Hm... I'm wondering if we also should move S390_FEAT_ZPCI from the max cpu model to the qemu cpu model (is there any reason not to turn it on by default in tcg?)

Migration compatibility? Wouldn't that cause problems when migrating back to older versions of QEMU?

Not sure. But we can still do that in a follow-on patch, if wanted.