Skip to contentSkip to conversation

ARM: xen: unexport HYPERVISOR_platform_op function

PATCH In review Arnd Bergmann 6 participants 11 messages Sep 6, 2019 – Oct 1, 2019

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 ARM: xen: unexport HYPERVISOR_platform_op function

HYPERVISOR_platform_op() is an inline function and should not be exported. Since commit 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions"), this causes a warning:

WARNING: "HYPERVISOR_platform_op" [vmlinux] is a static EXPORT_SYMBOL_GPL

Remove the extraneous export.

Fixes 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions")
---
 arch/arm/xen/enlighten.c | 1 -
 1 file changed, 1 deletion(-)
arch/arm/xen/enlighten.c +0 -1
index 1e57692552d9..845c528acf24 100644
@@ -437,7 +437,6 @@ EXPORT_SYMBOL_GPL(HYPERVISOR_memory_op);
 EXPORT_SYMBOL_GPL(HYPERVISOR_physdev_op);
 EXPORT_SYMBOL_GPL(HYPERVISOR_vcpu_op);
 EXPORT_SYMBOL_GPL(HYPERVISOR_tmem_op);
-EXPORT_SYMBOL_GPL(HYPERVISOR_platform_op);
 EXPORT_SYMBOL_GPL(HYPERVISOR_multicall);
 EXPORT_SYMBOL_GPL(HYPERVISOR_vm_assist);
 EXPORT_SYMBOL_GPL(HYPERVISOR_dm_op);
Signature · 1 line
2.20.0
#1 lore raw
Quoted 10 lines · Arnd Bergmann

HYPERVISOR_platform_op() is an inline function and should not be exported. Since commit 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions"), this causes a warning:

WARNING: "HYPERVISOR_platform_op" [vmlinux] is a static EXPORT_SYMBOL_GPL

Remove the extraneous export.

Fixes 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions")

Something is wonky.  That symbol is (/ really ought to be) in the hypercall page and most definitely not inline.

Which tree is that changeset from?  I can't find the SHA.

I hate to open a separate can of worms, but why are they tagged GPL? The Xen hypercall ABI, like the Linux syscall ABI, are specifically not GPL.  Xen has as something very similar to (and probably derived from) the Linux-syscall-note exception.

~Andrew

#2 lore raw
Quoted 7 lines · Andrew Cooper
Quoted 6 lines · Arnd Bergmann

HYPERVISOR_platform_op() is an inline function and should not be exported. Since commit 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions"), this causes a warning:

WARNING: "HYPERVISOR_platform_op" [vmlinux] is a static EXPORT_SYMBOL_GPL

Quoted 7 lines
Quoted 4 lines

Remove the extraneous export.

Fixes 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions")

Something is wonky.  That symbol is (/ really ought to be) in the hypercall page and most definitely not inline.

It's HYPERVISOR_platform_op_raw that's in the hypercall page afaics.

Jan

#3 lore raw
In reply to Andrew Cooper
Quoted 10 lines · Andrew Cooper
Quoted 8 lines · Arnd Bergmann

HYPERVISOR_platform_op() is an inline function and should not be exported. Since commit 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions"), this causes a warning:

WARNING: "HYPERVISOR_platform_op" [vmlinux] is a static EXPORT_SYMBOL_GPL

Remove the extraneous export.

Quoted 7 lines
Quoted 2 lines
Fixes 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions")

Something is wonky. That symbol is (/ really ought to be) in the hypercall page and most definitely not inline.

Which tree is that changeset from? I can't find the SHA.

This is from linux-next, I think from the kbuild tree.

Arnd

#4 lore raw
Quoted 10 lines · Arnd Bergmann
Quoted 9 lines · Andrew Cooper
Quoted 8 lines · Arnd Bergmann

HYPERVISOR_platform_op() is an inline function and should not be exported. Since commit 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions"), this causes a warning:

WARNING: "HYPERVISOR_platform_op" [vmlinux] is a static EXPORT_SYMBOL_GPL

Remove the extraneous export.

Quoted 7 lines
Quoted 6 lines
Quoted 2 lines
Fixes 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions")

Something is wonky. That symbol is (/ really ought to be) in the hypercall page and most definitely not inline.

Which tree is that changeset from? I can't find the SHA.

This is from linux-next, I think from the kbuild tree.

Thanks.

Julien/Stefano: Why are any of these hypercalls out-of-line?  ARM doesn't use the hypercall page, and there is no argument translation (not even in arm32 as there are no 5-argument hypercalls declared).

They'd surely be easier to implement with a few static inlines and some common code, than to try and replicate the x86 side hypercall_page interface ?

~Andrew

#5 lore raw

Hi Andrew,

Quoted 16 lines · Andrew Cooper
Quoted 15 lines · Arnd Bergmann
Quoted 14 lines · Andrew Cooper
Quoted 10 lines · Arnd Bergmann

HYPERVISOR_platform_op() is an inline function and should not be exported. Since commit 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions"), this causes a warning:

WARNING: "HYPERVISOR_platform_op" [vmlinux] is a static EXPORT_SYMBOL_GPL

Remove the extraneous export.

Fixes 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions")

Something is wonky. That symbol is (/ really ought to be) in the hypercall page and most definitely not inline.

Quoted 8 lines
Quoted 2 lines
Quoted 1 line

Which tree is that changeset from? I can't find the SHA.

This is from linux-next, I think from the kbuild tree.

Thanks.

Julien/Stefano: Why are any of these hypercalls out-of-line?  ARM doesn't use the hypercall page, and there is no argument translation (not even in arm32 as there are no 5-argument hypercalls declared).

I am not sure how the hypercall page makes things different. You still have to store the arguments in the correct register so...

Quoted 4 lines

They'd surely be easier to implement with a few static inlines and some common code, than to try and replicate the x86 side hypercall_page interface ?

... I don't think they will be easier to implement with a few static inlines. The implementation will likely end up to be similar to arch/x86/asm/xen/hypercall.h.

Furthermore, one of the downside of per-arch static inline is it is more difficult to ensure the prototype match for all the architectures. Although, it might be possible to make them common by only requesting per-arch to implement HYPERCALL_N(...).

So I think the code is better as it is.

While looking at the code, I also realized that the implementation of HYPERCALL_dm_op might be incorrect for Arm32. Similarly do privcmd call, I think dm_op call should enable user access as they will be used by userspace.

We don't use dm_op on Arm so far, hence why I think this was unnoticed. I will see if I can reproduce it and send a patch.

Cheers,

Signature · 1 line
Julien Grall
#6 lore raw

Hi Julien,

Quoted 45 lines · Julien Grall
Quoted 16 lines · Andrew Cooper
Quoted 15 lines · Arnd Bergmann
Quoted 14 lines · Andrew Cooper
Quoted 10 lines · Arnd Bergmann

HYPERVISOR_platform_op() is an inline function and should not be exported. Since commit 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions"), this causes a warning:

WARNING: "HYPERVISOR_platform_op" [vmlinux] is a static EXPORT_SYMBOL_GPL

Remove the extraneous export.

Fixes 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions")

Something is wonky. That symbol is (/ really ought to be) in the hypercall page and most definitely not inline.

Quoted 8 lines
Quoted 2 lines
Quoted 1 line

Which tree is that changeset from? I can't find the SHA.

This is from linux-next, I think from the kbuild tree.

Thanks.

Julien/Stefano: Why are any of these hypercalls out-of-line?  ARM doesn't use the hypercall page, and there is no argument translation (not even in arm32 as there are no 5-argument hypercalls declared).

I am not sure how the hypercall page makes things different. You still have to store the arguments in the correct register so...

Quoted 4 lines

They'd surely be easier to implement with a few static inlines and some common code, than to try and replicate the x86 side hypercall_page interface ?

... I don't think they will be easier to implement with a few static inlines. The implementation will likely end up to be similar to arch/x86/asm/xen/hypercall.h.

Furthermore, one of the downside of per-arch static inline is it is more difficult to ensure the prototype match for all the architectures. Although, it might be possible to make them common by only requesting per-arch to implement HYPERCALL_N(...).

So I think the code is better as it is.

Quoted 7 lines

While looking at the code, I also realized that the implementation of HYPERCALL_dm_op might be incorrect for Arm32. Similarly do privcmd call, I think dm_op call should enable user access as they will be used by userspace.

We don't use dm_op on Arm so far, hence why I think this was unnoticed. I will see if I can reproduce it and send a patch.

I'm seeing this when building arm64 defconfig v5.4-rc1:

| [mark@lakrids:~/src/linux]% usekorg 8.1.0  make ARCH=arm64 CROSS_COMPILE=aarch64-linux- -j56 -s
| arch/arm64/Makefile:62: CROSS_COMPILE_COMPAT not defined or empty, the compat vDSO will not be built
| WARNING: "HYPERVISOR_platform_op" [vmlinux] is a static EXPORT_SYMBOL_GPL
| WARNING: "HYPERVISOR_platform_op" [vmlinux] is a static EXPORT_SYMBOL_GPL

I couldn't see a follow-up; do you have a patch for this?

Thanks,
Mark.

#7 lore raw
Quoted 1 line · Mark Rutland

Hi Julien,

Hi Mark,

Quoted 55 lines · Julien Grall
Quoted 45 lines · Andrew Cooper
Quoted 16 lines · Arnd Bergmann
Quoted 15 lines · Andrew Cooper
Quoted 14 lines · Arnd Bergmann
Quoted 10 lines

HYPERVISOR_platform_op() is an inline function and should not be exported. Since commit 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions"), this causes a warning:

WARNING: "HYPERVISOR_platform_op" [vmlinux] is a static EXPORT_SYMBOL_GPL

Remove the extraneous export.

Fixes 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions")

Something is wonky. That symbol is (/ really ought to be) in the hypercall page and most definitely not inline.

Quoted 8 lines
Quoted 2 lines
Quoted 1 line

Which tree is that changeset from? I can't find the SHA.

This is from linux-next, I think from the kbuild tree.

Thanks.

Julien/Stefano: Why are any of these hypercalls out-of-line?  ARM doesn't use the hypercall page, and there is no argument translation (not even in arm32 as there are no 5-argument hypercalls declared).

I am not sure how the hypercall page makes things different. You still have to store the arguments in the correct register so...

Quoted 4 lines

They'd surely be easier to implement with a few static inlines and some common code, than to try and replicate the x86 side hypercall_page interface ?

... I don't think they will be easier to implement with a few static inlines. The implementation will likely end up to be similar to arch/x86/asm/xen/hypercall.h.

Furthermore, one of the downside of per-arch static inline is it is more difficult to ensure the prototype match for all the architectures. Although, it might be possible to make them common by only requesting per-arch to implement HYPERCALL_N(...).

So I think the code is better as it is.

Quoted 7 lines

While looking at the code, I also realized that the implementation of HYPERCALL_dm_op might be incorrect for Arm32. Similarly do privcmd call, I think dm_op call should enable user access as they will be used by userspace.

We don't use dm_op on Arm so far, hence why I think this was unnoticed. I will see if I can reproduce it and send a patch.

Quoted 8 lines

I'm seeing this when building arm64 defconfig v5.4-rc1:

| [mark@lakrids:~/src/linux]% usekorg 8.1.0  make ARCH=arm64 CROSS_COMPILE=aarch64-linux- -j56 -s
| arch/arm64/Makefile:62: CROSS_COMPILE_COMPAT not defined or empty, the compat vDSO will not be built
| WARNING: "HYPERVISOR_platform_op" [vmlinux] is a static EXPORT_SYMBOL_GPL
| WARNING: "HYPERVISOR_platform_op" [vmlinux] is a static EXPORT_SYMBOL_GPL

I couldn't see a follow-up; do you have a patch for this?

The first e-mail of the thread should contain a patch to address the warning (see [1]). But it is still waiting on an Ack from Stefano so it can get merged.

Cheers,

[1] https://patchwork.kernel.org/patch/11135601/

Signature · 1 line
Julien Grall
#8 lore raw
Quoted 16 lines · Julien Grall
Quoted 15 lines · Mark Rutland
Quoted 14 lines · Julien Grall
Quoted 13 lines · Andrew Cooper
Quoted 12 lines · Arnd Bergmann
Quoted 11 lines · Andrew Cooper
Quoted 10 lines · Arnd Bergmann

HYPERVISOR_platform_op() is an inline function and should not be exported. Since commit 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions"), this causes a warning:

WARNING: "HYPERVISOR_platform_op" [vmlinux] is a static EXPORT_SYMBOL_GPL

Remove the extraneous export.

Fixes 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions")

[...]

Quoted 15 lines
Quoted 15 lines
Quoted 7 lines

While looking at the code, I also realized that the implementation of HYPERCALL_dm_op might be incorrect for Arm32. Similarly do privcmd call, I think dm_op call should enable user access as they will be used by userspace.

We don't use dm_op on Arm so far, hence why I think this was unnoticed. I will see if I can reproduce it and send a patch.

I'm seeing this when building arm64 defconfig v5.4-rc1:

| [mark@lakrids:~/src/linux]% usekorg 8.1.0  make ARCH=arm64 CROSS_COMPILE=aarch64-linux- -j56 -s
| arch/arm64/Makefile:62: CROSS_COMPILE_COMPAT not defined or empty, the compat vDSO will not be built
| WARNING: "HYPERVISOR_platform_op" [vmlinux] is a static EXPORT_SYMBOL_GPL
| WARNING: "HYPERVISOR_platform_op" [vmlinux] is a static EXPORT_SYMBOL_GPL
Quoted 5 lines
Quoted 1 line

I couldn't see a follow-up; do you have a patch for this?

The first e-mail of the thread should contain a patch to address the warning (see [1]). But it is still waiting on an Ack from Stefano so it can get merged.

Ah, sorry. I misunderstood what you were planning to send a patch for, and assumed you were going to propose an alternative to Arnd's patch.

Stefano, do you see any problem with Arnd's patch? If not, it would be good to get this merged soon.

Thanks,
Mark.

Quoted 7 lines

Cheers,

[1] https://patchwork.kernel.org/patch/11135601/

Signature · 1 line
Julien Grall
#9 lore raw
In reply to Arnd Bergmann
Quoted 19 lines · Arnd Bergmann

HYPERVISOR_platform_op() is an inline function and should not be exported. Since commit 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions"), this causes a warning:

WARNING: "HYPERVISOR_platform_op" [vmlinux] is a static EXPORT_SYMBOL_GPL

Remove the extraneous export.

Fixes 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions")
---
 arch/arm/xen/enlighten.c | 1 -
 1 file changed, 1 deletion(-)
arch/arm/xen/enlighten.c +0 -0
index 1e57692552d9..845c528acf24 100644

arch/arm/xen/enlighten.c EXPORT_SYMBOL_GPL(HYPERVISOR_memory_op); · line 442

Quoted 7 lines
arch/arm/xen/enlighten.c +0 -1
@@ -437,7 +437,6 @@ EXPORT_SYMBOL_GPL(HYPERVISOR_memory_op);
 EXPORT_SYMBOL_GPL(HYPERVISOR_physdev_op);
 EXPORT_SYMBOL_GPL(HYPERVISOR_vcpu_op);
 EXPORT_SYMBOL_GPL(HYPERVISOR_tmem_op);
-EXPORT_SYMBOL_GPL(HYPERVISOR_platform_op);
 EXPORT_SYMBOL_GPL(HYPERVISOR_multicall);
 EXPORT_SYMBOL_GPL(HYPERVISOR_vm_assist);
 EXPORT_SYMBOL_GPL(HYPERVISOR_dm_op);

Hi Arnd,

Thank you for the patch. HYPERVISOR_platform_op() is an inline function, the underlying function that should be exported is HYPERVISOR_platform_op_raw. So, instead of removing HYPERVISOR_platform_op, we should change it to HYPERVISOR_platform_op_raw.

For convenience, and for testing I cooked up a patch. Arnd, if you are happy with it (in the sense that it solves your problem) we'll check it in the xentip tree, unless you would like to get it in your tree?

Cheers,

Stefano

---

From: Stefano Stabellini <[email protected]>

HYPERVISOR_platform_op() is an inline function and should not be exported. Since commit 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions"), this causes a warning:

WARNING: "HYPERVISOR_platform_op" [vmlinux] is a static EXPORT_SYMBOL_GPL

Instead, export the underlying function called by the static inline:
HYPERVISOR_platform_op_raw.

Fixes 15bfc2348d54 ("modpost: check for static EXPORT_SYMBOL* functions")
arch/arm/xen/enlighten.c +1 -1
index 1e57692552d9..522c97d43ef8 100644
@@ -437,7 +437,7 @@ EXPORT_SYMBOL_GPL(HYPERVISOR_memory_op);
 EXPORT_SYMBOL_GPL(HYPERVISOR_physdev_op);
 EXPORT_SYMBOL_GPL(HYPERVISOR_vcpu_op);
 EXPORT_SYMBOL_GPL(HYPERVISOR_tmem_op);
-EXPORT_SYMBOL_GPL(HYPERVISOR_platform_op);
+EXPORT_SYMBOL_GPL(HYPERVISOR_platform_op_raw);
 EXPORT_SYMBOL_GPL(HYPERVISOR_multicall);
 EXPORT_SYMBOL_GPL(HYPERVISOR_vm_assist);
 EXPORT_SYMBOL_GPL(HYPERVISOR_dm_op);
#10 lore raw
Quoted 5 lines · Stefano Stabellini

Thank you for the patch. HYPERVISOR_platform_op() is an inline function, the underlying function that should be exported is HYPERVISOR_platform_op_raw. So, instead of removing HYPERVISOR_platform_op, we should change it to HYPERVISOR_platform_op_raw.

Ok, that makes sense.

Quoted 4 lines

For convenience, and for testing I cooked up a patch. Arnd, if you are happy with it (in the sense that it solves your problem) we'll check it in the xentip tree, unless you would like to get it in your tree?

Please merge it through your tree.

arch/arm/xen/enlighten.c EXPORT_SYMBOL_GPL(HYPERVISOR_memory_op); · line 443

Quoted 9 lines
arch/arm/xen/enlighten.c +1 -1
@@ -437,7 +437,7 @@ EXPORT_SYMBOL_GPL(HYPERVISOR_memory_op);
 EXPORT_SYMBOL_GPL(HYPERVISOR_physdev_op);
 EXPORT_SYMBOL_GPL(HYPERVISOR_vcpu_op);
 EXPORT_SYMBOL_GPL(HYPERVISOR_tmem_op);
-EXPORT_SYMBOL_GPL(HYPERVISOR_platform_op);
+EXPORT_SYMBOL_GPL(HYPERVISOR_platform_op_raw);
 EXPORT_SYMBOL_GPL(HYPERVISOR_multicall);
 EXPORT_SYMBOL_GPL(HYPERVISOR_vm_assist);
 EXPORT_SYMBOL_GPL(HYPERVISOR_dm_op);

Note that there are obviously no callers from any loadable modules in the kernel, all users are in built-in code at the moment. As an API definition it still makes sense though.

Arnd