Link the relocatable x86 kernel as PIE

[RFC/RFT PATCH 11/19] x86/rethook: Use RIP-relative reference for fake return address

Posted by Ard Biesheuvel 1 month ago

Pushing an immediate absolute address to the stack is not permitted when
linking x86_64 code in PIE mode. Usually, the address can be taken using
a RIP-relative LEA instruction, but this is not possible here as there
are no available registers.

So instead, take the address into a static global, and push it onto the
stack using a RIP-relative memory operand.

Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
---
 arch/x86/kernel/rethook.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/arch/x86/kernel/rethook.c b/arch/x86/kernel/rethook.c
index 85e2f2d16a90..50812ac718b0 100644
--- a/arch/x86/kernel/rethook.c
+++ b/arch/x86/kernel/rethook.c
@@ -11,6 +11,10 @@
 
 __visible void arch_rethook_trampoline_callback(struct pt_regs *regs);
 
+#ifdef CONFIG_X86_64
+static __used void * const __arch_rethook_trampoline = &arch_rethook_trampoline;
+#endif
+
 #ifndef ANNOTATE_NOENDBR
 #define ANNOTATE_NOENDBR
 #endif
@@ -27,7 +31,7 @@ asm(
 #ifdef CONFIG_X86_64
 	ANNOTATE_NOENDBR "\n"	/* This is only jumped from ret instruction */
 	/* Push a fake return address to tell the unwinder it's a rethook. */
-	"	pushq $arch_rethook_trampoline\n"
+	"	pushq __arch_rethook_trampoline(%rip)\n"
 	UNWIND_HINT_FUNC
 	"       pushq $" __stringify(__KERNEL_DS) "\n"
 	/* Save the 'sp - 16', this will be fixed later. */
-- 
2.47.3

Re: [RFC/RFT PATCH 11/19] x86/rethook: Use RIP-relative reference for fake return address

Posted by David Laight 1 month ago

On Thu,  8 Jan 2026 09:25:38 +0000
Ard Biesheuvel <ardb@kernel.org> wrote:

> Pushing an immediate absolute address to the stack is not permitted when
> linking x86_64 code in PIE mode. Usually, the address can be taken using
> a RIP-relative LEA instruction, but this is not possible here as there
> are no available registers.
> 
> So instead, take the address into a static global, and push it onto the
> stack using a RIP-relative memory operand.

The comment implies the address is 'fake'.
Does that mean it could just be a constant?
Clearly the unwinder would need the same change.

	David

> 
> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
> ---
>  arch/x86/kernel/rethook.c | 6 +++++-
>  1 file changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/arch/x86/kernel/rethook.c b/arch/x86/kernel/rethook.c
> index 85e2f2d16a90..50812ac718b0 100644
> --- a/arch/x86/kernel/rethook.c
> +++ b/arch/x86/kernel/rethook.c
> @@ -11,6 +11,10 @@
>  
>  __visible void arch_rethook_trampoline_callback(struct pt_regs *regs);
>  
> +#ifdef CONFIG_X86_64
> +static __used void * const __arch_rethook_trampoline = &arch_rethook_trampoline;
> +#endif
> +
>  #ifndef ANNOTATE_NOENDBR
>  #define ANNOTATE_NOENDBR
>  #endif
> @@ -27,7 +31,7 @@ asm(
>  #ifdef CONFIG_X86_64
>  	ANNOTATE_NOENDBR "\n"	/* This is only jumped from ret instruction */
>  	/* Push a fake return address to tell the unwinder it's a rethook. */
> -	"	pushq $arch_rethook_trampoline\n"
> +	"	pushq __arch_rethook_trampoline(%rip)\n"
>  	UNWIND_HINT_FUNC
>  	"       pushq $" __stringify(__KERNEL_DS) "\n"
>  	/* Save the 'sp - 16', this will be fixed later. */

Re: [RFC/RFT PATCH 11/19] x86/rethook: Use RIP-relative reference for fake return address

Posted by Ard Biesheuvel 1 month ago

On Thu, 8 Jan 2026 at 13:08, David Laight <david.laight.linux@gmail.com> wrote:
>
> On Thu,  8 Jan 2026 09:25:38 +0000
> Ard Biesheuvel <ardb@kernel.org> wrote:
>
> > Pushing an immediate absolute address to the stack is not permitted when
> > linking x86_64 code in PIE mode. Usually, the address can be taken using
> > a RIP-relative LEA instruction, but this is not possible here as there
> > are no available registers.
> >
> > So instead, take the address into a static global, and push it onto the
> > stack using a RIP-relative memory operand.
>
> The comment implies the address is 'fake'.
> Does that mean it could just be a constant?

It could be, but it isn't, across all architectures.

> Clearly the unwinder would need the same change.
>

Why? The value being pushed is the same.

Re: [RFC/RFT PATCH 11/19] x86/rethook: Use RIP-relative reference for fake return address

Posted by Ard Biesheuvel 1 month ago

On Thu, 8 Jan 2026 at 13:10, Ard Biesheuvel <ardb@kernel.org> wrote:
>
> On Thu, 8 Jan 2026 at 13:08, David Laight <david.laight.linux@gmail.com> wrote:
> >
> > On Thu,  8 Jan 2026 09:25:38 +0000
> > Ard Biesheuvel <ardb@kernel.org> wrote:
> >
> > > Pushing an immediate absolute address to the stack is not permitted when
> > > linking x86_64 code in PIE mode. Usually, the address can be taken using
> > > a RIP-relative LEA instruction, but this is not possible here as there
> > > are no available registers.
> > >
> > > So instead, take the address into a static global, and push it onto the
> > > stack using a RIP-relative memory operand.
> >
> > The comment implies the address is 'fake'.
> > Does that mean it could just be a constant?
>
> It could be, but it isn't, across all architectures.
>
> > Clearly the unwinder would need the same change.
> >
>
> Why? The value being pushed is the same.

Never mind - I guess you meant 'the unwinder would need to use the constant too'