diff mbox series

arm64: kprobes: Drop ID map text from kprobes blacklist

Message ID 20230204101807.2862321-1-ardb@kernel.org (mailing list archive)
State New, archived
Headers show
Series arm64: kprobes: Drop ID map text from kprobes blacklist | expand

Commit Message

Ard Biesheuvel Feb. 4, 2023, 10:18 a.m. UTC
The ID mapped text region is never accessed via the normal kernel
mapping of text, and so it was moved into .rodata instead. This means it
is no longer considered as a suitable place for kprobes by default, and
the explicit blacklist is unnecessary, and actually results in an error
message at boot:

  kprobes: Failed to populate blacklist (error -22), kprobes not restricted, be careful using them!

So stop blacklisting the ID map text explicitly.

Fixes: af7249b317e4d0b3d ("arm64: kernel: move identity map out of .text mapping")
Reported-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
---
 arch/arm64/kernel/probes/kprobes.c | 4 ----
 1 file changed, 4 deletions(-)

Comments

Nathan Chancellor Feb. 5, 2023, 6:37 p.m. UTC | #1
On Sat, Feb 04, 2023 at 11:18:07AM +0100, Ard Biesheuvel wrote:
> The ID mapped text region is never accessed via the normal kernel
> mapping of text, and so it was moved into .rodata instead. This means it
> is no longer considered as a suitable place for kprobes by default, and
> the explicit blacklist is unnecessary, and actually results in an error
> message at boot:
> 
>   kprobes: Failed to populate blacklist (error -22), kprobes not restricted, be careful using them!
> 
> So stop blacklisting the ID map text explicitly.
> 
> Fixes: af7249b317e4d0b3d ("arm64: kernel: move identity map out of .text mapping")
> Reported-by: Nathan Chancellor <nathan@kernel.org>
> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>

Thank you for the quick fix!

Tested-by: Nathan Chancellor <nathan@kernel.org>

> ---
>  arch/arm64/kernel/probes/kprobes.c | 4 ----
>  1 file changed, 4 deletions(-)
> 
> diff --git a/arch/arm64/kernel/probes/kprobes.c b/arch/arm64/kernel/probes/kprobes.c
> index f35d059a9a366fa6..70b91a8c6bb3f358 100644
> --- a/arch/arm64/kernel/probes/kprobes.c
> +++ b/arch/arm64/kernel/probes/kprobes.c
> @@ -387,10 +387,6 @@ int __init arch_populate_kprobe_blacklist(void)
>  					(unsigned long)__irqentry_text_end);
>  	if (ret)
>  		return ret;
> -	ret = kprobe_add_area_blacklist((unsigned long)__idmap_text_start,
> -					(unsigned long)__idmap_text_end);
> -	if (ret)
> -		return ret;
>  	ret = kprobe_add_area_blacklist((unsigned long)__hyp_text_start,
>  					(unsigned long)__hyp_text_end);
>  	if (ret || is_kernel_in_hyp_mode())
> -- 
> 2.39.1
>
Mark Rutland Feb. 7, 2023, 10:15 a.m. UTC | #2
On Sat, Feb 04, 2023 at 11:18:07AM +0100, Ard Biesheuvel wrote:
> The ID mapped text region is never accessed via the normal kernel
> mapping of text, and so it was moved into .rodata instead. This means it
> is no longer considered as a suitable place for kprobes by default, and
> the explicit blacklist is unnecessary, and actually results in an error
> message at boot:
> 
>   kprobes: Failed to populate blacklist (error -22), kprobes not restricted, be careful using them!
> 
> So stop blacklisting the ID map text explicitly.
> 
> Fixes: af7249b317e4d0b3d ("arm64: kernel: move identity map out of .text mapping")
> Reported-by: Nathan Chancellor <nathan@kernel.org>
> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>

Acked-by: Mark Rutland <mark.rutland@arm.com>

Mark.

> ---
>  arch/arm64/kernel/probes/kprobes.c | 4 ----
>  1 file changed, 4 deletions(-)
> 
> diff --git a/arch/arm64/kernel/probes/kprobes.c b/arch/arm64/kernel/probes/kprobes.c
> index f35d059a9a366fa6..70b91a8c6bb3f358 100644
> --- a/arch/arm64/kernel/probes/kprobes.c
> +++ b/arch/arm64/kernel/probes/kprobes.c
> @@ -387,10 +387,6 @@ int __init arch_populate_kprobe_blacklist(void)
>  					(unsigned long)__irqentry_text_end);
>  	if (ret)
>  		return ret;
> -	ret = kprobe_add_area_blacklist((unsigned long)__idmap_text_start,
> -					(unsigned long)__idmap_text_end);
> -	if (ret)
> -		return ret;
>  	ret = kprobe_add_area_blacklist((unsigned long)__hyp_text_start,
>  					(unsigned long)__hyp_text_end);
>  	if (ret || is_kernel_in_hyp_mode())
> -- 
> 2.39.1
> 
> 
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Catalin Marinas Feb. 7, 2023, 6:54 p.m. UTC | #3
On Sat, 04 Feb 2023 11:18:07 +0100, Ard Biesheuvel wrote:
> The ID mapped text region is never accessed via the normal kernel
> mapping of text, and so it was moved into .rodata instead. This means it
> is no longer considered as a suitable place for kprobes by default, and
> the explicit blacklist is unnecessary, and actually results in an error
> message at boot:
> 
>   kprobes: Failed to populate blacklist (error -22), kprobes not restricted, be careful using them!
> 
> [...]

Applied to arm64 (for-next/efi-boot-mmu-on), thanks!

[1/1] arm64: kprobes: Drop ID map text from kprobes blacklist
      https://git.kernel.org/arm64/c/a088cf8eee12
diff mbox series

Patch

diff --git a/arch/arm64/kernel/probes/kprobes.c b/arch/arm64/kernel/probes/kprobes.c
index f35d059a9a366fa6..70b91a8c6bb3f358 100644
--- a/arch/arm64/kernel/probes/kprobes.c
+++ b/arch/arm64/kernel/probes/kprobes.c
@@ -387,10 +387,6 @@  int __init arch_populate_kprobe_blacklist(void)
 					(unsigned long)__irqentry_text_end);
 	if (ret)
 		return ret;
-	ret = kprobe_add_area_blacklist((unsigned long)__idmap_text_start,
-					(unsigned long)__idmap_text_end);
-	if (ret)
-		return ret;
 	ret = kprobe_add_area_blacklist((unsigned long)__hyp_text_start,
 					(unsigned long)__hyp_text_end);
 	if (ret || is_kernel_in_hyp_mode())