Restricted pointers ("%pK") are not meant to be used through printk().
It can unintentionally expose security sensitive, raw pointer values.

Use regular pointer formatting instead.

Link: https://lore.kernel.org/lkml/20250113171731-dc10e3c1-da64-4af0-b767-7c7070468023@linutronix.de/
Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de>
---
 arch/arm64/mm/physaddr.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/arm64/mm/physaddr.c b/arch/arm64/mm/physaddr.c
index cde44c13dda1bc0924d2d50369ad87f544a0a802..7d94e09b01b35094e883875a46b979666549095c 100644
--- a/arch/arm64/mm/physaddr.c
+++ b/arch/arm64/mm/physaddr.c
@@ -10,7 +10,7 @@
 phys_addr_t __virt_to_phys(unsigned long x)
 {
 	WARN(!__is_lm_address(__tag_reset(x)),
-	     "virt_to_phys used for non-linear address: %pK (%pS)\n",
+	     "virt_to_phys used for non-linear address: %p (%pS)\n",
 	      (void *)x,
 	      (void *)x);
 

---
base-commit: 0ad2507d5d93f39619fc42372c347d6006b64319
change-id: 20250217-restricted-pointers-arm64-74d54aef13f7

Best regards,
-- 
Thomas Weißschuh <thomas.weissschuh@linutronix.de>

On Mon, 17 Feb 2025 08:39:06 +0100, Thomas Weißschuh wrote:
> Restricted pointers ("%pK") are not meant to be used through printk().
> It can unintentionally expose security sensitive, raw pointer values.
> 
> Use regular pointer formatting instead.
> 
> 

Applied to arm64 (for-next/misc), thanks!

[1/1] arm64: mm: Don't use %pK through printk
      https://git.kernel.org/arm64/c/892d20acf36c

-- 
Catalin

On Mon, Feb 17, 2025 at 08:39:06AM +0100, Thomas Weißschuh wrote:
> Restricted pointers ("%pK") are not meant to be used through printk().
> It can unintentionally expose security sensitive, raw pointer values.
> 
> Use regular pointer formatting instead.
> 
> Link: https://lore.kernel.org/lkml/20250113171731-dc10e3c1-da64-4af0-b767-7c7070468023@linutronix.de/
> Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de>
> ---
>  arch/arm64/mm/physaddr.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/arch/arm64/mm/physaddr.c b/arch/arm64/mm/physaddr.c
> index cde44c13dda1bc0924d2d50369ad87f544a0a802..7d94e09b01b35094e883875a46b979666549095c 100644
> --- a/arch/arm64/mm/physaddr.c
> +++ b/arch/arm64/mm/physaddr.c
> @@ -10,7 +10,7 @@
>  phys_addr_t __virt_to_phys(unsigned long x)
>  {
>  	WARN(!__is_lm_address(__tag_reset(x)),
> -	     "virt_to_phys used for non-linear address: %pK (%pS)\n",
> +	     "virt_to_phys used for non-linear address: %p (%pS)\n",
>  	      (void *)x,
>  	      (void *)x);

This is part of a WARN() and knowing the actual address may well be
useful in debugging. I suspect it also ends up being dumped as part of
the register dump _anyway_.

Will

On Fri, Feb 21, 2025 at 02:55:14PM +0000, Will Deacon wrote:
> On Mon, Feb 17, 2025 at 08:39:06AM +0100, Thomas Weißschuh wrote:
> > Restricted pointers ("%pK") are not meant to be used through printk().
> > It can unintentionally expose security sensitive, raw pointer values.
> > 
> > Use regular pointer formatting instead.
> > 
> > Link: https://lore.kernel.org/lkml/20250113171731-dc10e3c1-da64-4af0-b767-7c7070468023@linutronix.de/
> > Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de>
> > ---
> >  arch/arm64/mm/physaddr.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/arch/arm64/mm/physaddr.c b/arch/arm64/mm/physaddr.c
> > index cde44c13dda1bc0924d2d50369ad87f544a0a802..7d94e09b01b35094e883875a46b979666549095c 100644
> > --- a/arch/arm64/mm/physaddr.c
> > +++ b/arch/arm64/mm/physaddr.c
> > @@ -10,7 +10,7 @@
> >  phys_addr_t __virt_to_phys(unsigned long x)
> >  {
> >  	WARN(!__is_lm_address(__tag_reset(x)),
> > -	     "virt_to_phys used for non-linear address: %pK (%pS)\n",
> > +	     "virt_to_phys used for non-linear address: %p (%pS)\n",
> >  	      (void *)x,
> >  	      (void *)x);
> 
> This is part of a WARN() and knowing the actual address may well be
> useful in debugging. I suspect it also ends up being dumped as part of
> the register dump _anyway_.

That makes sense, especially the value being part of the register dump.
OTOH in the default kernel configuration (kptr_restrict==0), %pK is an alias to
%p and also hashing the pointers. And so far nobody seems to have cared.

If you prefer I can also change it to %px, to specifically show the value.
My main goal is to remove %pK, as it's prone to misuse.


Thomas

On Mon, Feb 24, 2025 at 03:12:54PM +0100, Thomas Weißschuh wrote:
> On Fri, Feb 21, 2025 at 02:55:14PM +0000, Will Deacon wrote:
> > On Mon, Feb 17, 2025 at 08:39:06AM +0100, Thomas Weißschuh wrote:
> > > Restricted pointers ("%pK") are not meant to be used through printk().
> > > It can unintentionally expose security sensitive, raw pointer values.
> > > 
> > > Use regular pointer formatting instead.
> > > 
> > > Link: https://lore.kernel.org/lkml/20250113171731-dc10e3c1-da64-4af0-b767-7c7070468023@linutronix.de/
> > > Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de>
> > > ---
> > >  arch/arm64/mm/physaddr.c | 2 +-
> > >  1 file changed, 1 insertion(+), 1 deletion(-)
> > > 
> > > diff --git a/arch/arm64/mm/physaddr.c b/arch/arm64/mm/physaddr.c
> > > index cde44c13dda1bc0924d2d50369ad87f544a0a802..7d94e09b01b35094e883875a46b979666549095c 100644
> > > --- a/arch/arm64/mm/physaddr.c
> > > +++ b/arch/arm64/mm/physaddr.c
> > > @@ -10,7 +10,7 @@
> > >  phys_addr_t __virt_to_phys(unsigned long x)
> > >  {
> > >  	WARN(!__is_lm_address(__tag_reset(x)),
> > > -	     "virt_to_phys used for non-linear address: %pK (%pS)\n",
> > > +	     "virt_to_phys used for non-linear address: %p (%pS)\n",
> > >  	      (void *)x,
> > >  	      (void *)x);
> > 
> > This is part of a WARN() and knowing the actual address may well be
> > useful in debugging. I suspect it also ends up being dumped as part of
> > the register dump _anyway_.
> 
> That makes sense, especially the value being part of the register dump.
> OTOH in the default kernel configuration (kptr_restrict==0), %pK is an alias to
> %p and also hashing the pointers. And so far nobody seems to have cared.
> 
> If you prefer I can also change it to %px, to specifically show the value.
> My main goal is to remove %pK, as it's prone to misuse.

Probably best to ignore me, then, and keep the existing behaviour with
your change (as your patch is doing).

Thanks,

Will