Annotating a pointer from kernel to __user and then back again requires
an extra __force annotation to silent sparse warning. In call_undef_hook()
this unnecessary complexity can be avoided by modifying the intermediate
user pointer to unsigned long.
This way there is no inter-changeable use of user and kernel pointers
and the code is consistent.
Note: This patch adds no functional changes to code.
Cc: Catalin Marinas <[email protected]>
Cc: Will Deacon <[email protected]>
Signed-off-by: Amit Daniel Kachhap <[email protected]>
---
Changes from v1:
* Instead of directly using instruction_pointer() for kernel pointers,
modified the pc type from void __user * to unsigned long as suggested
by Mark Rutland.
arch/arm64/kernel/traps.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/arch/arm64/kernel/traps.c b/arch/arm64/kernel/traps.c
index b03e383d944a..09236751283e 100644
--- a/arch/arm64/kernel/traps.c
+++ b/arch/arm64/kernel/traps.c
@@ -400,11 +400,11 @@ static int call_undef_hook(struct pt_regs *regs)
unsigned long flags;
u32 instr;
int (*fn)(struct pt_regs *regs, u32 instr) = NULL;
- void __user *pc = (void __user *)instruction_pointer(regs);
+ unsigned long pc = instruction_pointer(regs);
if (!user_mode(regs)) {
__le32 instr_le;
- if (get_kernel_nofault(instr_le, (__force __le32 *)pc))
+ if (get_kernel_nofault(instr_le, (__le32 *)pc))
goto exit;
instr = le32_to_cpu(instr_le);
} else if (compat_thumb_mode(regs)) {
--
2.17.1
On Fri, Sep 17, 2021 at 11:28:11AM +0530, Amit Daniel Kachhap wrote:
> Annotating a pointer from kernel to __user and then back again requires
> an extra __force annotation to silent sparse warning. In call_undef_hook()
> this unnecessary complexity can be avoided by modifying the intermediate
> user pointer to unsigned long.
>
> This way there is no inter-changeable use of user and kernel pointers
> and the code is consistent.
>
> Note: This patch adds no functional changes to code.
>
> Cc: Catalin Marinas <[email protected]>
> Cc: Will Deacon <[email protected]>
> Signed-off-by: Amit Daniel Kachhap <[email protected]>
I like that this clarifies that `pc` is not necessarily a user pointer,
and this makes the relevant lines shorter and eaier to read, so FWIW:
Acked-by: Mark Rutland <[email protected]>
Thanks,
Mark.
> ---
> Changes from v1:
> * Instead of directly using instruction_pointer() for kernel pointers,
> modified the pc type from void __user * to unsigned long as suggested
> by Mark Rutland.
>
> arch/arm64/kernel/traps.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/arch/arm64/kernel/traps.c b/arch/arm64/kernel/traps.c
> index b03e383d944a..09236751283e 100644
> --- a/arch/arm64/kernel/traps.c
> +++ b/arch/arm64/kernel/traps.c
> @@ -400,11 +400,11 @@ static int call_undef_hook(struct pt_regs *regs)
> unsigned long flags;
> u32 instr;
> int (*fn)(struct pt_regs *regs, u32 instr) = NULL;
> - void __user *pc = (void __user *)instruction_pointer(regs);
> + unsigned long pc = instruction_pointer(regs);
>
> if (!user_mode(regs)) {
> __le32 instr_le;
> - if (get_kernel_nofault(instr_le, (__force __le32 *)pc))
> + if (get_kernel_nofault(instr_le, (__le32 *)pc))
> goto exit;
> instr = le32_to_cpu(instr_le);
> } else if (compat_thumb_mode(regs)) {
> --
> 2.17.1
>
On Fri, 17 Sep 2021 11:28:11 +0530, Amit Daniel Kachhap wrote:
> Annotating a pointer from kernel to __user and then back again requires
> an extra __force annotation to silent sparse warning. In call_undef_hook()
> this unnecessary complexity can be avoided by modifying the intermediate
> user pointer to unsigned long.
>
> This way there is no inter-changeable use of user and kernel pointers
> and the code is consistent.
>
> [...]
Applied to arm64 (for-next/misc), thanks!
[1/1] arm64/traps: Avoid unnecessary kernel/user pointer conversion
https://git.kernel.org/arm64/c/f5b650f887f3
Cheers,
--
Will
https://fixes.arm64.dev
https://next.arm64.dev
https://will.arm64.dev