[PATCH 3.16.y,3.18.y] kprobes/x86: Return correct length in __copy_instruction()

[PATCH 3.16.y,3.18.y] kprobes/x86: Return correct length in __copy_instruction()

[Eugene Shatokhin]

Commit c80e5c0c23ce2282476fdc64c4b5e3d3a40723fd upstream.
This is the backport of that fix for 3.16.y-3.18.y.

On x86-64, __copy_instruction() always returns 0 (error) if the
instruction uses %rip-relative addressing. This is because
kernel_insn_init() is called the second time for 'insn' instance
in such cases and sets all its fields to 0.

Because of this, trying to place a kprobe on such instruction
will fail, register_kprobe() will return -EINVAL.

This patch fixes the problem.

Signed-off-by: Eugene Shatokhin <eugene.shatokhin@rosalab.ru>
---
 arch/x86/kernel/kprobes/core.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/arch/x86/kernel/kprobes/core.c b/arch/x86/kernel/kprobes/core.c
index 93d2c04..f2e281c 100644
--- a/arch/x86/kernel/kprobes/core.c
+++ b/arch/x86/kernel/kprobes/core.c
@@ -330,13 +330,16 @@ int __copy_instruction(u8 *dest, u8 *src)
 {
 	struct insn insn;
 	kprobe_opcode_t buf[MAX_INSN_SIZE];
+	int length;
 
 	kernel_insn_init(&insn, (void *)recover_probed_instruction(buf, (unsigned long)src));
 	insn_get_length(&insn);
+	length = insn.length;
+
 	/* Another subsystem puts a breakpoint, failed to recover */
 	if (insn.opcode.bytes[0] == BREAKPOINT_INSTRUCTION)
 		return 0;
-	memcpy(dest, insn.kaddr, insn.length);
+	memcpy(dest, insn.kaddr, length);
 
 #ifdef CONFIG_X86_64
 	if (insn_rip_relative(&insn)) {
@@ -366,7 +369,7 @@ int __copy_instruction(u8 *dest, u8 *src)
 		*(s32 *) disp = (s32) newdisp;
 	}
 #endif
-	return insn.length;
+	return length;
 }
 
 static int arch_copy_kprobe(struct kprobe *p)
-- 
2.3.2

Re: [PATCH 3.16.y,3.18.y] kprobes/x86: Return correct length in __copy_instruction()

[Luis Henriques]

On Tue, May 19, 2015 at 04:58:14PM +0300, Eugene Shatokhin wrote:
> Commit c80e5c0c23ce2282476fdc64c4b5e3d3a40723fd upstream.
> This is the backport of that fix for 3.16.y-3.18.y.
>

Thank you, Eugene.  I'll queue this for the 3.16 kernel.

Cheers,
--
Lu�s

> On x86-64, __copy_instruction() always returns 0 (error) if the
> instruction uses %rip-relative addressing. This is because
> kernel_insn_init() is called the second time for 'insn' instance
> in such cases and sets all its fields to 0.
> 
> Because of this, trying to place a kprobe on such instruction
> will fail, register_kprobe() will return -EINVAL.
> 
> This patch fixes the problem.
> 
> Signed-off-by: Eugene Shatokhin <eugene.shatokhin@rosalab.ru>
> ---
>  arch/x86/kernel/kprobes/core.c | 7 +++++--
>  1 file changed, 5 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/x86/kernel/kprobes/core.c b/arch/x86/kernel/kprobes/core.c
> index 93d2c04..f2e281c 100644
> --- a/arch/x86/kernel/kprobes/core.c
> +++ b/arch/x86/kernel/kprobes/core.c
> @@ -330,13 +330,16 @@ int __copy_instruction(u8 *dest, u8 *src)
>  {
>  	struct insn insn;
>  	kprobe_opcode_t buf[MAX_INSN_SIZE];
> +	int length;
>  
>  	kernel_insn_init(&insn, (void *)recover_probed_instruction(buf, (unsigned long)src));
>  	insn_get_length(&insn);
> +	length = insn.length;
> +
>  	/* Another subsystem puts a breakpoint, failed to recover */
>  	if (insn.opcode.bytes[0] == BREAKPOINT_INSTRUCTION)
>  		return 0;
> -	memcpy(dest, insn.kaddr, insn.length);
> +	memcpy(dest, insn.kaddr, length);
>  
>  #ifdef CONFIG_X86_64
>  	if (insn_rip_relative(&insn)) {
> @@ -366,7 +369,7 @@ int __copy_instruction(u8 *dest, u8 *src)
>  		*(s32 *) disp = (s32) newdisp;
>  	}
>  #endif
> -	return insn.length;
> +	return length;
>  }
>  
>  static int arch_copy_kprobe(struct kprobe *p)
> -- 
> 2.3.2
> 
> --
> To unsubscribe from this list: send the line "unsubscribe stable" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html