[bpf,v2] bpf: fix narrower loads on s390
diff mbox series

Message ID 20190718150103.84837-1-iii@linux.ibm.com
State Changes Requested
Delegated to: BPF Maintainers
Headers show
Series
  • [bpf,v2] bpf: fix narrower loads on s390
Related show

Commit Message

Ilya Leoshkevich July 18, 2019, 3:01 p.m. UTC
The very first check in test_pkt_md_access is failing on s390, which
happens because loading a part of a struct __sk_buff field produces
an incorrect result.

The preprocessed code of the check is:

{
	__u8 tmp = *((volatile __u8 *)&skb->len +
		((sizeof(skb->len) - sizeof(__u8)) / sizeof(__u8)));
	if (tmp != ((*(volatile __u32 *)&skb->len) & 0xFF)) return 2;
};

clang generates the following code for it:

      0:	71 21 00 03 00 00 00 00	r2 = *(u8 *)(r1 + 3)
      1:	61 31 00 00 00 00 00 00	r3 = *(u32 *)(r1 + 0)
      2:	57 30 00 00 00 00 00 ff	r3 &= 255
      3:	5d 23 00 1d 00 00 00 00	if r2 != r3 goto +29 <LBB0_10>

Finally, verifier transforms it to:

  0: (61) r2 = *(u32 *)(r1 +104)
  1: (bc) w2 = w2
  2: (74) w2 >>= 24
  3: (bc) w2 = w2
  4: (54) w2 &= 255
  5: (bc) w2 = w2

The problem is that when verifier emits the code to replace a partial
load of a struct __sk_buff field (*(u8 *)(r1 + 3)) with a full load of
struct sk_buff field (*(u32 *)(r1 + 104)), an optional shift and a
bitwise AND, it assumes that the machine is little endian and
incorrectly decides to use a shift.

Adjust shift count calculation to account for endianness.

Fixes: 31fd85816dbe ("bpf: permits narrower load from bpf program context fields")
Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
---
 include/linux/filter.h | 13 +++++++++++++
 kernel/bpf/verifier.c  |  4 ++--
 2 files changed, 15 insertions(+), 2 deletions(-)

Comments

Y Song July 18, 2019, 3:24 p.m. UTC | #1
On Thu, Jul 18, 2019 at 8:01 AM Ilya Leoshkevich <iii@linux.ibm.com> wrote:
>
> The very first check in test_pkt_md_access is failing on s390, which
> happens because loading a part of a struct __sk_buff field produces
> an incorrect result.
>
> The preprocessed code of the check is:
>
> {
>         __u8 tmp = *((volatile __u8 *)&skb->len +
>                 ((sizeof(skb->len) - sizeof(__u8)) / sizeof(__u8)));
>         if (tmp != ((*(volatile __u32 *)&skb->len) & 0xFF)) return 2;
> };
>
> clang generates the following code for it:
>
>       0:        71 21 00 03 00 00 00 00 r2 = *(u8 *)(r1 + 3)
>       1:        61 31 00 00 00 00 00 00 r3 = *(u32 *)(r1 + 0)
>       2:        57 30 00 00 00 00 00 ff r3 &= 255
>       3:        5d 23 00 1d 00 00 00 00 if r2 != r3 goto +29 <LBB0_10>
>
> Finally, verifier transforms it to:
>
>   0: (61) r2 = *(u32 *)(r1 +104)
>   1: (bc) w2 = w2
>   2: (74) w2 >>= 24
>   3: (bc) w2 = w2
>   4: (54) w2 &= 255
>   5: (bc) w2 = w2
>
> The problem is that when verifier emits the code to replace a partial
> load of a struct __sk_buff field (*(u8 *)(r1 + 3)) with a full load of
> struct sk_buff field (*(u32 *)(r1 + 104)), an optional shift and a
> bitwise AND, it assumes that the machine is little endian and
> incorrectly decides to use a shift.
>
> Adjust shift count calculation to account for endianness.
>
> Fixes: 31fd85816dbe ("bpf: permits narrower load from bpf program context fields")
> Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>

Acked-by: Yonghong Song <yhs@fb.com>

> ---
>  include/linux/filter.h | 13 +++++++++++++
>  kernel/bpf/verifier.c  |  4 ++--
>  2 files changed, 15 insertions(+), 2 deletions(-)
>
> diff --git a/include/linux/filter.h b/include/linux/filter.h
> index ff65d22cf336..4fe88e43f0fe 100644
> --- a/include/linux/filter.h
> +++ b/include/linux/filter.h
> @@ -24,6 +24,8 @@
>
>  #include <net/sch_generic.h>
>
> +#include <asm/byteorder.h>
> +
>  #include <uapi/linux/filter.h>
>  #include <uapi/linux/bpf.h>
>
> @@ -1216,4 +1218,15 @@ struct bpf_sockopt_kern {
>         s32             retval;
>  };
>
> +static inline u8 bpf_narrower_load_shift(u32 size_default, u32 size, u32 off)
> +{
> +       u8 load_off = off & (size_default - 1);
> +
> +#ifdef __LITTLE_ENDIAN
> +       return load_off * 8;
> +#else
> +       return (size_default - (load_off + size)) * 8;
> +#endif
> +}
> +
>  #endif /* __LINUX_FILTER_H__ */
> diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
> index 5900cbb966b1..48edc9c9a879 100644
> --- a/kernel/bpf/verifier.c
> +++ b/kernel/bpf/verifier.c
> @@ -8616,8 +8616,8 @@ static int convert_ctx_accesses(struct bpf_verifier_env *env)
>                 }
>
>                 if (is_narrower_load && size < target_size) {
> -                       u8 shift = (off & (size_default - 1)) * 8;
> -
> +                       u8 shift = bpf_narrower_load_shift(size_default, size,
> +                                                          off);
>                         if (ctx_field_size <= 4) {
>                                 if (shift)
>                                         insn_buf[cnt++] = BPF_ALU32_IMM(BPF_RSH,
> --
> 2.21.0
>
Alexei Starovoitov July 18, 2019, 8:43 p.m. UTC | #2
On Thu, Jul 18, 2019 at 05:01:03PM +0200, Ilya Leoshkevich wrote:
> The very first check in test_pkt_md_access is failing on s390, which
> happens because loading a part of a struct __sk_buff field produces
> an incorrect result.
> 
> The preprocessed code of the check is:
> 
> {
> 	__u8 tmp = *((volatile __u8 *)&skb->len +
> 		((sizeof(skb->len) - sizeof(__u8)) / sizeof(__u8)));
> 	if (tmp != ((*(volatile __u32 *)&skb->len) & 0xFF)) return 2;
> };
> 
> clang generates the following code for it:
> 
>       0:	71 21 00 03 00 00 00 00	r2 = *(u8 *)(r1 + 3)
>       1:	61 31 00 00 00 00 00 00	r3 = *(u32 *)(r1 + 0)
>       2:	57 30 00 00 00 00 00 ff	r3 &= 255
>       3:	5d 23 00 1d 00 00 00 00	if r2 != r3 goto +29 <LBB0_10>
> 
> Finally, verifier transforms it to:
> 
>   0: (61) r2 = *(u32 *)(r1 +104)
>   1: (bc) w2 = w2
>   2: (74) w2 >>= 24
>   3: (bc) w2 = w2
>   4: (54) w2 &= 255
>   5: (bc) w2 = w2
> 
> The problem is that when verifier emits the code to replace a partial
> load of a struct __sk_buff field (*(u8 *)(r1 + 3)) with a full load of
> struct sk_buff field (*(u32 *)(r1 + 104)), an optional shift and a
> bitwise AND, it assumes that the machine is little endian and
> incorrectly decides to use a shift.
> 
> Adjust shift count calculation to account for endianness.
> 
> Fixes: 31fd85816dbe ("bpf: permits narrower load from bpf program context fields")
> Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
> ---
>  include/linux/filter.h | 13 +++++++++++++
>  kernel/bpf/verifier.c  |  4 ++--
>  2 files changed, 15 insertions(+), 2 deletions(-)
> 
> diff --git a/include/linux/filter.h b/include/linux/filter.h
> index ff65d22cf336..4fe88e43f0fe 100644
> --- a/include/linux/filter.h
> +++ b/include/linux/filter.h
> @@ -24,6 +24,8 @@
>  
>  #include <net/sch_generic.h>
>  
> +#include <asm/byteorder.h>
> +

unnecessary empty line

>  #include <uapi/linux/filter.h>
>  #include <uapi/linux/bpf.h>
>  
> @@ -1216,4 +1218,15 @@ struct bpf_sockopt_kern {
>  	s32		retval;
>  };
>  
> +static inline u8 bpf_narrower_load_shift(u32 size_default, u32 size, u32 off)

please place this function right after bpf_ctx_narrow_access_ok()
and order arguments the same way as bpf_ctx_narrow_access_ok does.

> +{
> +	u8 load_off = off & (size_default - 1);
> +
> +#ifdef __LITTLE_ENDIAN
> +	return load_off * 8;
> +#else
> +	return (size_default - (load_off + size)) * 8;
> +#endif
> +}
> +
>  #endif /* __LINUX_FILTER_H__ */
> diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
> index 5900cbb966b1..48edc9c9a879 100644
> --- a/kernel/bpf/verifier.c
> +++ b/kernel/bpf/verifier.c
> @@ -8616,8 +8616,8 @@ static int convert_ctx_accesses(struct bpf_verifier_env *env)
>  		}
>  
>  		if (is_narrower_load && size < target_size) {
> -			u8 shift = (off & (size_default - 1)) * 8;
> -
> +			u8 shift = bpf_narrower_load_shift(size_default, size,
> +							   off);
>  			if (ctx_field_size <= 4) {
>  				if (shift)
>  					insn_buf[cnt++] = BPF_ALU32_IMM(BPF_RSH,
> -- 
> 2.21.0
>

Patch
diff mbox series

diff --git a/include/linux/filter.h b/include/linux/filter.h
index ff65d22cf336..4fe88e43f0fe 100644
--- a/include/linux/filter.h
+++ b/include/linux/filter.h
@@ -24,6 +24,8 @@ 
 
 #include <net/sch_generic.h>
 
+#include <asm/byteorder.h>
+
 #include <uapi/linux/filter.h>
 #include <uapi/linux/bpf.h>
 
@@ -1216,4 +1218,15 @@  struct bpf_sockopt_kern {
 	s32		retval;
 };
 
+static inline u8 bpf_narrower_load_shift(u32 size_default, u32 size, u32 off)
+{
+	u8 load_off = off & (size_default - 1);
+
+#ifdef __LITTLE_ENDIAN
+	return load_off * 8;
+#else
+	return (size_default - (load_off + size)) * 8;
+#endif
+}
+
 #endif /* __LINUX_FILTER_H__ */
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index 5900cbb966b1..48edc9c9a879 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -8616,8 +8616,8 @@  static int convert_ctx_accesses(struct bpf_verifier_env *env)
 		}
 
 		if (is_narrower_load && size < target_size) {
-			u8 shift = (off & (size_default - 1)) * 8;
-
+			u8 shift = bpf_narrower_load_shift(size_default, size,
+							   off);
 			if (ctx_field_size <= 4) {
 				if (shift)
 					insn_buf[cnt++] = BPF_ALU32_IMM(BPF_RSH,