[PATCH bpf] s390/bpf: fix lcgr instruction encoding

[PATCH bpf] s390/bpf: fix lcgr instruction encoding

[Ilya Leoshkevich]

"masking, test in bounds 3" fails on s390, because
BPF_ALU64_IMM(BPF_NEG, BPF_REG_2, 0) ignores the top 32 bits of
BPF_REG_2. The reason is that JIT emits lcgfr instead of lcgr.
The associated comment indicates that the code was intended to emit lcgr
in the first place, it's just that the wrong opcode was used.

Fix by using the correct opcode.

Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
---
 arch/s390/net/bpf_jit_comp.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/s390/net/bpf_jit_comp.c b/arch/s390/net/bpf_jit_comp.c
index e636728ab452..6299156f9738 100644
--- a/arch/s390/net/bpf_jit_comp.c
+++ b/arch/s390/net/bpf_jit_comp.c
@@ -863,7 +863,7 @@ static noinline int bpf_jit_insn(struct bpf_jit *jit, struct bpf_prog *fp, int i
 		break;
 	case BPF_ALU64 | BPF_NEG: /* dst = -dst */
 		/* lcgr %dst,%dst */
-		EMIT4(0xb9130000, dst_reg, dst_reg);
+		EMIT4(0xb9030000, dst_reg, dst_reg);
 		break;
 	/*
 	 * BPF_FROM_BE/LE
-- 
2.21.0

Re: [PATCH bpf] s390/bpf: fix lcgr instruction encoding

[Vasily Gorbik]

On Mon, Aug 12, 2019 at 05:03:32PM +0200, Ilya Leoshkevich wrote:
> "masking, test in bounds 3" fails on s390, because
> BPF_ALU64_IMM(BPF_NEG, BPF_REG_2, 0) ignores the top 32 bits of
> BPF_REG_2. The reason is that JIT emits lcgfr instead of lcgr.
> The associated comment indicates that the code was intended to emit lcgr
> in the first place, it's just that the wrong opcode was used.
> 
> Fix by using the correct opcode.
> 
> Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
> ---
>  arch/s390/net/bpf_jit_comp.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/arch/s390/net/bpf_jit_comp.c b/arch/s390/net/bpf_jit_comp.c
> index e636728ab452..6299156f9738 100644
> --- a/arch/s390/net/bpf_jit_comp.c
> +++ b/arch/s390/net/bpf_jit_comp.c
> @@ -863,7 +863,7 @@ static noinline int bpf_jit_insn(struct bpf_jit *jit, struct bpf_prog *fp, int i
>  		break;
>  	case BPF_ALU64 | BPF_NEG: /* dst = -dst */
>  		/* lcgr %dst,%dst */
> -		EMIT4(0xb9130000, dst_reg, dst_reg);
> +		EMIT4(0xb9030000, dst_reg, dst_reg);
>  		break;
>  	/*
>  	 * BPF_FROM_BE/LE
> -- 
> 2.21.0
> 
Please add
Fixes: 054623105728 ("s390/bpf: Add s390x eBPF JIT compiler backend")
or whatever it should be. With that:
Acked-by: Vasily Gorbik <gor@linux.ibm.com>

Re: [PATCH bpf] s390/bpf: fix lcgr instruction encoding

[Daniel Borkmann]

On 8/12/19 5:03 PM, Ilya Leoshkevich wrote:
> "masking, test in bounds 3" fails on s390, because
> BPF_ALU64_IMM(BPF_NEG, BPF_REG_2, 0) ignores the top 32 bits of
> BPF_REG_2. The reason is that JIT emits lcgfr instead of lcgr.
> The associated comment indicates that the code was intended to emit lcgr
> in the first place, it's just that the wrong opcode was used.
> 
> Fix by using the correct opcode.
> 
> Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>

Applied, thanks!