Re: [PATCH bpf-next] bpf: Avoid unnecessary instruction in conver_bpf_ld_abs()

[Daniel Borkmann]

On 11/26/2018 10:42 PM, David Miller wrote:
> 
> 'offset' is constant and if it is zero, no need to subtract it
> from BPF_REG_TMP.
> 
> Signed-off-by: David S. Miller 

Applied to bpf-next, thanks!

[PATCH bpf-next] bpf: Avoid unnecessary instruction in conver_bpf_ld_abs()

[David Miller]

'offset' is constant and if it is zero, no need to subtract it
from BPF_REG_TMP.

Signed-off-by: David S. Miller 

diff --git a/net/core/filter.c b/net/core/filter.c
index aa274679965d..f50ea971f7a9 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -463,7 +463,8 @@ static bool convert_bpf_ld_abs(struct sock_filter *fp, 
struct bpf_insn **insnp)
bool ldx_off_ok = offset <= S16_MAX;
 
*insn++ = BPF_MOV64_REG(BPF_REG_TMP, BPF_REG_H);
-   *insn++ = BPF_ALU64_IMM(BPF_SUB, BPF_REG_TMP, offset);
+   if (offset)
+   *insn++ = BPF_ALU64_IMM(BPF_SUB, BPF_REG_TMP, offset);
*insn++ = BPF_JMP_IMM(BPF_JSLT, BPF_REG_TMP,
  size, 2 + endian + (!ldx_off_ok * 2));
if (ldx_off_ok) {