Vincent reported that running XDP synproxy program on LoongArch
results in the following error:
JIT doesn't support bpf-to-bpf calls
With dmesg:
multi-func JIT bug 1391 != 1390
The root cause is that verifier will refill the imm with the
correct addresses of bpf_calls for BPF_PSEUDO_FUNC instructions
and then run the last pass of JIT. So we generate different JIT
code for the same instruction in two passes (one for placeholder
and one for real address). Let's use move_addr() instead.
See commit 64f50f657572 ("LoongArch, bpf: Use 4 instructions for
function address in JIT") for a similar fix.
Fixes: 69c087ba6225 ("bpf: Add bpf_for_each_map_elem() helper")
Fixes: bb035ef0cc91 ("LoongArch: BPF: Support mixing bpf2bpf and tailcalls")
Reported-by: Vincent Li <vincent.mc.li@xxxxxxxxx>
Closes: https://lore.kernel.org/loongarch/CAK3+h2yfM9FTNiXvEQBkvtuoJrvzmN4c_NZsFXqEk4Cj1tsBNA@xxxxxxxxxxxxxx/T/#u
Signed-off-by: Hengqi Chen <hengqi.chen@xxxxxxxxx>
---
arch/loongarch/net/bpf_jit.c | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/arch/loongarch/net/bpf_jit.c b/arch/loongarch/net/bpf_jit.c
index ea357a3edc09..b25b0bb43428 100644
--- a/arch/loongarch/net/bpf_jit.c
+++ b/arch/loongarch/net/bpf_jit.c
@@ -930,7 +930,10 @@ static int build_insn(const struct bpf_insn *insn, struct jit_ctx *ctx, bool ext
{
const u64 imm64 = (u64)(insn + 1)->imm << 32 | (u32)insn->imm;
- move_imm(ctx, dst, imm64, is32);
+ if (bpf_pseudo_func(insn))
+ move_addr(ctx, dst, imm64);
+ else
+ move_imm(ctx, dst, imm64, is32);
return 1;
}
--
2.43.5
