On Wed, Nov 24, 2021 at 2:43 PM Andrii Nakryiko
[off-list ref] wrote:

quoted

+               /* Implement bpf_arg inline. */
+               if (prog_type == BPF_PROG_TYPE_TRACING &&
+                   insn->imm == BPF_FUNC_arg) {
+                       /* Load nr_args from ctx - 8 */
+                       insn_buf[0] = BPF_LDX_MEM(BPF_DW, BPF_REG_0, BPF_REG_1, -8);
+                       insn_buf[1] = BPF_JMP32_REG(BPF_JGE, BPF_REG_2, BPF_REG_0, 4);
+                       insn_buf[2] = BPF_ALU64_IMM(BPF_MUL, BPF_REG_2, 8);
+                       insn_buf[3] = BPF_ALU64_REG(BPF_ADD, BPF_REG_2, BPF_REG_1);
+                       insn_buf[4] = BPF_LDX_MEM(BPF_DW, BPF_REG_0, BPF_REG_2, 0);
+                       insn_buf[5] = BPF_JMP_A(1);
+                       insn_buf[6] = BPF_MOV64_IMM(BPF_REG_0, 0);
+
+                       new_prog = bpf_patch_insn_data(env, i + delta, insn_buf, 7);
+                       if (!new_prog)
+                               return -ENOMEM;
+
+                       delta    += 6;
+                       env->prog = prog = new_prog;
+                       insn      = new_prog->insnsi + i + delta;
+                       continue;

nit: this whole sequence of steps and calculations seems like
something that might be abstracted and hidden behind a macro or helper
func? Not related to your change, though. But wouldn't it be easier to
understand if it was just written as:

PATCH_INSNS(
    BPF_LDX_MEM(BPF_DW, BPF_REG_0, BPF_REG_1, -8);
    BPF_JMP32_REG(BPF_JGE, BPF_REG_2, BPF_REG_0, 4);
    BPF_ALU64_IMM(BPF_MUL, BPF_REG_2, 8);
    BPF_ALU64_REG(BPF_ADD, BPF_REG_2, BPF_REG_1);
    BPF_LDX_MEM(BPF_DW, BPF_REG_0, BPF_REG_2, 0);
    BPF_JMP_A(1);
    BPF_MOV64_IMM(BPF_REG_0, 0));

Daniel and myself tried to do similar macro magic in the past,
but it suffers unnecessary stack increase and extra copies.
So eventually we got rid of it.
I suggest staying with Jiri's approach.

Independent from anything else...
Just noticed BPF_MUL in the above...
Please use BPF_LSH instead. JITs don't optimize such things.
It's a job of gcc/llvm to do so. JITs assume that all
normal optimizations were done by the compiler.