Re: [PATCH 03/13] powerpc/bpf: Update ldimm64 instructions during extra pass

[Christophe Leroy <christophe.leroy@csgroup.eu>]

Le 06/01/2022 à 12:45, Naveen N. Rao a écrit :
> These instructions are updated after the initial JIT, so redo codegen
> during the extra pass. Rename bpf_jit_fixup_subprog_calls() to clarify
> that this is more than just subprog calls.
> 
> Fixes: 69c087ba6225b5 ("bpf: Add bpf_for_each_map_elem() helper")
> Cc: stable@vger.kernel.org # v5.15
> Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
> ---
>   arch/powerpc/net/bpf_jit_comp.c   | 29 +++++++++++++++++++++++------
>   arch/powerpc/net/bpf_jit_comp32.c |  6 ++++++
>   arch/powerpc/net/bpf_jit_comp64.c |  7 ++++++-
>   3 files changed, 35 insertions(+), 7 deletions(-)
> 
> diff --git a/arch/powerpc/net/bpf_jit_comp.c b/arch/powerpc/net/bpf_jit_comp.c
> index d6ffdd0f2309d0..56dd1f4e3e4447 100644
> --- a/arch/powerpc/net/bpf_jit_comp.c
> +++ b/arch/powerpc/net/bpf_jit_comp.c
> @@ -23,15 +23,15 @@ static void bpf_jit_fill_ill_insns(void *area, unsigned int size)
>   	memset32(area, BREAKPOINT_INSTRUCTION, size / 4);
>   }
>   
> -/* Fix the branch target addresses for subprog calls */
> -static int bpf_jit_fixup_subprog_calls(struct bpf_prog *fp, u32 *image,
> -				       struct codegen_context *ctx, u32 *addrs)
> +/* Fix updated addresses (for subprog calls, ldimm64, et al) during extra pass */
> +static int bpf_jit_fixup_addresses(struct bpf_prog *fp, u32 *image,
> +				   struct codegen_context *ctx, u32 *addrs)
>   {
>   	const struct bpf_insn *insn = fp->insnsi;
>   	bool func_addr_fixed;
>   	u64 func_addr;
>   	u32 tmp_idx;
> -	int i, ret;
> +	int i, j, ret;
>   
>   	for (i = 0; i < fp->len; i++) {
>   		/*
> @@ -66,6 +66,23 @@ static int bpf_jit_fixup_subprog_calls(struct bpf_prog *fp, u32 *image,
>   			 * of the JITed sequence remains unchanged.
>   			 */
>   			ctx->idx = tmp_idx;
> +		} else if (insn[i].code == (BPF_LD | BPF_IMM | BPF_DW)) {
> +			tmp_idx = ctx->idx;
> +			ctx->idx = addrs[i] / 4;
> +#ifdef CONFIG_PPC32
> +			PPC_LI32(ctx->b2p[insn[i].dst_reg] - 1, (u32)insn[i + 1].imm);
> +			PPC_LI32(ctx->b2p[insn[i].dst_reg], (u32)insn[i].imm);
> +			for (j = ctx->idx - addrs[i] / 4; j < 4; j++)
> +				EMIT(PPC_RAW_NOP());
> +#else
> +			func_addr = ((u64)(u32)insn[i].imm) | (((u64)(u32)insn[i + 1].imm) << 32);
> +			PPC_LI64(b2p[insn[i].dst_reg], func_addr);
> +			/* overwrite rest with nops */
> +			for (j = ctx->idx - addrs[i] / 4; j < 5; j++)
> +				EMIT(PPC_RAW_NOP());
> +#endif

#ifdefs should be avoided as much as possible.

Here it seems we could easily do an

	if (IS_ENABLED(CONFIG_PPC32)) {
	} else {
	}

And it looks like the CONFIG_PPC64 alternative would in fact also work 
on PPC32, wouldn't it ?


> +			ctx->idx = tmp_idx;
> +			i++;
>   		}
>   	}
>   
> @@ -200,13 +217,13 @@ struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *fp)
>   		/*
>   		 * Do not touch the prologue and epilogue as they will remain
>   		 * unchanged. Only fix the branch target address for subprog
> -		 * calls in the body.
> +		 * calls in the body, and ldimm64 instructions.
>   		 *
>   		 * This does not change the offsets and lengths of the subprog
>   		 * call instruction sequences and hence, the size of the JITed
>   		 * image as well.
>   		 */
> -		bpf_jit_fixup_subprog_calls(fp, code_base, &cgctx, addrs);
> +		bpf_jit_fixup_addresses(fp, code_base, &cgctx, addrs);
>   
>   		/* There is no need to perform the usual passes. */
>   		goto skip_codegen_passes;
> diff --git a/arch/powerpc/net/bpf_jit_comp32.c b/arch/powerpc/net/bpf_jit_comp32.c
> index 997a47fa615b30..2258d3886d02ec 100644
> --- a/arch/powerpc/net/bpf_jit_comp32.c
> +++ b/arch/powerpc/net/bpf_jit_comp32.c
> @@ -293,6 +293,8 @@ int bpf_jit_build_body(struct bpf_prog *fp, u32 *image, struct codegen_context *
>   		bool func_addr_fixed;
>   		u64 func_addr;
>   		u32 true_cond;
> +		u32 tmp_idx;
> +		int j;
>   
>   		/*
>   		 * addrs[] maps a BPF bytecode address into a real offset from
> @@ -908,8 +910,12 @@ int bpf_jit_build_body(struct bpf_prog *fp, u32 *image, struct codegen_context *
>   		 * 16 byte instruction that uses two 'struct bpf_insn'
>   		 */
>   		case BPF_LD | BPF_IMM | BPF_DW: /* dst = (u64) imm */
> +			tmp_idx = ctx->idx;
>   			PPC_LI32(dst_reg_h, (u32)insn[i + 1].imm);
>   			PPC_LI32(dst_reg, (u32)insn[i].imm);
> +			/* padding to allow full 4 instructions for later patching */
> +			for (j = ctx->idx - tmp_idx; j < 4; j++)
> +				EMIT(PPC_RAW_NOP());
>   			/* Adjust for two bpf instructions */
>   			addrs[++i] = ctx->idx * 4;
>   			break;
> diff --git a/arch/powerpc/net/bpf_jit_comp64.c b/arch/powerpc/net/bpf_jit_comp64.c
> index 472d4a551945dd..3d018ecc475b2b 100644
> --- a/arch/powerpc/net/bpf_jit_comp64.c
> +++ b/arch/powerpc/net/bpf_jit_comp64.c
> @@ -319,6 +319,7 @@ int bpf_jit_build_body(struct bpf_prog *fp, u32 *image, struct codegen_context *
>   		u64 imm64;
>   		u32 true_cond;
>   		u32 tmp_idx;
> +		int j;
>   
>   		/*
>   		 * addrs[] maps a BPF bytecode address into a real offset from
> @@ -848,9 +849,13 @@ int bpf_jit_build_body(struct bpf_prog *fp, u32 *image, struct codegen_context *
>   		case BPF_LD | BPF_IMM | BPF_DW: /* dst = (u64) imm */
>   			imm64 = ((u64)(u32) insn[i].imm) |
>   				    (((u64)(u32) insn[i+1].imm) << 32);
> +			tmp_idx = ctx->idx;
> +			PPC_LI64(dst_reg, imm64);
> +			/* padding to allow full 5 instructions for later patching */
> +			for (j = ctx->idx - tmp_idx; j < 5; j++)
> +				EMIT(PPC_RAW_NOP());
>   			/* Adjust for two bpf instructions */
>   			addrs[++i] = ctx->idx * 4;
> -			PPC_LI64(dst_reg, imm64);
>   			break;
>   
>   		/*