LoongArch: BPF: Add dynamic code modification support

This commit adds support for BPF dynamic code modification on the
LoongArch architecture:
1. Add bpf_arch_text_copy() for instruction block copying.
2. Add bpf_arch_text_poke() for runtime instruction patching.
3. Add bpf_arch_text_invalidate() for code invalidation.

On LoongArch, since symbol addresses in the direct mapping region can't
be reached via relative jump instructions from the paged mapping region,
we use the move_imm+jirl instruction pair as absolute jump instructions.
These require 2-5 instructions, so we reserve 5 NOP instructions in the
program as placeholders for function jumps.

The larch_insn_text_copy() function is solely used for BPF. And the use
of larch_insn_text_copy() requires PAGE_SIZE alignment. Currently, only
the size of the BPF trampoline is page-aligned.

Co-developed-by: George Guo <guodongtai@kylinos.cn>
Signed-off-by: George Guo <guodongtai@kylinos.cn>
Signed-off-by: Chenghao Duan <duanchenghao@kylinos.cn>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>

arch/loongarch/include/asm/inst.h

@@ -497,6 +497,7 @@ void arch_simulate_insn(union loongarch_instruction insn, struct pt_regs *regs);
 int larch_insn_read(void *addr, u32 *insnp);
 int larch_insn_write(void *addr, u32 insn);
 int larch_insn_patch_text(void *addr, u32 insn);
+int larch_insn_text_copy(void *dst, void *src, size_t len);
 
 u32 larch_insn_gen_nop(void);
 u32 larch_insn_gen_b(unsigned long pc, unsigned long dest);

arch/loongarch/kernel/inst.c

@@ -4,6 +4,8 @@
  */
 #include <linux/sizes.h>
 #include <linux/uaccess.h>
+#include <linux/set_memory.h>
+#include <linux/stop_machine.h>
 
 #include <asm/cacheflush.h>
 #include <asm/inst.h>
@@ -218,6 +220,50 @@ int larch_insn_patch_text(void *addr, u32 insn)
 	return ret;
 }
 
+struct insn_copy {
+	void *dst;
+	void *src;
+	size_t len;
+	unsigned int cpu;
+};
+
+static int text_copy_cb(void *data)
+{
+	int ret = 0;
+	struct insn_copy *copy = data;
+
+	if (smp_processor_id() == copy->cpu) {
+		ret = copy_to_kernel_nofault(copy->dst, copy->src, copy->len);
+		if (ret)
+			pr_err("%s: operation failed\n", __func__);
+	}
+
+	flush_icache_range((unsigned long)copy->dst, (unsigned long)copy->dst + copy->len);
+
+	return ret;
+}
+
+int larch_insn_text_copy(void *dst, void *src, size_t len)
+{
+	int ret = 0;
+	size_t start, end;
+	struct insn_copy copy = {
+		.dst = dst,
+		.src = src,
+		.len = len,
+		.cpu = smp_processor_id(),
+	};
+
+	start = round_down((size_t)dst, PAGE_SIZE);
+	end   = round_up((size_t)dst + len, PAGE_SIZE);
+
+	set_memory_rw(start, (end - start) / PAGE_SIZE);
+	ret = stop_machine(text_copy_cb, &copy, cpu_online_mask);
+	set_memory_rox(start, (end - start) / PAGE_SIZE);
+
+	return ret;
+}
+
 u32 larch_insn_gen_nop(void)
 {
 	return INSN_NOP;

arch/loongarch/net/bpf_jit.c

@@ -4,8 +4,12 @@
  *
  * Copyright (C) 2022 Loongson Technology Corporation Limited
  */
+#include <linux/memory.h>
 #include "bpf_jit.h"
 
+#define LOONGARCH_LONG_JUMP_NINSNS 5
+#define LOONGARCH_LONG_JUMP_NBYTES (LOONGARCH_LONG_JUMP_NINSNS * 4)
+
 #define REG_TCC		LOONGARCH_GPR_A6
 #define TCC_SAVED	LOONGARCH_GPR_S5
 
@@ -88,7 +92,7 @@ static u8 tail_call_reg(struct jit_ctx *ctx)
  */
 static void build_prologue(struct jit_ctx *ctx)
 {
-	int stack_adjust = 0, store_offset, bpf_stack_adjust;
+	int i, stack_adjust = 0, store_offset, bpf_stack_adjust;
 
 	bpf_stack_adjust = round_up(ctx->prog->aux->stack_depth, 16);
 
@@ -98,6 +102,10 @@ static void build_prologue(struct jit_ctx *ctx)
 	stack_adjust = round_up(stack_adjust, 16);
 	stack_adjust += bpf_stack_adjust;
 
+	/* Reserve space for the move_imm + jirl instruction */
+	for (i = 0; i < LOONGARCH_LONG_JUMP_NINSNS; i++)
+		emit_insn(ctx, nop);
+
 	/*
 	 * First instruction initializes the tail call count (TCC).
 	 * On tail call we skip this instruction, and the TCC is
@@ -1194,6 +1202,101 @@ static int validate_ctx(struct jit_ctx *ctx)
 	return 0;
 }
 
+static int emit_jump_and_link(struct jit_ctx *ctx, u8 rd, u64 target)
+{
+	if (!target) {
+		pr_err("bpf_jit: jump target address is error\n");
+		return -EFAULT;
+	}
+
+	move_imm(ctx, LOONGARCH_GPR_T1, target, false);
+	emit_insn(ctx, jirl, rd, LOONGARCH_GPR_T1, 0);
+
+	return 0;
+}
+
+static int emit_jump_or_nops(void *target, void *ip, u32 *insns, bool is_call)
+{
+	int i;
+	struct jit_ctx ctx;
+
+	ctx.idx = 0;
+	ctx.image = (union loongarch_instruction *)insns;
+
+	if (!target) {
+		for (i = 0; i < LOONGARCH_LONG_JUMP_NINSNS; i++)
+			emit_insn((&ctx), nop);
+		return 0;
+	}
+
+	return emit_jump_and_link(&ctx, is_call ? LOONGARCH_GPR_T0 : LOONGARCH_GPR_ZERO, (u64)target);
+}
+
+void *bpf_arch_text_copy(void *dst, void *src, size_t len)
+{
+	int ret;
+
+	mutex_lock(&text_mutex);
+	ret = larch_insn_text_copy(dst, src, len);
+	mutex_unlock(&text_mutex);
+
+	return ret ? ERR_PTR(-EINVAL) : dst;
+}
+
+int bpf_arch_text_poke(void *ip, enum bpf_text_poke_type poke_type,
+		       void *old_addr, void *new_addr)
+{
+	int ret;
+	bool is_call = (poke_type == BPF_MOD_CALL);
+	u32 old_insns[LOONGARCH_LONG_JUMP_NINSNS] = {[0 ... 4] = INSN_NOP};
+	u32 new_insns[LOONGARCH_LONG_JUMP_NINSNS] = {[0 ... 4] = INSN_NOP};
+
+	if (!is_kernel_text((unsigned long)ip) &&
+		!is_bpf_text_address((unsigned long)ip))
+		return -ENOTSUPP;
+
+	ret = emit_jump_or_nops(old_addr, ip, old_insns, is_call);
+	if (ret)
+		return ret;
+
+	if (memcmp(ip, old_insns, LOONGARCH_LONG_JUMP_NBYTES))
+		return -EFAULT;
+
+	ret = emit_jump_or_nops(new_addr, ip, new_insns, is_call);
+	if (ret)
+		return ret;
+
+	mutex_lock(&text_mutex);
+	if (memcmp(ip, new_insns, LOONGARCH_LONG_JUMP_NBYTES))
+		ret = larch_insn_text_copy(ip, new_insns, LOONGARCH_LONG_JUMP_NBYTES);
+	mutex_unlock(&text_mutex);
+
+	return ret;
+}
+
+int bpf_arch_text_invalidate(void *dst, size_t len)
+{
+	int i;
+	int ret = 0;
+	u32 *inst;
+
+	inst = kvmalloc(len, GFP_KERNEL);
+	if (!inst)
+		return -ENOMEM;
+
+	for (i = 0; i < (len / sizeof(u32)); i++)
+		inst[i] = INSN_BREAK;
+
+	mutex_lock(&text_mutex);
+	if (larch_insn_text_copy(dst, inst, len))
+		ret = -EINVAL;
+	mutex_unlock(&text_mutex);
+
+	kvfree(inst);
+
+	return ret;
+}
+
 struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *prog)
 {
 	bool tmp_blinded = false, extra_pass = false;