linux-loongson/arch/loongarch/kernel/inst.c
Chenghao Duan 9fbd18cf4c 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>
2025-08-05 19:00:18 +08:00

413 lines
8.9 KiB
C

// SPDX-License-Identifier: GPL-2.0
/*
* Copyright (C) 2020-2022 Loongson Technology Corporation Limited
*/
#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>
static DEFINE_RAW_SPINLOCK(patch_lock);
void simu_pc(struct pt_regs *regs, union loongarch_instruction insn)
{
unsigned long pc = regs->csr_era;
unsigned int rd = insn.reg1i20_format.rd;
unsigned int imm = insn.reg1i20_format.immediate;
if (pc & 3) {
pr_warn("%s: invalid pc 0x%lx\n", __func__, pc);
return;
}
switch (insn.reg1i20_format.opcode) {
case pcaddi_op:
regs->regs[rd] = pc + sign_extend64(imm << 2, 21);
break;
case pcaddu12i_op:
regs->regs[rd] = pc + sign_extend64(imm << 12, 31);
break;
case pcaddu18i_op:
regs->regs[rd] = pc + sign_extend64(imm << 18, 37);
break;
case pcalau12i_op:
regs->regs[rd] = pc + sign_extend64(imm << 12, 31);
regs->regs[rd] &= ~((1 << 12) - 1);
break;
default:
pr_info("%s: unknown opcode\n", __func__);
return;
}
regs->csr_era += LOONGARCH_INSN_SIZE;
}
void simu_branch(struct pt_regs *regs, union loongarch_instruction insn)
{
unsigned int imm, imm_l, imm_h, rd, rj;
unsigned long pc = regs->csr_era;
if (pc & 3) {
pr_warn("%s: invalid pc 0x%lx\n", __func__, pc);
return;
}
imm_l = insn.reg0i26_format.immediate_l;
imm_h = insn.reg0i26_format.immediate_h;
switch (insn.reg0i26_format.opcode) {
case b_op:
regs->csr_era = pc + sign_extend64((imm_h << 16 | imm_l) << 2, 27);
return;
case bl_op:
regs->csr_era = pc + sign_extend64((imm_h << 16 | imm_l) << 2, 27);
regs->regs[1] = pc + LOONGARCH_INSN_SIZE;
return;
}
imm_l = insn.reg1i21_format.immediate_l;
imm_h = insn.reg1i21_format.immediate_h;
rj = insn.reg1i21_format.rj;
switch (insn.reg1i21_format.opcode) {
case beqz_op:
if (regs->regs[rj] == 0)
regs->csr_era = pc + sign_extend64((imm_h << 16 | imm_l) << 2, 22);
else
regs->csr_era = pc + LOONGARCH_INSN_SIZE;
return;
case bnez_op:
if (regs->regs[rj] != 0)
regs->csr_era = pc + sign_extend64((imm_h << 16 | imm_l) << 2, 22);
else
regs->csr_era = pc + LOONGARCH_INSN_SIZE;
return;
}
imm = insn.reg2i16_format.immediate;
rj = insn.reg2i16_format.rj;
rd = insn.reg2i16_format.rd;
switch (insn.reg2i16_format.opcode) {
case beq_op:
if (regs->regs[rj] == regs->regs[rd])
regs->csr_era = pc + sign_extend64(imm << 2, 17);
else
regs->csr_era = pc + LOONGARCH_INSN_SIZE;
break;
case bne_op:
if (regs->regs[rj] != regs->regs[rd])
regs->csr_era = pc + sign_extend64(imm << 2, 17);
else
regs->csr_era = pc + LOONGARCH_INSN_SIZE;
break;
case blt_op:
if ((long)regs->regs[rj] < (long)regs->regs[rd])
regs->csr_era = pc + sign_extend64(imm << 2, 17);
else
regs->csr_era = pc + LOONGARCH_INSN_SIZE;
break;
case bge_op:
if ((long)regs->regs[rj] >= (long)regs->regs[rd])
regs->csr_era = pc + sign_extend64(imm << 2, 17);
else
regs->csr_era = pc + LOONGARCH_INSN_SIZE;
break;
case bltu_op:
if (regs->regs[rj] < regs->regs[rd])
regs->csr_era = pc + sign_extend64(imm << 2, 17);
else
regs->csr_era = pc + LOONGARCH_INSN_SIZE;
break;
case bgeu_op:
if (regs->regs[rj] >= regs->regs[rd])
regs->csr_era = pc + sign_extend64(imm << 2, 17);
else
regs->csr_era = pc + LOONGARCH_INSN_SIZE;
break;
case jirl_op:
regs->csr_era = regs->regs[rj] + sign_extend64(imm << 2, 17);
regs->regs[rd] = pc + LOONGARCH_INSN_SIZE;
break;
default:
pr_info("%s: unknown opcode\n", __func__);
return;
}
}
bool insns_not_supported(union loongarch_instruction insn)
{
switch (insn.reg3_format.opcode) {
case amswapw_op ... ammindbdu_op:
pr_notice("atomic memory access instructions are not supported\n");
return true;
}
switch (insn.reg2i14_format.opcode) {
case llw_op:
case lld_op:
case scw_op:
case scd_op:
pr_notice("ll and sc instructions are not supported\n");
return true;
}
switch (insn.reg1i21_format.opcode) {
case bceqz_op:
pr_notice("bceqz and bcnez instructions are not supported\n");
return true;
}
return false;
}
bool insns_need_simulation(union loongarch_instruction insn)
{
if (is_pc_ins(&insn))
return true;
if (is_branch_ins(&insn))
return true;
return false;
}
void arch_simulate_insn(union loongarch_instruction insn, struct pt_regs *regs)
{
if (is_pc_ins(&insn))
simu_pc(regs, insn);
else if (is_branch_ins(&insn))
simu_branch(regs, insn);
}
int larch_insn_read(void *addr, u32 *insnp)
{
int ret;
u32 val;
ret = copy_from_kernel_nofault(&val, addr, LOONGARCH_INSN_SIZE);
if (!ret)
*insnp = val;
return ret;
}
int larch_insn_write(void *addr, u32 insn)
{
int ret;
unsigned long flags = 0;
raw_spin_lock_irqsave(&patch_lock, flags);
ret = copy_to_kernel_nofault(addr, &insn, LOONGARCH_INSN_SIZE);
raw_spin_unlock_irqrestore(&patch_lock, flags);
return ret;
}
int larch_insn_patch_text(void *addr, u32 insn)
{
int ret;
u32 *tp = addr;
if ((unsigned long)tp & 3)
return -EINVAL;
ret = larch_insn_write(tp, insn);
if (!ret)
flush_icache_range((unsigned long)tp,
(unsigned long)tp + LOONGARCH_INSN_SIZE);
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;
}
u32 larch_insn_gen_b(unsigned long pc, unsigned long dest)
{
long offset = dest - pc;
union loongarch_instruction insn;
if ((offset & 3) || offset < -SZ_128M || offset >= SZ_128M) {
pr_warn("The generated b instruction is out of range.\n");
return INSN_BREAK;
}
emit_b(&insn, offset >> 2);
return insn.word;
}
u32 larch_insn_gen_bl(unsigned long pc, unsigned long dest)
{
long offset = dest - pc;
union loongarch_instruction insn;
if ((offset & 3) || offset < -SZ_128M || offset >= SZ_128M) {
pr_warn("The generated bl instruction is out of range.\n");
return INSN_BREAK;
}
emit_bl(&insn, offset >> 2);
return insn.word;
}
u32 larch_insn_gen_break(int imm)
{
union loongarch_instruction insn;
if (imm < 0 || imm >= SZ_32K) {
pr_warn("The generated break instruction is out of range.\n");
return INSN_BREAK;
}
emit_break(&insn, imm);
return insn.word;
}
u32 larch_insn_gen_or(enum loongarch_gpr rd, enum loongarch_gpr rj, enum loongarch_gpr rk)
{
union loongarch_instruction insn;
emit_or(&insn, rd, rj, rk);
return insn.word;
}
u32 larch_insn_gen_move(enum loongarch_gpr rd, enum loongarch_gpr rj)
{
return larch_insn_gen_or(rd, rj, 0);
}
u32 larch_insn_gen_lu12iw(enum loongarch_gpr rd, int imm)
{
union loongarch_instruction insn;
if (imm < -SZ_512K || imm >= SZ_512K) {
pr_warn("The generated lu12i.w instruction is out of range.\n");
return INSN_BREAK;
}
emit_lu12iw(&insn, rd, imm);
return insn.word;
}
u32 larch_insn_gen_lu32id(enum loongarch_gpr rd, int imm)
{
union loongarch_instruction insn;
if (imm < -SZ_512K || imm >= SZ_512K) {
pr_warn("The generated lu32i.d instruction is out of range.\n");
return INSN_BREAK;
}
emit_lu32id(&insn, rd, imm);
return insn.word;
}
u32 larch_insn_gen_lu52id(enum loongarch_gpr rd, enum loongarch_gpr rj, int imm)
{
union loongarch_instruction insn;
if (imm < -SZ_2K || imm >= SZ_2K) {
pr_warn("The generated lu52i.d instruction is out of range.\n");
return INSN_BREAK;
}
emit_lu52id(&insn, rd, rj, imm);
return insn.word;
}
u32 larch_insn_gen_beq(enum loongarch_gpr rd, enum loongarch_gpr rj, int imm)
{
union loongarch_instruction insn;
if ((imm & 3) || imm < -SZ_128K || imm >= SZ_128K) {
pr_warn("The generated beq instruction is out of range.\n");
return INSN_BREAK;
}
emit_beq(&insn, rj, rd, imm >> 2);
return insn.word;
}
u32 larch_insn_gen_bne(enum loongarch_gpr rd, enum loongarch_gpr rj, int imm)
{
union loongarch_instruction insn;
if ((imm & 3) || imm < -SZ_128K || imm >= SZ_128K) {
pr_warn("The generated bne instruction is out of range.\n");
return INSN_BREAK;
}
emit_bne(&insn, rj, rd, imm >> 2);
return insn.word;
}
u32 larch_insn_gen_jirl(enum loongarch_gpr rd, enum loongarch_gpr rj, int imm)
{
union loongarch_instruction insn;
if ((imm & 3) || imm < -SZ_128K || imm >= SZ_128K) {
pr_warn("The generated jirl instruction is out of range.\n");
return INSN_BREAK;
}
emit_jirl(&insn, rd, rj, imm >> 2);
return insn.word;
}