bpf: Support 64-bit pointers to kfuncs
Subsystems (4)
Ilya Leoshkevich
maintainerPATCH v4 1/1 bpf: Support 64-bit pointers to kfuncs
test_ksyms_module fails to emit a kfunc call targeting a module on s390x, because the verifier stores the difference between kfunc address and __bpf_call_base in bpf_insn.imm, which is s32, and modules are roughly (1 << 42) bytes away from the kernel on s390x.
Fix by keeping BTF id in bpf_insn.imm for BPF_PSEUDO_KFUNC_CALLs, and storing the absolute address in bpf_kfunc_desc.
Introduce bpf_jit_supports_far_kfunc_call() in order to limit this new behavior to the s390x JIT. Otherwise other JITs need to be modified, which is not desired.
Introduce bpf_get_kfunc_addr() instead of exposing both find_kfunc_desc() and struct bpf_kfunc_desc.
In addition to sorting kfuncs by imm, also sort them by offset, in order to handle conflicting imms from different modules. Do this on all architectures in order to simplify code.
Factor out resolving specialized kfuncs (XPD and dynptr) from fixup_kfunc_call(). This was required in the first place, because fixup_kfunc_call() uses find_kfunc_desc(), which returns a const pointer, so it's not possible to modify kfunc addr without stripping const, which is not nice. It also removes repetition of code like:
if (bpf_jit_supports_far_kfunc_call()) desc->addr = func; else insn->imm = BPF_CALL_IMM(func);
and separates kfunc_desc_tab fixups from kfunc_call fixups.
---
v3: https://lore.kernel.org/bpf/[email protected]/
v3 -> v4: Use Jiri's proposal and make it work on s390x.
arch/s390/net/bpf_jit_comp.c | 5 ++ include/linux/bpf.h | 2 + include/linux/filter.h | 1 + kernel/bpf/core.c | 11 ++++ kernel/bpf/verifier.c | 116 +++++++++++++++++++++++++---------- 5 files changed, 101 insertions(+), 34 deletions(-)
arch/s390/net/bpf_jit_comp.c +5 -0
@@ -2001,6 +2001,11 @@ bool bpf_jit_supports_kfunc_call(void)return true; } +bool bpf_jit_supports_far_kfunc_call (void ) +{ +return true; +} +int bpf_arch_text_poke (void *ip,enum bpf_text_poke_type t,void *old_addr,void *new_addr) {
include/linux/bpf.h +2 -0
@@ -2296,6 +2296,8 @@ bool bpf_prog_has_kfunc_call(const struct bpf_prog *prog);const struct btf_func_model *bpf_jit_find_kfunc_model (const struct bpf_prog *prog,const struct bpf_insn *insn); +int bpf_get_kfunc_addr (const struct bpf_prog *prog,u32 func_id,u16 offset, +u8 **func_addr);struct bpf_core_ctx {struct bpf_verifier_log *log;const struct btf *btf;
include/linux/filter.h +1 -0
@@ -921,6 +921,7 @@ void bpf_jit_compile(struct bpf_prog *prog);bool bpf_jit_needs_zext (void );bool bpf_jit_supports_subprog_tailcalls (void );bool bpf_jit_supports_kfunc_call (void ); +bool bpf_jit_supports_far_kfunc_call (void );bool bpf_helper_changes_pkt_data (void *func);static inline bool bpf_dump_raw_ok (const struct cred *cred)
kernel/bpf/core.c +11 -0
@@ -1187,6 +1187,7 @@ int bpf_jit_get_func_addr(const struct bpf_prog *prog,s16 off = insn->off;s32 imm = insn->imm;u8 *addr; +int err; *func_addr_fixed = insn->src_reg !=BPF_PSEUDO_CALL ;if (!*func_addr_fixed) {
@@ -1201,6 +1202,11 @@ int bpf_jit_get_func_addr(const struct bpf_prog *prog, addr = (u8 *)prog->aux->func[off]->bpf_func; else return -EINVAL; + }else if (insn->src_reg ==BPF_PSEUDO_KFUNC_CALL && +bpf_jit_supports_far_kfunc_call ()) { + err =bpf_get_kfunc_addr (prog, insn->imm, insn->off, &addr); +if (err) +return err; }else {/* Address of a BPF helper call. Since part of the core * kernel, it's always at a fixed location. __bpf_call_base
@@ -2732,6 +2738,11 @@ bool __weak bpf_jit_supports_kfunc_call(void)return false; } +bool __weak bpf_jit_supports_far_kfunc_call (void ) +{ +return false; +} +/* To execute LD_ABS/LD_IND instructions __bpf_prog_run() may call * skb_copy_bits(), so provide a weak definition of it for NET-less config. */
kernel/bpf/verifier.c +82 -34
216 lines not on this page · raw message
Signature · 1 line
2.39.2
Quoted 47 lines · Ilya Leoshkevich
test_ksyms_modulefails to emit a kfunc call targeting a module on s390x, because the verifier stores the difference between kfunc address and__bpf_call_basein bpf_insn.imm, which is s32, and modules are roughly (1 << 42) bytes away from the kernel on s390x.Fix by keeping BTF id in bpf_insn.imm for
BPF_PSEUDO_KFUNC_CALLs, and storing the absolute address inbpf_kfunc_desc.Introduce
bpf_jit_supports_far_kfunc_call()in order to limit this new behavior to the s390x JIT. Otherwise other JITs need to be modified, which is not desired.Introduce
bpf_get_kfunc_addr()instead of exposing bothfind_kfunc_desc()and structbpf_kfunc_desc.In addition to sorting kfuncs by imm, also sort them by offset, in order to handle conflicting imms from different modules. Do this on all architectures in order to simplify code.
Factor out resolving specialized kfuncs (XPD and dynptr) from
fixup_kfunc_call(). This was required in the first place, becausefixup_kfunc_call()usesfind_kfunc_desc(), which returns a const pointer, so it's not possible to modify kfunc addr without stripping const, which is not nice. It also removes repetition of code like:if (bpf_jit_supports_far_kfunc_call()) desc->addr = func; else insn->imm = BPF_CALL_IMM(func);and separates
kfunc_desc_tabfixups fromkfunc_callfixups.Suggested-by Jiri Olsa <[email protected]>Signed-off-by Ilya Leoshkevich <[email protected]>---
v3: https://lore.kernel.org/bpf/[email protected]/
v3 -> v4: Use Jiri's proposal and make it work on s390x.arch/s390/net/bpf_jit_comp.c | 5 ++ include/linux/bpf.h | 2 + include/linux/filter.h | 1 + kernel/bpf/core.c | 11 ++++ kernel/bpf/verifier.c | 116 +++++++++++++++++++++++++---------- 5 files changed, 101 insertions(+), 34 deletions(-)
[...]
Quoted 13 lines
kernel/bpf/verifier.c +5
-0@@ -2452,6 +2453,11 @@ struct bpf_kfunc_btf { };struct bpf_kfunc_desc_tab { +/* Sorted by func_id (BTF ID) and offset (fd_array offset) during +* verification. JITs do lookups by bpf_insn, where func_id may not be +* available, therefore at the end of verification do_misc_fixups() +* sorts this by imm and offset. +*/ struct bpf_kfunc_desc descs[MAX_KFUNC_DESCS ];u32 nr_descs; };
kernel/bpf/verifier.c find_kfunc_desc(const struct bpf_prog *prog, u32 func_id, u16 offset) · line 2501
Quoted 4 lines
kernel/bpf/verifier.c +1
-0@@ -2492,6 +2498,19 @@ find_kfunc_desc(const struct bpf_prog *prog, u32 func_id, u16 offset) sizeof(tab->descs[0]), kfunc_desc_cmp_by_id_off); } +int bpf_get_kfunc_addr(const struct bpf_prog *prog, u32 func_id, u16 offset,
nit: offset is so generic, if it's fd array index, should we call it
that: btf_fd_idx or something like that?
Quoted 16 lines
kernel/bpf/verifier.c +12
-0+ u8 **func_addr) +{ +const struct bpf_kfunc_desc *desc; + + desc =find_kfunc_desc (prog, func_id, offset); +if (!desc) +return -EFAULT ; + + *func_addr = (u8 *)desc->addr; +return 0 ; +} +static struct btf *__find_kfunc_desc_btf (struct bpf_verifier_env *env,s16 offset) {
kernel/bpf/verifier.c static int add_kfunc_call(struct bpf_verifier_env *env, u32 func_id, s16 offset) · line 2698
Quoted 9 lines
kernel/bpf/verifier.c +2
-1@@ -2672,7 +2691,8 @@ static int add_kfunc_call(struct bpf_verifier_env *env, u32 func_id, s16 offset)return -EINVAL ; } - call_imm =BPF_CALL_IMM (addr); + call_imm =bpf_jit_supports_far_kfunc_call () ? func_id : +BPF_CALL_IMM (addr);/* Check whether or not the relative offset overflows desc->imm */ if ((unsigned long )(s32 )call_imm != call_imm) { verbose(env, "address of kernel function %s is out of range\n",
this check makes no sense and would have misleading message if bpf_jit_supports_far_kfunc_call(), so how about actually making it a proper if/else and doing check only if call_imm is a 32-bit address offset?
Quoted 14 lines
kernel/bpf/verifier.c +2
-1@@ -2690,6 +2710,7 @@ static int add_kfunc_call(struct bpf_verifier_env *env, u32 func_id, s16 offset) desc->func_id = func_id; desc->imm = call_imm; desc->offset = offset; + desc->addr = addr; err =btf_distill_func_proto (&env->log, desc_btf, func_proto, func_name, &desc->func_model);@@ -2699,19 +2720,15 @@ static int add_kfunc_call(struct bpf_verifier_env *env, u32 func_id, s16 offset)return err; } -static int kfunc_desc_cmp_by_imm (const void *a,const void *b) +static int kfunc_desc_cmp_by_imm_off (const void *a,const void *b)
kernel/bpf/verifier.c static int add_kfunc_call(struct bpf_verifier_env *env, u32 func_id, s16 offset) · line 2728
Quoted 10 lines
kernel/bpf/verifier.c +1
-5{const struct bpf_kfunc_desc *d0 = a;const struct bpf_kfunc_desc *d1 = b; -if (d0->imm > d1->imm) -return 1 ; -else if (d0->imm < d1->imm) -return -1 ; -return 0 ; +return d0->imm - d1->imm ?: d0->offset - d1->offset;
very succinct, but does it work for all possible inputs? let's see:
$ cat t.c #include <stdio.h> #include <limits.h>
int main() {
int x1 = INT_MAX;
int x2 = INT_MIN;
int d1 = x1 - x2;
int d2 = x2 - x1;
printf("x1 %d x2 %d d1 %d d2 %d\n", x1, x2, d1, d2);
return 0;
}
$ cc t.c && ./a.out
x1 2147483647 x2 -2147483648 d1 -1 d2 1I believe d1 should be positive and d2 should be negative, though, right? So maybe let's not be too clever here and just do:
if (d0->imm != d1->imm)
return d0->imm < d10>imm ? -1 : 1:
if (d0->off != d1->off)
return d0->off < d10>off ? -1 : 1;return 0;
Still succinct enough, I think.
Quoted 7 lines
kernel/bpf/verifier.c +1
-1} -static void sort_kfunc_descs_by_imm (struct bpf_prog *prog) +static void sort_kfunc_descs_by_imm_off (struct bpf_prog *prog) {struct bpf_kfunc_desc_tab *tab;
[...]
Quoted 14 lines
static int fixup_kfunc_call(struct
bpf_verifier_env*env, structbpf_insn*insn,
structbpf_insn*insn_buf, intinsn_idx, int *cnt){ const struct bpf_kfunc_desc *desc;kernel/bpf/verifier.c +0
-3-void *xdp_kfunc;if (!insn->imm) {verbose (env,"invalid kernel function call not eliminated in verifier pass\n" );@@ -17306,16 +17372,6 @@ static int fixup_kfunc_call(struct bpf_verifier_env *env, struct bpf_insn *insn, *cnt =0 ; -if (bpf_dev_bound_kfunc_id (insn->imm)) { - xdp_kfunc =bpf_dev_bound_resolve_kfunc (env->prog, insn->imm);
kernel/bpf/verifier.c static int fixup_kfunc_call(struct bpf_verifier_env *env, struct bpf_insn *insn, · line 17376
Quoted 10 lines
kernel/bpf/verifier.c +0
-8-if (xdp_kfunc) { - insn->imm =BPF_CALL_IMM (xdp_kfunc); -return 0 ; - } - -/* fallback to default kfunc when not supported by netdev */ - } -/* insn->imm has the btf func_id. Replace it with * an address (relative to __bpf_call_base).
update comment mentioning bpf_jit_supports_far_kfunc_call() ?
Quoted 50 lines
kernel/bpf/verifier.c +5
-14*/ @@ -17326,7 +17382,8 @@ static int fixup_kfunc_call(struct bpf_verifier_env *env, struct bpf_insn *insn,return -EFAULT ; } - insn->imm = desc->imm; +if (!bpf_jit_supports_far_kfunc_call ()) + insn->imm =BPF_CALL_IMM (desc->addr);if (insn->off)return 0 ;if (desc->func_id == special_kfunc_list[KF_bpf_obj_new_impl]) {@@ -17351,17 +17408,6 @@ static int fixup_kfunc_call(struct bpf_verifier_env *env, struct bpf_insn *insn, desc->func_id == special_kfunc_list[KF_bpf_rdonly_cast]) { insn_buf[0 ] =BPF_MOV64_REG (BPF_REG_0 ,BPF_REG_1 ); *cnt =1 ; - } else if (desc->func_id == special_kfunc_list[KF_bpf_dynptr_from_skb]) { -bool seen_direct_write = env->seen_direct_write; -bool is_rdonly = !may_access_direct_pkt_data (env,NULL ,BPF_WRITE ); - -if (is_rdonly) - insn->imm =BPF_CALL_IMM (bpf_dynptr_from_skb_rdonly); - -/* restore env->seen_direct_write to its original value, since -* may_access_direct_pkt_data mutates it -*/ - env->seen_direct_write = seen_direct_write; }return 0 ; }@@ -17384,6 +17430,8 @@ static int do_misc_fixups(struct bpf_verifier_env *env)struct bpf_map *map_ptr;int i, ret, cnt, delta =0 ; +fixup_kfunc_desc_tab (env); +for (i =0 ; i < insn_cnt; i++, insn++) {/* Make divide-by-zero exceptions impossible. */ if (insn->code == (BPF_ALU64 | BPF_MOD | BPF_X) ||@@ -17891,7 +17939,7 @@ static int do_misc_fixups(struct bpf_verifier_env *env) } } -sort_kfunc_descs_by_imm (env->prog); +sort_kfunc_descs_by_imm_off (env->prog);return 0 ; } --2.39.2