[PATCH bpf-next v1 0/4] bpf: allow void* cast using bpf_rdonly_cast()

[PATCH bpf-next v1 0/4] bpf: allow void* cast using bpf_rdonly_cast()

[Eduard Zingerman]

At the moment pointers returned by bpf_rdonly_cast() have type
"pointer to btf id", and only casts to structure types are allowed.
Access to memory pointed to by returned pointers is done through
BPF_PROBE_{MEM,MEMSX} instructions and does not produce errors on
invalid memory access.

This patch-set extends bpf_rdonly_cast() to allow casts to an
equivalent of 'void *', effectively replacing bpf_probe_read_kernel()
calls in situations when access to individual bytes or integers is
necessary.

The mechanism was suggested and explored by Andrii Nakryiko in [1].

To help with detecting support for this feature 'enum bpf_features' is
added with intended usage like below:

  if (bpf_core_enum_value_exists(enum bpf_features,
                                 BPF_FEAT_RDONLY_CAST_TO_VOID))
    ...

[1] https://github.com/anakryiko/linux/tree/bpf-mem-cast

Eduard Zingerman (4):
  bpf: allow void* cast using bpf_rdonly_cast()
  bpf: add bpf_features enum
  selftests/bpf: allow tests from verifier.c not to drop CAP_SYS_ADMIN
  selftests/bpf: check operations on untrusted ro pointers to mem

 kernel/bpf/verifier.c                         |  79 ++++++++--
 .../selftests/bpf/prog_tests/verifier.c       |  19 ++-
 .../bpf/progs/verifier_mem_rdonly_untrusted.c | 136 ++++++++++++++++++
 3 files changed, 216 insertions(+), 18 deletions(-)
 create mode 100644 tools/testing/selftests/bpf/progs/verifier_mem_rdonly_untrusted.c

-- 
2.47.1

[PATCH bpf-next v1 1/4] bpf: allow void* cast using bpf_rdonly_cast()

[Eduard Zingerman]

Introduce support for `bpf_rdonly_cast(v, 0)`, which casts the value
`v` to an untyped, untrusted pointer, logically similar to a `void *`.
The memory pointed to by such a pointer is treated as read-only.
As with other untrusted pointers, memory access violations on loads
return zero instead of causing a fault.

Technically:
- The resulting pointer is represented as a register of type
  `PTR_TO_MEM | MEM_RDONLY | PTR_UNTRUSTED` with size zero.
- Offsets within such pointers are not tracked.
- Same load instructions are allowed to have both
  `PTR_TO_MEM | MEM_RDONLY | PTR_UNTRUSTED` and `PTR_TO_BTF_ID`
  as the base pointer types.
  In such cases, `bpf_insn_aux_data->ptr_type` is considered the
  weaker of the two: `PTR_TO_MEM | MEM_RDONLY | PTR_UNTRUSTED`.

The following constraints apply to the new pointer type:
- can be used as a base for LDX instructions;
- can't be used as a base for ST/STX or atomic instructions;
- can't be used as parameter for kfuncs or helpers.

These constraints are enforced by existing handling of `MEM_RDONLY`
flag and `PTR_TO_MEM` of size zero.

Suggested-by: Alexei Starovoitov <alexei.starovoitov@gmail.com>
Suggested-by: Andrii Nakryiko <andrii.nakryiko@gmail.com>
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
---
 kernel/bpf/verifier.c | 72 +++++++++++++++++++++++++++++++++++--------
 1 file changed, 60 insertions(+), 12 deletions(-)

diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index 279a64933262..8fd65eb74051 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -7535,6 +7535,7 @@ static int check_mem_access(struct bpf_verifier_env *env, int insn_idx, u32 regn
 		}
 	} else if (base_type(reg->type) == PTR_TO_MEM) {
 		bool rdonly_mem = type_is_rdonly_mem(reg->type);
+		bool rdonly_untrusted = rdonly_mem && (reg->type & PTR_UNTRUSTED);
 
 		if (type_may_be_null(reg->type)) {
 			verbose(env, "R%d invalid mem access '%s'\n", regno,
@@ -7554,8 +7555,13 @@ static int check_mem_access(struct bpf_verifier_env *env, int insn_idx, u32 regn
 			return -EACCES;
 		}
 
-		err = check_mem_region_access(env, regno, off, size,
-					      reg->mem_size, false);
+		/*
+		 * Accesses to untrusted PTR_TO_MEM are done through probe
+		 * instructions, hence no need to check bounds in that case.
+		 */
+		if (!rdonly_untrusted)
+			err = check_mem_region_access(env, regno, off, size,
+						      reg->mem_size, false);
 		if (!err && value_regno >= 0 && (t == BPF_READ || rdonly_mem))
 			mark_reg_unknown(env, regs, value_regno);
 	} else if (reg->type == PTR_TO_CTX) {
@@ -13602,16 +13608,24 @@ static int check_special_kfunc(struct bpf_verifier_env *env, struct bpf_kfunc_ca
 		regs[BPF_REG_0].btf_id = meta->ret_btf_id;
 	} else if (meta->func_id == special_kfunc_list[KF_bpf_rdonly_cast]) {
 		ret_t = btf_type_by_id(desc_btf, meta->arg_constant.value);
-		if (!ret_t || !btf_type_is_struct(ret_t)) {
+		if (!ret_t) {
+			verbose(env, "Unknown type ID %lld passed to kfunc bpf_rdonly_cast\n",
+				meta->arg_constant.value);
+			return -EINVAL;
+		} else if (btf_type_is_struct(ret_t)) {
+			mark_reg_known_zero(env, regs, BPF_REG_0);
+			regs[BPF_REG_0].type = PTR_TO_BTF_ID | PTR_UNTRUSTED;
+			regs[BPF_REG_0].btf = desc_btf;
+			regs[BPF_REG_0].btf_id = meta->arg_constant.value;
+		} else if (btf_type_is_void(ret_t)) {
+			mark_reg_known_zero(env, regs, BPF_REG_0);
+			regs[BPF_REG_0].type = PTR_TO_MEM | MEM_RDONLY | PTR_UNTRUSTED;
+			regs[BPF_REG_0].mem_size = 0;
+		} else {
 			verbose(env,
-				"kfunc bpf_rdonly_cast type ID argument must be of a struct\n");
+				"kfunc bpf_rdonly_cast type ID argument must be of a struct or void\n");
 			return -EINVAL;
 		}
-
-		mark_reg_known_zero(env, regs, BPF_REG_0);
-		regs[BPF_REG_0].type = PTR_TO_BTF_ID | PTR_UNTRUSTED;
-		regs[BPF_REG_0].btf = desc_btf;
-		regs[BPF_REG_0].btf_id = meta->arg_constant.value;
 	} else if (meta->func_id == special_kfunc_list[KF_bpf_dynptr_slice] ||
 		   meta->func_id == special_kfunc_list[KF_bpf_dynptr_slice_rdwr]) {
 		enum bpf_type_flag type_flag = get_dynptr_type_flag(meta->initialized_dynptr.type);
@@ -14410,6 +14424,13 @@ static int adjust_ptr_min_max_vals(struct bpf_verifier_env *env,
 		return -EACCES;
 	}
 
+	/*
+	 * Accesses to untrusted PTR_TO_MEM are done through probe
+	 * instructions, hence no need to track offsets.
+	 */
+	if (base_type(ptr_reg->type) == PTR_TO_MEM && (ptr_reg->type & PTR_UNTRUSTED))
+		return 0;
+
 	switch (base_type(ptr_reg->type)) {
 	case PTR_TO_CTX:
 	case PTR_TO_MAP_VALUE:
@@ -19567,10 +19588,27 @@ static bool reg_type_mismatch(enum bpf_reg_type src, enum bpf_reg_type prev)
 			       !reg_type_mismatch_ok(prev));
 }
 
+static bool is_ptr_to_mem_or_btf_id(enum bpf_reg_type type)
+{
+	switch (base_type(type)) {
+	case PTR_TO_MEM:
+	case PTR_TO_BTF_ID:
+		return true;
+	default:
+		return false;
+	}
+}
+
+static bool is_ptr_to_mem(enum bpf_reg_type type)
+{
+	return base_type(type) == PTR_TO_MEM;
+}
+
 static int save_aux_ptr_type(struct bpf_verifier_env *env, enum bpf_reg_type type,
 			     bool allow_trust_mismatch)
 {
 	enum bpf_reg_type *prev_type = &env->insn_aux_data[env->insn_idx].ptr_type;
+	enum bpf_reg_type merged_type;
 
 	if (*prev_type == NOT_INIT) {
 		/* Saw a valid insn
@@ -19587,15 +19625,24 @@ static int save_aux_ptr_type(struct bpf_verifier_env *env, enum bpf_reg_type typ
 		 * Reject it.
 		 */
 		if (allow_trust_mismatch &&
-		    base_type(type) == PTR_TO_BTF_ID &&
-		    base_type(*prev_type) == PTR_TO_BTF_ID) {
+		    is_ptr_to_mem_or_btf_id(type) &&
+		    is_ptr_to_mem_or_btf_id(*prev_type)) {
 			/*
 			 * Have to support a use case when one path through
 			 * the program yields TRUSTED pointer while another
 			 * is UNTRUSTED. Fallback to UNTRUSTED to generate
 			 * BPF_PROBE_MEM/BPF_PROBE_MEMSX.
+			 * Same behavior of MEM_RDONLY flag.
 			 */
-			*prev_type = PTR_TO_BTF_ID | PTR_UNTRUSTED;
+			if (is_ptr_to_mem(type) || is_ptr_to_mem(*prev_type))
+				merged_type = PTR_TO_MEM;
+			else
+				merged_type = PTR_TO_BTF_ID;
+			if ((type & PTR_UNTRUSTED) || (*prev_type & PTR_UNTRUSTED))
+				merged_type |= PTR_UNTRUSTED;
+			if ((type & MEM_RDONLY) || (*prev_type & MEM_RDONLY))
+				merged_type |= MEM_RDONLY;
+			*prev_type = merged_type;
 		} else {
 			verbose(env, "same insn cannot be used with different pointers\n");
 			return -EINVAL;
@@ -21203,6 +21250,7 @@ static int convert_ctx_accesses(struct bpf_verifier_env *env)
 		 * for this case.
 		 */
 		case PTR_TO_BTF_ID | MEM_ALLOC | PTR_UNTRUSTED:
+		case PTR_TO_MEM | MEM_RDONLY | PTR_UNTRUSTED:
 			if (type == BPF_READ) {
 				if (BPF_MODE(insn->code) == BPF_MEM)
 					insn->code = BPF_LDX | BPF_PROBE_MEM |
-- 
2.47.1

[PATCH bpf-next v1 2/4] bpf: add bpf_features enum

[Eduard Zingerman]

This commit adds a kernel side enum for use in conjucntion with BTF
CO-RE bpf_core_enum_value_exists. The goal of the enum is to assist
with available BPF features detection.

Support for bpf_rdonly_cast to void* is the first feature listed in
the enum. Here is an example usage:

  if (bpf_core_enum_value_exists(enum bpf_features,
                                 BPF_FEAT_RDONLY_CAST_TO_VOID))
     ... bpf_rdonly_cast(..., 0) ...

Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
---
 kernel/bpf/verifier.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index 8fd65eb74051..01050d1f7389 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -44,6 +44,11 @@ static const struct bpf_verifier_ops * const bpf_verifier_ops[] = {
 #undef BPF_LINK_TYPE
 };
 
+enum bpf_features {
+	BPF_FEAT_RDONLY_CAST_TO_VOID = 0,
+	BPF_FEAT_TOTAL,
+};
+
 struct bpf_mem_alloc bpf_global_percpu_ma;
 static bool bpf_global_percpu_ma_set;
 
@@ -24436,6 +24441,8 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr, bpfptr_t uattr, __u3
 	u32 log_true_size;
 	bool is_priv;
 
+	BTF_TYPE_EMIT(enum bpf_features);
+
 	/* no program is valid */
 	if (ARRAY_SIZE(bpf_verifier_ops) == 0)
 		return -EINVAL;
-- 
2.47.1

[PATCH bpf-next v1 3/4] selftests/bpf: allow tests from verifier.c not to drop CAP_SYS_ADMIN

[Eduard Zingerman]

Originally prog_tests/verifier.c was developed to run tests ported
from test_verifier binary. test_verifier runs tests with CAP_SYS_ADMIN
dropped, hence this behaviour was copied in prog_tests/verifier.c.
BPF_OBJ_GET_NEXT_ID BPF syscall command fails w/o CAP_SYS_ADMIN and
this prevents libbpf from loading module BTFs.
This commit adds an optout from capability drop.

Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
---
 .../testing/selftests/bpf/prog_tests/verifier.c | 17 +++++++++++------
 1 file changed, 11 insertions(+), 6 deletions(-)

diff --git a/tools/testing/selftests/bpf/prog_tests/verifier.c b/tools/testing/selftests/bpf/prog_tests/verifier.c
index c9da06741104..cedb86d8f717 100644
--- a/tools/testing/selftests/bpf/prog_tests/verifier.c
+++ b/tools/testing/selftests/bpf/prog_tests/verifier.c
@@ -115,14 +115,16 @@ struct test_val {
 __maybe_unused
 static void run_tests_aux(const char *skel_name,
 			  skel_elf_bytes_fn elf_bytes_factory,
-			  pre_execution_cb pre_execution_cb)
+			  pre_execution_cb pre_execution_cb,
+			  bool drop_sysadmin)
 {
 	struct test_loader tester = {};
-	__u64 old_caps;
+	__u64 caps_to_drop, old_caps;
 	int err;
 
 	/* test_verifier tests are executed w/o CAP_SYS_ADMIN, do the same here */
-	err = cap_disable_effective(1ULL << CAP_SYS_ADMIN, &old_caps);
+	caps_to_drop = drop_sysadmin ? 1ULL << CAP_SYS_ADMIN : 0;
+	err = cap_disable_effective(caps_to_drop, &old_caps);
 	if (err) {
 		PRINT_FAIL("failed to drop CAP_SYS_ADMIN: %i, %s\n", err, strerror(-err));
 		return;
@@ -137,7 +139,8 @@ static void run_tests_aux(const char *skel_name,
 		PRINT_FAIL("failed to restore CAP_SYS_ADMIN: %i, %s\n", err, strerror(-err));
 }
 
-#define RUN(skel) run_tests_aux(#skel, skel##__elf_bytes, NULL)
+#define RUN(skel) run_tests_aux(#skel, skel##__elf_bytes, NULL, true)
+#define RUN_FULL_CAPS(skel) run_tests_aux(#skel, skel##__elf_bytes, NULL, false)
 
 void test_verifier_and(void)                  { RUN(verifier_and); }
 void test_verifier_arena(void)                { RUN(verifier_arena); }
@@ -272,7 +275,8 @@ void test_verifier_array_access(void)
 {
 	run_tests_aux("verifier_array_access",
 		      verifier_array_access__elf_bytes,
-		      init_array_access_maps);
+		      init_array_access_maps,
+		      true);
 }
 
 static int init_value_ptr_arith_maps(struct bpf_object *obj)
@@ -284,5 +288,6 @@ void test_verifier_value_ptr_arith(void)
 {
 	run_tests_aux("verifier_value_ptr_arith",
 		      verifier_value_ptr_arith__elf_bytes,
-		      init_value_ptr_arith_maps);
+		      init_value_ptr_arith_maps,
+		      true);
 }
-- 
2.47.1

[PATCH bpf-next v1 4/4] selftests/bpf: check operations on untrusted ro pointers to mem

[Eduard Zingerman]

The following cases are tested:
- it is ok to load memory at any offset from rdonly_untrusted_mem;
- rdonly_untrusted_mem offset/bounds are not tracked;
- writes into rdonly_untrusted_mem are forbidden;
- atomic operations on rdonly_untrusted_mem are forbidden;
- rdonly_untrusted_mem can't be passed as a memory argument of a
  helper of kfunc;
- it is ok to use PTR_TO_MEM and PTR_TO_BTF_ID in a same load
  instruction.

Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
---
 .../selftests/bpf/prog_tests/verifier.c       |   2 +
 .../bpf/progs/verifier_mem_rdonly_untrusted.c | 136 ++++++++++++++++++
 2 files changed, 138 insertions(+)
 create mode 100644 tools/testing/selftests/bpf/progs/verifier_mem_rdonly_untrusted.c

diff --git a/tools/testing/selftests/bpf/prog_tests/verifier.c b/tools/testing/selftests/bpf/prog_tests/verifier.c
index cedb86d8f717..5cb49ba089d2 100644
--- a/tools/testing/selftests/bpf/prog_tests/verifier.c
+++ b/tools/testing/selftests/bpf/prog_tests/verifier.c
@@ -57,6 +57,7 @@
 #include "verifier_may_goto_1.skel.h"
 #include "verifier_may_goto_2.skel.h"
 #include "verifier_meta_access.skel.h"
+#include "verifier_mem_rdonly_untrusted.skel.h"
 #include "verifier_movsx.skel.h"
 #include "verifier_mtu.skel.h"
 #include "verifier_netfilter_ctx.skel.h"
@@ -205,6 +206,7 @@ void test_verifier_prevent_map_lookup(void)   { RUN(verifier_prevent_map_lookup)
 void test_verifier_private_stack(void)        { RUN(verifier_private_stack); }
 void test_verifier_raw_stack(void)            { RUN(verifier_raw_stack); }
 void test_verifier_raw_tp_writable(void)      { RUN(verifier_raw_tp_writable); }
+void test_verifier_mem_rdonly_untrusted(void) { RUN_FULL_CAPS(verifier_mem_rdonly_untrusted); }
 void test_verifier_reg_equal(void)            { RUN(verifier_reg_equal); }
 void test_verifier_ref_tracking(void)         { RUN(verifier_ref_tracking); }
 void test_verifier_regalloc(void)             { RUN(verifier_regalloc); }
diff --git a/tools/testing/selftests/bpf/progs/verifier_mem_rdonly_untrusted.c b/tools/testing/selftests/bpf/progs/verifier_mem_rdonly_untrusted.c
new file mode 100644
index 000000000000..00604755e698
--- /dev/null
+++ b/tools/testing/selftests/bpf/progs/verifier_mem_rdonly_untrusted.c
@@ -0,0 +1,136 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#include <vmlinux.h>
+#include <bpf/bpf_core_read.h>
+#include "bpf_misc.h"
+#include "../test_kmods/bpf_testmod_kfunc.h"
+
+SEC("socket")
+__success
+__retval(0)
+int ldx_is_ok_bad_addr(void *ctx)
+{
+	char *p;
+
+	if (!bpf_core_enum_value_exists(enum bpf_features, BPF_FEAT_RDONLY_CAST_TO_VOID))
+		return 42;
+
+	p = bpf_rdonly_cast(0, 0);
+	return p[0x7fff];
+}
+
+SEC("socket")
+__success
+__retval(1)
+int ldx_is_ok_good_addr(void *ctx)
+{
+	int v, *p;
+
+	v = 1;
+	p = bpf_rdonly_cast(&v, 0);
+	return *p;
+}
+
+SEC("socket")
+__success
+int offset_not_tracked(void *ctx)
+{
+	int *p, i, s;
+
+	p = bpf_rdonly_cast(0, 0);
+	s = 0;
+	bpf_for(i, 0, 1000 * 1000 * 1000) {
+		p++;
+		s += *p;
+	}
+	return s;
+}
+
+SEC("socket")
+__failure
+__msg("cannot write into rdonly_untrusted_mem")
+int stx_not_ok(void *ctx)
+{
+	int v, *p;
+
+	v = 1;
+	p = bpf_rdonly_cast(&v, 0);
+	*p = 1;
+	return 0;
+}
+
+SEC("socket")
+__failure
+__msg("cannot write into rdonly_untrusted_mem")
+int atomic_not_ok(void *ctx)
+{
+	int v, *p;
+
+	v = 1;
+	p = bpf_rdonly_cast(&v, 0);
+	__sync_fetch_and_add(p, 1);
+	return 0;
+}
+
+SEC("socket")
+__failure
+__msg("cannot write into rdonly_untrusted_mem")
+int atomic_rmw_not_ok(void *ctx)
+{
+	long v, *p;
+
+	v = 1;
+	p = bpf_rdonly_cast(&v, 0);
+	return __sync_val_compare_and_swap(p, 0, 42);
+}
+
+SEC("socket")
+__failure
+__msg("invalid access to memory, mem_size=0 off=0 size=4")
+__msg("R1 min value is outside of the allowed memory range")
+int kfunc_param_not_ok(void *ctx)
+{
+	int *p;
+
+	p = bpf_rdonly_cast(0, 0);
+	bpf_kfunc_trusted_num_test(p);
+	return 0;
+}
+
+SEC("?fentry.s/" SYS_PREFIX "sys_getpgid")
+__failure
+__msg("R1 type=rdonly_untrusted_mem expected=")
+int helper_param_not_ok(void *ctx)
+{
+	char *p;
+
+	p = bpf_rdonly_cast(0, 0);
+	/*
+	 * Any helper with ARG_CONST_SIZE_OR_ZERO constraint will do,
+	 * the most permissive constraint
+	 */
+	bpf_copy_from_user(p, 0, (void *)42);
+	return 0;
+}
+
+static __noinline u64 *get_some_addr(void)
+{
+	if (bpf_get_prandom_u32())
+		return bpf_rdonly_cast(0, bpf_core_type_id_kernel(struct sock));
+	else
+		return bpf_rdonly_cast(0, 0);
+}
+
+SEC("socket")
+__success
+__retval(0)
+int mixed_mem_type(void *ctx)
+{
+	u64 *p;
+
+	/* Try to avoid compiler hoisting load to if branches by using __noinline func. */
+	p = get_some_addr();
+	return *p;
+}
+
+char _license[] SEC("license") = "GPL";
-- 
2.47.1

Re: [PATCH bpf-next v1 3/4] selftests/bpf: allow tests from verifier.c not to drop CAP_SYS_ADMIN

[Alexei Starovoitov]

On Tue, Jun 24, 2025 at 12:10 PM Eduard Zingerman <eddyz87@gmail.com> wrote:
>
> Originally prog_tests/verifier.c was developed to run tests ported
> from test_verifier binary. test_verifier runs tests with CAP_SYS_ADMIN
> dropped, hence this behaviour was copied in prog_tests/verifier.c.
> BPF_OBJ_GET_NEXT_ID BPF syscall command fails w/o CAP_SYS_ADMIN and
> this prevents libbpf from loading module BTFs.

You need this only because of 'bpf_kfunc_trusted_num_test' access
in patch 4?
Can you use kernel kfunc instead?
This needs more thought.
s/RUN/RUN_FULL_CAPS/ just because of kfunc in the bpf_testmod
doesn't look like a good long term approach.

I thought we agreed to relax BPF_OBJ_GET_NEXT_ID to allow for CAP_BPF.
Probably even unpriv can do it.
Just knowing a set of prog, map, bpf IDs is not a security threat.

BPF_BTF_GET_FD_BY_ID can also be allowed for unpriv,
since one can do it already from /sys/kernel/btf/

> This commit adds an optout from capability drop.
>
> Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
> ---
>  .../testing/selftests/bpf/prog_tests/verifier.c | 17 +++++++++++------
>  1 file changed, 11 insertions(+), 6 deletions(-)
>
> diff --git a/tools/testing/selftests/bpf/prog_tests/verifier.c b/tools/testing/selftests/bpf/prog_tests/verifier.c
> index c9da06741104..cedb86d8f717 100644
> --- a/tools/testing/selftests/bpf/prog_tests/verifier.c
> +++ b/tools/testing/selftests/bpf/prog_tests/verifier.c
> @@ -115,14 +115,16 @@ struct test_val {
>  __maybe_unused
>  static void run_tests_aux(const char *skel_name,
>                           skel_elf_bytes_fn elf_bytes_factory,
> -                         pre_execution_cb pre_execution_cb)
> +                         pre_execution_cb pre_execution_cb,
> +                         bool drop_sysadmin)

I have an allergic reaction to bool arguments.

>         run_tests_aux("verifier_array_access",
>                       verifier_array_access__elf_bytes,
> -                     init_array_access_maps);
> +                     init_array_access_maps,
> +                     true);

This is not readable without looking at the argument name.

--
pw-bot: cr

Re: [PATCH bpf-next v1 2/4] bpf: add bpf_features enum

[Alexei Starovoitov]

On Tue, Jun 24, 2025 at 12:10 PM Eduard Zingerman <eddyz87@gmail.com> wrote:
>
> This commit adds a kernel side enum for use in conjucntion with BTF
> CO-RE bpf_core_enum_value_exists. The goal of the enum is to assist
> with available BPF features detection.
>
> Support for bpf_rdonly_cast to void* is the first feature listed in
> the enum. Here is an example usage:
>
>   if (bpf_core_enum_value_exists(enum bpf_features,
>                                  BPF_FEAT_RDONLY_CAST_TO_VOID))
>      ... bpf_rdonly_cast(..., 0) ...
>
> Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
> ---
>  kernel/bpf/verifier.c | 7 +++++++
>  1 file changed, 7 insertions(+)
>
> diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
> index 8fd65eb74051..01050d1f7389 100644
> --- a/kernel/bpf/verifier.c
> +++ b/kernel/bpf/verifier.c
> @@ -44,6 +44,11 @@ static const struct bpf_verifier_ops * const bpf_verifier_ops[] = {
>  #undef BPF_LINK_TYPE
>  };
>
> +enum bpf_features {
> +       BPF_FEAT_RDONLY_CAST_TO_VOID = 0,
> +       BPF_FEAT_TOTAL,

I don't see the value of 'total', but not strongly against it.
But pls be consistent with __MAX_BPF_CMD, __MAX_BPF_MAP_TYPE, ...
Say, __MAX_BPF_FEAT ?


Also it's better to introduce this enum in some earlier patch,
and then always add BTF_FEAT_... to this enum
in the same patch that adds the feature to make
sure backports won't screw it up.
Another rule should be to always assign a number to it.

At the end with random backports the __MAX_BPF_FEAT
won't be accurate, but whatever.

> +};
> +
>  struct bpf_mem_alloc bpf_global_percpu_ma;
>  static bool bpf_global_percpu_ma_set;
>
> @@ -24436,6 +24441,8 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr, bpfptr_t uattr, __u3
>         u32 log_true_size;
>         bool is_priv;
>
> +       BTF_TYPE_EMIT(enum bpf_features);
> +
>         /* no program is valid */
>         if (ARRAY_SIZE(bpf_verifier_ops) == 0)
>                 return -EINVAL;
> --
> 2.47.1
>

Re: [PATCH bpf-next v1 3/4] selftests/bpf: allow tests from verifier.c not to drop CAP_SYS_ADMIN

[Eduard Zingerman]

On Tue, 2025-06-24 at 14:55 -0700, Alexei Starovoitov wrote:
> On Tue, Jun 24, 2025 at 12:10 PM Eduard Zingerman <eddyz87@gmail.com> wrote:
> > 
> > Originally prog_tests/verifier.c was developed to run tests ported
> > from test_verifier binary. test_verifier runs tests with CAP_SYS_ADMIN
> > dropped, hence this behaviour was copied in prog_tests/verifier.c.
> > BPF_OBJ_GET_NEXT_ID BPF syscall command fails w/o CAP_SYS_ADMIN and
> > this prevents libbpf from loading module BTFs.
> 
> You need this only because of 'bpf_kfunc_trusted_num_test' access
> in patch 4?

Yes.

> Can you use kernel kfunc instead?

Should be able to.

> This needs more thought.
> s/RUN/RUN_FULL_CAPS/ just because of kfunc in the bpf_testmod
> doesn't look like a good long term approach.
> 
> I thought we agreed to relax BPF_OBJ_GET_NEXT_ID to allow for CAP_BPF.
> Probably even unpriv can do it.
> Just knowing a set of prog, map, bpf IDs is not a security threat.
> 
> BPF_BTF_GET_FD_BY_ID can also be allowed for unpriv,
> since one can do it already from /sys/kernel/btf/

Makes sense to me.

> > This commit adds an optout from capability drop.
> > 
> > Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
> > ---
> >  .../testing/selftests/bpf/prog_tests/verifier.c | 17 +++++++++++------
> >  1 file changed, 11 insertions(+), 6 deletions(-)
> > 
> > diff --git a/tools/testing/selftests/bpf/prog_tests/verifier.c b/tools/testing/selftests/bpf/prog_tests/verifier.c
> > index c9da06741104..cedb86d8f717 100644
> > --- a/tools/testing/selftests/bpf/prog_tests/verifier.c
> > +++ b/tools/testing/selftests/bpf/prog_tests/verifier.c
> > @@ -115,14 +115,16 @@ struct test_val {
> >  __maybe_unused
> >  static void run_tests_aux(const char *skel_name,
> >                           skel_elf_bytes_fn elf_bytes_factory,
> > -                         pre_execution_cb pre_execution_cb)
> > +                         pre_execution_cb pre_execution_cb,
> > +                         bool drop_sysadmin)
> 
> I have an allergic reaction to bool arguments.
> 
> >         run_tests_aux("verifier_array_access",
> >                       verifier_array_access__elf_bytes,
> > -                     init_array_access_maps);
> > +                     init_array_access_maps,
> > +                     true);
> 
> This is not readable without looking at the argument name.

I'll drop this change in v2.

Re: [PATCH bpf-next v1 2/4] bpf: add bpf_features enum

[Eduard Zingerman]

On Tue, 2025-06-24 at 14:59 -0700, Alexei Starovoitov wrote:

[...]

> > diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
> > index 8fd65eb74051..01050d1f7389 100644
> > --- a/kernel/bpf/verifier.c
> > +++ b/kernel/bpf/verifier.c
> > @@ -44,6 +44,11 @@ static const struct bpf_verifier_ops * const bpf_verifier_ops[] = {
> >  #undef BPF_LINK_TYPE
> >  };
> > 
> > +enum bpf_features {
> > +       BPF_FEAT_RDONLY_CAST_TO_VOID = 0,
> > +       BPF_FEAT_TOTAL,
> 
> I don't see the value of 'total', but not strongly against it.
> But pls be consistent with __MAX_BPF_CMD, __MAX_BPF_MAP_TYPE, ...
> Say, __MAX_BPF_FEAT ?
> 
> 
> Also it's better to introduce this enum in some earlier patch,
> and then always add BTF_FEAT_... to this enum
> in the same patch that adds the feature to make
> sure backports won't screw it up.
> Another rule should be to always assign a number to it.
> 
> At the end with random backports the __MAX_BPF_FEAT
> won't be accurate, but whatever.

Ack. Andrii asked to add MAX for people willing to do broken kind of
feature detection and just in case.

Re: [PATCH bpf-next v1 3/4] selftests/bpf: allow tests from verifier.c not to drop CAP_SYS_ADMIN

[Eduard Zingerman]

On Tue, 2025-06-24 at 14:55 -0700, Alexei Starovoitov wrote:
> On Tue, Jun 24, 2025 at 12:10 PM Eduard Zingerman <eddyz87@gmail.com> wrote:
> > 
> > Originally prog_tests/verifier.c was developed to run tests ported
> > from test_verifier binary. test_verifier runs tests with CAP_SYS_ADMIN
> > dropped, hence this behaviour was copied in prog_tests/verifier.c.
> > BPF_OBJ_GET_NEXT_ID BPF syscall command fails w/o CAP_SYS_ADMIN and
> > this prevents libbpf from loading module BTFs.
> 
> You need this only because of 'bpf_kfunc_trusted_num_test' access
> in patch 4?
> Can you use kernel kfunc instead?

This turned out non-trivial, not many kernel kfuncs take pointers to
primitive types, and those that do are either STRUCT_OPS or need
device bound program or have special checks requiring stack pointers.

I declared a separate prog_tests/mem_rdonly_untrusted.c runner.

[...]

Re: [PATCH bpf-next v1 3/4] selftests/bpf: allow tests from verifier.c not to drop CAP_SYS_ADMIN

[Alexei Starovoitov]

On Tue, Jun 24, 2025 at 4:31 PM Eduard Zingerman <eddyz87@gmail.com> wrote:
>
> On Tue, 2025-06-24 at 14:55 -0700, Alexei Starovoitov wrote:
> > On Tue, Jun 24, 2025 at 12:10 PM Eduard Zingerman <eddyz87@gmail.com> wrote:
> > >
> > > Originally prog_tests/verifier.c was developed to run tests ported
> > > from test_verifier binary. test_verifier runs tests with CAP_SYS_ADMIN
> > > dropped, hence this behaviour was copied in prog_tests/verifier.c.
> > > BPF_OBJ_GET_NEXT_ID BPF syscall command fails w/o CAP_SYS_ADMIN and
> > > this prevents libbpf from loading module BTFs.
> >
> > You need this only because of 'bpf_kfunc_trusted_num_test' access
> > in patch 4?
> > Can you use kernel kfunc instead?
>
> This turned out non-trivial, not many kernel kfuncs take pointers to
> primitive types, and those that do are either STRUCT_OPS or need
> device bound program or have special checks requiring stack pointers.
>
> I declared a separate prog_tests/mem_rdonly_untrusted.c runner.

Just skip it for now.
Adding unpriv btf access can be a follow up.