[PATCH v7 bpf-next 0/2] bpf: Add a generic bits iterator

[PATCH v7 bpf-next 0/2] bpf: Add a generic bits iterator

[Yafang Shao]

hree new kfuncs, namely bpf_iter_bits_{new,next,destroy}, have been
added for the new bpf_iter_bits functionality. These kfuncs enable the
iteration of the bits from a given address and a given number of bits.

- bpf_iter_bits_new
  Initialize a new bits iterator for a given memory area. Due to the
  limitation of bpf memalloc, the max number of bits to be iterated
  over is (4096 * 8).
- bpf_iter_bits_next
  Get the next bit in a bpf_iter_bits
- bpf_iter_bits_destroy
  Destroy a bpf_iter_bits

The bits iterator can be used in any context and on any address.

Changes:
- v6->v7:
  - Fix endianness error for non-long-aligned data (Andrii)
- v5->v6:
  - Add positive tests (Andrii)
- v4->v5:
  - Simplify test cases (Andrii)
- v3->v4:
  - Fix endianness error on s390x (Andrii)
  - zero-initialize kit->bits_copy and zero out nr_bits (Andrii)
- v2->v3:
  - Optimization for u64/u32 mask (Andrii)
- v1->v2:
  - Simplify the CPU number verification code to avoid the failure on s390x
    (Eduard)
- bpf: Add bpf_iter_cpumask
  https://lwn.net/Articles/961104/
- bpf: Add new bpf helper bpf_for_each_cpu
  https://lwn.net/Articles/939939/

Yafang Shao (2):
  bpf: Add bits iterator
  selftests/bpf: Add selftest for bits iter

 kernel/bpf/helpers.c                          | 140 +++++++++++++++
 .../selftests/bpf/prog_tests/verifier.c       |   2 +
 .../selftests/bpf/progs/verifier_bits_iter.c  | 160 ++++++++++++++++++
 3 files changed, 302 insertions(+)
 create mode 100644 tools/testing/selftests/bpf/progs/verifier_bits_iter.c

-- 
2.30.1 (Apple Git-130)

[PATCH v7 bpf-next 1/2] bpf: Add bits iterator

[Yafang Shao]

Add three new kfuncs for the bits iterator:
- bpf_iter_bits_new
  Initialize a new bits iterator for a given memory area. Due to the
  limitation of bpf memalloc, the max number of bits that can be iterated
  over is limited to (4096 * 8).
- bpf_iter_bits_next
  Get the next bit in a bpf_iter_bits
- bpf_iter_bits_destroy
  Destroy a bpf_iter_bits

The bits iterator facilitates the iteration of the bits of a memory area,
such as cpumask. It can be used in any context and on any address.

Signed-off-by: Yafang Shao <laoar.shao@gmail.com>
---
 kernel/bpf/helpers.c | 140 +++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 140 insertions(+)

diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
index 2a69a9a36c0f..83b2a02f795f 100644
--- a/kernel/bpf/helpers.c
+++ b/kernel/bpf/helpers.c
@@ -2744,6 +2744,143 @@ __bpf_kfunc void bpf_preempt_enable(void)
 	preempt_enable();
 }
 
+struct bpf_iter_bits {
+	__u64 __opaque[2];
+} __aligned(8);
+
+struct bpf_iter_bits_kern {
+	union {
+		unsigned long *bits;
+		unsigned long bits_copy;
+	};
+	u32 nr_bits;
+	int bit;
+} __aligned(8);
+
+/**
+ * bpf_iter_bits_new() - Initialize a new bits iterator for a given memory area
+ * @it: The new bpf_iter_bits to be created
+ * @unsafe_ptr__ign: A ponter pointing to a memory area to be iterated over
+ * @nr_bits: The number of bits to be iterated over. Due to the limitation of
+ * memalloc, it can't greater than (4096 * 8).
+ *
+ * This function initializes a new bpf_iter_bits structure for iterating over
+ * a memory area which is specified by the @unsafe_ptr__ign and @nr_bits. It
+ * copy the data of the memory area to the newly created bpf_iter_bits @it for
+ * subsequent iteration operations.
+ *
+ * On success, 0 is returned. On failure, ERR is returned.
+ */
+__bpf_kfunc int
+bpf_iter_bits_new(struct bpf_iter_bits *it, const void *unsafe_ptr__ign, u32 nr_bits)
+{
+	struct bpf_iter_bits_kern *kit = (void *)it;
+	u32 words = BITS_TO_LONGS(nr_bits);
+	u32 size = BITS_TO_BYTES(nr_bits);
+	u32 left, offset;
+	int err;
+
+	BUILD_BUG_ON(sizeof(struct bpf_iter_bits_kern) != sizeof(struct bpf_iter_bits));
+	BUILD_BUG_ON(__alignof__(struct bpf_iter_bits_kern) !=
+		     __alignof__(struct bpf_iter_bits));
+
+	if (!unsafe_ptr__ign || !nr_bits) {
+		kit->bits = NULL;
+		return -EINVAL;
+	}
+
+	kit->nr_bits = 0;
+	kit->bits_copy = 0;
+	/* Optimization for u64/u32 mask */
+	if (nr_bits <= 64) {
+		/* For big-endian, we must calculate the offset */
+		offset = IS_ENABLED(CONFIG_S390) ? sizeof(u64) - size : 0;
+
+		err = bpf_probe_read_kernel_common(((char *)&kit->bits_copy) + offset,
+						   size, unsafe_ptr__ign);
+		if (err)
+			return -EFAULT;
+
+		kit->nr_bits = nr_bits;
+		kit->bit = -1;
+		return 0;
+	}
+
+	/* Fallback to memalloc */
+	kit->bits = bpf_mem_alloc(&bpf_global_ma, size);
+	if (!kit->bits)
+		return -ENOMEM;
+
+	err = bpf_probe_read_kernel_common(kit->bits, words * sizeof(u64), unsafe_ptr__ign);
+	if (err) {
+		bpf_mem_free(&bpf_global_ma, kit->bits);
+		return err;
+	}
+
+	/* long-aligned */
+	left = size & (sizeof(u64) - 1);
+	if (!left)
+		goto out;
+
+	offset = IS_ENABLED(CONFIG_S390) ? sizeof(u64) - left : 0;
+	err = bpf_probe_read_kernel_common((char *)(kit->bits + words - 1) + offset, left,
+					   unsafe_ptr__ign + (words - 1) * sizeof(u64));
+	if (err) {
+		bpf_mem_free(&bpf_global_ma, kit->bits);
+		return err;
+	}
+
+out:
+	kit->nr_bits = nr_bits;
+	kit->bit = -1;
+	return 0;
+}
+
+/**
+ * bpf_iter_bits_next() - Get the next bit in a bpf_iter_bits
+ * @it: The bpf_iter_bits to be checked
+ *
+ * This function returns a pointer to a number representing the value of the
+ * next bit in the bits.
+ *
+ * If there are no further bit available, it returns NULL.
+ */
+__bpf_kfunc int *bpf_iter_bits_next(struct bpf_iter_bits *it)
+{
+	struct bpf_iter_bits_kern *kit = (void *)it;
+	u32 nr_bits = kit->nr_bits;
+	const unsigned long *bits;
+	int bit;
+
+	if (nr_bits == 0)
+		return NULL;
+
+	bits = nr_bits <= 64 ? &kit->bits_copy : kit->bits;
+	bit = find_next_bit(bits, nr_bits, kit->bit + 1);
+	if (bit >= nr_bits) {
+		kit->nr_bits = 0;
+		return NULL;
+	}
+
+	kit->bit = bit;
+	return &kit->bit;
+}
+
+/**
+ * bpf_iter_bits_destroy() - Destroy a bpf_iter_bits
+ * @it: The bpf_iter_bits to be destroyed
+ *
+ * Destroy the resource associated with the bpf_iter_bits.
+ */
+__bpf_kfunc void bpf_iter_bits_destroy(struct bpf_iter_bits *it)
+{
+	struct bpf_iter_bits_kern *kit = (void *)it;
+
+	if (kit->nr_bits <= 64)
+		return;
+	bpf_mem_free(&bpf_global_ma, kit->bits);
+}
+
 __bpf_kfunc_end_defs();
 
 BTF_KFUNCS_START(generic_btf_ids)
@@ -2826,6 +2963,9 @@ BTF_ID_FLAGS(func, bpf_wq_set_callback_impl)
 BTF_ID_FLAGS(func, bpf_wq_start)
 BTF_ID_FLAGS(func, bpf_preempt_disable)
 BTF_ID_FLAGS(func, bpf_preempt_enable)
+BTF_ID_FLAGS(func, bpf_iter_bits_new, KF_ITER_NEW)
+BTF_ID_FLAGS(func, bpf_iter_bits_next, KF_ITER_NEXT | KF_RET_NULL)
+BTF_ID_FLAGS(func, bpf_iter_bits_destroy, KF_ITER_DESTROY)
 BTF_KFUNCS_END(common_btf_ids)
 
 static const struct btf_kfunc_id_set common_kfunc_set = {
-- 
2.30.1 (Apple Git-130)

[PATCH v7 bpf-next 2/2] selftests/bpf: Add selftest for bits iter

[Yafang Shao]

Add test cases for the bits iter:
- positive case
  - bit mask smaller than 8 bytes
  - a typical case of having 8-byte bit mask
  - another typical case where bit mask is > 8 bytes
  - the index of set bit

- nagative cases
  - bpf_iter_bits_destroy() is required after calling
    bpf_iter_bits_new()
  - bpf_iter_bits_destroy() can only destroy an initialized iter
  - bpf_iter_bits_next() must use an initialized iter

Signed-off-by: Yafang Shao <laoar.shao@gmail.com>
---
 .../selftests/bpf/prog_tests/verifier.c       |   2 +
 .../selftests/bpf/progs/verifier_bits_iter.c  | 160 ++++++++++++++++++
 2 files changed, 162 insertions(+)
 create mode 100644 tools/testing/selftests/bpf/progs/verifier_bits_iter.c

diff --git a/tools/testing/selftests/bpf/prog_tests/verifier.c b/tools/testing/selftests/bpf/prog_tests/verifier.c
index c4f9f306646e..7e04ecaaa20a 100644
--- a/tools/testing/selftests/bpf/prog_tests/verifier.c
+++ b/tools/testing/selftests/bpf/prog_tests/verifier.c
@@ -84,6 +84,7 @@
 #include "verifier_xadd.skel.h"
 #include "verifier_xdp.skel.h"
 #include "verifier_xdp_direct_packet_access.skel.h"
+#include "verifier_bits_iter.skel.h"
 
 #define MAX_ENTRIES 11
 
@@ -198,6 +199,7 @@ void test_verifier_var_off(void)              { RUN(verifier_var_off); }
 void test_verifier_xadd(void)                 { RUN(verifier_xadd); }
 void test_verifier_xdp(void)                  { RUN(verifier_xdp); }
 void test_verifier_xdp_direct_packet_access(void) { RUN(verifier_xdp_direct_packet_access); }
+void test_verifier_bits_iter(void) { RUN(verifier_bits_iter); }
 
 static int init_test_val_map(struct bpf_object *obj, char *map_name)
 {
diff --git a/tools/testing/selftests/bpf/progs/verifier_bits_iter.c b/tools/testing/selftests/bpf/progs/verifier_bits_iter.c
new file mode 100644
index 000000000000..2f7b62b25638
--- /dev/null
+++ b/tools/testing/selftests/bpf/progs/verifier_bits_iter.c
@@ -0,0 +1,160 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* Copyright (c) 2024 Yafang Shao <laoar.shao@gmail.com> */
+
+#include "vmlinux.h"
+#include <bpf/bpf_helpers.h>
+#include <bpf/bpf_tracing.h>
+
+#include "bpf_misc.h"
+#include "task_kfunc_common.h"
+
+char _license[] SEC("license") = "GPL";
+
+int bpf_iter_bits_new(struct bpf_iter_bits *it, const void *unsafe_ptr__ign,
+		      u32 nr_bits) __ksym __weak;
+int *bpf_iter_bits_next(struct bpf_iter_bits *it) __ksym __weak;
+void bpf_iter_bits_destroy(struct bpf_iter_bits *it) __ksym __weak;
+
+SEC("iter.s/cgroup")
+__description("bits iter without destroy")
+__failure __msg("Unreleased reference")
+int BPF_PROG(no_destroy, struct bpf_iter_meta *meta, struct cgroup *cgrp)
+{
+	struct bpf_iter_bits it;
+	struct task_struct *p;
+
+	p = bpf_task_from_pid(1);
+	if (!p)
+		return 1;
+
+	bpf_iter_bits_new(&it, p->cpus_ptr, 8192);
+
+	bpf_iter_bits_next(&it);
+	bpf_task_release(p);
+	return 0;
+}
+
+SEC("iter/cgroup")
+__description("bits iter with uninitialized iter in ->next()")
+__failure __msg("expected an initialized iter_bits as arg #1")
+int BPF_PROG(next_uninit, struct bpf_iter_meta *meta, struct cgroup *cgrp)
+{
+	struct bpf_iter_bits *it = NULL;
+
+	bpf_iter_bits_next(it);
+	return 0;
+}
+
+SEC("iter/cgroup")
+__description("bits iter with uninitialized iter in ->destroy()")
+__failure __msg("expected an initialized iter_bits as arg #1")
+int BPF_PROG(destroy_uninit, struct bpf_iter_meta *meta, struct cgroup *cgrp)
+{
+	struct bpf_iter_bits it = {};
+
+	bpf_iter_bits_destroy(&it);
+	return 0;
+}
+
+SEC("syscall")
+__description("bits copy 32")
+__success __retval(10)
+int bits_copy32(void)
+{
+	/* 21 bits:             --------------------- */
+	u32 data = 0b11111101111101111100001000100101U;
+	int nr = 0, offset = 0;
+	int *bit;
+
+#if defined(__TARGET_ARCH_s390)
+	offset = sizeof(u32) - (21 + 7) / 8;
+#endif
+	bpf_for_each(bits, bit, ((char *)&data) + offset, 21)
+		nr++;
+	return nr;
+}
+
+SEC("syscall")
+__description("bits copy 64")
+__success __retval(18)
+int bits_copy64(void)
+{
+	/* 34 bits:         ~-------- */
+	u64 data = 0xffffefdf0f0f0f0fUL;
+	int nr = 0, offset = 0;
+	int *bit;
+
+#if defined(__TARGET_ARCH_s390)
+	offset = sizeof(u64) - (34 + 7) / 8;
+#endif
+
+	bpf_for_each(bits, bit, ((char *)&data) + offset, 34)
+		nr++;
+	return nr;
+}
+
+SEC("syscall")
+__description("bits memalloc long-aligned")
+__success __retval(32) /* 16 * 2 */
+int bits_memalloc(void)
+{
+	char data[16];
+	int nr = 0;
+	int *bit;
+
+	__builtin_memset(&data, 0x48, sizeof(data));
+	bpf_for_each(bits, bit, &data, sizeof(data) * 8)
+		nr++;
+	return nr;
+}
+
+SEC("syscall")
+__description("bits memalloc non-long-aligned")
+__success __retval(85) /* 17 * 5*/
+int bits_memalloc_non_aligned(void)
+{
+	char data[17];
+	int nr = 0;
+	int *bit;
+
+	__builtin_memset(&data, 0x1f, sizeof(data));
+	bpf_for_each(bits, bit, &data, sizeof(data) * 8)
+		nr++;
+	return nr;
+}
+
+SEC("syscall")
+__description("bits memalloc non-aligned-bits")
+__success __retval(27) /* 8 * 3 + 3 */
+int bits_memalloc_non_aligned_bits(void)
+{
+	char data[16];
+	int nr = 0;
+	int *bit;
+
+	__builtin_memset(&data, 0x31, sizeof(data));
+	/* Different with all other bytes */
+	data[8] = 0xf7;
+
+	bpf_for_each(bits, bit, &data,  68)
+		nr++;
+	return nr;
+}
+
+
+SEC("syscall")
+__description("bit index")
+__success __retval(8)
+int bit_index(void)
+{
+	u64 data = 0x100;
+	int bit_idx = 0;
+	int *bit;
+
+	bpf_for_each(bits, bit, &data, 64) {
+		if (*bit == 0)
+			continue;
+		bit_idx = *bit;
+	}
+	return bit_idx;
+}
-- 
2.30.1 (Apple Git-130)

Re: [PATCH v7 bpf-next 1/2] bpf: Add bits iterator

[Andrii Nakryiko]

On Sun, May 5, 2024 at 8:35 PM Yafang Shao <laoar.shao@gmail.com> wrote:
>
> Add three new kfuncs for the bits iterator:
> - bpf_iter_bits_new
>   Initialize a new bits iterator for a given memory area. Due to the
>   limitation of bpf memalloc, the max number of bits that can be iterated
>   over is limited to (4096 * 8).
> - bpf_iter_bits_next
>   Get the next bit in a bpf_iter_bits
> - bpf_iter_bits_destroy
>   Destroy a bpf_iter_bits
>
> The bits iterator facilitates the iteration of the bits of a memory area,
> such as cpumask. It can be used in any context and on any address.
>
> Signed-off-by: Yafang Shao <laoar.shao@gmail.com>
> ---
>  kernel/bpf/helpers.c | 140 +++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 140 insertions(+)
>
> diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
> index 2a69a9a36c0f..83b2a02f795f 100644
> --- a/kernel/bpf/helpers.c
> +++ b/kernel/bpf/helpers.c
> @@ -2744,6 +2744,143 @@ __bpf_kfunc void bpf_preempt_enable(void)
>         preempt_enable();
>  }
>
> +struct bpf_iter_bits {
> +       __u64 __opaque[2];
> +} __aligned(8);
> +
> +struct bpf_iter_bits_kern {
> +       union {
> +               unsigned long *bits;
> +               unsigned long bits_copy;
> +       };
> +       u32 nr_bits;
> +       int bit;
> +} __aligned(8);
> +
> +/**
> + * bpf_iter_bits_new() - Initialize a new bits iterator for a given memory area
> + * @it: The new bpf_iter_bits to be created
> + * @unsafe_ptr__ign: A ponter pointing to a memory area to be iterated over

typo: pointer

> + * @nr_bits: The number of bits to be iterated over. Due to the limitation of
> + * memalloc, it can't greater than (4096 * 8).

typo: can't be greater

> + *
> + * This function initializes a new bpf_iter_bits structure for iterating over
> + * a memory area which is specified by the @unsafe_ptr__ign and @nr_bits. It
> + * copy the data of the memory area to the newly created bpf_iter_bits @it for

s/copy/copies/

> + * subsequent iteration operations.
> + *
> + * On success, 0 is returned. On failure, ERR is returned.
> + */
> +__bpf_kfunc int
> +bpf_iter_bits_new(struct bpf_iter_bits *it, const void *unsafe_ptr__ign, u32 nr_bits)
> +{
> +       struct bpf_iter_bits_kern *kit = (void *)it;
> +       u32 words = BITS_TO_LONGS(nr_bits);
> +       u32 size = BITS_TO_BYTES(nr_bits);
> +       u32 left, offset;
> +       int err;
> +
> +       BUILD_BUG_ON(sizeof(struct bpf_iter_bits_kern) != sizeof(struct bpf_iter_bits));
> +       BUILD_BUG_ON(__alignof__(struct bpf_iter_bits_kern) !=
> +                    __alignof__(struct bpf_iter_bits));
> +
> +       if (!unsafe_ptr__ign || !nr_bits) {
> +               kit->bits = NULL;
> +               return -EINVAL;
> +       }
> +
> +       kit->nr_bits = 0;
> +       kit->bits_copy = 0;
> +       /* Optimization for u64/u32 mask */
> +       if (nr_bits <= 64) {
> +               /* For big-endian, we must calculate the offset */
> +               offset = IS_ENABLED(CONFIG_S390) ? sizeof(u64) - size : 0;

S390 isn't the only big-endian architecture, it's wrong to hard-code just S390

there is __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__ check throughout the
kernel to do this detection

> +
> +               err = bpf_probe_read_kernel_common(((char *)&kit->bits_copy) + offset,
> +                                                  size, unsafe_ptr__ign);
> +               if (err)
> +                       return -EFAULT;

I'd rewrite the above to something like (not tested, but should give
the right idea):

long bits = 0;

err = bpf_probe_read_kernel_common(&bits, size, unsafe_ptr__ign);
if (err)
    return -EFAULT;

#if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
bits = __swab64(bits);
#endif

/* deal with bit mask of weird size, ensuring upper bits are zero */
bits <<= 64 - nr_bits;
bits >>= 64 - nr_bits;

kit->bits_copy = bits;


This should take care of both big-endianness, and non-multiple-of-8
sized bitmasks (I think, we need tests).

pw-bot: cr


> +
> +               kit->nr_bits = nr_bits;
> +               kit->bit = -1;
> +               return 0;
> +       }
> +
> +       /* Fallback to memalloc */
> +       kit->bits = bpf_mem_alloc(&bpf_global_ma, size);
> +       if (!kit->bits)
> +               return -ENOMEM;
> +
> +       err = bpf_probe_read_kernel_common(kit->bits, words * sizeof(u64), unsafe_ptr__ign);
> +       if (err) {
> +               bpf_mem_free(&bpf_global_ma, kit->bits);
> +               return err;
> +       }
> +
> +       /* long-aligned */
> +       left = size & (sizeof(u64) - 1);
> +       if (!left)
> +               goto out;
> +
> +       offset = IS_ENABLED(CONFIG_S390) ? sizeof(u64) - left : 0;
> +       err = bpf_probe_read_kernel_common((char *)(kit->bits + words - 1) + offset, left,
> +                                          unsafe_ptr__ign + (words - 1) * sizeof(u64));
> +       if (err) {
> +               bpf_mem_free(&bpf_global_ma, kit->bits);
> +               return err;
> +       }

tbh, I'm not sure what's the desired behavior here is. David (cc'ed),
you were dealing with cpumasks, how is the bit mask specified there?
Is it considered to be an long[] array or byte[] array? And how is
that working on big-endian, because I think it makes a difference?
Please take a look, thanks.

> +
> +out:
> +       kit->nr_bits = nr_bits;
> +       kit->bit = -1;
> +       return 0;
> +}
> +
> +/**
> + * bpf_iter_bits_next() - Get the next bit in a bpf_iter_bits
> + * @it: The bpf_iter_bits to be checked
> + *
> + * This function returns a pointer to a number representing the value of the
> + * next bit in the bits.
> + *
> + * If there are no further bit available, it returns NULL.
> + */
> +__bpf_kfunc int *bpf_iter_bits_next(struct bpf_iter_bits *it)
> +{
> +       struct bpf_iter_bits_kern *kit = (void *)it;
> +       u32 nr_bits = kit->nr_bits;
> +       const unsigned long *bits;
> +       int bit;
> +
> +       if (nr_bits == 0)
> +               return NULL;
> +
> +       bits = nr_bits <= 64 ? &kit->bits_copy : kit->bits;
> +       bit = find_next_bit(bits, nr_bits, kit->bit + 1);
> +       if (bit >= nr_bits) {
> +               kit->nr_bits = 0;
> +               return NULL;
> +       }
> +
> +       kit->bit = bit;
> +       return &kit->bit;
> +}
> +
> +/**
> + * bpf_iter_bits_destroy() - Destroy a bpf_iter_bits
> + * @it: The bpf_iter_bits to be destroyed
> + *
> + * Destroy the resource associated with the bpf_iter_bits.
> + */
> +__bpf_kfunc void bpf_iter_bits_destroy(struct bpf_iter_bits *it)
> +{
> +       struct bpf_iter_bits_kern *kit = (void *)it;
> +
> +       if (kit->nr_bits <= 64)
> +               return;
> +       bpf_mem_free(&bpf_global_ma, kit->bits);
> +}
> +
>  __bpf_kfunc_end_defs();
>
>  BTF_KFUNCS_START(generic_btf_ids)
> @@ -2826,6 +2963,9 @@ BTF_ID_FLAGS(func, bpf_wq_set_callback_impl)
>  BTF_ID_FLAGS(func, bpf_wq_start)
>  BTF_ID_FLAGS(func, bpf_preempt_disable)
>  BTF_ID_FLAGS(func, bpf_preempt_enable)
> +BTF_ID_FLAGS(func, bpf_iter_bits_new, KF_ITER_NEW)
> +BTF_ID_FLAGS(func, bpf_iter_bits_next, KF_ITER_NEXT | KF_RET_NULL)
> +BTF_ID_FLAGS(func, bpf_iter_bits_destroy, KF_ITER_DESTROY)
>  BTF_KFUNCS_END(common_btf_ids)
>
>  static const struct btf_kfunc_id_set common_kfunc_set = {
> --
> 2.30.1 (Apple Git-130)
>

Re: [PATCH v7 bpf-next 2/2] selftests/bpf: Add selftest for bits iter

[Andrii Nakryiko]

On Sun, May 5, 2024 at 8:35 PM Yafang Shao <laoar.shao@gmail.com> wrote:
>
> Add test cases for the bits iter:
> - positive case
>   - bit mask smaller than 8 bytes
>   - a typical case of having 8-byte bit mask
>   - another typical case where bit mask is > 8 bytes
>   - the index of set bit
>
> - nagative cases
>   - bpf_iter_bits_destroy() is required after calling
>     bpf_iter_bits_new()
>   - bpf_iter_bits_destroy() can only destroy an initialized iter
>   - bpf_iter_bits_next() must use an initialized iter
>
> Signed-off-by: Yafang Shao <laoar.shao@gmail.com>
> ---
>  .../selftests/bpf/prog_tests/verifier.c       |   2 +
>  .../selftests/bpf/progs/verifier_bits_iter.c  | 160 ++++++++++++++++++
>  2 files changed, 162 insertions(+)
>  create mode 100644 tools/testing/selftests/bpf/progs/verifier_bits_iter.c
>
> diff --git a/tools/testing/selftests/bpf/prog_tests/verifier.c b/tools/testing/selftests/bpf/prog_tests/verifier.c
> index c4f9f306646e..7e04ecaaa20a 100644
> --- a/tools/testing/selftests/bpf/prog_tests/verifier.c
> +++ b/tools/testing/selftests/bpf/prog_tests/verifier.c
> @@ -84,6 +84,7 @@
>  #include "verifier_xadd.skel.h"
>  #include "verifier_xdp.skel.h"
>  #include "verifier_xdp_direct_packet_access.skel.h"
> +#include "verifier_bits_iter.skel.h"
>
>  #define MAX_ENTRIES 11
>
> @@ -198,6 +199,7 @@ void test_verifier_var_off(void)              { RUN(verifier_var_off); }
>  void test_verifier_xadd(void)                 { RUN(verifier_xadd); }
>  void test_verifier_xdp(void)                  { RUN(verifier_xdp); }
>  void test_verifier_xdp_direct_packet_access(void) { RUN(verifier_xdp_direct_packet_access); }
> +void test_verifier_bits_iter(void) { RUN(verifier_bits_iter); }
>
>  static int init_test_val_map(struct bpf_object *obj, char *map_name)
>  {
> diff --git a/tools/testing/selftests/bpf/progs/verifier_bits_iter.c b/tools/testing/selftests/bpf/progs/verifier_bits_iter.c
> new file mode 100644
> index 000000000000..2f7b62b25638
> --- /dev/null
> +++ b/tools/testing/selftests/bpf/progs/verifier_bits_iter.c
> @@ -0,0 +1,160 @@
> +// SPDX-License-Identifier: GPL-2.0-only
> +/* Copyright (c) 2024 Yafang Shao <laoar.shao@gmail.com> */
> +
> +#include "vmlinux.h"
> +#include <bpf/bpf_helpers.h>
> +#include <bpf/bpf_tracing.h>
> +
> +#include "bpf_misc.h"
> +#include "task_kfunc_common.h"
> +
> +char _license[] SEC("license") = "GPL";
> +
> +int bpf_iter_bits_new(struct bpf_iter_bits *it, const void *unsafe_ptr__ign,
> +                     u32 nr_bits) __ksym __weak;
> +int *bpf_iter_bits_next(struct bpf_iter_bits *it) __ksym __weak;
> +void bpf_iter_bits_destroy(struct bpf_iter_bits *it) __ksym __weak;
> +
> +SEC("iter.s/cgroup")
> +__description("bits iter without destroy")
> +__failure __msg("Unreleased reference")
> +int BPF_PROG(no_destroy, struct bpf_iter_meta *meta, struct cgroup *cgrp)
> +{
> +       struct bpf_iter_bits it;
> +       struct task_struct *p;
> +
> +       p = bpf_task_from_pid(1);
> +       if (!p)
> +               return 1;
> +
> +       bpf_iter_bits_new(&it, p->cpus_ptr, 8192);
> +
> +       bpf_iter_bits_next(&it);
> +       bpf_task_release(p);
> +       return 0;
> +}
> +
> +SEC("iter/cgroup")
> +__description("bits iter with uninitialized iter in ->next()")
> +__failure __msg("expected an initialized iter_bits as arg #1")
> +int BPF_PROG(next_uninit, struct bpf_iter_meta *meta, struct cgroup *cgrp)
> +{
> +       struct bpf_iter_bits *it = NULL;
> +
> +       bpf_iter_bits_next(it);
> +       return 0;
> +}
> +
> +SEC("iter/cgroup")
> +__description("bits iter with uninitialized iter in ->destroy()")
> +__failure __msg("expected an initialized iter_bits as arg #1")
> +int BPF_PROG(destroy_uninit, struct bpf_iter_meta *meta, struct cgroup *cgrp)
> +{
> +       struct bpf_iter_bits it = {};
> +
> +       bpf_iter_bits_destroy(&it);
> +       return 0;
> +}
> +
> +SEC("syscall")
> +__description("bits copy 32")
> +__success __retval(10)
> +int bits_copy32(void)
> +{
> +       /* 21 bits:             --------------------- */
> +       u32 data = 0b11111101111101111100001000100101U;

if you define this bit mask as an array of bytes, then you won't have
to handle big-endian in the tests at all


> +       int nr = 0, offset = 0;
> +       int *bit;
> +
> +#if defined(__TARGET_ARCH_s390)
> +       offset = sizeof(u32) - (21 + 7) / 8;
> +#endif
> +       bpf_for_each(bits, bit, ((char *)&data) + offset, 21)
> +               nr++;
> +       return nr;
> +}
> +
> +SEC("syscall")
> +__description("bits copy 64")
> +__success __retval(18)
> +int bits_copy64(void)
> +{
> +       /* 34 bits:         ~-------- */
> +       u64 data = 0xffffefdf0f0f0f0fUL;
> +       int nr = 0, offset = 0;
> +       int *bit;
> +
> +#if defined(__TARGET_ARCH_s390)
> +       offset = sizeof(u64) - (34 + 7) / 8;
> +#endif
> +
> +       bpf_for_each(bits, bit, ((char *)&data) + offset, 34)

see above about byte array, but if we define different (not as byte
array but long[]), it would be cleaner to have

#if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
u64 data = 0x......UL;
#else
u64 data = 0x......UL;
#endif

wherer we'd hard-code bit masks in proper endianness in one place and
then just do a clean `bpf_for_each(bits, bit, &data, <len>) {}` calls

> +               nr++;
> +       return nr;
> +}
> +
> +SEC("syscall")
> +__description("bits memalloc long-aligned")
> +__success __retval(32) /* 16 * 2 */
> +int bits_memalloc(void)
> +{
> +       char data[16];
> +       int nr = 0;
> +       int *bit;
> +
> +       __builtin_memset(&data, 0x48, sizeof(data));
> +       bpf_for_each(bits, bit, &data, sizeof(data) * 8)
> +               nr++;
> +       return nr;
> +}
> +
> +SEC("syscall")
> +__description("bits memalloc non-long-aligned")
> +__success __retval(85) /* 17 * 5*/
> +int bits_memalloc_non_aligned(void)
> +{
> +       char data[17];
> +       int nr = 0;
> +       int *bit;
> +
> +       __builtin_memset(&data, 0x1f, sizeof(data));
> +       bpf_for_each(bits, bit, &data, sizeof(data) * 8)
> +               nr++;
> +       return nr;
> +}
> +
> +SEC("syscall")
> +__description("bits memalloc non-aligned-bits")
> +__success __retval(27) /* 8 * 3 + 3 */
> +int bits_memalloc_non_aligned_bits(void)
> +{
> +       char data[16];
> +       int nr = 0;
> +       int *bit;
> +
> +       __builtin_memset(&data, 0x31, sizeof(data));
> +       /* Different with all other bytes */
> +       data[8] = 0xf7;
> +
> +       bpf_for_each(bits, bit, &data,  68)
> +               nr++;
> +       return nr;
> +}
> +
> +
> +SEC("syscall")
> +__description("bit index")
> +__success __retval(8)
> +int bit_index(void)
> +{
> +       u64 data = 0x100;
> +       int bit_idx = 0;
> +       int *bit;
> +
> +       bpf_for_each(bits, bit, &data, 64) {
> +               if (*bit == 0)
> +                       continue;
> +               bit_idx = *bit;
> +       }
> +       return bit_idx;
> +}
> --
> 2.30.1 (Apple Git-130)
>

Re: [PATCH v7 bpf-next 1/2] bpf: Add bits iterator

[Yafang Shao]

On Tue, May 7, 2024 at 11:38 AM Andrii Nakryiko
<andrii.nakryiko@gmail.com> wrote:
>
> On Sun, May 5, 2024 at 8:35 PM Yafang Shao <laoar.shao@gmail.com> wrote:
> >
> > Add three new kfuncs for the bits iterator:
> > - bpf_iter_bits_new
> >   Initialize a new bits iterator for a given memory area. Due to the
> >   limitation of bpf memalloc, the max number of bits that can be iterated
> >   over is limited to (4096 * 8).
> > - bpf_iter_bits_next
> >   Get the next bit in a bpf_iter_bits
> > - bpf_iter_bits_destroy
> >   Destroy a bpf_iter_bits
> >
> > The bits iterator facilitates the iteration of the bits of a memory area,
> > such as cpumask. It can be used in any context and on any address.
> >
> > Signed-off-by: Yafang Shao <laoar.shao@gmail.com>
> > ---
> >  kernel/bpf/helpers.c | 140 +++++++++++++++++++++++++++++++++++++++++++
> >  1 file changed, 140 insertions(+)
> >
> > diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
> > index 2a69a9a36c0f..83b2a02f795f 100644
> > --- a/kernel/bpf/helpers.c
> > +++ b/kernel/bpf/helpers.c
> > @@ -2744,6 +2744,143 @@ __bpf_kfunc void bpf_preempt_enable(void)
> >         preempt_enable();
> >  }
> >
> > +struct bpf_iter_bits {
> > +       __u64 __opaque[2];
> > +} __aligned(8);
> > +
> > +struct bpf_iter_bits_kern {
> > +       union {
> > +               unsigned long *bits;
> > +               unsigned long bits_copy;
> > +       };
> > +       u32 nr_bits;
> > +       int bit;
> > +} __aligned(8);
> > +
> > +/**
> > + * bpf_iter_bits_new() - Initialize a new bits iterator for a given memory area
> > + * @it: The new bpf_iter_bits to be created
> > + * @unsafe_ptr__ign: A ponter pointing to a memory area to be iterated over
>
> typo: pointer

Thanks for the fix and the other fixes.

>
> > + * @nr_bits: The number of bits to be iterated over. Due to the limitation of
> > + * memalloc, it can't greater than (4096 * 8).
>
> typo: can't be greater
>
> > + *
> > + * This function initializes a new bpf_iter_bits structure for iterating over
> > + * a memory area which is specified by the @unsafe_ptr__ign and @nr_bits. It
> > + * copy the data of the memory area to the newly created bpf_iter_bits @it for
>
> s/copy/copies/
>
> > + * subsequent iteration operations.
> > + *
> > + * On success, 0 is returned. On failure, ERR is returned.
> > + */
> > +__bpf_kfunc int
> > +bpf_iter_bits_new(struct bpf_iter_bits *it, const void *unsafe_ptr__ign, u32 nr_bits)
> > +{
> > +       struct bpf_iter_bits_kern *kit = (void *)it;
> > +       u32 words = BITS_TO_LONGS(nr_bits);
> > +       u32 size = BITS_TO_BYTES(nr_bits);
> > +       u32 left, offset;
> > +       int err;
> > +
> > +       BUILD_BUG_ON(sizeof(struct bpf_iter_bits_kern) != sizeof(struct bpf_iter_bits));
> > +       BUILD_BUG_ON(__alignof__(struct bpf_iter_bits_kern) !=
> > +                    __alignof__(struct bpf_iter_bits));
> > +
> > +       if (!unsafe_ptr__ign || !nr_bits) {
> > +               kit->bits = NULL;
> > +               return -EINVAL;
> > +       }
> > +
> > +       kit->nr_bits = 0;
> > +       kit->bits_copy = 0;
> > +       /* Optimization for u64/u32 mask */
> > +       if (nr_bits <= 64) {
> > +               /* For big-endian, we must calculate the offset */
> > +               offset = IS_ENABLED(CONFIG_S390) ? sizeof(u64) - size : 0;
>
> S390 isn't the only big-endian architecture, it's wrong to hard-code just S390
>
> there is __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__ check throughout the
> kernel to do this detection

I missed that. will check it.

>
> > +
> > +               err = bpf_probe_read_kernel_common(((char *)&kit->bits_copy) + offset,
> > +                                                  size, unsafe_ptr__ign);
> > +               if (err)
> > +                       return -EFAULT;
>
> I'd rewrite the above to something like (not tested, but should give
> the right idea):
>
> long bits = 0;
>
> err = bpf_probe_read_kernel_common(&bits, size, unsafe_ptr__ign);
> if (err)
>     return -EFAULT;
>
> #if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
> bits = __swab64(bits);
> #endif
>
> /* deal with bit mask of weird size, ensuring upper bits are zero */
> bits <<= 64 - nr_bits;
> bits >>= 64 - nr_bits;
>
> kit->bits_copy = bits;
>
>
> This should take care of both big-endianness, and non-multiple-of-8
> sized bitmasks (I think, we need tests).

looks good, will change it.

>
> pw-bot: cr
>
>
> > +
> > +               kit->nr_bits = nr_bits;
> > +               kit->bit = -1;
> > +               return 0;
> > +       }
> > +
> > +       /* Fallback to memalloc */
> > +       kit->bits = bpf_mem_alloc(&bpf_global_ma, size);
> > +       if (!kit->bits)
> > +               return -ENOMEM;
> > +
> > +       err = bpf_probe_read_kernel_common(kit->bits, words * sizeof(u64), unsafe_ptr__ign);
> > +       if (err) {
> > +               bpf_mem_free(&bpf_global_ma, kit->bits);
> > +               return err;
> > +       }
> > +
> > +       /* long-aligned */
> > +       left = size & (sizeof(u64) - 1);
> > +       if (!left)
> > +               goto out;
> > +
> > +       offset = IS_ENABLED(CONFIG_S390) ? sizeof(u64) - left : 0;
> > +       err = bpf_probe_read_kernel_common((char *)(kit->bits + words - 1) + offset, left,
> > +                                          unsafe_ptr__ign + (words - 1) * sizeof(u64));
> > +       if (err) {
> > +               bpf_mem_free(&bpf_global_ma, kit->bits);
> > +               return err;
> > +       }
>
> tbh, I'm not sure what's the desired behavior here is. David (cc'ed),
> you were dealing with cpumasks, how is the bit mask specified there?
> Is it considered to be an long[] array or byte[] array? And how is
> that working on big-endian, because I think it makes a difference?
> Please take a look, thanks.

The function find_next_bit() requires the pointer to be of type
"unsigned long *", hence, we must ensure consistency by converting it
here as well. As cpumask represents a bitmap and is always of type
"unsigned long *", it remains unaffected by endianness considerations.

>
> > +
> > +out:
> > +       kit->nr_bits = nr_bits;
> > +       kit->bit = -1;
> > +       return 0;
> > +}
> > +
> > +/**
> > + * bpf_iter_bits_next() - Get the next bit in a bpf_iter_bits
> > + * @it: The bpf_iter_bits to be checked
> > + *
> > + * This function returns a pointer to a number representing the value of the
> > + * next bit in the bits.
> > + *
> > + * If there are no further bit available, it returns NULL.
> > + */
> > +__bpf_kfunc int *bpf_iter_bits_next(struct bpf_iter_bits *it)
> > +{
> > +       struct bpf_iter_bits_kern *kit = (void *)it;
> > +       u32 nr_bits = kit->nr_bits;
> > +       const unsigned long *bits;
> > +       int bit;
> > +
> > +       if (nr_bits == 0)
> > +               return NULL;
> > +
> > +       bits = nr_bits <= 64 ? &kit->bits_copy : kit->bits;
> > +       bit = find_next_bit(bits, nr_bits, kit->bit + 1);
> > +       if (bit >= nr_bits) {
> > +               kit->nr_bits = 0;
> > +               return NULL;
> > +       }
> > +
> > +       kit->bit = bit;
> > +       return &kit->bit;
> > +}
> > +
> > +/**
> > + * bpf_iter_bits_destroy() - Destroy a bpf_iter_bits
> > + * @it: The bpf_iter_bits to be destroyed
> > + *
> > + * Destroy the resource associated with the bpf_iter_bits.
> > + */
> > +__bpf_kfunc void bpf_iter_bits_destroy(struct bpf_iter_bits *it)
> > +{
> > +       struct bpf_iter_bits_kern *kit = (void *)it;
> > +
> > +       if (kit->nr_bits <= 64)
> > +               return;
> > +       bpf_mem_free(&bpf_global_ma, kit->bits);
> > +}
> > +
> >  __bpf_kfunc_end_defs();
> >
> >  BTF_KFUNCS_START(generic_btf_ids)
> > @@ -2826,6 +2963,9 @@ BTF_ID_FLAGS(func, bpf_wq_set_callback_impl)
> >  BTF_ID_FLAGS(func, bpf_wq_start)
> >  BTF_ID_FLAGS(func, bpf_preempt_disable)
> >  BTF_ID_FLAGS(func, bpf_preempt_enable)
> > +BTF_ID_FLAGS(func, bpf_iter_bits_new, KF_ITER_NEW)
> > +BTF_ID_FLAGS(func, bpf_iter_bits_next, KF_ITER_NEXT | KF_RET_NULL)
> > +BTF_ID_FLAGS(func, bpf_iter_bits_destroy, KF_ITER_DESTROY)
> >  BTF_KFUNCS_END(common_btf_ids)
> >
> >  static const struct btf_kfunc_id_set common_kfunc_set = {
> > --
> > 2.30.1 (Apple Git-130)
> >



-- 
Regards
Yafang

Re: [PATCH v7 bpf-next 2/2] selftests/bpf: Add selftest for bits iter

[Yafang Shao]

On Tue, May 7, 2024 at 11:42 AM Andrii Nakryiko
<andrii.nakryiko@gmail.com> wrote:
>
> On Sun, May 5, 2024 at 8:35 PM Yafang Shao <laoar.shao@gmail.com> wrote:
> >
> > Add test cases for the bits iter:
> > - positive case
> >   - bit mask smaller than 8 bytes
> >   - a typical case of having 8-byte bit mask
> >   - another typical case where bit mask is > 8 bytes
> >   - the index of set bit
> >
> > - nagative cases
> >   - bpf_iter_bits_destroy() is required after calling
> >     bpf_iter_bits_new()
> >   - bpf_iter_bits_destroy() can only destroy an initialized iter
> >   - bpf_iter_bits_next() must use an initialized iter
> >
> > Signed-off-by: Yafang Shao <laoar.shao@gmail.com>
> > ---
> >  .../selftests/bpf/prog_tests/verifier.c       |   2 +
> >  .../selftests/bpf/progs/verifier_bits_iter.c  | 160 ++++++++++++++++++
> >  2 files changed, 162 insertions(+)
> >  create mode 100644 tools/testing/selftests/bpf/progs/verifier_bits_iter.c
> >
> > diff --git a/tools/testing/selftests/bpf/prog_tests/verifier.c b/tools/testing/selftests/bpf/prog_tests/verifier.c
> > index c4f9f306646e..7e04ecaaa20a 100644
> > --- a/tools/testing/selftests/bpf/prog_tests/verifier.c
> > +++ b/tools/testing/selftests/bpf/prog_tests/verifier.c
> > @@ -84,6 +84,7 @@
> >  #include "verifier_xadd.skel.h"
> >  #include "verifier_xdp.skel.h"
> >  #include "verifier_xdp_direct_packet_access.skel.h"
> > +#include "verifier_bits_iter.skel.h"
> >
> >  #define MAX_ENTRIES 11
> >
> > @@ -198,6 +199,7 @@ void test_verifier_var_off(void)              { RUN(verifier_var_off); }
> >  void test_verifier_xadd(void)                 { RUN(verifier_xadd); }
> >  void test_verifier_xdp(void)                  { RUN(verifier_xdp); }
> >  void test_verifier_xdp_direct_packet_access(void) { RUN(verifier_xdp_direct_packet_access); }
> > +void test_verifier_bits_iter(void) { RUN(verifier_bits_iter); }
> >
> >  static int init_test_val_map(struct bpf_object *obj, char *map_name)
> >  {
> > diff --git a/tools/testing/selftests/bpf/progs/verifier_bits_iter.c b/tools/testing/selftests/bpf/progs/verifier_bits_iter.c
> > new file mode 100644
> > index 000000000000..2f7b62b25638
> > --- /dev/null
> > +++ b/tools/testing/selftests/bpf/progs/verifier_bits_iter.c
> > @@ -0,0 +1,160 @@
> > +// SPDX-License-Identifier: GPL-2.0-only
> > +/* Copyright (c) 2024 Yafang Shao <laoar.shao@gmail.com> */
> > +
> > +#include "vmlinux.h"
> > +#include <bpf/bpf_helpers.h>
> > +#include <bpf/bpf_tracing.h>
> > +
> > +#include "bpf_misc.h"
> > +#include "task_kfunc_common.h"
> > +
> > +char _license[] SEC("license") = "GPL";
> > +
> > +int bpf_iter_bits_new(struct bpf_iter_bits *it, const void *unsafe_ptr__ign,
> > +                     u32 nr_bits) __ksym __weak;
> > +int *bpf_iter_bits_next(struct bpf_iter_bits *it) __ksym __weak;
> > +void bpf_iter_bits_destroy(struct bpf_iter_bits *it) __ksym __weak;
> > +
> > +SEC("iter.s/cgroup")
> > +__description("bits iter without destroy")
> > +__failure __msg("Unreleased reference")
> > +int BPF_PROG(no_destroy, struct bpf_iter_meta *meta, struct cgroup *cgrp)
> > +{
> > +       struct bpf_iter_bits it;
> > +       struct task_struct *p;
> > +
> > +       p = bpf_task_from_pid(1);
> > +       if (!p)
> > +               return 1;
> > +
> > +       bpf_iter_bits_new(&it, p->cpus_ptr, 8192);
> > +
> > +       bpf_iter_bits_next(&it);
> > +       bpf_task_release(p);
> > +       return 0;
> > +}
> > +
> > +SEC("iter/cgroup")
> > +__description("bits iter with uninitialized iter in ->next()")
> > +__failure __msg("expected an initialized iter_bits as arg #1")
> > +int BPF_PROG(next_uninit, struct bpf_iter_meta *meta, struct cgroup *cgrp)
> > +{
> > +       struct bpf_iter_bits *it = NULL;
> > +
> > +       bpf_iter_bits_next(it);
> > +       return 0;
> > +}
> > +
> > +SEC("iter/cgroup")
> > +__description("bits iter with uninitialized iter in ->destroy()")
> > +__failure __msg("expected an initialized iter_bits as arg #1")
> > +int BPF_PROG(destroy_uninit, struct bpf_iter_meta *meta, struct cgroup *cgrp)
> > +{
> > +       struct bpf_iter_bits it = {};
> > +
> > +       bpf_iter_bits_destroy(&it);
> > +       return 0;
> > +}
> > +
> > +SEC("syscall")
> > +__description("bits copy 32")
> > +__success __retval(10)
> > +int bits_copy32(void)
> > +{
> > +       /* 21 bits:             --------------------- */
> > +       u32 data = 0b11111101111101111100001000100101U;
>
> if you define this bit mask as an array of bytes, then you won't have
> to handle big-endian in the tests at all

This test case provides a clear example of iterating over data of type
u32, offering valuable guidance for users who need to perform such
iterations.

>
>
> > +       int nr = 0, offset = 0;
> > +       int *bit;
> > +
> > +#if defined(__TARGET_ARCH_s390)
> > +       offset = sizeof(u32) - (21 + 7) / 8;
> > +#endif
> > +       bpf_for_each(bits, bit, ((char *)&data) + offset, 21)
> > +               nr++;
> > +       return nr;
> > +}
> > +
> > +SEC("syscall")
> > +__description("bits copy 64")
> > +__success __retval(18)
> > +int bits_copy64(void)
> > +{
> > +       /* 34 bits:         ~-------- */
> > +       u64 data = 0xffffefdf0f0f0f0fUL;
> > +       int nr = 0, offset = 0;
> > +       int *bit;
> > +
> > +#if defined(__TARGET_ARCH_s390)
> > +       offset = sizeof(u64) - (34 + 7) / 8;
> > +#endif
> > +
> > +       bpf_for_each(bits, bit, ((char *)&data) + offset, 34)
>
> see above about byte array, but if we define different (not as byte
> array but long[]), it would be cleaner to have

This test case demonstrates how to iterate over data of type u64.

>
> #if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
> u64 data = 0x......UL;
> #else
> u64 data = 0x......UL;
> #endif

looks good.

>
> wherer we'd hard-code bit masks in proper endianness in one place and
> then just do a clean `bpf_for_each(bits, bit, &data, <len>) {}` calls
>
> > +               nr++;
> > +       return nr;
> > +}
> > +
> > +SEC("syscall")
> > +__description("bits memalloc long-aligned")
> > +__success __retval(32) /* 16 * 2 */
> > +int bits_memalloc(void)
> > +{
> > +       char data[16];
> > +       int nr = 0;
> > +       int *bit;
> > +
> > +       __builtin_memset(&data, 0x48, sizeof(data));
> > +       bpf_for_each(bits, bit, &data, sizeof(data) * 8)
> > +               nr++;
> > +       return nr;
> > +}
> > +
> > +SEC("syscall")
> > +__description("bits memalloc non-long-aligned")
> > +__success __retval(85) /* 17 * 5*/
> > +int bits_memalloc_non_aligned(void)
> > +{
> > +       char data[17];
> > +       int nr = 0;
> > +       int *bit;
> > +
> > +       __builtin_memset(&data, 0x1f, sizeof(data));
> > +       bpf_for_each(bits, bit, &data, sizeof(data) * 8)
> > +               nr++;
> > +       return nr;
> > +}
> > +
> > +SEC("syscall")
> > +__description("bits memalloc non-aligned-bits")
> > +__success __retval(27) /* 8 * 3 + 3 */
> > +int bits_memalloc_non_aligned_bits(void)
> > +{
> > +       char data[16];
> > +       int nr = 0;
> > +       int *bit;
> > +
> > +       __builtin_memset(&data, 0x31, sizeof(data));
> > +       /* Different with all other bytes */
> > +       data[8] = 0xf7;
> > +
> > +       bpf_for_each(bits, bit, &data,  68)
> > +               nr++;
> > +       return nr;
> > +}
> > +
> > +
> > +SEC("syscall")
> > +__description("bit index")
> > +__success __retval(8)
> > +int bit_index(void)
> > +{
> > +       u64 data = 0x100;
> > +       int bit_idx = 0;
> > +       int *bit;
> > +
> > +       bpf_for_each(bits, bit, &data, 64) {
> > +               if (*bit == 0)
> > +                       continue;
> > +               bit_idx = *bit;
> > +       }
> > +       return bit_idx;
> > +}
> > --
> > 2.30.1 (Apple Git-130)
> >



-- 
Regards
Yafang

Re: [PATCH v7 bpf-next 1/2] bpf: Add bits iterator

[Andrii Nakryiko]

On Tue, May 7, 2024 at 6:32 AM Yafang Shao <laoar.shao@gmail.com> wrote:
>
> On Tue, May 7, 2024 at 11:38 AM Andrii Nakryiko
> <andrii.nakryiko@gmail.com> wrote:
> >
> > On Sun, May 5, 2024 at 8:35 PM Yafang Shao <laoar.shao@gmail.com> wrote:
> > >
> > > Add three new kfuncs for the bits iterator:
> > > - bpf_iter_bits_new
> > >   Initialize a new bits iterator for a given memory area. Due to the
> > >   limitation of bpf memalloc, the max number of bits that can be iterated
> > >   over is limited to (4096 * 8).
> > > - bpf_iter_bits_next
> > >   Get the next bit in a bpf_iter_bits
> > > - bpf_iter_bits_destroy
> > >   Destroy a bpf_iter_bits
> > >
> > > The bits iterator facilitates the iteration of the bits of a memory area,
> > > such as cpumask. It can be used in any context and on any address.
> > >
> > > Signed-off-by: Yafang Shao <laoar.shao@gmail.com>
> > > ---
> > >  kernel/bpf/helpers.c | 140 +++++++++++++++++++++++++++++++++++++++++++
> > >  1 file changed, 140 insertions(+)
> > >
> > > diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
> > > index 2a69a9a36c0f..83b2a02f795f 100644
> > > --- a/kernel/bpf/helpers.c
> > > +++ b/kernel/bpf/helpers.c
> > > @@ -2744,6 +2744,143 @@ __bpf_kfunc void bpf_preempt_enable(void)
> > >         preempt_enable();
> > >  }
> > >
> > > +struct bpf_iter_bits {
> > > +       __u64 __opaque[2];
> > > +} __aligned(8);
> > > +
> > > +struct bpf_iter_bits_kern {
> > > +       union {
> > > +               unsigned long *bits;
> > > +               unsigned long bits_copy;
> > > +       };
> > > +       u32 nr_bits;
> > > +       int bit;
> > > +} __aligned(8);
> > > +
> > > +/**
> > > + * bpf_iter_bits_new() - Initialize a new bits iterator for a given memory area
> > > + * @it: The new bpf_iter_bits to be created
> > > + * @unsafe_ptr__ign: A ponter pointing to a memory area to be iterated over
> >
> > typo: pointer
>
> Thanks for the fix and the other fixes.
>
> >
> > > + * @nr_bits: The number of bits to be iterated over. Due to the limitation of
> > > + * memalloc, it can't greater than (4096 * 8).
> >
> > typo: can't be greater
> >
> > > + *
> > > + * This function initializes a new bpf_iter_bits structure for iterating over
> > > + * a memory area which is specified by the @unsafe_ptr__ign and @nr_bits. It
> > > + * copy the data of the memory area to the newly created bpf_iter_bits @it for
> >
> > s/copy/copies/
> >
> > > + * subsequent iteration operations.
> > > + *
> > > + * On success, 0 is returned. On failure, ERR is returned.
> > > + */
> > > +__bpf_kfunc int
> > > +bpf_iter_bits_new(struct bpf_iter_bits *it, const void *unsafe_ptr__ign, u32 nr_bits)
> > > +{
> > > +       struct bpf_iter_bits_kern *kit = (void *)it;
> > > +       u32 words = BITS_TO_LONGS(nr_bits);
> > > +       u32 size = BITS_TO_BYTES(nr_bits);
> > > +       u32 left, offset;
> > > +       int err;
> > > +
> > > +       BUILD_BUG_ON(sizeof(struct bpf_iter_bits_kern) != sizeof(struct bpf_iter_bits));
> > > +       BUILD_BUG_ON(__alignof__(struct bpf_iter_bits_kern) !=
> > > +                    __alignof__(struct bpf_iter_bits));
> > > +
> > > +       if (!unsafe_ptr__ign || !nr_bits) {
> > > +               kit->bits = NULL;
> > > +               return -EINVAL;
> > > +       }
> > > +
> > > +       kit->nr_bits = 0;
> > > +       kit->bits_copy = 0;
> > > +       /* Optimization for u64/u32 mask */
> > > +       if (nr_bits <= 64) {
> > > +               /* For big-endian, we must calculate the offset */
> > > +               offset = IS_ENABLED(CONFIG_S390) ? sizeof(u64) - size : 0;
> >
> > S390 isn't the only big-endian architecture, it's wrong to hard-code just S390
> >
> > there is __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__ check throughout the
> > kernel to do this detection
>
> I missed that. will check it.
>
> >
> > > +
> > > +               err = bpf_probe_read_kernel_common(((char *)&kit->bits_copy) + offset,
> > > +                                                  size, unsafe_ptr__ign);
> > > +               if (err)
> > > +                       return -EFAULT;
> >
> > I'd rewrite the above to something like (not tested, but should give
> > the right idea):
> >
> > long bits = 0;
> >
> > err = bpf_probe_read_kernel_common(&bits, size, unsafe_ptr__ign);
> > if (err)
> >     return -EFAULT;
> >
> > #if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
> > bits = __swab64(bits);
> > #endif
> >
> > /* deal with bit mask of weird size, ensuring upper bits are zero */
> > bits <<= 64 - nr_bits;
> > bits >>= 64 - nr_bits;
> >
> > kit->bits_copy = bits;
> >
> >
> > This should take care of both big-endianness, and non-multiple-of-8
> > sized bitmasks (I think, we need tests).
>
> looks good, will change it.
>
> >
> > pw-bot: cr
> >
> >
> > > +
> > > +               kit->nr_bits = nr_bits;
> > > +               kit->bit = -1;
> > > +               return 0;
> > > +       }
> > > +
> > > +       /* Fallback to memalloc */
> > > +       kit->bits = bpf_mem_alloc(&bpf_global_ma, size);
> > > +       if (!kit->bits)
> > > +               return -ENOMEM;
> > > +
> > > +       err = bpf_probe_read_kernel_common(kit->bits, words * sizeof(u64), unsafe_ptr__ign);
> > > +       if (err) {
> > > +               bpf_mem_free(&bpf_global_ma, kit->bits);
> > > +               return err;
> > > +       }
> > > +
> > > +       /* long-aligned */
> > > +       left = size & (sizeof(u64) - 1);
> > > +       if (!left)
> > > +               goto out;
> > > +
> > > +       offset = IS_ENABLED(CONFIG_S390) ? sizeof(u64) - left : 0;
> > > +       err = bpf_probe_read_kernel_common((char *)(kit->bits + words - 1) + offset, left,
> > > +                                          unsafe_ptr__ign + (words - 1) * sizeof(u64));
> > > +       if (err) {
> > > +               bpf_mem_free(&bpf_global_ma, kit->bits);
> > > +               return err;
> > > +       }
> >
> > tbh, I'm not sure what's the desired behavior here is. David (cc'ed),
> > you were dealing with cpumasks, how is the bit mask specified there?
> > Is it considered to be an long[] array or byte[] array? And how is
> > that working on big-endian, because I think it makes a difference?
> > Please take a look, thanks.
>
> The function find_next_bit() requires the pointer to be of type
> "unsigned long *", hence, we must ensure consistency by converting it
> here as well. As cpumask represents a bitmap and is always of type
> "unsigned long *", it remains unaffected by endianness considerations.
>

Right, but the question is whether this iterator should make the same
simplifying assumption or not? I think the motivation for this
iterator was the ability to iterate over CPU masks, so I'm asking (and
that's why I cc'ed David) what we should do to make it work well for
CPU masks.

> >
> > > +
> > > +out:
> > > +       kit->nr_bits = nr_bits;
> > > +       kit->bit = -1;
> > > +       return 0;
> > > +}
> > > +
> > > +/**
> > > + * bpf_iter_bits_next() - Get the next bit in a bpf_iter_bits
> > > + * @it: The bpf_iter_bits to be checked
> > > + *
> > > + * This function returns a pointer to a number representing the value of the
> > > + * next bit in the bits.
> > > + *
> > > + * If there are no further bit available, it returns NULL.
> > > + */
> > > +__bpf_kfunc int *bpf_iter_bits_next(struct bpf_iter_bits *it)
> > > +{
> > > +       struct bpf_iter_bits_kern *kit = (void *)it;
> > > +       u32 nr_bits = kit->nr_bits;
> > > +       const unsigned long *bits;
> > > +       int bit;
> > > +
> > > +       if (nr_bits == 0)
> > > +               return NULL;
> > > +
> > > +       bits = nr_bits <= 64 ? &kit->bits_copy : kit->bits;
> > > +       bit = find_next_bit(bits, nr_bits, kit->bit + 1);
> > > +       if (bit >= nr_bits) {
> > > +               kit->nr_bits = 0;
> > > +               return NULL;
> > > +       }
> > > +
> > > +       kit->bit = bit;
> > > +       return &kit->bit;
> > > +}
> > > +
> > > +/**
> > > + * bpf_iter_bits_destroy() - Destroy a bpf_iter_bits
> > > + * @it: The bpf_iter_bits to be destroyed
> > > + *
> > > + * Destroy the resource associated with the bpf_iter_bits.
> > > + */
> > > +__bpf_kfunc void bpf_iter_bits_destroy(struct bpf_iter_bits *it)
> > > +{
> > > +       struct bpf_iter_bits_kern *kit = (void *)it;
> > > +
> > > +       if (kit->nr_bits <= 64)
> > > +               return;
> > > +       bpf_mem_free(&bpf_global_ma, kit->bits);
> > > +}
> > > +
> > >  __bpf_kfunc_end_defs();
> > >
> > >  BTF_KFUNCS_START(generic_btf_ids)
> > > @@ -2826,6 +2963,9 @@ BTF_ID_FLAGS(func, bpf_wq_set_callback_impl)
> > >  BTF_ID_FLAGS(func, bpf_wq_start)
> > >  BTF_ID_FLAGS(func, bpf_preempt_disable)
> > >  BTF_ID_FLAGS(func, bpf_preempt_enable)
> > > +BTF_ID_FLAGS(func, bpf_iter_bits_new, KF_ITER_NEW)
> > > +BTF_ID_FLAGS(func, bpf_iter_bits_next, KF_ITER_NEXT | KF_RET_NULL)
> > > +BTF_ID_FLAGS(func, bpf_iter_bits_destroy, KF_ITER_DESTROY)
> > >  BTF_KFUNCS_END(common_btf_ids)
> > >
> > >  static const struct btf_kfunc_id_set common_kfunc_set = {
> > > --
> > > 2.30.1 (Apple Git-130)
> > >
>
>
>
> --
> Regards
> Yafang

Re: [PATCH v7 bpf-next 2/2] selftests/bpf: Add selftest for bits iter

[Andrii Nakryiko]

On Tue, May 7, 2024 at 6:39 AM Yafang Shao <laoar.shao@gmail.com> wrote:
>
> On Tue, May 7, 2024 at 11:42 AM Andrii Nakryiko
> <andrii.nakryiko@gmail.com> wrote:
> >
> > On Sun, May 5, 2024 at 8:35 PM Yafang Shao <laoar.shao@gmail.com> wrote:
> > >
> > > Add test cases for the bits iter:
> > > - positive case
> > >   - bit mask smaller than 8 bytes
> > >   - a typical case of having 8-byte bit mask
> > >   - another typical case where bit mask is > 8 bytes
> > >   - the index of set bit
> > >
> > > - nagative cases
> > >   - bpf_iter_bits_destroy() is required after calling
> > >     bpf_iter_bits_new()
> > >   - bpf_iter_bits_destroy() can only destroy an initialized iter
> > >   - bpf_iter_bits_next() must use an initialized iter
> > >
> > > Signed-off-by: Yafang Shao <laoar.shao@gmail.com>
> > > ---
> > >  .../selftests/bpf/prog_tests/verifier.c       |   2 +
> > >  .../selftests/bpf/progs/verifier_bits_iter.c  | 160 ++++++++++++++++++
> > >  2 files changed, 162 insertions(+)
> > >  create mode 100644 tools/testing/selftests/bpf/progs/verifier_bits_iter.c
> > >
> > > diff --git a/tools/testing/selftests/bpf/prog_tests/verifier.c b/tools/testing/selftests/bpf/prog_tests/verifier.c
> > > index c4f9f306646e..7e04ecaaa20a 100644
> > > --- a/tools/testing/selftests/bpf/prog_tests/verifier.c
> > > +++ b/tools/testing/selftests/bpf/prog_tests/verifier.c
> > > @@ -84,6 +84,7 @@
> > >  #include "verifier_xadd.skel.h"
> > >  #include "verifier_xdp.skel.h"
> > >  #include "verifier_xdp_direct_packet_access.skel.h"
> > > +#include "verifier_bits_iter.skel.h"
> > >
> > >  #define MAX_ENTRIES 11
> > >
> > > @@ -198,6 +199,7 @@ void test_verifier_var_off(void)              { RUN(verifier_var_off); }
> > >  void test_verifier_xadd(void)                 { RUN(verifier_xadd); }
> > >  void test_verifier_xdp(void)                  { RUN(verifier_xdp); }
> > >  void test_verifier_xdp_direct_packet_access(void) { RUN(verifier_xdp_direct_packet_access); }
> > > +void test_verifier_bits_iter(void) { RUN(verifier_bits_iter); }
> > >
> > >  static int init_test_val_map(struct bpf_object *obj, char *map_name)
> > >  {
> > > diff --git a/tools/testing/selftests/bpf/progs/verifier_bits_iter.c b/tools/testing/selftests/bpf/progs/verifier_bits_iter.c
> > > new file mode 100644
> > > index 000000000000..2f7b62b25638
> > > --- /dev/null
> > > +++ b/tools/testing/selftests/bpf/progs/verifier_bits_iter.c
> > > @@ -0,0 +1,160 @@
> > > +// SPDX-License-Identifier: GPL-2.0-only
> > > +/* Copyright (c) 2024 Yafang Shao <laoar.shao@gmail.com> */
> > > +
> > > +#include "vmlinux.h"
> > > +#include <bpf/bpf_helpers.h>
> > > +#include <bpf/bpf_tracing.h>
> > > +
> > > +#include "bpf_misc.h"
> > > +#include "task_kfunc_common.h"
> > > +
> > > +char _license[] SEC("license") = "GPL";
> > > +
> > > +int bpf_iter_bits_new(struct bpf_iter_bits *it, const void *unsafe_ptr__ign,
> > > +                     u32 nr_bits) __ksym __weak;
> > > +int *bpf_iter_bits_next(struct bpf_iter_bits *it) __ksym __weak;
> > > +void bpf_iter_bits_destroy(struct bpf_iter_bits *it) __ksym __weak;
> > > +
> > > +SEC("iter.s/cgroup")
> > > +__description("bits iter without destroy")
> > > +__failure __msg("Unreleased reference")
> > > +int BPF_PROG(no_destroy, struct bpf_iter_meta *meta, struct cgroup *cgrp)
> > > +{
> > > +       struct bpf_iter_bits it;
> > > +       struct task_struct *p;
> > > +
> > > +       p = bpf_task_from_pid(1);
> > > +       if (!p)
> > > +               return 1;
> > > +
> > > +       bpf_iter_bits_new(&it, p->cpus_ptr, 8192);
> > > +
> > > +       bpf_iter_bits_next(&it);
> > > +       bpf_task_release(p);
> > > +       return 0;
> > > +}
> > > +
> > > +SEC("iter/cgroup")
> > > +__description("bits iter with uninitialized iter in ->next()")
> > > +__failure __msg("expected an initialized iter_bits as arg #1")
> > > +int BPF_PROG(next_uninit, struct bpf_iter_meta *meta, struct cgroup *cgrp)
> > > +{
> > > +       struct bpf_iter_bits *it = NULL;
> > > +
> > > +       bpf_iter_bits_next(it);
> > > +       return 0;
> > > +}
> > > +
> > > +SEC("iter/cgroup")
> > > +__description("bits iter with uninitialized iter in ->destroy()")
> > > +__failure __msg("expected an initialized iter_bits as arg #1")
> > > +int BPF_PROG(destroy_uninit, struct bpf_iter_meta *meta, struct cgroup *cgrp)
> > > +{
> > > +       struct bpf_iter_bits it = {};
> > > +
> > > +       bpf_iter_bits_destroy(&it);
> > > +       return 0;
> > > +}
> > > +
> > > +SEC("syscall")
> > > +__description("bits copy 32")
> > > +__success __retval(10)
> > > +int bits_copy32(void)
> > > +{
> > > +       /* 21 bits:             --------------------- */
> > > +       u32 data = 0b11111101111101111100001000100101U;
> >
> > if you define this bit mask as an array of bytes, then you won't have
> > to handle big-endian in the tests at all
>
> This test case provides a clear example of iterating over data of type
> u32, offering valuable guidance for users who need to perform such
> iterations.
>
> >
> >
> > > +       int nr = 0, offset = 0;
> > > +       int *bit;
> > > +
> > > +#if defined(__TARGET_ARCH_s390)
> > > +       offset = sizeof(u32) - (21 + 7) / 8;
> > > +#endif
> > > +       bpf_for_each(bits, bit, ((char *)&data) + offset, 21)
> > > +               nr++;
> > > +       return nr;
> > > +}
> > > +
> > > +SEC("syscall")
> > > +__description("bits copy 64")
> > > +__success __retval(18)
> > > +int bits_copy64(void)
> > > +{
> > > +       /* 34 bits:         ~-------- */
> > > +       u64 data = 0xffffefdf0f0f0f0fUL;
> > > +       int nr = 0, offset = 0;
> > > +       int *bit;
> > > +
> > > +#if defined(__TARGET_ARCH_s390)
> > > +       offset = sizeof(u64) - (34 + 7) / 8;
> > > +#endif
> > > +
> > > +       bpf_for_each(bits, bit, ((char *)&data) + offset, 34)
> >
> > see above about byte array, but if we define different (not as byte
> > array but long[]), it would be cleaner to have
>
> This test case demonstrates how to iterate over data of type u64.
>
> >
> > #if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
> > u64 data = 0x......UL;
> > #else
> > u64 data = 0x......UL;
> > #endif
>
> looks good.
>

Please hold off on sending a new revision until we figure out what the
contract should be. Because I feel like it's a (relatively) big
decision whether a bit mask is treated as an array of bytes or as an
array of longs. For little-endian it makes no difference, but for
big-endian it's a big difference and has usability and performance
implications.

> >
> > wherer we'd hard-code bit masks in proper endianness in one place and
> > then just do a clean `bpf_for_each(bits, bit, &data, <len>) {}` calls
> >
> > > +               nr++;
> > > +       return nr;
> > > +}
> > > +
> > > +SEC("syscall")
> > > +__description("bits memalloc long-aligned")
> > > +__success __retval(32) /* 16 * 2 */
> > > +int bits_memalloc(void)
> > > +{
> > > +       char data[16];
> > > +       int nr = 0;
> > > +       int *bit;
> > > +
> > > +       __builtin_memset(&data, 0x48, sizeof(data));
> > > +       bpf_for_each(bits, bit, &data, sizeof(data) * 8)
> > > +               nr++;
> > > +       return nr;
> > > +}
> > > +
> > > +SEC("syscall")
> > > +__description("bits memalloc non-long-aligned")
> > > +__success __retval(85) /* 17 * 5*/
> > > +int bits_memalloc_non_aligned(void)
> > > +{
> > > +       char data[17];
> > > +       int nr = 0;
> > > +       int *bit;
> > > +
> > > +       __builtin_memset(&data, 0x1f, sizeof(data));
> > > +       bpf_for_each(bits, bit, &data, sizeof(data) * 8)
> > > +               nr++;
> > > +       return nr;
> > > +}
> > > +
> > > +SEC("syscall")
> > > +__description("bits memalloc non-aligned-bits")
> > > +__success __retval(27) /* 8 * 3 + 3 */
> > > +int bits_memalloc_non_aligned_bits(void)
> > > +{
> > > +       char data[16];
> > > +       int nr = 0;
> > > +       int *bit;
> > > +
> > > +       __builtin_memset(&data, 0x31, sizeof(data));
> > > +       /* Different with all other bytes */
> > > +       data[8] = 0xf7;
> > > +
> > > +       bpf_for_each(bits, bit, &data,  68)
> > > +               nr++;
> > > +       return nr;
> > > +}
> > > +
> > > +
> > > +SEC("syscall")
> > > +__description("bit index")
> > > +__success __retval(8)
> > > +int bit_index(void)
> > > +{
> > > +       u64 data = 0x100;
> > > +       int bit_idx = 0;
> > > +       int *bit;
> > > +
> > > +       bpf_for_each(bits, bit, &data, 64) {
> > > +               if (*bit == 0)
> > > +                       continue;
> > > +               bit_idx = *bit;
> > > +       }
> > > +       return bit_idx;
> > > +}
> > > --
> > > 2.30.1 (Apple Git-130)
> > >
>
>
>
> --
> Regards
> Yafang

Re: [PATCH v7 bpf-next 2/2] selftests/bpf: Add selftest for bits iter

[Yafang Shao]

On Wed, May 8, 2024 at 1:12 AM Andrii Nakryiko
<andrii.nakryiko@gmail.com> wrote:
>
> On Tue, May 7, 2024 at 6:39 AM Yafang Shao <laoar.shao@gmail.com> wrote:
> >
> > On Tue, May 7, 2024 at 11:42 AM Andrii Nakryiko
> > <andrii.nakryiko@gmail.com> wrote:
> > >
> > > On Sun, May 5, 2024 at 8:35 PM Yafang Shao <laoar.shao@gmail.com> wrote:
> > > >
> > > > Add test cases for the bits iter:
> > > > - positive case
> > > >   - bit mask smaller than 8 bytes
> > > >   - a typical case of having 8-byte bit mask
> > > >   - another typical case where bit mask is > 8 bytes
> > > >   - the index of set bit
> > > >
> > > > - nagative cases
> > > >   - bpf_iter_bits_destroy() is required after calling
> > > >     bpf_iter_bits_new()
> > > >   - bpf_iter_bits_destroy() can only destroy an initialized iter
> > > >   - bpf_iter_bits_next() must use an initialized iter
> > > >
> > > > Signed-off-by: Yafang Shao <laoar.shao@gmail.com>
> > > > ---
> > > >  .../selftests/bpf/prog_tests/verifier.c       |   2 +
> > > >  .../selftests/bpf/progs/verifier_bits_iter.c  | 160 ++++++++++++++++++
> > > >  2 files changed, 162 insertions(+)
> > > >  create mode 100644 tools/testing/selftests/bpf/progs/verifier_bits_iter.c
> > > >
> > > > diff --git a/tools/testing/selftests/bpf/prog_tests/verifier.c b/tools/testing/selftests/bpf/prog_tests/verifier.c
> > > > index c4f9f306646e..7e04ecaaa20a 100644
> > > > --- a/tools/testing/selftests/bpf/prog_tests/verifier.c
> > > > +++ b/tools/testing/selftests/bpf/prog_tests/verifier.c
> > > > @@ -84,6 +84,7 @@
> > > >  #include "verifier_xadd.skel.h"
> > > >  #include "verifier_xdp.skel.h"
> > > >  #include "verifier_xdp_direct_packet_access.skel.h"
> > > > +#include "verifier_bits_iter.skel.h"
> > > >
> > > >  #define MAX_ENTRIES 11
> > > >
> > > > @@ -198,6 +199,7 @@ void test_verifier_var_off(void)              { RUN(verifier_var_off); }
> > > >  void test_verifier_xadd(void)                 { RUN(verifier_xadd); }
> > > >  void test_verifier_xdp(void)                  { RUN(verifier_xdp); }
> > > >  void test_verifier_xdp_direct_packet_access(void) { RUN(verifier_xdp_direct_packet_access); }
> > > > +void test_verifier_bits_iter(void) { RUN(verifier_bits_iter); }
> > > >
> > > >  static int init_test_val_map(struct bpf_object *obj, char *map_name)
> > > >  {
> > > > diff --git a/tools/testing/selftests/bpf/progs/verifier_bits_iter.c b/tools/testing/selftests/bpf/progs/verifier_bits_iter.c
> > > > new file mode 100644
> > > > index 000000000000..2f7b62b25638
> > > > --- /dev/null
> > > > +++ b/tools/testing/selftests/bpf/progs/verifier_bits_iter.c
> > > > @@ -0,0 +1,160 @@
> > > > +// SPDX-License-Identifier: GPL-2.0-only
> > > > +/* Copyright (c) 2024 Yafang Shao <laoar.shao@gmail.com> */
> > > > +
> > > > +#include "vmlinux.h"
> > > > +#include <bpf/bpf_helpers.h>
> > > > +#include <bpf/bpf_tracing.h>
> > > > +
> > > > +#include "bpf_misc.h"
> > > > +#include "task_kfunc_common.h"
> > > > +
> > > > +char _license[] SEC("license") = "GPL";
> > > > +
> > > > +int bpf_iter_bits_new(struct bpf_iter_bits *it, const void *unsafe_ptr__ign,
> > > > +                     u32 nr_bits) __ksym __weak;
> > > > +int *bpf_iter_bits_next(struct bpf_iter_bits *it) __ksym __weak;
> > > > +void bpf_iter_bits_destroy(struct bpf_iter_bits *it) __ksym __weak;
> > > > +
> > > > +SEC("iter.s/cgroup")
> > > > +__description("bits iter without destroy")
> > > > +__failure __msg("Unreleased reference")
> > > > +int BPF_PROG(no_destroy, struct bpf_iter_meta *meta, struct cgroup *cgrp)
> > > > +{
> > > > +       struct bpf_iter_bits it;
> > > > +       struct task_struct *p;
> > > > +
> > > > +       p = bpf_task_from_pid(1);
> > > > +       if (!p)
> > > > +               return 1;
> > > > +
> > > > +       bpf_iter_bits_new(&it, p->cpus_ptr, 8192);
> > > > +
> > > > +       bpf_iter_bits_next(&it);
> > > > +       bpf_task_release(p);
> > > > +       return 0;
> > > > +}
> > > > +
> > > > +SEC("iter/cgroup")
> > > > +__description("bits iter with uninitialized iter in ->next()")
> > > > +__failure __msg("expected an initialized iter_bits as arg #1")
> > > > +int BPF_PROG(next_uninit, struct bpf_iter_meta *meta, struct cgroup *cgrp)
> > > > +{
> > > > +       struct bpf_iter_bits *it = NULL;
> > > > +
> > > > +       bpf_iter_bits_next(it);
> > > > +       return 0;
> > > > +}
> > > > +
> > > > +SEC("iter/cgroup")
> > > > +__description("bits iter with uninitialized iter in ->destroy()")
> > > > +__failure __msg("expected an initialized iter_bits as arg #1")
> > > > +int BPF_PROG(destroy_uninit, struct bpf_iter_meta *meta, struct cgroup *cgrp)
> > > > +{
> > > > +       struct bpf_iter_bits it = {};
> > > > +
> > > > +       bpf_iter_bits_destroy(&it);
> > > > +       return 0;
> > > > +}
> > > > +
> > > > +SEC("syscall")
> > > > +__description("bits copy 32")
> > > > +__success __retval(10)
> > > > +int bits_copy32(void)
> > > > +{
> > > > +       /* 21 bits:             --------------------- */
> > > > +       u32 data = 0b11111101111101111100001000100101U;
> > >
> > > if you define this bit mask as an array of bytes, then you won't have
> > > to handle big-endian in the tests at all
> >
> > This test case provides a clear example of iterating over data of type
> > u32, offering valuable guidance for users who need to perform such
> > iterations.
> >
> > >
> > >
> > > > +       int nr = 0, offset = 0;
> > > > +       int *bit;
> > > > +
> > > > +#if defined(__TARGET_ARCH_s390)
> > > > +       offset = sizeof(u32) - (21 + 7) / 8;
> > > > +#endif
> > > > +       bpf_for_each(bits, bit, ((char *)&data) + offset, 21)
> > > > +               nr++;
> > > > +       return nr;
> > > > +}
> > > > +
> > > > +SEC("syscall")
> > > > +__description("bits copy 64")
> > > > +__success __retval(18)
> > > > +int bits_copy64(void)
> > > > +{
> > > > +       /* 34 bits:         ~-------- */
> > > > +       u64 data = 0xffffefdf0f0f0f0fUL;
> > > > +       int nr = 0, offset = 0;
> > > > +       int *bit;
> > > > +
> > > > +#if defined(__TARGET_ARCH_s390)
> > > > +       offset = sizeof(u64) - (34 + 7) / 8;
> > > > +#endif
> > > > +
> > > > +       bpf_for_each(bits, bit, ((char *)&data) + offset, 34)
> > >
> > > see above about byte array, but if we define different (not as byte
> > > array but long[]), it would be cleaner to have
> >
> > This test case demonstrates how to iterate over data of type u64.
> >
> > >
> > > #if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
> > > u64 data = 0x......UL;
> > > #else
> > > u64 data = 0x......UL;
> > > #endif
> >
> > looks good.
> >
>
> Please hold off on sending a new revision until we figure out what the
> contract should be. Because I feel like it's a (relatively) big
> decision whether a bit mask is treated as an array of bytes or as an
> array of longs. For little-endian it makes no difference, but for
> big-endian it's a big difference and has usability and performance
> implications.

Perhaps it would be advantageous to define the interface as follows:

bpf_iter_bits_new(struct bpf_iter_bits *it, const u64
*unsafe_ptr__ign, u32 words)

This approach eliminates the need to account for endianness.

-- 
Regards
Yafang

Re: [PATCH v7 bpf-next 2/2] selftests/bpf: Add selftest for bits iter

[Andrii Nakryiko]

On Wed, May 8, 2024 at 7:11 PM Yafang Shao <laoar.shao@gmail.com> wrote:
>
> On Wed, May 8, 2024 at 1:12 AM Andrii Nakryiko
> <andrii.nakryiko@gmail.com> wrote:
> >
> > On Tue, May 7, 2024 at 6:39 AM Yafang Shao <laoar.shao@gmail.com> wrote:
> > >
> > > On Tue, May 7, 2024 at 11:42 AM Andrii Nakryiko
> > > <andrii.nakryiko@gmail.com> wrote:
> > > >
> > > > On Sun, May 5, 2024 at 8:35 PM Yafang Shao <laoar.shao@gmail.com> wrote:
> > > > >
> > > > > Add test cases for the bits iter:
> > > > > - positive case
> > > > >   - bit mask smaller than 8 bytes
> > > > >   - a typical case of having 8-byte bit mask
> > > > >   - another typical case where bit mask is > 8 bytes
> > > > >   - the index of set bit
> > > > >
> > > > > - nagative cases
> > > > >   - bpf_iter_bits_destroy() is required after calling
> > > > >     bpf_iter_bits_new()
> > > > >   - bpf_iter_bits_destroy() can only destroy an initialized iter
> > > > >   - bpf_iter_bits_next() must use an initialized iter
> > > > >
> > > > > Signed-off-by: Yafang Shao <laoar.shao@gmail.com>
> > > > > ---
> > > > >  .../selftests/bpf/prog_tests/verifier.c       |   2 +
> > > > >  .../selftests/bpf/progs/verifier_bits_iter.c  | 160 ++++++++++++++++++
> > > > >  2 files changed, 162 insertions(+)
> > > > >  create mode 100644 tools/testing/selftests/bpf/progs/verifier_bits_iter.c
> > > > >
> > > > > diff --git a/tools/testing/selftests/bpf/prog_tests/verifier.c b/tools/testing/selftests/bpf/prog_tests/verifier.c
> > > > > index c4f9f306646e..7e04ecaaa20a 100644
> > > > > --- a/tools/testing/selftests/bpf/prog_tests/verifier.c
> > > > > +++ b/tools/testing/selftests/bpf/prog_tests/verifier.c
> > > > > @@ -84,6 +84,7 @@
> > > > >  #include "verifier_xadd.skel.h"
> > > > >  #include "verifier_xdp.skel.h"
> > > > >  #include "verifier_xdp_direct_packet_access.skel.h"
> > > > > +#include "verifier_bits_iter.skel.h"
> > > > >
> > > > >  #define MAX_ENTRIES 11
> > > > >
> > > > > @@ -198,6 +199,7 @@ void test_verifier_var_off(void)              { RUN(verifier_var_off); }
> > > > >  void test_verifier_xadd(void)                 { RUN(verifier_xadd); }
> > > > >  void test_verifier_xdp(void)                  { RUN(verifier_xdp); }
> > > > >  void test_verifier_xdp_direct_packet_access(void) { RUN(verifier_xdp_direct_packet_access); }
> > > > > +void test_verifier_bits_iter(void) { RUN(verifier_bits_iter); }
> > > > >
> > > > >  static int init_test_val_map(struct bpf_object *obj, char *map_name)
> > > > >  {
> > > > > diff --git a/tools/testing/selftests/bpf/progs/verifier_bits_iter.c b/tools/testing/selftests/bpf/progs/verifier_bits_iter.c
> > > > > new file mode 100644
> > > > > index 000000000000..2f7b62b25638
> > > > > --- /dev/null
> > > > > +++ b/tools/testing/selftests/bpf/progs/verifier_bits_iter.c
> > > > > @@ -0,0 +1,160 @@
> > > > > +// SPDX-License-Identifier: GPL-2.0-only
> > > > > +/* Copyright (c) 2024 Yafang Shao <laoar.shao@gmail.com> */
> > > > > +
> > > > > +#include "vmlinux.h"
> > > > > +#include <bpf/bpf_helpers.h>
> > > > > +#include <bpf/bpf_tracing.h>
> > > > > +
> > > > > +#include "bpf_misc.h"
> > > > > +#include "task_kfunc_common.h"
> > > > > +
> > > > > +char _license[] SEC("license") = "GPL";
> > > > > +
> > > > > +int bpf_iter_bits_new(struct bpf_iter_bits *it, const void *unsafe_ptr__ign,
> > > > > +                     u32 nr_bits) __ksym __weak;
> > > > > +int *bpf_iter_bits_next(struct bpf_iter_bits *it) __ksym __weak;
> > > > > +void bpf_iter_bits_destroy(struct bpf_iter_bits *it) __ksym __weak;
> > > > > +
> > > > > +SEC("iter.s/cgroup")
> > > > > +__description("bits iter without destroy")
> > > > > +__failure __msg("Unreleased reference")
> > > > > +int BPF_PROG(no_destroy, struct bpf_iter_meta *meta, struct cgroup *cgrp)
> > > > > +{
> > > > > +       struct bpf_iter_bits it;
> > > > > +       struct task_struct *p;
> > > > > +
> > > > > +       p = bpf_task_from_pid(1);
> > > > > +       if (!p)
> > > > > +               return 1;
> > > > > +
> > > > > +       bpf_iter_bits_new(&it, p->cpus_ptr, 8192);
> > > > > +
> > > > > +       bpf_iter_bits_next(&it);
> > > > > +       bpf_task_release(p);
> > > > > +       return 0;
> > > > > +}
> > > > > +
> > > > > +SEC("iter/cgroup")
> > > > > +__description("bits iter with uninitialized iter in ->next()")
> > > > > +__failure __msg("expected an initialized iter_bits as arg #1")
> > > > > +int BPF_PROG(next_uninit, struct bpf_iter_meta *meta, struct cgroup *cgrp)
> > > > > +{
> > > > > +       struct bpf_iter_bits *it = NULL;
> > > > > +
> > > > > +       bpf_iter_bits_next(it);
> > > > > +       return 0;
> > > > > +}
> > > > > +
> > > > > +SEC("iter/cgroup")
> > > > > +__description("bits iter with uninitialized iter in ->destroy()")
> > > > > +__failure __msg("expected an initialized iter_bits as arg #1")
> > > > > +int BPF_PROG(destroy_uninit, struct bpf_iter_meta *meta, struct cgroup *cgrp)
> > > > > +{
> > > > > +       struct bpf_iter_bits it = {};
> > > > > +
> > > > > +       bpf_iter_bits_destroy(&it);
> > > > > +       return 0;
> > > > > +}
> > > > > +
> > > > > +SEC("syscall")
> > > > > +__description("bits copy 32")
> > > > > +__success __retval(10)
> > > > > +int bits_copy32(void)
> > > > > +{
> > > > > +       /* 21 bits:             --------------------- */
> > > > > +       u32 data = 0b11111101111101111100001000100101U;
> > > >
> > > > if you define this bit mask as an array of bytes, then you won't have
> > > > to handle big-endian in the tests at all
> > >
> > > This test case provides a clear example of iterating over data of type
> > > u32, offering valuable guidance for users who need to perform such
> > > iterations.
> > >
> > > >
> > > >
> > > > > +       int nr = 0, offset = 0;
> > > > > +       int *bit;
> > > > > +
> > > > > +#if defined(__TARGET_ARCH_s390)
> > > > > +       offset = sizeof(u32) - (21 + 7) / 8;
> > > > > +#endif
> > > > > +       bpf_for_each(bits, bit, ((char *)&data) + offset, 21)
> > > > > +               nr++;
> > > > > +       return nr;
> > > > > +}
> > > > > +
> > > > > +SEC("syscall")
> > > > > +__description("bits copy 64")
> > > > > +__success __retval(18)
> > > > > +int bits_copy64(void)
> > > > > +{
> > > > > +       /* 34 bits:         ~-------- */
> > > > > +       u64 data = 0xffffefdf0f0f0f0fUL;
> > > > > +       int nr = 0, offset = 0;
> > > > > +       int *bit;
> > > > > +
> > > > > +#if defined(__TARGET_ARCH_s390)
> > > > > +       offset = sizeof(u64) - (34 + 7) / 8;
> > > > > +#endif
> > > > > +
> > > > > +       bpf_for_each(bits, bit, ((char *)&data) + offset, 34)
> > > >
> > > > see above about byte array, but if we define different (not as byte
> > > > array but long[]), it would be cleaner to have
> > >
> > > This test case demonstrates how to iterate over data of type u64.
> > >
> > > >
> > > > #if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
> > > > u64 data = 0x......UL;
> > > > #else
> > > > u64 data = 0x......UL;
> > > > #endif
> > >
> > > looks good.
> > >
> >
> > Please hold off on sending a new revision until we figure out what the
> > contract should be. Because I feel like it's a (relatively) big
> > decision whether a bit mask is treated as an array of bytes or as an
> > array of longs. For little-endian it makes no difference, but for
> > big-endian it's a big difference and has usability and performance
> > implications.
>
> Perhaps it would be advantageous to define the interface as follows:
>
> bpf_iter_bits_new(struct bpf_iter_bits *it, const u64
> *unsafe_ptr__ign, u32 words)
>
> This approach eliminates the need to account for endianness.

I don't mind that, if others don't have any opinion. Let's just
document that by "words" we mean 8-byte integers.

>
> --
> Regards
> Yafang