[v2,bpf-next,1/2] bpf: Add bits iterator

Commit Message

Yafang Shao Feb. 25, 2024, 10:06 a.m. UTC

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 <[e-mail]>
---
 kernel/bpf/helpers.c | 100 +++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 100 insertions(+)

Comments

Andrii Nakryiko Feb. 28, 2024, 1:24 a.m. UTC | #1

On Sun, Feb 25, 2024 at 2:07 AM Yafang Shao <[e-mail]> 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 <[e-mail]>
> ---
>  kernel/bpf/helpers.c | 100 +++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 100 insertions(+)
>
> diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
> index 93edf730d288..052f63891834 100644
> --- a/kernel/bpf/helpers.c
> +++ b/kernel/bpf/helpers.c
> @@ -2542,6 +2542,103 @@ __bpf_kfunc void bpf_throw(u64 cookie)
>         WARN(1, "A call to BPF exception callback should never return\n");
>  }
>
> +struct bpf_iter_bits {
> +       __u64 __opaque[2];
> +} __aligned(8);
> +
> +struct bpf_iter_bits_kern {
> +       unsigned long *bits;
> +       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 size = BITS_TO_BYTES(nr_bits);
> +       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->bits = bpf_mem_alloc(&bpf_global_ma, size);
> +       if (!kit->bits)
> +               return -ENOMEM;

it's probably going to be a pretty common case to do bits iteration
for nr_bits<=64, right? So as an optimization, instead of doing
bpf_mem_alloc() for this case, you can just copy up to 8 bytes and
store it in a union of `unsigned long *bits` and `unsigned long
bits_copy`. As a performance optimization (and to reduce dependency on
memory allocation). WDYT?

> +
> +       err = bpf_probe_read_kernel_common(kit->bits, size, unsafe_ptr__ign);
> +       if (err) {
> +               bpf_mem_free(&bpf_global_ma, kit->bits);
> +               kit->bits = NULL;
> +               return err;
> +       }
> +
> +       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;
> +       const unsigned long *bits = kit->bits;
> +       int bit;
> +
> +       if (!bits)
> +               return NULL;
> +
> +       bit = find_next_bit(bits, kit->nr_bits, kit->bit + 1);
> +       if (bit >= kit->nr_bits)
> +               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->bits)
> +               return;
> +       bpf_mem_free(&bpf_global_ma, kit->bits);
> +}
> +
>  __bpf_kfunc_end_defs();
>
>  BTF_KFUNCS_START(generic_btf_ids)
> @@ -2618,6 +2715,9 @@ BTF_ID_FLAGS(func, bpf_dynptr_is_null)
>  BTF_ID_FLAGS(func, bpf_dynptr_is_rdonly)
>  BTF_ID_FLAGS(func, bpf_dynptr_size)
>  BTF_ID_FLAGS(func, bpf_dynptr_clone)
> +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 = {
> --
> 2.39.1
>

Yafang Shao Feb. 28, 2024, 2:25 a.m. UTC | #2

On Wed, Feb 28, 2024 at 9:24 AM Andrii Nakryiko
<[e-mail]> wrote:
>
> On Sun, Feb 25, 2024 at 2:07 AM Yafang Shao <[e-mail]> 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 <[e-mail]>
> > ---
> >  kernel/bpf/helpers.c | 100 +++++++++++++++++++++++++++++++++++++++++++
> >  1 file changed, 100 insertions(+)
> >
> > diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
> > index 93edf730d288..052f63891834 100644
> > --- a/kernel/bpf/helpers.c
> > +++ b/kernel/bpf/helpers.c
> > @@ -2542,6 +2542,103 @@ __bpf_kfunc void bpf_throw(u64 cookie)
> >         WARN(1, "A call to BPF exception callback should never return\n");
> >  }
> >
> > +struct bpf_iter_bits {
> > +       __u64 __opaque[2];
> > +} __aligned(8);
> > +
> > +struct bpf_iter_bits_kern {
> > +       unsigned long *bits;
> > +       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 size = BITS_TO_BYTES(nr_bits);
> > +       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->bits = bpf_mem_alloc(&bpf_global_ma, size);
> > +       if (!kit->bits)
> > +               return -ENOMEM;
>
> it's probably going to be a pretty common case to do bits iteration
> for nr_bits<=64, right?

It's highly unlikely.
Consider the CPU count as an example; There are 256 CPUs on our AMD
EPYC servers.

>  So as an optimization, instead of doing
> bpf_mem_alloc() for this case, you can just copy up to 8 bytes and
> store it in a union of `unsigned long *bits` and `unsigned long
> bits_copy`. As a performance optimization (and to reduce dependency on
> memory allocation). WDYT?
>

Andrii Nakryiko Feb. 28, 2024, 6:04 a.m. UTC | #3

On Tue, Feb 27, 2024 at 6:25 PM Yafang Shao <[e-mail]> wrote:
>
> On Wed, Feb 28, 2024 at 9:24 AM Andrii Nakryiko
> <[e-mail]> wrote:
> >
> > On Sun, Feb 25, 2024 at 2:07 AM Yafang Shao <[e-mail]> 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 <[e-mail]>
> > > ---
> > >  kernel/bpf/helpers.c | 100 +++++++++++++++++++++++++++++++++++++++++++
> > >  1 file changed, 100 insertions(+)
> > >
> > > diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
> > > index 93edf730d288..052f63891834 100644
> > > --- a/kernel/bpf/helpers.c
> > > +++ b/kernel/bpf/helpers.c
> > > @@ -2542,6 +2542,103 @@ __bpf_kfunc void bpf_throw(u64 cookie)
> > >         WARN(1, "A call to BPF exception callback should never return\n");
> > >  }
> > >
> > > +struct bpf_iter_bits {
> > > +       __u64 __opaque[2];
> > > +} __aligned(8);
> > > +
> > > +struct bpf_iter_bits_kern {
> > > +       unsigned long *bits;
> > > +       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 size = BITS_TO_BYTES(nr_bits);
> > > +       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->bits = bpf_mem_alloc(&bpf_global_ma, size);
> > > +       if (!kit->bits)
> > > +               return -ENOMEM;
> >
> > it's probably going to be a pretty common case to do bits iteration
> > for nr_bits<=64, right?
>
> It's highly unlikely.
> Consider the CPU count as an example; There are 256 CPUs on our AMD
> EPYC servers.

Also consider u64-based bit masks (like struct backtrack_state in
verifier code, which has u32 reg_mask and u64 stack_mask). This
iterator is a generic bits iterator, there are tons of cases of
u64/u32 masks in practice.

>
> >  So as an optimization, instead of doing
> > bpf_mem_alloc() for this case, you can just copy up to 8 bytes and
> > store it in a union of `unsigned long *bits` and `unsigned long
> > bits_copy`. As a performance optimization (and to reduce dependency on
> > memory allocation). WDYT?
> >
>
> --
> Regards
> Yafang

Yafang Shao Feb. 29, 2024, 2:15 a.m. UTC | #4

On Wed, Feb 28, 2024 at 2:04 PM Andrii Nakryiko
<[e-mail]> wrote:
>
> On Tue, Feb 27, 2024 at 6:25 PM Yafang Shao <[e-mail]> wrote:
> >
> > On Wed, Feb 28, 2024 at 9:24 AM Andrii Nakryiko
> > <[e-mail]> wrote:
> > >
> > > On Sun, Feb 25, 2024 at 2:07 AM Yafang Shao <[e-mail]> 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 <[e-mail]>
> > > > ---
> > > >  kernel/bpf/helpers.c | 100 +++++++++++++++++++++++++++++++++++++++++++
> > > >  1 file changed, 100 insertions(+)
> > > >
> > > > diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
> > > > index 93edf730d288..052f63891834 100644
> > > > --- a/kernel/bpf/helpers.c
> > > > +++ b/kernel/bpf/helpers.c
> > > > @@ -2542,6 +2542,103 @@ __bpf_kfunc void bpf_throw(u64 cookie)
> > > >         WARN(1, "A call to BPF exception callback should never return\n");
> > > >  }
> > > >
> > > > +struct bpf_iter_bits {
> > > > +       __u64 __opaque[2];
> > > > +} __aligned(8);
> > > > +
> > > > +struct bpf_iter_bits_kern {
> > > > +       unsigned long *bits;
> > > > +       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 size = BITS_TO_BYTES(nr_bits);
> > > > +       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->bits = bpf_mem_alloc(&bpf_global_ma, size);
> > > > +       if (!kit->bits)
> > > > +               return -ENOMEM;
> > >
> > > it's probably going to be a pretty common case to do bits iteration
> > > for nr_bits<=64, right?
> >
> > It's highly unlikely.
> > Consider the CPU count as an example; There are 256 CPUs on our AMD
> > EPYC servers.
>
> Also consider u64-based bit masks (like struct backtrack_state in
> verifier code, which has u32 reg_mask and u64 stack_mask). This
> iterator is a generic bits iterator, there are tons of cases of
> u64/u32 masks in practice.

Should we optimize it as follows?

    if (nr_bits <= 64) {
        // do the optimization
    } else {
        // fallback to memalloc
    }

>
> >
> > >  So as an optimization, instead of doing
> > > bpf_mem_alloc() for this case, you can just copy up to 8 bytes and
> > > store it in a union of `unsigned long *bits` and `unsigned long
> > > bits_copy`. As a performance optimization (and to reduce dependency on
> > > memory allocation). WDYT?
> > >
> >
> > --
> > Regards
> > Yafang

Andrii Nakryiko Feb. 29, 2024, 5:26 p.m. UTC | #5

On Wed, Feb 28, 2024 at 6:16 PM Yafang Shao <[e-mail]> wrote:
>
> On Wed, Feb 28, 2024 at 2:04 PM Andrii Nakryiko
> <[e-mail]> wrote:
> >
> > On Tue, Feb 27, 2024 at 6:25 PM Yafang Shao <[e-mail]> wrote:
> > >
> > > On Wed, Feb 28, 2024 at 9:24 AM Andrii Nakryiko
> > > <[e-mail]> wrote:
> > > >
> > > > On Sun, Feb 25, 2024 at 2:07 AM Yafang Shao <[e-mail]> 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 <[e-mail]>
> > > > > ---
> > > > >  kernel/bpf/helpers.c | 100 +++++++++++++++++++++++++++++++++++++++++++
> > > > >  1 file changed, 100 insertions(+)
> > > > >
> > > > > diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
> > > > > index 93edf730d288..052f63891834 100644
> > > > > --- a/kernel/bpf/helpers.c
> > > > > +++ b/kernel/bpf/helpers.c
> > > > > @@ -2542,6 +2542,103 @@ __bpf_kfunc void bpf_throw(u64 cookie)
> > > > >         WARN(1, "A call to BPF exception callback should never return\n");
> > > > >  }
> > > > >
> > > > > +struct bpf_iter_bits {
> > > > > +       __u64 __opaque[2];
> > > > > +} __aligned(8);
> > > > > +
> > > > > +struct bpf_iter_bits_kern {
> > > > > +       unsigned long *bits;
> > > > > +       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 size = BITS_TO_BYTES(nr_bits);
> > > > > +       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->bits = bpf_mem_alloc(&bpf_global_ma, size);
> > > > > +       if (!kit->bits)
> > > > > +               return -ENOMEM;
> > > >
> > > > it's probably going to be a pretty common case to do bits iteration
> > > > for nr_bits<=64, right?
> > >
> > > It's highly unlikely.
> > > Consider the CPU count as an example; There are 256 CPUs on our AMD
> > > EPYC servers.
> >
> > Also consider u64-based bit masks (like struct backtrack_state in
> > verifier code, which has u32 reg_mask and u64 stack_mask). This
> > iterator is a generic bits iterator, there are tons of cases of
> > u64/u32 masks in practice.
>
> Should we optimize it as follows?
>
>     if (nr_bits <= 64) {
>         // do the optimization
>     } else {
>         // fallback to memalloc
>     }
>

Yep, that's what I'm proposing


> >
> > >
> > > >  So as an optimization, instead of doing
> > > > bpf_mem_alloc() for this case, you can just copy up to 8 bytes and
> > > > store it in a union of `unsigned long *bits` and `unsigned long
> > > > bits_copy`. As a performance optimization (and to reduce dependency on
> > > > memory allocation). WDYT?
> > > >
> > >
> > > --
> > > Regards
> > > Yafang
>
>
>
> --
> Regards
> Yafang

John Fastabend Feb. 29, 2024, 10:19 p.m. UTC | #6

Andrii Nakryiko wrote:
> On Wed, Feb 28, 2024 at 6:16 PM Yafang Shao <[e-mail]> wrote:
> >
> > On Wed, Feb 28, 2024 at 2:04 PM Andrii Nakryiko
> > <[e-mail]> wrote:
> > >
> > > On Tue, Feb 27, 2024 at 6:25 PM Yafang Shao <[e-mail]> wrote:
> > > >
> > > > On Wed, Feb 28, 2024 at 9:24 AM Andrii Nakryiko
> > > > <[e-mail]> wrote:
> > > > >
> > > > > On Sun, Feb 25, 2024 at 2:07 AM Yafang Shao <[e-mail]> 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 <[e-mail]>
> > > > > > ---
> > > > > >  kernel/bpf/helpers.c | 100 +++++++++++++++++++++++++++++++++++++++++++
> > > > > >  1 file changed, 100 insertions(+)
> > > > > >
> > > > > > diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
> > > > > > index 93edf730d288..052f63891834 100644
> > > > > > --- a/kernel/bpf/helpers.c
> > > > > > +++ b/kernel/bpf/helpers.c
> > > > > > @@ -2542,6 +2542,103 @@ __bpf_kfunc void bpf_throw(u64 cookie)
> > > > > >         WARN(1, "A call to BPF exception callback should never return\n");
> > > > > >  }
> > > > > >
> > > > > > +struct bpf_iter_bits {
> > > > > > +       __u64 __opaque[2];
> > > > > > +} __aligned(8);
> > > > > > +
> > > > > > +struct bpf_iter_bits_kern {
> > > > > > +       unsigned long *bits;
> > > > > > +       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 size = BITS_TO_BYTES(nr_bits);
> > > > > > +       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->bits = bpf_mem_alloc(&bpf_global_ma, size);
> > > > > > +       if (!kit->bits)
> > > > > > +               return -ENOMEM;
> > > > >
> > > > > it's probably going to be a pretty common case to do bits iteration
> > > > > for nr_bits<=64, right?
> > > >
> > > > It's highly unlikely.
> > > > Consider the CPU count as an example; There are 256 CPUs on our AMD
> > > > EPYC servers.
> > >
> > > Also consider u64-based bit masks (like struct backtrack_state in
> > > verifier code, which has u32 reg_mask and u64 stack_mask). This
> > > iterator is a generic bits iterator, there are tons of cases of
> > > u64/u32 masks in practice.
> >
> > Should we optimize it as follows?
> >
> >     if (nr_bits <= 64) {
> >         // do the optimization
> >     } else {
> >         // fallback to memalloc
> >     }
> >
> 
> Yep, that's what I'm proposing

When I suggested why not just open code this in BPF earlier I was
mostly thinking of these u64 and u32 masks we have lots of them
in our code base as well.

I have something like this which might be even better than 3
calls depending on your use case,

 int find_next_bit(uint64_t bits, int last_bit)
 {
    int i = last_bit;
    for (i = 0; i < sizeof(uint64_t) * 8; i++) {
        if (bits & (1 << i))
           return i;
    }
    return -1;
  }

Verifier seems plenty happy with above.

Thanks,
John

Yafang Shao March 1, 2024, 6:40 a.m. UTC | #7

On Fri, Mar 1, 2024 at 6:19 AM John Fastabend <[e-mail]> wrote:
>
> Andrii Nakryiko wrote:
> > On Wed, Feb 28, 2024 at 6:16 PM Yafang Shao <[e-mail]> wrote:
> > >
> > > On Wed, Feb 28, 2024 at 2:04 PM Andrii Nakryiko
> > > <[e-mail]> wrote:
> > > >
> > > > On Tue, Feb 27, 2024 at 6:25 PM Yafang Shao <[e-mail]> wrote:
> > > > >
> > > > > On Wed, Feb 28, 2024 at 9:24 AM Andrii Nakryiko
> > > > > <[e-mail]> wrote:
> > > > > >
> > > > > > On Sun, Feb 25, 2024 at 2:07 AM Yafang Shao <[e-mail]> 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 <[e-mail]>
> > > > > > > ---
> > > > > > >  kernel/bpf/helpers.c | 100 +++++++++++++++++++++++++++++++++++++++++++
> > > > > > >  1 file changed, 100 insertions(+)
> > > > > > >
> > > > > > > diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
> > > > > > > index 93edf730d288..052f63891834 100644
> > > > > > > --- a/kernel/bpf/helpers.c
> > > > > > > +++ b/kernel/bpf/helpers.c
> > > > > > > @@ -2542,6 +2542,103 @@ __bpf_kfunc void bpf_throw(u64 cookie)
> > > > > > >         WARN(1, "A call to BPF exception callback should never return\n");
> > > > > > >  }
> > > > > > >
> > > > > > > +struct bpf_iter_bits {
> > > > > > > +       __u64 __opaque[2];
> > > > > > > +} __aligned(8);
> > > > > > > +
> > > > > > > +struct bpf_iter_bits_kern {
> > > > > > > +       unsigned long *bits;
> > > > > > > +       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 size = BITS_TO_BYTES(nr_bits);
> > > > > > > +       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->bits = bpf_mem_alloc(&bpf_global_ma, size);
> > > > > > > +       if (!kit->bits)
> > > > > > > +               return -ENOMEM;
> > > > > >
> > > > > > it's probably going to be a pretty common case to do bits iteration
> > > > > > for nr_bits<=64, right?
> > > > >
> > > > > It's highly unlikely.
> > > > > Consider the CPU count as an example; There are 256 CPUs on our AMD
> > > > > EPYC servers.
> > > >
> > > > Also consider u64-based bit masks (like struct backtrack_state in
> > > > verifier code, which has u32 reg_mask and u64 stack_mask). This
> > > > iterator is a generic bits iterator, there are tons of cases of
> > > > u64/u32 masks in practice.
> > >
> > > Should we optimize it as follows?
> > >
> > >     if (nr_bits <= 64) {
> > >         // do the optimization
> > >     } else {
> > >         // fallback to memalloc
> > >     }
> > >
> >
> > Yep, that's what I'm proposing
>
> When I suggested why not just open code this in BPF earlier I was
> mostly thinking of these u64 and u32 masks we have lots of them
> in our code base as well.
>
> I have something like this which might be even better than 3
> calls depending on your use case,
>
>  int find_next_bit(uint64_t bits, int last_bit)
>  {
>     int i = last_bit;
>     for (i = 0; i < sizeof(uint64_t) * 8; i++) {
>         if (bits & (1 << i))
>            return i;
>     }
>     return -1;
>   }
>
> Verifier seems plenty happy with above.

I'm not quite following.
Regarding the find_next_bit() function you mentioned, it seems it only
retrieves one bit at a time, necessitating a for loop for execution,
correct? Consequently, the verifier will likely fail the for loop.

John Fastabend March 1, 2024, 6:01 p.m. UTC | #8

Yafang Shao wrote:
> On Fri, Mar 1, 2024 at 6:19 AM John Fastabend <[e-mail]> wrote:
> >
> > Andrii Nakryiko wrote:
> > > On Wed, Feb 28, 2024 at 6:16 PM Yafang Shao <[e-mail]> wrote:
> > > >
> > > > On Wed, Feb 28, 2024 at 2:04 PM Andrii Nakryiko
> > > > <[e-mail]> wrote:
> > > > >
> > > > > On Tue, Feb 27, 2024 at 6:25 PM Yafang Shao <[e-mail]> wrote:
> > > > > >
> > > > > > On Wed, Feb 28, 2024 at 9:24 AM Andrii Nakryiko
> > > > > > <[e-mail]> wrote:
> > > > > > >
> > > > > > > On Sun, Feb 25, 2024 at 2:07 AM Yafang Shao <[e-mail]> 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 <[e-mail]>
> > > > > > > > ---

[...]

> > > > > > > > +/**
> > > > > > > > + * 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 size = BITS_TO_BYTES(nr_bits);
> > > > > > > > +       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->bits = bpf_mem_alloc(&bpf_global_ma, size);
> > > > > > > > +       if (!kit->bits)
> > > > > > > > +               return -ENOMEM;
> > > > > > >
> > > > > > > it's probably going to be a pretty common case to do bits iteration
> > > > > > > for nr_bits<=64, right?
> > > > > >
> > > > > > It's highly unlikely.
> > > > > > Consider the CPU count as an example; There are 256 CPUs on our AMD
> > > > > > EPYC servers.
> > > > >
> > > > > Also consider u64-based bit masks (like struct backtrack_state in
> > > > > verifier code, which has u32 reg_mask and u64 stack_mask). This
> > > > > iterator is a generic bits iterator, there are tons of cases of
> > > > > u64/u32 masks in practice.
> > > >
> > > > Should we optimize it as follows?
> > > >
> > > >     if (nr_bits <= 64) {
> > > >         // do the optimization
> > > >     } else {
> > > >         // fallback to memalloc
> > > >     }
> > > >
> > >
> > > Yep, that's what I'm proposing
> >
> > When I suggested why not just open code this in BPF earlier I was
> > mostly thinking of these u64 and u32 masks we have lots of them
> > in our code base as well.
> >
> > I have something like this which might be even better than 3
> > calls depending on your use case,
> >
> >  int find_next_bit(uint64_t bits, int last_bit)
> >  {
> >     int i = last_bit;
> >     for (i = 0; i < sizeof(uint64_t) * 8; i++) {
> >         if (bits & (1 << i))
> >            return i;
> >     }
> >     return -1;
> >   }
> >
> > Verifier seems plenty happy with above.
> 
> I'm not quite following.
> Regarding the find_next_bit() function you mentioned, it seems it only
> retrieves one bit at a time, necessitating a for loop for execution,
> correct? Consequently, the verifier will likely fail the for loop.

In practice for small sizes uint64_t and uint32_t we don't see any
issue. Just a comment that this can be open coded without much
trouble in many cases. Not against the helper at all.

> 
> -- 
> Regards
> Yafang

Patch

diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
index 93edf730d288..052f63891834 100644
--- a/kernel/bpf/helpers.c
+++ b/kernel/bpf/helpers.c
@@ -2542,6 +2542,103 @@  __bpf_kfunc void bpf_throw(u64 cookie)
 	WARN(1, "A call to BPF exception callback should never return\n");
 }
 
+struct bpf_iter_bits {
+	__u64 __opaque[2];
+} __aligned(8);
+
+struct bpf_iter_bits_kern {
+	unsigned long *bits;
+	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 size = BITS_TO_BYTES(nr_bits);
+	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->bits = bpf_mem_alloc(&bpf_global_ma, size);
+	if (!kit->bits)
+		return -ENOMEM;
+
+	err = bpf_probe_read_kernel_common(kit->bits, size, unsafe_ptr__ign);
+	if (err) {
+		bpf_mem_free(&bpf_global_ma, kit->bits);
+		kit->bits = NULL;
+		return err;
+	}
+
+	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;
+	const unsigned long *bits = kit->bits;
+	int bit;
+
+	if (!bits)
+		return NULL;
+
+	bit = find_next_bit(bits, kit->nr_bits, kit->bit + 1);
+	if (bit >= kit->nr_bits)
+		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->bits)
+		return;
+	bpf_mem_free(&bpf_global_ma, kit->bits);
+}
+
 __bpf_kfunc_end_defs();
 
 BTF_KFUNCS_START(generic_btf_ids)
@@ -2618,6 +2715,9 @@  BTF_ID_FLAGS(func, bpf_dynptr_is_null)
 BTF_ID_FLAGS(func, bpf_dynptr_is_rdonly)
 BTF_ID_FLAGS(func, bpf_dynptr_size)
 BTF_ID_FLAGS(func, bpf_dynptr_clone)
+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 = {