On 10/9/2024 10:45 AM, Hou Tao wrote: > > On 10/9/2024 2:30 AM, Andrii Nakryiko wrote: >> On Tue, Oct 8, 2024 at 2:05 AM Hou Tao <houtao@xxxxxxxxxxxxxxx> wrote: >>> From: Hou Tao <houtao1@xxxxxxxxxx> >>> >>> Under 32-bits host (e.g, arm32) , when a bpf program passes an u64 to >>> bpf_iter_bits_new(), bpf_iter_bits_new() will use bits_copy to save the >>> content of the u64, but the size of bits_copy is only 4-bytes, and there >>> will be stack corruption. >>> >>> Fix it by change the type of unsafe_ptr from u64 * to unsigned long *. >>> >> This will be confusing as BPF-side long is always 64-bit. So why not >> instead make sure it's u64 throughout (i.e., bits_copy is u64 >> explicitly), even on 32-bit architectures? > Just learn about the size of BPF-side long is always 64-bits. I had > considered to change bits_copy to u64. The main obstacle is that the > pointer type of find_next_bit is unsigned long *, if it is used on an > u64 under big-endian host, it may return invalid result. I think doing the following swap for big endian and 32-bits host will let find_next_bit return the correct result: +static void swap_bits(u64 *bits, unsigned int nr) +{ +#if defined(__BIG_ENDIAN) && !defined(CONFIG_64BIT) + unsigned int i; + + for (i = 0; i < nr; i++) + bits[i] = (bits[i] >> 32) | ((u64)(u32)bits[i] << 32); +#endif +} + Will try to get some test environment to test it. >>> Signed-off-by: Hou Tao <houtao1@xxxxxxxxxx> >>> --- >>> kernel/bpf/helpers.c | 18 ++++++++++-------- >>> 1 file changed, 10 insertions(+), 8 deletions(-) >>> >>> diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c >>> index 6c0205d5018c..dee69c3904a0 100644 >>> --- a/kernel/bpf/helpers.c >>> +++ b/kernel/bpf/helpers.c >>> @@ -2852,7 +2852,7 @@ struct bpf_iter_bits { >>> } __aligned(8); >>> >>> /* nr_bits only has 31 bits */ >>> -#define BITS_ITER_NR_WORDS_MAX ((1U << 31) / BITS_PER_TYPE(u64)) >>> +#define BITS_ITER_NR_WORDS_MAX ((1U << 31) / BITS_PER_TYPE(unsigned long)) >>> >>> struct bpf_iter_bits_kern { >>> union { >>> @@ -2868,8 +2868,9 @@ struct bpf_iter_bits_kern { >>> * 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 pointer pointing to a memory area to be iterated over >>> - * @nr_words: The size of the specified memory area, measured in 8-byte units. >>> - * Due to the limitation of memalloc, it can't be greater than 512. >>> + * @nr_words: The size of the specified memory area, measured in units of >>> + * sizeof(unsigned long). Due to the limitation of memalloc, it can't be >>> + * greater than 512. >>> * >>> * 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_words. It >>> @@ -2879,17 +2880,18 @@ struct bpf_iter_bits_kern { >>> * On success, 0 is returned. On failure, ERR is returned. >>> */ >>> __bpf_kfunc int >>> -bpf_iter_bits_new(struct bpf_iter_bits *it, const u64 *unsafe_ptr__ign, u32 nr_words) >>> +bpf_iter_bits_new(struct bpf_iter_bits *it, const unsigned long *unsafe_ptr__ign, u32 nr_words) >>> { >>> - struct bpf_iter_bits_kern *kit = (void *)it; >>> - u32 nr_bytes = nr_words * sizeof(u64); >>> + u32 nr_bytes = nr_words * sizeof(*unsafe_ptr__ign); >>> u32 nr_bits = BYTES_TO_BITS(nr_bytes); >>> + struct bpf_iter_bits_kern *kit; >>> 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)); >>> >>> + kit = (void *)it; >>> kit->allocated = 0; >>> kit->nr_bits = 0; >>> kit->bits_copy = 0; >>> @@ -2900,8 +2902,8 @@ bpf_iter_bits_new(struct bpf_iter_bits *it, const u64 *unsafe_ptr__ign, u32 nr_w >>> if (nr_words > BITS_ITER_NR_WORDS_MAX) >>> return -E2BIG; >>> >>> - /* Optimization for u64 mask */ >>> - if (nr_bits == 64) { >>> + /* Optimization for unsigned long mask */ >>> + if (nr_words == 1) { >>> err = bpf_probe_read_kernel_common(&kit->bits_copy, nr_bytes, unsafe_ptr__ign); >>> if (err) >>> return -EFAULT; >>> -- >>> 2.29.2 >>> >> . > .