From: Yiming Qian <yimingqian591@gmail.com> stable inclusion from stable-v5.10.260 commit f126eed589eec6f201405abbc398844042ef6d57 category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/16159 CVE: CVE-2026-63830 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=... -------------------------------- [ Upstream commit 406e8a651a7b854c41fecd5117bb282b3a6c2c6b ] The sk_msg sg.copy bitmap is part of the scatterlist entry ownership state. A set bit tells sk_msg_compute_data_pointers() not to expose the entry through writable BPF ctx->data. This protects entries backed by pages that are not private to the sk_msg, such as splice-backed file page-cache pages. Several sk_msg transform paths move, copy, split, or compact msg->sg.data[] entries without moving the matching sg.copy bit. This can make an externally backed entry arrive at a new slot with a clear copy bit. A later SK_MSG verdict can then expose sg_virt(sge) as writable ctx->data and BPF stores can modify the original page cache. Keep sg.copy synchronized with sg.data[] whenever entries are transferred, shifted, split, or copied into a new sk_msg. Clear the bit when an entry is replaced by a newly allocated private page or freed. This covers the BPF pull/push/pop helpers, sk_msg_shift_left/right(), sk_msg_xfer(), and tls_split_open_record(), including the partial tail entry created during TLS open-record splitting. Fixes: d3b18ad31f93 ("tls: add bpf support to sk_msg handling") Cc: stable@vger.kernel.org Reported-by: Yiming Qian <yimingqian591@gmail.com> Reported-by: Keenan Dong <keenanat2000@gmail.com> Signed-off-by: Yiming Qian <yimingqian591@gmail.com> Link: https://patch.msgid.link/20260610062137.49075-1-yimingqian591@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Dong Chenchen <dongchenchen2@huawei.com> --- include/linux/skmsg.h | 15 +++++++++++---- net/core/filter.c | 27 +++++++++++++++++++++++++++ net/core/skmsg.c | 2 ++ net/tls/tls_sw.c | 4 ++++ 4 files changed, 44 insertions(+), 4 deletions(-) diff --git a/include/linux/skmsg.h b/include/linux/skmsg.h index 594524e87e58..2b3d7f773951 100644 --- a/include/linux/skmsg.h +++ b/include/linux/skmsg.h @@ -4,6 +4,7 @@ #ifndef _LINUX_SKMSG_H #define _LINUX_SKMSG_H +#include <linux/bitops.h> #include <linux/bpf.h> #include <linux/filter.h> #include <linux/scatterlist.h> @@ -199,11 +200,14 @@ static inline void sk_msg_xfer(struct sk_msg *dst, struct sk_msg *src, int which, u32 size) { dst->sg.data[which] = src->sg.data[which]; + __assign_bit(which, dst->sg.copy, test_bit(which, src->sg.copy)); dst->sg.data[which].length = size; dst->sg.size += size; src->sg.size -= size; src->sg.data[which].length -= size; src->sg.data[which].offset += size; + if (!src->sg.data[which].length) + __clear_bit(which, src->sg.copy); } static inline void sk_msg_xfer_full(struct sk_msg *dst, struct sk_msg *src) @@ -273,16 +277,19 @@ static inline void sk_msg_page_add(struct sk_msg *msg, struct page *page, static inline void sk_msg_sg_copy(struct sk_msg *msg, u32 i, bool copy_state) { do { - if (copy_state) - __set_bit(i, msg->sg.copy); - else - __clear_bit(i, msg->sg.copy); + __assign_bit(i, msg->sg.copy, copy_state); sk_msg_iter_var_next(i); if (i == msg->sg.end) break; } while (1); } +static inline void sk_msg_sg_copy_assign(struct sk_msg *dst, u32 dst_i, + const struct sk_msg *src, u32 src_i) +{ + __assign_bit(dst_i, dst->sg.copy, test_bit(src_i, src->sg.copy)); +} + static inline void sk_msg_sg_copy_set(struct sk_msg *msg, u32 start) { sk_msg_sg_copy(msg, start, true); diff --git a/net/core/filter.c b/net/core/filter.c index 553b9e6fb16e..6b597e6c8dd4 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -2692,11 +2692,13 @@ BPF_CALL_4(bpf_msg_pull_data, struct sk_msg *, msg, u32, start, poffset += len; sge->length = 0; put_page(sg_page(sge)); + __clear_bit(i, msg->sg.copy); sk_msg_iter_var_next(i); } while (i != last_sge); sg_set_page(&msg->sg.data[first_sge], page, copy, 0); + __clear_bit(first_sge, msg->sg.copy); /* To repair sg ring we need to shift entries. If we only * had a single entry though we can just replace it and @@ -2722,9 +2724,11 @@ BPF_CALL_4(bpf_msg_pull_data, struct sk_msg *, msg, u32, start, break; msg->sg.data[i] = msg->sg.data[move_from]; + sk_msg_sg_copy_assign(msg, i, msg, move_from); msg->sg.data[move_from].length = 0; msg->sg.data[move_from].page_link = 0; msg->sg.data[move_from].offset = 0; + __clear_bit(move_from, msg->sg.copy); sk_msg_iter_var_next(i); } while (1); @@ -2753,6 +2757,7 @@ BPF_CALL_4(bpf_msg_push_data, struct sk_msg *, msg, u32, start, { struct scatterlist sge, nsge, nnsge, rsge = {0}, *psge; u32 new, i = 0, l = 0, space, copy = 0, offset = 0; + bool sge_copy, nsge_copy, nnsge_copy, rsge_copy = false; u8 *raw, *to, *from; struct page *page; @@ -2825,6 +2830,7 @@ BPF_CALL_4(bpf_msg_push_data, struct sk_msg *, msg, u32, start, sk_msg_iter_var_prev(i); psge = sk_msg_elem(msg, i); rsge = sk_msg_elem_cpy(msg, i); + rsge_copy = test_bit(i, msg->sg.copy); psge->length = start - offset; rsge.length -= psge->length; @@ -2849,24 +2855,32 @@ BPF_CALL_4(bpf_msg_push_data, struct sk_msg *, msg, u32, start, /* Shift one or two slots as needed */ sge = sk_msg_elem_cpy(msg, new); + sge_copy = test_bit(new, msg->sg.copy); sg_unmark_end(&sge); nsge = sk_msg_elem_cpy(msg, i); + nsge_copy = test_bit(i, msg->sg.copy); if (rsge.length) { sk_msg_iter_var_next(i); nnsge = sk_msg_elem_cpy(msg, i); + nnsge_copy = test_bit(i, msg->sg.copy); sk_msg_iter_next(msg, end); } while (i != msg->sg.end) { msg->sg.data[i] = sge; + __assign_bit(i, msg->sg.copy, sge_copy); sge = nsge; + sge_copy = nsge_copy; sk_msg_iter_var_next(i); if (rsge.length) { nsge = nnsge; + nsge_copy = nnsge_copy; nnsge = sk_msg_elem_cpy(msg, i); + nnsge_copy = test_bit(i, msg->sg.copy); } else { nsge = sk_msg_elem_cpy(msg, i); + nsge_copy = test_bit(i, msg->sg.copy); } } @@ -2880,6 +2894,7 @@ BPF_CALL_4(bpf_msg_push_data, struct sk_msg *, msg, u32, start, get_page(sg_page(&rsge)); sk_msg_iter_var_next(new); msg->sg.data[new] = rsge; + __assign_bit(new, msg->sg.copy, rsge_copy); } sk_msg_reset_curr(msg); @@ -2907,25 +2922,33 @@ static void sk_msg_shift_left(struct sk_msg *msg, int i) prev = i; sk_msg_iter_var_next(i); msg->sg.data[prev] = msg->sg.data[i]; + sk_msg_sg_copy_assign(msg, prev, msg, i); } while (i != msg->sg.end); sk_msg_iter_prev(msg, end); + __clear_bit(msg->sg.end, msg->sg.copy); } static void sk_msg_shift_right(struct sk_msg *msg, int i) { struct scatterlist tmp, sge; + bool tmp_copy, sge_copy; sk_msg_iter_next(msg, end); sge = sk_msg_elem_cpy(msg, i); + sge_copy = test_bit(i, msg->sg.copy); sk_msg_iter_var_next(i); tmp = sk_msg_elem_cpy(msg, i); + tmp_copy = test_bit(i, msg->sg.copy); while (i != msg->sg.end) { msg->sg.data[i] = sge; + __assign_bit(i, msg->sg.copy, sge_copy); sk_msg_iter_var_next(i); sge = tmp; + sge_copy = tmp_copy; tmp = sk_msg_elem_cpy(msg, i); + tmp_copy = test_bit(i, msg->sg.copy); } } @@ -2985,6 +3008,8 @@ BPF_CALL_4(bpf_msg_pop_data, struct sk_msg *, msg, u32, start, struct scatterlist *nsge, *sge = sk_msg_elem(msg, i); int a = start - offset; int b = sge->length - pop - a; + u32 sge_i = i; + bool sge_copy = test_bit(i, msg->sg.copy); sk_msg_iter_var_next(i); @@ -2997,6 +3022,7 @@ BPF_CALL_4(bpf_msg_pop_data, struct sk_msg *, msg, u32, start, sg_set_page(nsge, sg_page(sge), b, sge->offset + pop + a); + __assign_bit(i, msg->sg.copy, sge_copy); } else { struct page *page, *orig; u8 *to, *from; @@ -3013,6 +3039,7 @@ BPF_CALL_4(bpf_msg_pop_data, struct sk_msg *, msg, u32, start, memcpy(to, from, a); memcpy(to + a, from + a + pop, b); sg_set_page(sge, page, a + b, 0); + __clear_bit(sge_i, msg->sg.copy); put_page(orig); } pop = 0; diff --git a/net/core/skmsg.c b/net/core/skmsg.c index b979a4cba4d2..11d771266635 100644 --- a/net/core/skmsg.c +++ b/net/core/skmsg.c @@ -65,6 +65,7 @@ int sk_msg_alloc(struct sock *sk, struct sk_msg *msg, int len, sge = &msg->sg.data[msg->sg.end]; sg_unmark_end(sge); sg_set_page(sge, pfrag->page, use, orig_offset); + __clear_bit(msg->sg.end, msg->sg.copy); get_page(pfrag->page); sk_msg_iter_next(msg, end); } @@ -185,6 +186,7 @@ static int sk_msg_free_elem(struct sock *sk, struct sk_msg *msg, u32 i, sk_mem_uncharge(sk, len); put_page(sg_page(sge)); } + __clear_bit(i, msg->sg.copy); memset(sge, 0, sizeof(*sge)); return len; } diff --git a/net/tls/tls_sw.c b/net/tls/tls_sw.c index de3303baef70..98a7621d57f7 100644 --- a/net/tls/tls_sw.c +++ b/net/tls/tls_sw.c @@ -574,6 +574,7 @@ static int tls_split_open_record(struct sock *sk, struct tls_rec *from, struct scatterlist *sge, *osge, *nsge; u32 orig_size = msg_opl->sg.size; struct scatterlist tmp = { }; + u32 tmp_i = 0; struct sk_msg *msg_npl; struct tls_rec *new; int ret; @@ -595,6 +596,7 @@ static int tls_split_open_record(struct sock *sk, struct tls_rec *from, if (sge->length > apply) { u32 len = sge->length - apply; + tmp_i = i; get_page(sg_page(sge)); sg_set_page(&tmp, sg_page(sge), len, sge->offset + apply); @@ -626,6 +628,7 @@ static int tls_split_open_record(struct sock *sk, struct tls_rec *from, nsge = sk_msg_elem(msg_npl, j); if (tmp.length) { memcpy(nsge, &tmp, sizeof(*nsge)); + sk_msg_sg_copy_assign(msg_npl, j, msg_opl, tmp_i); sk_msg_iter_var_next(j); nsge = sk_msg_elem(msg_npl, j); } @@ -633,6 +636,7 @@ static int tls_split_open_record(struct sock *sk, struct tls_rec *from, osge = sk_msg_elem(msg_opl, i); while (osge->length) { memcpy(nsge, osge, sizeof(*nsge)); + sk_msg_sg_copy_assign(msg_npl, j, msg_opl, i); sg_unmark_end(nsge); sk_msg_iter_var_next(i); sk_msg_iter_var_next(j); -- 2.43.0