From 15ab0548e3107665c34579ae523b2b6e7c22082a Mon Sep 17 00:00:00 2001 From: Zijian Zhang Date: Wed, 6 Nov 2024 22:25:18 +0000 Subject: bpf, sockmap: Several fixes to bpf_msg_push_data Several fixes to bpf_msg_push_data, 1. test_sockmap has tests where bpf_msg_push_data is invoked to push some data at the end of a message, but -EINVAL is returned. In this case, in bpf_msg_push_data, after the first loop, i will be set to msg->sg.end, add the logic to handle it. 2. In the code block of "if (start - offset)", it's possible that "i" points to the last of sk_msg_elem. In this case, "sk_msg_iter_next(msg, end)" might still be called twice, another invoking is in "if (!copy)" code block, but actually only one is needed. Add the logic to handle it, and reconstruct the code to make the logic more clear. Fixes: 6fff607e2f14 ("bpf: sk_msg program helper bpf_msg_push_data") Signed-off-by: Zijian Zhang Link: https://lore.kernel.org/r/20241106222520.527076-7-zijianzhang@bytedance.com Signed-off-by: Martin KaFai Lau --- net/core/filter.c | 53 +++++++++++++++++++++++++++++++++-------------------- 1 file changed, 33 insertions(+), 20 deletions(-) (limited to 'net') diff --git a/net/core/filter.c b/net/core/filter.c index 82f92ed0dc72..255d58bae2a9 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -2778,7 +2778,7 @@ BPF_CALL_4(bpf_msg_push_data, struct sk_msg *, msg, u32, start, sk_msg_iter_var_next(i); } while (i != msg->sg.end); - if (start >= offset + l) + if (start > offset + l) return -EINVAL; space = MAX_MSG_FRAGS - sk_msg_elem_used(msg); @@ -2803,6 +2803,8 @@ BPF_CALL_4(bpf_msg_push_data, struct sk_msg *, msg, u32, start, raw = page_address(page); + if (i == msg->sg.end) + sk_msg_iter_var_prev(i); psge = sk_msg_elem(msg, i); front = start - offset; back = psge->length - front; @@ -2819,7 +2821,13 @@ BPF_CALL_4(bpf_msg_push_data, struct sk_msg *, msg, u32, start, } put_page(sg_page(psge)); - } else if (start - offset) { + new = i; + goto place_new; + } + + if (start - offset) { + if (i == msg->sg.end) + sk_msg_iter_var_prev(i); psge = sk_msg_elem(msg, i); rsge = sk_msg_elem_cpy(msg, i); @@ -2830,39 +2838,44 @@ BPF_CALL_4(bpf_msg_push_data, struct sk_msg *, msg, u32, start, sk_msg_iter_var_next(i); sg_unmark_end(psge); sg_unmark_end(&rsge); - sk_msg_iter_next(msg, end); } /* Slot(s) to place newly allocated data */ + sk_msg_iter_next(msg, end); new = i; + sk_msg_iter_var_next(i); + + if (i == msg->sg.end) { + if (!rsge.length) + goto place_new; + sk_msg_iter_next(msg, end); + goto place_new; + } /* Shift one or two slots as needed */ - if (!copy) { - sge = sk_msg_elem_cpy(msg, i); + sge = sk_msg_elem_cpy(msg, new); + sg_unmark_end(&sge); + nsge = sk_msg_elem_cpy(msg, i); + if (rsge.length) { sk_msg_iter_var_next(i); - sg_unmark_end(&sge); + nnsge = sk_msg_elem_cpy(msg, i); sk_msg_iter_next(msg, end); + } - nsge = sk_msg_elem_cpy(msg, i); + while (i != msg->sg.end) { + msg->sg.data[i] = sge; + sge = nsge; + sk_msg_iter_var_next(i); if (rsge.length) { - sk_msg_iter_var_next(i); + nsge = nnsge; nnsge = sk_msg_elem_cpy(msg, i); - } - - while (i != msg->sg.end) { - msg->sg.data[i] = sge; - sge = nsge; - sk_msg_iter_var_next(i); - if (rsge.length) { - nsge = nnsge; - nnsge = sk_msg_elem_cpy(msg, i); - } else { - nsge = sk_msg_elem_cpy(msg, i); - } + } else { + nsge = sk_msg_elem_cpy(msg, i); } } +place_new: /* Place newly allocated data buffer */ sk_mem_charge(msg->sk, len); msg->sg.size += len; -- cgit v1.2.3 From 5d609ba262475db450ba69b8e8a557bd768ac07a Mon Sep 17 00:00:00 2001 From: Zijian Zhang Date: Wed, 6 Nov 2024 22:25:19 +0000 Subject: bpf, sockmap: Several fixes to bpf_msg_pop_data Several fixes to bpf_msg_pop_data, 1. In sk_msg_shift_left, we should put_page 2. if (len == 0), return early is better 3. pop the entire sk_msg (last == msg->sg.size) should be supported 4. Fix for the value of variable "a" 5. In sk_msg_shift_left, after shifting, i has already pointed to the next element. Addtional sk_msg_iter_var_next may result in BUG. Fixes: 7246d8ed4dcc ("bpf: helper to pop data from messages") Signed-off-by: Zijian Zhang Reviewed-by: John Fastabend Link: https://lore.kernel.org/r/20241106222520.527076-8-zijianzhang@bytedance.com Signed-off-by: Martin KaFai Lau --- net/core/filter.c | 15 +++++++++------ 1 file changed, 9 insertions(+), 6 deletions(-) (limited to 'net') diff --git a/net/core/filter.c b/net/core/filter.c index 255d58bae2a9..2fdba950b575 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -2904,8 +2904,10 @@ static const struct bpf_func_proto bpf_msg_push_data_proto = { static void sk_msg_shift_left(struct sk_msg *msg, int i) { + struct scatterlist *sge = sk_msg_elem(msg, i); int prev; + put_page(sg_page(sge)); do { prev = i; sk_msg_iter_var_next(i); @@ -2942,6 +2944,9 @@ BPF_CALL_4(bpf_msg_pop_data, struct sk_msg *, msg, u32, start, if (unlikely(flags)) return -EINVAL; + if (unlikely(len == 0)) + return 0; + /* First find the starting scatterlist element */ i = msg->sg.start; do { @@ -2954,7 +2959,7 @@ BPF_CALL_4(bpf_msg_pop_data, struct sk_msg *, msg, u32, start, } while (i != msg->sg.end); /* Bounds checks: start and pop must be inside message */ - if (start >= offset + l || last >= msg->sg.size) + if (start >= offset + l || last > msg->sg.size) return -EINVAL; space = MAX_MSG_FRAGS - sk_msg_elem_used(msg); @@ -2983,12 +2988,12 @@ BPF_CALL_4(bpf_msg_pop_data, struct sk_msg *, msg, u32, start, */ if (start != offset) { struct scatterlist *nsge, *sge = sk_msg_elem(msg, i); - int a = start; + int a = start - offset; int b = sge->length - pop - a; sk_msg_iter_var_next(i); - if (pop < sge->length - a) { + if (b > 0) { if (space) { sge->length = a; sk_msg_shift_right(msg, i); @@ -3007,7 +3012,6 @@ BPF_CALL_4(bpf_msg_pop_data, struct sk_msg *, msg, u32, start, if (unlikely(!page)) return -ENOMEM; - sge->length = a; orig = sg_page(sge); from = sg_virt(sge); to = page_address(page); @@ -3017,7 +3021,7 @@ BPF_CALL_4(bpf_msg_pop_data, struct sk_msg *, msg, u32, start, put_page(orig); } pop = 0; - } else if (pop >= sge->length - a) { + } else { pop -= (sge->length - a); sge->length = a; } @@ -3051,7 +3055,6 @@ BPF_CALL_4(bpf_msg_pop_data, struct sk_msg *, msg, u32, start, pop -= sge->length; sk_msg_shift_left(msg, i); } - sk_msg_iter_var_next(i); } sk_mem_uncharge(msg->sk, len - pop); -- cgit v1.2.3 From 955afd57dc4bf7e8c620a0a9e3af3c881c2c6dff Mon Sep 17 00:00:00 2001 From: Zijian Zhang Date: Wed, 6 Nov 2024 22:25:20 +0000 Subject: bpf, sockmap: Fix sk_msg_reset_curr Found in the test_txmsg_pull in test_sockmap, ``` txmsg_cork = 512; // corking is importrant here opt->iov_length = 3; opt->iov_count = 1; opt->rate = 512; // sendmsg will be invoked 512 times ``` The first sendmsg will send an sk_msg with size 3, and bpf_msg_pull_data will be invoked the first time. sk_msg_reset_curr will reset the copybreak from 3 to 0. In the second sendmsg, since we are in the stage of corking, psock->cork will be reused in func sk_msg_alloc. msg->sg.copybreak is 0 now, the second msg will overwrite the first msg. As a result, we could not pass the data integrity test. The same problem happens in push and pop test. Thus, fix sk_msg_reset_curr to restore the correct copybreak. Fixes: bb9aefde5bba ("bpf: sockmap, updating the sg structure should also update curr") Signed-off-by: Zijian Zhang Link: https://lore.kernel.org/r/20241106222520.527076-9-zijianzhang@bytedance.com Signed-off-by: Martin KaFai Lau --- net/core/filter.c | 20 +++++++++----------- 1 file changed, 9 insertions(+), 11 deletions(-) (limited to 'net') diff --git a/net/core/filter.c b/net/core/filter.c index 2fdba950b575..64248d0ac4ad 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -2604,18 +2604,16 @@ BPF_CALL_2(bpf_msg_cork_bytes, struct sk_msg *, msg, u32, bytes) static void sk_msg_reset_curr(struct sk_msg *msg) { - u32 i = msg->sg.start; - u32 len = 0; - - do { - len += sk_msg_elem(msg, i)->length; - sk_msg_iter_var_next(i); - if (len >= msg->sg.size) - break; - } while (i != msg->sg.end); + if (!msg->sg.size) { + msg->sg.curr = msg->sg.start; + msg->sg.copybreak = 0; + } else { + u32 i = msg->sg.end; - msg->sg.curr = i; - msg->sg.copybreak = 0; + sk_msg_iter_var_prev(i); + msg->sg.curr = i; + msg->sg.copybreak = msg->sg.data[i].length; + } } static const struct bpf_func_proto bpf_msg_cork_bytes_proto = { -- cgit v1.2.3