In the Linux kernel, the following vulnerability has been resolved:<\/p>\n
ppp_async: drop the errored frame instead of resetting its headroom<\/p>\n
ppp_receive_nonmp_frame() prepends a two-byte direction tag before running
\nthe pass\/active BPF filters:<\/p>\n
\t*(__be16 *)skb_push(skb, 2) = htons(PPP_FILTER_INBOUND_TAG);<\/p>\n
Nothing on the receive path guarantees those two bytes of headroom. The
\nframe-error path in ppp_async’s process_input_packet() resets a reused skb’s
\nheadroom to zero while claiming to restore it to a freshly allocated state –
\nbut a fresh skb from dev_alloc_skb() carries NET_SKB_PAD:<\/p>\n
\terr:
\n\t\tif (skb) {
\n\t\t\t\/* make skb appear as freshly allocated *\/
\n\t\t\tskb_trim(skb, 0);
\n\t\t\tskb_reserve(skb, – skb_headroom(skb));
\n\t\t}<\/p>\n
ap->rpkt still points at that skb, so the next frame is reassembled into it
\nwith no headroom at all. A peer that sends a bad-FCS frame followed by one
\nbeginning ff 03 then leaves a single byte of headroom by the time the filter
\ntag is pushed, which lands one byte below skb->head:<\/p>\n
skbuff: skb_under_panic: len:49 put:2 head:ffff888003c10000
\n data:ffff888003c0ffff tail:0x30 end:0x640 dev:
\n kernel BUG at net\/core\/skbuff.c:214!
\n RIP: 0010:skb_panic+0x13e\/0x230
\n Call Trace:
\n skb_push+0xbd\/0x100
\n ppp_receive_nonmp_frame+0x48a\/0x1d10
\n ppp_input+0x4e9\/0x2f80
\n ppp_async_process+0x2a\/0xe0
\n tasklet_action_common+0x20f\/0x8a0
\n handle_softirqs+0x18e\/0x590
\n Kernel panic – not syncing: Fatal exception in interrupt<\/p>\n
Zeroing the headroom violates the NET_SKB_PAD guarantee that dev_alloc_skb()
\ngives the rest of the receive path. Besides the filter panic above, when CCP
\ncompression is enabled ppp_decompress_frame() hands skb->data – 2 to
\n->decompress()\/->incomp(), which then reads out of bounds before skb->head
\nfor the same reason.<\/p>\n
Rather than restore the headroom, drop the errored frame – as ppp_synctty
\nalready does on its error path – and clear ap->rpkt so the next frame is
\nreassembled into a fresh skb with proper headroom. This is simpler and fixes
\nboth the filter under-panic and the CCP out-of-bounds read.<\/p>\n
The original V1 of this patch made room in ppp_receive_nonmp_frame() with
\nskb_cow_head(); Eric pointed out that fixing the root cause in the transport
\nis the right approach.<\/p>\n
Found by fuzzing the PPP receive path with a mutating peer on a pty; it is an
\ninteresting (remote) DoS: root configures PPP, the peer supplies two crashing
\nframes. The reproducer (repro-ppp-skb.c, unchanged from v1) panics in about a
\nsecond, and returns cleanly with this applied.<\/p>\n","protected":false},"excerpt":{"rendered":"
In the Linux kernel, the following vulnerability has been resolved:<\/p>\n
ppp_async: drop the errored frame instead of resetting its headroom<\/p>\n
ppp_receive_nonmp_frame() prepends a two-byte direction tag be<\/p>\n","protected":false},"template":"","meta":[],"categories":[],"tags":[],"class_list":["post-29312","vuln","type-vuln","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/www.54iter.cn\/index.php?rest_route=\/wp\/v2\/vuln\/29312","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.54iter.cn\/index.php?rest_route=\/wp\/v2\/vuln"}],"about":[{"href":"https:\/\/www.54iter.cn\/index.php?rest_route=\/wp\/v2\/types\/vuln"}],"wp:attachment":[{"href":"https:\/\/www.54iter.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=29312"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.54iter.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=29312"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.54iter.cn\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=29312"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}