ila: call nf_unregister_net_hooks() sooner

In the Linux kernel, the following vulnerability has been resolved:

ila: call nf_unregister_net_hooks() sooner

syzbot found an use-after-free Read in ila_nf_input [1]

Issue here is that ila_xlat_exit_net() frees the rhashtable,

then call nf_unregister_net_hooks().

It should be done in the reverse way, with a synchronize_rcu().

This is a good match for a pre_exit() method.

[1]

BUG: KASAN: use-after-free in rht_key_hashfn include/linux/rhashtable.h:159 [inline]

BUG: KASAN: use-after-free in __rhashtable_lookup include/linux/rhashtable.h:604 [inline]

BUG: KASAN: use-after-free in rhashtable_lookup include/linux/rhashtable.h:646 [inline]

BUG: KASAN: use-after-free in rhashtable_lookup_fast+0x77a/0x9b0 include/linux/rhashtable.h:672

Read of size 4 at addr ffff888064620008 by task ksoftirqd/0/16

CPU: 0 UID: 0 PID: 16 Comm: ksoftirqd/0 Not tainted 6.11.0-rc4-syzkaller-00238-g2ad6d23f465a #0

Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 08/06/2024

Call Trace:

<TASK>

__dump_stack lib/dump_stack.c:93 [inline]

dump_stack_lvl+0x241/0x360 lib/dump_stack.c:119

print_address_description mm/kasan/report.c:377 [inline]

print_report+0x169/0x550 mm/kasan/report.c:488

kasan_report+0x143/0x180 mm/kasan/report.c:601

rht_key_hashfn include/linux/rhashtable.h:159 [inline]

__rhashtable_lookup include/linux/rhashtable.h:604 [inline]

rhashtable_lookup include/linux/rhashtable.h:646 [inline]

rhashtable_lookup_fast+0x77a/0x9b0 include/linux/rhashtable.h:672

ila_lookup_wildcards net/ipv6/ila/ila_xlat.c:132 [inline]

ila_xlat_addr net/ipv6/ila/ila_xlat.c:652 [inline]

ila_nf_input+0x1fe/0x3c0 net/ipv6/ila/ila_xlat.c:190

nf_hook_entry_hookfn include/linux/netfilter.h:154 [inline]

nf_hook_slow+0xc3/0x220 net/netfilter/core.c:626

nf_hook include/linux/netfilter.h:269 [inline]

NF_HOOK+0x29e/0x450 include/linux/netfilter.h:312

__netif_receive_skb_one_core net/core/dev.c:5661 [inline]

__netif_receive_skb+0x1ea/0x650 net/core/dev.c:5775

process_backlog+0x662/0x15b0 net/core/dev.c:6108

__napi_poll+0xcb/0x490 net/core/dev.c:6772

napi_poll net/core/dev.c:6841 [inline]

net_rx_action+0x89b/0x1240 net/core/dev.c:6963

handle_softirqs+0x2c4/0x970 kernel/softirq.c:554

run_ksoftirqd+0xca/0x130 kernel/softirq.c:928

smpboot_thread_fn+0x544/0xa30 kernel/smpboot.c:164

kthread+0x2f0/0x390 kernel/kthread.c:389

ret_from_fork+0x4b/0x80 arch/x86/kernel/process.c:147

ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:244

</TASK>

The buggy address belongs to the physical page:

page: refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x64620

flags: 0xfff00000000000(node=0|zone=1|lastcpupid=0x7ff)

page_type: 0xbfffffff(buddy)

raw: 00fff00000000000 ffffea0000959608 ffffea00019d9408 0000000000000000

raw: 0000000000000000 0000000000000003 00000000bfffffff 0000000000000000

page dumped because: kasan: bad access detected

page_owner tracks the page as freed

page last allocated via order 3, migratetype Unmovable, gfp_mask 0x52dc0(GFP_KERNEL|__GFP_NOWARN|__GFP_NORETRY|__GFP_COMP|__GFP_ZERO), pid 5242, tgid 5242 (syz-executor), ts 73611328570, free_ts 618981657187

set_page_owner include/linux/page_owner.h:32 [inline]

post_alloc_hook+0x1f3/0x230 mm/page_alloc.c:1493

prep_new_page mm/page_alloc.c:1501 [inline]

get_page_from_freelist+0x2e4c/0x2f10 mm/page_alloc.c:3439

__alloc_pages_noprof+0x256/0x6c0 mm/page_alloc.c:4695

__alloc_pages_node_noprof include/linux/gfp.h:269 [inline]

alloc_pages_node_noprof include/linux/gfp.h:296 [inline]

___kmalloc_large_node+0x8b/0x1d0 mm/slub.c:4103

__kmalloc_large_node_noprof+0x1a/0x80 mm/slub.c:4130

__do_kmalloc_node mm/slub.c:4146 [inline]

__kmalloc_node_noprof+0x2d2/0x440 mm/slub.c:4164

__kvmalloc_node_noprof+0x72/0x190 mm/util.c:650

bucket_table_alloc lib/rhashtable.c:186 [inline]

rhashtable_init_noprof+0x534/0xa60 lib/rhashtable.c:1071

ila_xlat_init_net+0xa0/0x110 net/ipv6/ila/ila_xlat.c:613

ops_ini

---truncated---