| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <CANn89iJ63Luv43WrgqD75c5oYmSrGX-qxYwDVcqpjP8005jTew@mail.gmail.com>
Date: Fri, 7 Feb 2025 07:25:28 +0100
From: Eric Dumazet <edumazet@...gle.com>
To: Jiacheng Xu <stitch@....edu.cn>
Cc: andrew+netdev@...n.ch, davem@...emloft.net, kuba@...nel.org,
pabeni@...hat.com, netdev@...r.kernel.org, wolffd@...p.nus.edu.sg
Subject: Re: [BUG] KASAN: slab-use-after-free in slip_open
On Fri, Feb 7, 2025 at 6:08 AM Jiacheng Xu <stitch@....edu.cn> wrote:
>
> Hi developers,
>
> We update patch as follows. In this version, we avoid to call `dev_close` inside an RCU read-side critical section while still preventing the use of freed memory.
>
This makes no sense for so many reasons.
At this point RTNL is held (drivers/net/slip/slip.c line 803)
rcu_read_lock() is not adding anything in a writer path.
This is really contrary to RCU rules to think that rcu_read_lock()
would be needed here.
Also what happens if kmalloc_array() fails ?
Please, find the root cause, explain what other part of the kernel is
not holding RTNL while it should,
then fix it ?
> --- slip.c 2025-02-07 02:29:27.551293235 +0000
> +++ slip.c 2025-02-07 02:46:57.156076645 +0000
> @@ -717,21 +717,41 @@
> /* Collect hanged up channels */
> static void sl_sync(void)
> {
> - int i;
> + int i, count;
> struct net_device *dev;
> + struct net_device **devs;
> struct slip *sl;
>
> + devs = kmalloc_array(slip_maxdev, sizeof(*devs), GFP_KERNEL);
> + if (!devs)
> + return;
> +
> + rcu_read_lock();
> + for (i = 0; i< slip_maxdev; i++) {
> + dev = rcu_dereference(slip_devs[i]);
> + if (!dev)
> + break;
> + dev_hold(dev);
> + devs[count++] = dev;
> + }
> + rcu_read_unlock();
> +
> for (i = 0; i < slip_maxdev; i++) {
> - dev = slip_devs[i];
> + dev = devs[i];
> if (dev == NULL)
> break;
>
> sl = netdev_priv(dev);
> - if (sl->tty || sl->leased)
> + if (sl->tty || sl->leased) {
> + dev_put(dev);
> continue;
> + }
> if (dev->flags & IFF_UP)
> dev_close(dev);
> + dev_put(dev);
> }
> +
> + kfree(devs);
> }
>
> Best regards,
> Jiacheng
>
> > -----原始邮件-----
> > 发件人: "Jiacheng Xu" <stitch@....edu.cn>
> > 发送时间:2025-02-06 20:30:09 (星期四)
> > 收件人: andrew+netdev@...n.ch, davem@...emloft.net, edumazet@...gle.com, kuba@...nel.org, pabeni@...hat.com, netdev@...r.kernel.org
> > 抄送: wolffd@...p.nus.edu.sg
> > 主题: [BUG] KASAN: slab-use-after-free in slip_open
> >
> > Hi developers:
> >
> > We are reporting a Linux issue using a modified version of Syzkaller.
> >
> > HEAD commit: 4bbf9020 6.13.0-rc4
> > git tree: upstream
> > kernel config: https://github.com/google/syzkaller/blob/master/dashboard/config/linux/upstream-apparmor-kasan.config
> > syz repro: https://drive.google.com/file/d/1XV9WHwMvK4eZIPuyjiWoTpyQ0FpDaUpT/view?usp=sharing
> > C reproducer: As this is a race condition, a stable C reproducer is not available yet.
> >
> > Environment:
> > Ubuntu 22.04 on Linux 5.15
> > QEMU emulator version 6.2.0
> > qemu-system-x86_64 \
> > -m 2G \
> > -smp 2 \
> > -kernel /home/wd/bzImage \
> > -append "console=ttyS0 root=/dev/sda earlyprintk=serial net.ifnames=0" \
> > -drive file=/home/wd/bullseye.img,format=raw \
> > -net user,host=10.0.2.10,hostfwd=tcp:127.0.0.1:10021-:22 \
> > -net nic,model=e1000 \
> > -enable-kvm \
> > -nographic \
> > -pidfile vm.pid \
> > 2>&1 | tee vm.log
> >
> > Steps to reproduce:
> > 1. Setup a vm on given kernel.
> > 2. Execute the syz reproducer with `./syz-execprog -executor=./syz-executor -repeat=20000 -procs=8 -cover=0 ~/repro.prog`
> >
> > Description:
> > The SLIP driver suffers from a Use-After-Free issue in sl_sync() due to unsynchronized access to net_device pointers stored in slip_devs[]. When a device is being freed (for example, via tty_release()/netdev_run_todo()), sl_sync() may access it after it has been freed.
> >
> > Patch:
> > We try to write a patch for the crash. And after applying the patch, the crash can no longer be triggered with the PoC.
> >
> > --- slip.c 2025-02-06 12:26:25.690890378 +0000
> > +++ slip.c 2025-02-06 12:29:30.554820899 +0000
> > @@ -721,8 +721,9 @@
> > struct net_device *dev;
> > struct slip *sl;
> >
> > + rcu_read_lock();
> > for (i = 0; i < slip_maxdev; i++) {
> > - dev = slip_devs[i];
> > + dev = rcu_dereference(slip_devs[i]);
> > if (dev == NULL)
> > break;
> >
> > @@ -732,6 +733,7 @@
> > if (dev->flags & IFF_UP)
> > dev_close(dev);
> > }
> > + rcu_read_unlock();
> > }
> >
> >
> > If you fix this issue, please add the following tag to the commit:
> > Reported-by: Jiacheng Xu <stitch@....edu.cn>, Dylan Wolff <wolffd@...p.nus.edu.sg>
> > ==================================================================
> > BUG: KASAN: slab-use-after-free in sl_sync drivers/net/slip/slip.c:732 [inline]
> > BUG: KASAN: slab-use-after-free in slip_open+0x293/0x1330 drivers/net/slip/slip.c:806
> > Read of size 4 at addr ffff88805cfca0b0 by task syz-executor.2/46673
> >
> >
> > CPU: 0 UID: 0 PID: 46673 Comm: syz-executor.2 Not tainted 6.13.0-rc4 #2
> >
> >
> > Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.15.0-1 04/01/2014
> >
> > Sched_ext: serialise (enabled+all), task: runnable_at=-30ms
> > Call Trace:
> > <TASK>
> > __dump_stack lib/dump_stack.c:94 [inline]
> > dump_stack_lvl+0x229/0x350 lib/dump_stack.c:120
> > print_address_description mm/kasan/report.c:378 [inline]
> > print_report+0x164/0x530 mm/kasan/report.c:489
> > kasan_report+0x147/0x180 mm/kasan/report.c:602
> > sl_sync drivers/net/slip/slip.c:732 [inline]
> > slip_open+0x293/0x1330 drivers/net/slip/slip.c:806
> > tty_ldisc_open+0xbd/0x120 drivers/tty/tty_ldisc.c:432
> > tty_set_ldisc+0x32b/0x5c0 drivers/tty/tty_ldisc.c:563
> > tiocsetd+0x126/0x170 drivers/tty/tty_io.c:2439
> > tty_ioctl+0xe5c/0x1090 drivers/tty/tty_io.c:2739
> > vfs_ioctl fs/ioctl.c:51 [inline]
> > __do_sys_ioctl fs/ioctl.c:906 [inline]
> > __se_sys_ioctl+0x266/0x350 fs/ioctl.c:892
> > do_syscall_x64 arch/x86/entry/common.c:52 [inline]
> > do_syscall_64+0xf6/0x210 arch/x86/entry/common.c:83
> > entry_SYSCALL_64_after_hwframe+0x77/0x7f
> > RIP: 0033:0x7fed56a903ad
> > Code: c3 e8 a7 2b 00 00 0f 1f 80 00 00 00 00 f3 0f 1e fa 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 b0 ff ff ff f7 d8 64 89 01 48
> > RSP: 002b:00007fed578b20c8 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
> > RAX: ffffffffffffffda RBX: 00007fed56bcbf80 RCX: 00007fed56a903ad
> > RDX: 0000000020000080 RSI: 0000000000005423 RDI: 0000000000000003
> > RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000
> > R10: 0000000000000000 R11: 0000000000000246 R12: 00007fed578b2640
> > R13: 000000000000000e R14: 00007fed56a4fc90 R15: 00007fed578aa000
> > </TASK>
> >
> >
> > Allocated by task 46683:
> >
> >
> > kasan_save_stack mm/kasan/common.c:47 [inline]
> >
> > kasan_save_track+0x3f/0x80 mm/kasan/common.c:68
> > poison_kmalloc_redzone mm/kasan/common.c:377 [inline]
> > __kasan_kmalloc+0x89/0xa0 mm/kasan/common.c:394
> > kasan_kmalloc include/linux/kasan.h:260 [inline]
> > __do_kmalloc_node mm/slub.c:4298 [inline]
> > __kmalloc_node_noprof+0x28c/0x530 mm/slub.c:4304
> > __kvmalloc_node_noprof+0x70/0x180 mm/util.c:650
> > alloc_netdev_mqs+0xa7/0x1870 net/core/dev.c:11209
> > sl_alloc drivers/net/slip/slip.c:756 [inline]
> > slip_open+0x483/0x1330 drivers/net/slip/slip.c:817
> > tty_ldisc_open+0xbd/0x120 drivers/tty/tty_ldisc.c:432
> > tty_set_ldisc+0x32b/0x5c0 drivers/tty/tty_ldisc.c:563
> > tiocsetd+0x126/0x170 drivers/tty/tty_io.c:2439
> > tty_ioctl+0xe5c/0x1090 drivers/tty/tty_io.c:2739
> > vfs_ioctl fs/ioctl.c:51 [inline]
> > __do_sys_ioctl fs/ioctl.c:906 [inline]
> > __se_sys_ioctl+0x266/0x350 fs/ioctl.c:892
> > do_syscall_x64 arch/x86/entry/common.c:52 [inline]
> > do_syscall_64+0xf6/0x210 arch/x86/entry/common.c:83
> > entry_SYSCALL_64_after_hwframe+0x77/0x7f
> >
> >
> > Freed by task 46682:
> >
> >
> > kasan_save_stack mm/kasan/common.c:47 [inline]
> >
> > kasan_save_track+0x3f/0x80 mm/kasan/common.c:68
> > kasan_save_free_info+0x40/0x50 mm/kasan/generic.c:582
> > poison_slab_object mm/kasan/common.c:247 [inline]
> > __kasan_slab_free+0x5a/0x70 mm/kasan/common.c:264
> > kasan_slab_free include/linux/kasan.h:233 [inline]
> > slab_free_hook mm/slub.c:2353 [inline]
> > slab_free mm/slub.c:4613 [inline]
> > kfree+0x196/0x450 mm/slub.c:4761
> > device_release+0xcd/0x240
> > kobject_cleanup lib/kobject.c:689 [inline]
> > kobject_release lib/kobject.c:720 [inline]
> > kref_put include/linux/kref.h:65 [inline]
> > kobject_put+0x248/0x490 lib/kobject.c:737
> > netdev_run_todo+0x10e0/0x1280 net/core/dev.c:10924
> > tty_ldisc_kill+0xbf/0x150 drivers/tty/tty_ldisc.c:613
> > tty_ldisc_release+0x1ae/0x210 drivers/tty/tty_ldisc.c:781
> > tty_release_struct+0x2a/0x100 drivers/tty/tty_io.c:1690
> > tty_release+0xe4d/0x1460 drivers/tty/tty_io.c:1861
> > __fput+0x2ba/0xa80 fs/file_table.c:450
> > __fput_sync+0x180/0x1e0 fs/file_table.c:535
> > __do_sys_close fs/open.c:1554 [inline]
> > __se_sys_close fs/open.c:1539 [inline]
> > __x64_sys_close+0x93/0x120 fs/open.c:1539
> > do_syscall_x64 arch/x86/entry/common.c:52 [inline]
> > do_syscall_64+0xf6/0x210 arch/x86/entry/common.c:83
> > entry_SYSCALL_64_after_hwframe+0x77/0x7f
> >
> >
> > The buggy address belongs to the object at ffff88805cfca000
> >
> >
> > which belongs to the cache kmalloc-cg-4k of size 4096
> >
> > The buggy address is located 176 bytes inside of
> > freed 4096-byte region [ffff88805cfca000, ffff88805cfcb000)
> >
> >
> > The buggy address belongs to the physical page:
> >
> >
> > page: refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x5cfc8
> >
> > head: order:3 mapcount:0 entire_mapcount:0 nr_pages_mapped:0 pincount:0
> > memcg:ffff888075c64a01
> > flags: 0x4fff00000000040(head|node=1|zone=1|lastcpupid=0x7ff)
> > page_type: f5(slab)
> > raw: 04fff00000000040 ffff88801d44f500 dead000000000100 dead000000000122
> > raw: 0000000000000000 0000000000040004 00000001f5000000 ffff888075c64a01
> > head: 04fff00000000040 ffff88801d44f500 dead000000000100 dead000000000122
> > head: 0000000000000000 0000000000040004 00000001f5000000 ffff888075c64a01
> > head: 04fff00000000003 ffffea000173f201 ffffffffffffffff 0000000000000000
> > head: 0000000000000008 0000000000000000 00000000ffffffff 0000000000000000
> > page dumped because: kasan: bad access detected
> > page_owner tracks the page as allocated
> > page last allocated via order 3, migratetype Unmovable, gfp_mask 0xd20c0(__GFP_IO|__GFP_FS|__GFP_NOWARN|__GFP_NORETRY|__GFP_COMP|__GFP_NOMEMALLOC), pid 9660, tgid 9660 (syz-executor.1), ts 93383578089, free_ts 93076565836
> > set_page_owner include/linux/page_owner.h:32 [inline]
> > post_alloc_hook+0x1f6/0x240 mm/page_alloc.c:1558
> > prep_new_page mm/page_alloc.c:1566 [inline]
> > get_page_from_freelist+0x3586/0x36d0 mm/page_alloc.c:3476
> > __alloc_pages_noprof+0x260/0x680 mm/page_alloc.c:4753
> > alloc_pages_mpol_noprof+0x3c8/0x650 mm/mempolicy.c:2269
> > alloc_slab_page+0x6a/0x110 mm/slub.c:2423
> > allocate_slab+0x5f/0x2b0 mm/slub.c:2589
> > new_slab mm/slub.c:2642 [inline]
> > ___slab_alloc+0xbdf/0x1490 mm/slub.c:3830
> > __slab_alloc mm/slub.c:3920 [inline]
> > __slab_alloc_node mm/slub.c:3995 [inline]
> > slab_alloc_node mm/slub.c:4156 [inline]
> > __do_kmalloc_node mm/slub.c:4297 [inline]
> > __kmalloc_node_track_caller_noprof+0x30f/0x520 mm/slub.c:4317
> > kmemdup_noprof+0x2b/0x60 mm/util.c:135
> > __addrconf_sysctl_register+0xb1/0x430 net/ipv6/addrconf.c:7221
> > addrconf_sysctl_register+0x1bd/0x220 net/ipv6/addrconf.c:7287
> > ipv6_add_dev+0xe13/0x13e0 net/ipv6/addrconf.c:456
> > addrconf_notify+0x6d8/0x1170 net/ipv6/addrconf.c:3674
> > notifier_call_chain+0x1c6/0x410 kernel/notifier.c:85
> > call_netdevice_notifiers_extack net/core/dev.c:2034 [inline]
> > call_netdevice_notifiers+0xd3/0x110 net/core/dev.c:2048
> > register_netdevice+0x190c/0x1da0 net/core/dev.c:10632
> > page last free pid 100 tgid 100 stack trace:
> > reset_page_owner include/linux/page_owner.h:25 [inline]
> > free_pages_prepare mm/page_alloc.c:1127 [inline]
> > free_unref_folios+0xe03/0x1860 mm/page_alloc.c:2706
> > shrink_folio_list+0x4698/0x5c80 mm/vmscan.c:1483
> > evict_folios+0x3b12/0x5610 mm/vmscan.c:4593
> > try_to_shrink_lruvec+0x941/0xc10 mm/vmscan.c:4789
> > shrink_one+0x20e/0x870 mm/vmscan.c:4834
> > shrink_many mm/vmscan.c:4897 [inline]
> > lru_gen_shrink_node mm/vmscan.c:4975 [inline]
> > shrink_node+0x3862/0x3f20 mm/vmscan.c:5956
> > kswapd_shrink_node mm/vmscan.c:6785 [inline]
> > balance_pgdat mm/vmscan.c:6977 [inline]
> > kswapd+0x1c9f/0x36f0 mm/vmscan.c:7246
> > kthread+0x2c3/0x360 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
> >
> >
> > Memory state around the buggy address:
> >
> >
> > ffff88805cfc9f80: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
> >
> > ffff88805cfca000: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> > >ffff88805cfca080: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> > ^
> > ffff88805cfca100: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> > ffff88805cfca180: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> > ==================================================================
> >
> > Hi developers:
> >
> >
> > We are reporting a Linux issue using a modified version of Syzkaller.
> >
> >
> > HEAD commit: 4bbf9020 6.13.0-rc4
> >
> > git tree: upstream
> > kernel config: https://github.com/google/syzkaller/blob/master/dashboard/config/linux/upstream-apparmor-kasan.config
> > syz repro: https://drive.google.com/file/d/1XV9WHwMvK4eZIPuyjiWoTpyQ0FpDaUpT/view?usp=sharing
> > C reproducer: As this is a race condition, we cannot obtain a stable C reproducer.
> >
> >
> > Description:
> >
> > The SLIP driver suffers from a concurrent Use-After-Free issue in sl_sync() due to
> > unsynchronized access to net_device pointers stored in slip_devs[]. When a
> > device is being freed (for example, via tty_release()/netdev_run_todo()),
> > sl_sync() may access it after it has been freed.
> >
> > Steps to reproducer:
> >
> > 1. Setup a vm on the given kernel.
> > 2. Use syzkaller to run syz repro with `./syz-execprog -executor=./syz-executor -repeat=20000 -procs=8 -cover=0 ~/repro.prog`.
> >
> >
> > Environment:
> >
> > Ubuntu 22.04 on Linux 5.15
> > QEMU emulator version 6.2.0
> > qemu-system-x86_64 \
> > -m 2G \
> > -smp 2 \
> > -kernel /home/wd/bzImage \
> > -append "console=ttyS0 root=/dev/sda earlyprintk=serial net.ifnames=0" \
> > -drive file=/home/wd/bullseye.img,format=raw \
> > -net user,host=10.0.2.10,hostfwd=tcp:127.0.0.1:10021-:22 \
> > -net nic,model=e1000 \
> > -enable-kvm \
> > -nographic \
> > -pidfile vm.pid \
> > 2>&1 | tee vm.log
> >
> > If you fix this issue, please add the following tag to the commit:
> >
> > Reported-by: Jiacheng Xu <stitch@....edu.cn>
> >
Powered by blists - more mailing lists