RE: [syzbot] KASAN: use-after-free Read in dev_uevent

["Zhang, Qiang1" <qiang1.zhang@intel.com>]

On Wed, Feb 23, 2022 at 06:51:10AM +0000, Zhang, Qiang1 wrote:
> 
> HEAD commit:    4f12b742eb2b Merge tag 'nfs-for-5.17-3' of git://git.linux..
> git tree:       upstream
> console output: 
> https://syzkaller.appspot.com/x/log.txt?x=110a6df2700000
> kernel config:  
> https://syzkaller.appspot.com/x/.config?x=f6a069ed94a1ed1d
> dashboard link: https://syzkaller.appspot.com/bug?extid=348b571beb5eeb70a582
> compiler:       gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.2
> syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=12377296700000
> 
> IMPORTANT: if you fix the issue, please add the following tag to the commit:
> Reported-by: syzbot+348b571beb5eeb70a582@syzkaller.appspotmail.com
> 
> ==================================================================
> BUG: KASAN: use-after-free in dev_uevent+0x712/0x780 
> drivers/base/core.c:2320 Read of size 8 at addr ffff88802b934098 by 
> task udevd/3689
> 
> CPU: 2 PID: 3689 Comm: udevd Not tainted 5.17.0-rc4-syzkaller-00229-g4f12b742eb2b #0 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-2 04/01/2014 Call Trace:
>  <TASK>
>  __dump_stack lib/dump_stack.c:88 [inline]
>  dump_stack_lvl+0xcd/0x134 lib/dump_stack.c:106
>  print_address_description.constprop.0.cold+0x8d/0x303 
> mm/kasan/report.c:255  __kasan_report mm/kasan/report.c:442 [inline]  
> kasan_report.cold+0x83/0xdf mm/kasan/report.c:459
>  dev_uevent+0x712/0x780 drivers/base/core.c:2320
>  uevent_show+0x1b8/0x380 drivers/base/core.c:2391
>  dev_attr_show+0x4b/0x90 drivers/base/core.c:2094
>  sysfs_kf_seq_show+0x219/0x3d0 fs/sysfs/file.c:59
>  seq_read_iter+0x4f5/0x1280 fs/seq_file.c:230
>  kernfs_fop_read_iter+0x514/0x6f0 fs/kernfs/file.c:241  call_read_iter 
> include/linux/fs.h:2068 [inline]
>  new_sync_read+0x429/0x6e0 fs/read_write.c:400
>  vfs_read+0x35c/0x600 fs/read_write.c:481
>  ksys_read+0x12d/0x250 fs/read_write.c:619
>  do_syscall_x64 arch/x86/entry/common.c:50 [inline]
>  do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80  
> entry_SYSCALL_64_after_hwframe+0x44/0xae
> RIP: 0033:0x7f964cc558fe
> Code: c0 e9 e6 fe ff ff 50 48 8d 3d 0e c7 09 00 e8 c9 cf 01 00 66 0f 
> 1f 84 00 00 00 00 00 64 8b 04 25 18 00 00 00 85 c0 75 14 0f 05 <48> 3d 
> 00 f0 ff ff 77 5a c3 66 0f 1f 84 00 00 00 00 00 48 83 ec 28
> RSP: 002b:00007ffc0133d258 EFLAGS: 00000246 ORIG_RAX: 0000000000000000
> RAX: ffffffffffffffda RBX: 000056497b21a140 RCX: 00007f964cc558fe
> RDX: 0000000000001000 RSI: 000056497b218650 RDI: 0000000000000008
> RBP: 00007f964cd22380 R08: 0000000000000008 R09: 00007f964cd25a60
> R10: 0000000000000008 R11: 0000000000000246 R12: 000056497b21a140
> R13: 0000000000000d68 R14: 00007f964cd21780 R15: 0000000000000d68  
> </TASK>
> 
> 
> Hi All
> 
> This should be because when the raw_dev is released, the 'driver' 
> address has expired, although the usb_gadget_remove_driver() empty 'dev.driver ' NULL, but UAF cannot be avoided.
> 
> static int dev_uevent(struct kobject *kobj, struct kobj_uevent_env 
> *env) { .....
>          if (dev->driver)
> 2320                 add_uevent_var(env, "DRIVER=%s", dev->driver->name);
> .....
> }
> 
> Whether protection can be added when operating 'dev->driver'?
>
>When the driver structure is unbound, this should be set to NULL.  Why isn't that happening?

Yes, it should be set NULL at:

close
raw_release
   ->usb_gadget_unregister_driver
      	-> usb_gadget_remove_driver
         	    {
                  ......
                   udc->driver = NULL;
                   udc->dev.driver = NULL;
                   udc->gadget->dev.driver = NULL;
           	    }

kref_put(&dev->count, dev_free)
   -> dev_free
         kfree(dev)   <------------------------------------------ raw_dev be freed


if the udev process not meet it. The UAF maybe can trigger.
Did I miss something?

thanks,
Zqiang

>
>thanks,
>
>greg k-h