Re: [REVIEW][PATCH 0/3] Fixing ptrace vs exec vs userns interactions

[Eric W. Biederman]

Willy Tarreau <w@xxxxxx> writes:

> On Sat, Nov 19, 2016 at 08:17:00AM +0100, Willy Tarreau wrote:
>> Hi Eric,
>> 
>> On Thu, Nov 17, 2016 at 11:02:47AM -0600, Eric W. Biederman wrote:
>> > 
>> > With everyone heading to Kernel Summit and Plumbers I put this set of
>> > patches down temporarily.   Now is the time to take it back up and to
>> > make certain I am not missing something stupid in this set of patches.
>> 
>> I couldn't get your patch set to apply to any of the kernels I tried,
>> I manually adjusted some parts but the second one has too many rejects.
>> What kernel should I apply this to ? Or maybe some preliminary patches
>> are needed ?
>
> OK I finally managed to get it to work on top of 4.8.9 (required less changes
> than master). I also had to drop the user_ns changes since there's no such
> user_ns in mm_struct there.
>
> I could run a test on it, that looks reasonable :
>
> FS:
>
> admin@vm:~$ strace -e trace=fstat,uname,ioctl,open uname
> open(0x7ffd01bbeeb0, O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> open(0x7ffd01bbeeb0, O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> open(0x7ffd01bbeeb0, O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> open(0x7ffd01bbeeb0, O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> open(0x7f3f9a1663e3, O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> open(0x7ffd01bbeeb0, O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> open(0x7ffd01bbeeb0, O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> open(0x7ffd01bbeeb0, O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> open(0x7ffd01bbeeb0, O_RDONLY|O_CLOEXEC) = 3
> fstat(3, {...})                         = 0
> open(0x7ffd01bbee80, O_RDONLY|O_CLOEXEC) = 3
> fstat(3, {...})                         = 0
> uname({...})                            = 0
> fstat(1, {...})                         = 0
> ioctl(1, SNDCTL_TMR_TIMEBASE or TCGETS, 0x7ffd01bbf400) = 0
>
> admin@vm:~$ sudo strace -e trace=fstat,uname,ioctl,open uname
> open("/etc/ld.so.cache", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> open("/lib64/tls/x86_64/libpthread.so.0", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> open("/lib64/tls/libpthread.so.0", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> open("/lib64/x86_64/libpthread.so.0", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> open("/lib64/libpthread.so.0", O_RDONLY|O_CLOEXEC) = 3
> fstat(3, {st_mode=S_IFREG|0555, st_size=101312, ...}) = 0
> open("/lib64/libc.so.6", O_RDONLY|O_CLOEXEC) = 3
> fstat(3, {st_mode=S_IFREG|0555, st_size=1479016, ...}) = 0
> uname({sys="Linux", node="vm", ...})    = 0
> fstat(1, {st_mode=S_IFCHR|0620, st_rdev=makedev(4, 64), ...}) = 0
> ioctl(1, SNDCTL_TMR_TIMEBASE or TCGETS, {B9600 opost isig icanon echo ...}) = 0
>
> Network:
>
> admin@vm:~$ strace -e trace=socket,setsockopt,connect /tmp/nc 198.18.3 22
> socket(PF_FILE, SOCK_STREAM|SOCK_CLOEXEC|SOCK_NONBLOCK, 0) = 3
> connect(3, {...}, 110)                  = -1 ENOENT (No such file or directory)
> socket(PF_FILE, SOCK_STREAM|SOCK_CLOEXEC|SOCK_NONBLOCK, 0) = 3
> connect(3, {...}, 110)                  = -1 ENOENT (No such file or directory)
> socket(PF_INET, SOCK_STREAM, IPPROTO_TCP) = 3
> setsockopt(3, SOL_SOCKET, SO_REUSEADDR, 0x7ffd2c26bdbc, 4) = 0
> connect(3, {...}, 16)                   = 0
>
> admin@vm:~$ sudo strace -e trace=socket,setsockopt,connect /tmp/nc 198.18.3 22
> socket(PF_FILE, SOCK_STREAM|SOCK_CLOEXEC|SOCK_NONBLOCK, 0) = 3
> connect(3, {sa_family=AF_FILE, path="/var/run/nscd/socket"}, 110) = -1 ENOENT (No such file or directory)
> socket(PF_FILE, SOCK_STREAM|SOCK_CLOEXEC|SOCK_NONBLOCK, 0) = 3
> connect(3, {sa_family=AF_FILE, path="/var/run/nscd/socket"}, 110) = -1 ENOENT (No such file or directory)
> socket(PF_INET, SOCK_STREAM, IPPROTO_TCP) = 3
> setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
> connect(3, {sa_family=AF_INET, sin_port=htons(22), sin_addr=inet_addr("198.18.0.3")}, 16) = 0
>
> So in short now we can at least see what syscall fails eventhough we can't
> know why. I think it can be an acceptable trade-off.

Thanks for testing, and thanks for you acceptance even if I didn't make
it easy for you.


Eric