Skip to content

K
Kernel

openeuler / Kernel
接近 2 年 前同步成功

通知 8
Star 0
Fork 0
K
Kernel
切换分支/标签
  1. 26 11月, 2019 1 次提交
    • L
      vfs: mark pipes and sockets as stream-like file descriptors · d8e464ec
      由 Linus Torvalds 提交于 
      In commit 3975b097 ("convert stream-like files -> stream_open, even
if they use noop_llseek") Kirill used a coccinelle script to change
"nonseekable_open()" to "stream_open()", which changed the trivial cases
of stream-like file descriptors to the new model with FMODE_STREAM.

However, the two big cases - sockets and pipes - don't actually have
that trivial pattern at all, and were thus never converted to
FMODE_STREAM even though it makes lots of sense to do so.

That's particularly true when looking forward to the next change:
getting rid of FMODE_ATOMIC_POS entirely, and just using FMODE_STREAM to
decide whether f_pos updates are needed or not.  And if they are, we'll
always do them atomically.

This came up because KCSAN (correctly) noted that the non-locked f_pos
updates are data races: they are clearly benign for the case where we
don't care, but it would be good to just not have that issue exist at
all.

Note that the reason we used FMODE_ATOMIC_POS originally is that only
doing it for the minimal required case is "safer" in that it's possible
that the f_pos locking can cause unnecessary serialization across the
whole write() call.  And in the worst case, that kind of serialization
can cause deadlock issues: think writers that need readers to empty the
state using the same file descriptor.

[ Note that the locking is per-file descriptor - because it protects
  "f_pos", which is obviously per-file descriptor - so it only affects
  cases where you literally use the same file descriptor to both read
  and write.

  So a regular pipe that has separate reading and writing file
  descriptors doesn't really have this situation even though it's the
  obvious case of "reader empties what a bit writer concurrently fills"

  But we want to make pipes as being stream-line anyway, because we
  don't want the unnecessary overhead of locking, and because a named
  pipe can be (ab-)used by reading and writing to the same file
  descriptor. ]

There are likely a lot of other cases that might want FMODE_STREAM, and
looking for ".llseek = no_llseek" users and other cases that don't have
an lseek file operation at all and making them use "stream_open()" might
be a good idea.  But pipes and sockets are likely to be the two main
cases.

Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Eic Dumazet <edumazet@google.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Marco Elver <elver@google.com>
Cc: Andrea Parri <parri.andrea@gmail.com>
Cc: Paul McKenney <paulmck@kernel.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      d8e464ec
  3. 10 7月, 2019 2 次提交
  5. 09 7月, 2019 2 次提交
    • A
      coallocate socket_wq with socket itself · 333f7909
      由 Al Viro 提交于 
      socket->wq is assign-once, set when we are initializing both
struct socket it's in and struct socket_wq it points to.  As the
matter of fact, the only reason for separate allocation was the
ability to RCU-delay freeing of socket_wq.  RCU-delaying the
freeing of socket itself gets rid of that need, so we can just
fold struct socket_wq into the end of struct socket and simplify
the life both for sock_alloc_inode() (one allocation instead of
two) and for tun/tap oddballs, where we used to embed struct socket
and struct socket_wq into the same structure (now - embedding just
the struct socket).

Note that reference to struct socket_wq in struct sock does remain
a reference - that's unchanged.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      333f7909
    • A
      sockfs: switch to ->free_inode() · 6d7855c5
      由 Al Viro 提交于 
      we do have an RCU-delayed part there already (freeing the wq),
so it's not like the pipe situation; moreover, it might be
worth considering coallocating wq with the rest of struct sock_alloc.
->sk_wq in struct sock would remain a pointer as it is, but
the object it normally points to would be coallocated with
struct socket...
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      6d7855c5
  7. 04 7月, 2019 1 次提交
  9. 28 6月, 2019 1 次提交
    • S
      bpf: implement getsockopt and setsockopt hooks · 0d01da6a
      由 Stanislav Fomichev 提交于 
      Implement new BPF_PROG_TYPE_CGROUP_SOCKOPT program type and
BPF_CGROUP_{G,S}ETSOCKOPT cgroup hooks.

BPF_CGROUP_SETSOCKOPT can modify user setsockopt arguments before
passing them down to the kernel or bypass kernel completely.
BPF_CGROUP_GETSOCKOPT can can inspect/modify getsockopt arguments that
kernel returns.
Both hooks reuse existing PTR_TO_PACKET{,_END} infrastructure.

The buffer memory is pre-allocated (because I don't think there is
a precedent for working with __user memory from bpf). This might be
slow to do for each {s,g}etsockopt call, that's why I've added
__cgroup_bpf_prog_array_is_empty that exits early if there is nothing
attached to a cgroup. Note, however, that there is a race between
__cgroup_bpf_prog_array_is_empty and BPF_PROG_RUN_ARRAY where cgroup
program layout might have changed; this should not be a problem
because in general there is a race between multiple calls to
{s,g}etsocktop and user adding/removing bpf progs from a cgroup.

The return code of the BPF program is handled as follows:
* 0: EPERM
* 1: success, continue with next BPF program in the cgroup chain

v9:
* allow overwriting setsockopt arguments (Alexei Starovoitov):
  * use set_fs (same as kernel_setsockopt)
  * buffer is always kzalloc'd (no small on-stack buffer)

v8:
* use s32 for optlen (Andrii Nakryiko)

v7:
* return only 0 or 1 (Alexei Starovoitov)
* always run all progs (Alexei Starovoitov)
* use optval=0 as kernel bypass in setsockopt (Alexei Starovoitov)
  (decided to use optval=-1 instead, optval=0 might be a valid input)
* call getsockopt hook after kernel handlers (Alexei Starovoitov)

v6:
* rework cgroup chaining; stop as soon as bpf program returns
  0 or 2; see patch with the documentation for the details
* drop Andrii's and Martin's Acked-by (not sure they are comfortable
  with the new state of things)

v5:
* skip copy_to_user() and put_user() when ret == 0 (Martin Lau)

v4:
* don't export bpf_sk_fullsock helper (Martin Lau)
* size != sizeof(__u64) for uapi pointers (Martin Lau)
* offsetof instead of bpf_ctx_range when checking ctx access (Martin Lau)

v3:
* typos in BPF_PROG_CGROUP_SOCKOPT_RUN_ARRAY comments (Andrii Nakryiko)
* reverse christmas tree in BPF_PROG_CGROUP_SOCKOPT_RUN_ARRAY (Andrii
  Nakryiko)
* use __bpf_md_ptr instead of __u32 for optval{,_end} (Martin Lau)
* use BPF_FIELD_SIZEOF() for consistency (Martin Lau)
* new CG_SOCKOPT_ACCESS macro to wrap repeated parts

v2:
* moved bpf_sockopt_kern fields around to remove a hole (Martin Lau)
* aligned bpf_sockopt_kern->buf to 8 bytes (Martin Lau)
* bpf_prog_array_is_empty instead of bpf_prog_array_length (Martin Lau)
* added [0,2] return code check to verifier (Martin Lau)
* dropped unused buf[64] from the stack (Martin Lau)
* use PTR_TO_SOCKET for bpf_sockopt->sk (Martin Lau)
* dropped bpf_target_off from ctx rewrites (Martin Lau)
* use return code for kernel bypass (Martin Lau & Andrii Nakryiko)

Cc: Andrii Nakryiko <andriin@fb.com>
Cc: Martin Lau <kafai@fb.com>
Signed-off-by: NStanislav Fomichev <sdf@google.com>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
      0d01da6a
  11. 06 6月, 2019 1 次提交
  13. 01 6月, 2019 1 次提交
  15. 31 5月, 2019 1 次提交
  17. 26 5月, 2019 2 次提交
    • D
      vfs: Convert sockfs to use the new mount API · fba9be49
      由 David Howells 提交于 
      Convert the sockfs filesystem to the new internal mount API as the old
one will be obsoleted and removed.  This allows greater flexibility in
communication of mount parameters between userspace, the VFS and the
filesystem.

See Documentation/filesystems/mount_api.txt for more information.
Signed-off-by: NDavid Howells <dhowells@redhat.com>
cc: netdev@vger.kernel.org
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>
      fba9be49
    • A
      mount_pseudo(): drop 'name' argument, switch to d_make_root() · 1f58bb18
      由 Al Viro 提交于 
      Once upon a time we used to set ->d_name of e.g. pipefs root
so that d_path() on pipes would work.  These days it's
completely pointless - dentries of pipes are not even connected
to pipefs root.  However, mount_pseudo() had set the root
dentry name (passed as the second argument) and callers
kept inventing names to pass to it.  Including those that
didn't *have* any non-root dentries to start with...

All of that had been pointless for about 8 years now; it's
time to get rid of that cargo-culting...
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>
      1f58bb18
  19. 20 5月, 2019 1 次提交
    • R
      net: fix kernel-doc warnings for socket.c · 85806af0
      由 Randy Dunlap 提交于 
      Fix kernel-doc warnings by moving the kernel-doc notation to be
immediately above the functions that it describes.

Fixes these warnings for sock_sendmsg() and sock_recvmsg():

../net/socket.c:658: warning: Excess function parameter 'sock' description in 'INDIRECT_CALLABLE_DECLARE'
../net/socket.c:658: warning: Excess function parameter 'msg' description in 'INDIRECT_CALLABLE_DECLARE'
../net/socket.c:889: warning: Excess function parameter 'sock' description in 'INDIRECT_CALLABLE_DECLARE'
../net/socket.c:889: warning: Excess function parameter 'msg' description in 'INDIRECT_CALLABLE_DECLARE'
../net/socket.c:889: warning: Excess function parameter 'flags' description in 'INDIRECT_CALLABLE_DECLARE'
Signed-off-by: NRandy Dunlap <rdunlap@infradead.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      85806af0
  21. 06 5月, 2019 1 次提交
  23. 26 4月, 2019 1 次提交
  25. 20 4月, 2019 2 次提交
    • A
      net: socket: implement 64-bit timestamps · 0768e170
      由 Arnd Bergmann 提交于 
      The 'timeval' and 'timespec' data structures used for socket timestamps
are going to be redefined in user space based on 64-bit time_t in future
versions of the C library to deal with the y2038 overflow problem,
which breaks the ABI definition.

Unlike many modern ioctl commands, SIOCGSTAMP and SIOCGSTAMPNS do not
use the _IOR() macro to encode the size of the transferred data, so it
remains ambiguous whether the application uses the old or new layout.

The best workaround I could find is rather ugly: we redefine the command
code based on the size of the respective data structure with a ternary
operator. This lets it get evaluated as late as possible, hopefully after
that structure is visible to the caller. We cannot use an #ifdef here,
because inux/sockios.h might have been included before any libc header
that could determine the size of time_t.

The ioctl implementation now interprets the new command codes as always
referring to the 64-bit structure on all architectures, while the old
architecture specific command code still refers to the old architecture
specific layout. The new command number is only used when they are
actually different.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      0768e170
    • A
      net: rework SIOCGSTAMP ioctl handling · c7cbdbf2
      由 Arnd Bergmann 提交于 
      The SIOCGSTAMP/SIOCGSTAMPNS ioctl commands are implemented by many
socket protocol handlers, and all of those end up calling the same
sock_get_timestamp()/sock_get_timestampns() helper functions, which
results in a lot of duplicate code.

With the introduction of 64-bit time_t on 32-bit architectures, this
gets worse, as we then need four different ioctl commands in each
socket protocol implementation.

To simplify that, let's add a new .gettstamp() operation in
struct proto_ops, and move ioctl implementation into the common
sock_ioctl()/compat_sock_ioctl_trans() functions that these all go
through.

We can reuse the sock_get_timestamp() implementation, but generalize
it so it can deal with both native and compat mode, as well as
timeval and timespec structures.
Acked-by: NStefan Schmidt <stefan@datenfreihafen.org>
Acked-by: NNeil Horman <nhorman@tuxdriver.com>
Acked-by: NMarc Kleine-Budde <mkl@pengutronix.de>
Link: https://lore.kernel.org/lkml/CAK8P3a038aDQQotzua_QtKGhq8O9n+rdiz2=WDCp82ys8eUT+A@mail.gmail.com/Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Acked-by: NWillem de Bruijn <willemb@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      c7cbdbf2
  27. 16 3月, 2019 1 次提交
  29. 26 2月, 2019 1 次提交
    • E
      net: socket: set sock->sk to NULL after calling proto_ops::release() · ff7b11aa
      由 Eric Biggers 提交于 
      Commit 9060cb71 ("net: crypto set sk to NULL when af_alg_release.")
fixed a use-after-free in sockfs_setattr() when an AF_ALG socket is
closed concurrently with fchownat().  However, it ignored that many
other proto_ops::release() methods don't set sock->sk to NULL and
therefore allow the same use-after-free:

    - base_sock_release
    - bnep_sock_release
    - cmtp_sock_release
    - data_sock_release
    - dn_release
    - hci_sock_release
    - hidp_sock_release
    - iucv_sock_release
    - l2cap_sock_release
    - llcp_sock_release
    - llc_ui_release
    - rawsock_release
    - rfcomm_sock_release
    - sco_sock_release
    - svc_release
    - vcc_release
    - x25_release

Rather than fixing all these and relying on every socket type to get
this right forever, just make __sock_release() set sock->sk to NULL
itself after calling proto_ops::release().

Reproducer that produces the KASAN splat when any of these socket types
are configured into the kernel:

    #include <pthread.h>
    #include <stdlib.h>
    #include <sys/socket.h>
    #include <unistd.h>

    pthread_t t;
    volatile int fd;

    void *close_thread(void *arg)
    {
        for (;;) {
            usleep(rand() % 100);
            close(fd);
        }
    }

    int main()
    {
        pthread_create(&t, NULL, close_thread, NULL);
        for (;;) {
            fd = socket(rand() % 50, rand() % 11, 0);
            fchownat(fd, "", 1000, 1000, 0x1000);
            close(fd);
        }
    }

Fixes: 86741ec2 ("net: core: Add a UID field to struct sock.")
Signed-off-by: NEric Biggers <ebiggers@google.com>
Acked-by: NCong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      ff7b11aa
  31. 04 2月, 2019 4 次提交
  33. 31 1月, 2019 4 次提交
  35. 18 12月, 2018 1 次提交
    • A
      y2038: socket: Add compat_sys_recvmmsg_time64 · e11d4284
      由 Arnd Bergmann 提交于 
      recvmmsg() takes two arguments to pointers of structures that differ
between 32-bit and 64-bit architectures: mmsghdr and timespec.

For y2038 compatbility, we are changing the native system call from
timespec to __kernel_timespec with a 64-bit time_t (in another patch),
and use the existing compat system call on both 32-bit and 64-bit
architectures for compatibility with traditional 32-bit user space.

As we now have two variants of recvmmsg() for 32-bit tasks that are both
different from the variant that we use on 64-bit tasks, this means we
also require two compat system calls!

The solution I picked is to flip things around: The existing
compat_sys_recvmmsg() call gets moved from net/compat.c into net/socket.c
and now handles the case for old user space on all architectures that
have set CONFIG_COMPAT_32BIT_TIME.  A new compat_sys_recvmmsg_time64()
call gets added in the old place for 64-bit architectures only, this
one handles the case of a compat mmsghdr structure combined with
__kernel_timespec.

In the indirect sys_socketcall(), we now need to call either
do_sys_recvmmsg() or __compat_sys_recvmmsg(), depending on what kind of
architecture we are on. For compat_sys_socketcall(), no such change is
needed, we always call __compat_sys_recvmmsg().

I decided to not add a new SYS_RECVMMSG_TIME64 socketcall: Any libc
implementation for 64-bit time_t will need significant changes including
an updated asm/unistd.h, and it seems better to consistently use the
separate syscalls that configuration, leaving the socketcall only for
backward compatibility with 32-bit time_t based libc.

The naming is asymmetric for the moment, so both existing syscalls
entry points keep their names, while the new ones are recvmmsg_time32
and compat_recvmmsg_time64 respectively. I expect that we will rename
the compat syscalls later as we start using generated syscall tables
everywhere and add these entry points.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
      e11d4284
  37. 18 11月, 2018 1 次提交
  39. 24 10月, 2018 1 次提交
    • D
      iov_iter: Separate type from direction and use accessor functions · aa563d7b
      由 David Howells 提交于 
      In the iov_iter struct, separate the iterator type from the iterator
direction and use accessor functions to access them in most places.

Convert a bunch of places to use switch-statements to access them rather
then chains of bitwise-AND statements.  This makes it easier to add further
iterator types.  Also, this can be more efficient as to implement a switch
of small contiguous integers, the compiler can use ~50% fewer compare
instructions than it has to use bitwise-and instructions.

Further, cease passing the iterator type into the iterator setup function.
The iterator function can set that itself.  Only the direction is required.
Signed-off-by: NDavid Howells <dhowells@redhat.com>
      aa563d7b
  41. 19 10月, 2018 1 次提交
    • W
      net: socket: fix a missing-check bug · b6168562
      由 Wenwen Wang 提交于 
      In ethtool_ioctl(), the ioctl command 'ethcmd' is checked through a switch
statement to see whether it is necessary to pre-process the ethtool
structure, because, as mentioned in the comment, the structure
ethtool_rxnfc is defined with padding. If yes, a user-space buffer 'rxnfc'
is allocated through compat_alloc_user_space(). One thing to note here is
that, if 'ethcmd' is ETHTOOL_GRXCLSRLALL, the size of the buffer 'rxnfc' is
partially determined by 'rule_cnt', which is actually acquired from the
user-space buffer 'compat_rxnfc', i.e., 'compat_rxnfc->rule_cnt', through
get_user(). After 'rxnfc' is allocated, the data in the original user-space
buffer 'compat_rxnfc' is then copied to 'rxnfc' through copy_in_user(),
including the 'rule_cnt' field. However, after this copy, no check is
re-enforced on 'rxnfc->rule_cnt'. So it is possible that a malicious user
race to change the value in the 'compat_rxnfc->rule_cnt' between these two
copies. Through this way, the attacker can bypass the previous check on
'rule_cnt' and inject malicious data. This can cause undefined behavior of
the kernel and introduce potential security risk.

This patch avoids the above issue via copying the value acquired by
get_user() to 'rxnfc->rule_cn', if 'ethcmd' is ETHTOOL_GRXCLSRLALL.
Signed-off-by: NWenwen Wang <wang6495@umn.edu>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      b6168562
  43. 06 10月, 2018 1 次提交
  45. 14 9月, 2018 1 次提交
  47. 29 8月, 2018 1 次提交
    • A
      y2038: socket: Change recvmmsg to use __kernel_timespec · c2e6c856
      由 Arnd Bergmann 提交于 
      This converts the recvmmsg() system call in all its variations to use
'timespec64' internally for its timeout, and have a __kernel_timespec64
argument in the native entry point. This lets us change the type to use
64-bit time_t at a later point while using the 32-bit compat system call
emulation for existing user space.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
      c2e6c856
  49. 15 8月, 2018 1 次提交
  51. 01 8月, 2018 1 次提交
  53. 31 7月, 2018 2 次提交
  55. 29 7月, 2018 2 次提交