The poll() changes were not well thought out, and completely
unexplained. They also caused a huge performance regression, because
"->poll()" was no longer a trivial file operation that just called down
to the underlying file operations, but instead did at least two indirect
calls.
Indirect calls are sadly slow now with the Spectre mitigation, but the
performance problem could at least be largely mitigated by changing the
"->get_poll_head()" operation to just have a per-file-descriptor pointer
to the poll head instead. That gets rid of one of the new indirections.
But that doesn't fix the new complexity that is completely unwarranted
for the regular case. The (undocumented) reason for the poll() changes
was some alleged AIO poll race fixing, but we don't make the common case
slower and more complex for some uncommon special case, so this all
really needs way more explanations and most likely a fundamental
redesign.
[ This revert is a revert of about 30 different commits, not reverted
individually because that would just be unnecessarily messy - Linus ]
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Changes since v1:
Added changes in these files:
drivers/infiniband/hw/usnic/usnic_transport.c
drivers/staging/lustre/lnet/lnet/lib-socket.c
drivers/target/iscsi/iscsi_target_login.c
drivers/vhost/net.c
fs/dlm/lowcomms.c
fs/ocfs2/cluster/tcp.c
security/tomoyo/network.c
Before:
All these functions either return a negative error indicator,
or store length of sockaddr into "int *socklen" parameter
and return zero on success.
"int *socklen" parameter is awkward. For example, if caller does not
care, it still needs to provide on-stack storage for the value
it does not need.
None of the many FOO_getname() functions of various protocols
ever used old value of *socklen. They always just overwrite it.
This change drops this parameter, and makes all these functions, on success,
return length of sockaddr. It's always >= 0 and can be differentiated
from an error.
Tests in callers are changed from "if (err)" to "if (err < 0)", where needed.
rpc_sockname() lost "int buflen" parameter, since its only use was
to be passed to kernel_getsockname() as &buflen and subsequently
not used in any way.
Userspace API is not changed.
text data bss dec hex filename
30108430 2633624 873672 33615726 200ef6e vmlinux.before.o
30108109 2633612 873672 33615393 200ee21 vmlinux.o
Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com>
CC: David S. Miller <davem@davemloft.net>
CC: linux-kernel@vger.kernel.org
CC: netdev@vger.kernel.org
CC: linux-bluetooth@vger.kernel.org
CC: linux-decnet-user@lists.sourceforge.net
CC: linux-wireless@vger.kernel.org
CC: linux-rdma@vger.kernel.org
CC: linux-sctp@vger.kernel.org
CC: linux-nfs@vger.kernel.org
CC: linux-x25@vger.kernel.org
Signed-off-by: David S. Miller <davem@davemloft.net>
Mart reported a deadlock in -RT in the call path:
hci_send_monitor_ctrl_event() -> hci_send_to_channel()
because both functions acquire the same read lock hci_sk_list.lock. This
is also a mainline issue because the qrwlock implementation is writer
fair (the traditional rwlock implementation is reader biased).
To avoid the deadlock there is now __hci_send_to_channel() which expects
the readlock to be held.
Fixes: 38ceaa00d0 ("Bluetooth: Add support for sending MGMT commands and events to monitor")
Reported-by: Mart van de Wege <mvdwege@gmail.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
This reverts commit dbbccdc4ce.
It turns out that the "legacy" users aren't so legacy at all, and that
turning off the legacy ioctl will break the current Qt bluetooth stack
for bluetooth LE devices that were released just a couple of months ago.
So it's simply not true that this was a legacy interface that hasn't
been needed and is only limited to old legacy BT devices. Because I
actually read Kconfig help messages, and actively try to turn off
features that I don't need, I turned the option off.
Then I spent _way_ too much time debugging BLE issues until I realized
that it wasn't the Qt and subsurface development that had broken one of
my dive computer BLE downloads, but simply my broken kernel config.
Maybe in a decade it will be true that this is a legacy interface. And
maybe with a better help-text and correct dependencies, this kind of
legacy removal might be acceptable. But as things are right now both
the commit message and the Kconfig help text were misleading, and the
Kconfig option had the wrong dependenencies.
There's no reason to keep that broken Kconfig option in the tree.
Cc: Marcel Holtmann <marcel@holtmann.org>
Cc: Johan Hedberg <johan.hedberg@intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The legacy ioctl interfaces are only useful for BR/EDR operation and
since Linux 3.4 no longer needed anyway. This options allows disabling
them alltogether and use only management interfaces for setup and
control.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Joe and Bjørn suggested that it'd be nicer to not have the
cast in the fairly common case of doing
*(u8 *)skb_put(skb, 1) = c;
Add skb_put_u8() for this case, and use it across the code,
using the following spatch:
@@
expression SKB, C, S;
typedef u8;
identifier fn = {skb_put};
fresh identifier fn2 = fn ## "_u8";
@@
- *(u8 *)fn(SKB, S) = C;
+ fn2(SKB, C);
Note that due to the "S", the spatch isn't perfect, it should
have checked that S is 1, but there's also places that use a
sizeof expression like sizeof(var) or sizeof(u8) etc. Turns
out that nobody ever did something like
*(u8 *)skb_put(skb, 2) = c;
which would be wrong anyway since the second byte wouldn't be
initialized.
Suggested-by: Joe Perches <joe@perches.com>
Suggested-by: Bjørn Mork <bjorn@mork.no>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
It seems like a historic accident that these return unsigned char *,
and in many places that means casts are required, more often than not.
Make these functions return void * and remove all the casts across
the tree, adding a (u8 *) cast only where the unsigned char pointer
was used directly, all done with the following spatch:
@@
expression SKB, LEN;
typedef u8;
identifier fn = { skb_push, __skb_push, skb_push_rcsum };
@@
- *(fn(SKB, LEN))
+ *(u8 *)fn(SKB, LEN)
@@
expression E, SKB, LEN;
identifier fn = { skb_push, __skb_push, skb_push_rcsum };
type T;
@@
- E = ((T *)(fn(SKB, LEN)))
+ E = fn(SKB, LEN)
@@
expression SKB, LEN;
identifier fn = { skb_push, __skb_push, skb_push_rcsum };
@@
- fn(SKB, LEN)[0]
+ *(u8 *)fn(SKB, LEN)
Note that the last part there converts from push(...)[0] to the
more idiomatic *(u8 *)push(...).
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
It seems like a historic accident that these return unsigned char *,
and in many places that means casts are required, more often than not.
Make these functions (skb_put, __skb_put and pskb_put) return void *
and remove all the casts across the tree, adding a (u8 *) cast only
where the unsigned char pointer was used directly, all done with the
following spatch:
@@
expression SKB, LEN;
typedef u8;
identifier fn = { skb_put, __skb_put };
@@
- *(fn(SKB, LEN))
+ *(u8 *)fn(SKB, LEN)
@@
expression E, SKB, LEN;
identifier fn = { skb_put, __skb_put };
type T;
@@
- E = ((T *)(fn(SKB, LEN)))
+ E = fn(SKB, LEN)
which actually doesn't cover pskb_put since there are only three
users overall.
A handful of stragglers were converted manually, notably a macro in
drivers/isdn/i4l/isdn_bsdcomp.c and, oddly enough, one of the many
instances in net/bluetooth/hci_sock.c. In the former file, I also
had to fix one whitespace problem spatch introduced.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
A common pattern with skb_put() is to just want to memcpy()
some data into the new space, introduce skb_put_data() for
this.
An spatch similar to the one for skb_put_zero() converts many
of the places using it:
@@
identifier p, p2;
expression len, skb, data;
type t, t2;
@@
(
-p = skb_put(skb, len);
+p = skb_put_data(skb, data, len);
|
-p = (t)skb_put(skb, len);
+p = skb_put_data(skb, data, len);
)
(
p2 = (t2)p;
-memcpy(p2, data, len);
|
-memcpy(p, data, len);
)
@@
type t, t2;
identifier p, p2;
expression skb, data;
@@
t *p;
...
(
-p = skb_put(skb, sizeof(t));
+p = skb_put_data(skb, data, sizeof(t));
|
-p = (t *)skb_put(skb, sizeof(t));
+p = skb_put_data(skb, data, sizeof(t));
)
(
p2 = (t2)p;
-memcpy(p2, data, sizeof(*p));
|
-memcpy(p, data, sizeof(*p));
)
@@
expression skb, len, data;
@@
-memcpy(skb_put(skb, len), data, len);
+skb_put_data(skb, data, len);
(again, manually post-processed to retain some comments)
Reviewed-by: Stephen Hemminger <stephen@networkplumber.org>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Running 32bit userspace on 64bit kernel results in MSG_CMSG_COMPAT being
defined as 0x80000000. This results in sendmsg failure if used from 32bit
userspace running on 64bit kernel. Fix this by accounting for MSG_CMSG_COMPAT
in flags check in hci_sock_sendmsg.
Signed-off-by: Szymon Janc <szymon.janc@codecoup.pl>
Signed-off-by: Marko Kiiskila <marko@runtime.io>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Cc: stable@vger.kernel.org
Fix typos and add the following to the scripts/spelling.txt:
an user||a user
an userspace||a userspace
I also added "userspace" to the list since it is a common word in Linux.
I found some instances for "an userfaultfd", but I did not add it to the
list. I felt it is endless to find words that start with "user" such as
"userland" etc., so must draw a line somewhere.
Link: http://lkml.kernel.org/r/1481573103-11329-4-git-send-email-yamada.masahiro@socionext.com
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When opening and closing HCI user channel, send monitoring messages to
be able to trace its behavior.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
In case an unbound HCI raw socket is later on bound, ensure that the
monitor notification messages indicate a close and re-open. None of
the userspace tools use the socket this, but it is actually possible
to use an ioctl on an unbound socket and then later bind it.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
When opening and closing HCI raw sockets their main usage is for legacy
userspace. To track interaction with the modern mgmt interface, send
open and close monitoring messages for these action.
The HCI raw sockets is special since it supports unbound ioctl operation
and for that special case delay the notification message until at least
one ioctl has been executed. The difference between a bound and unbound
socket will be detailed by the fact the HCI index is present or not.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
The control open and close monitoring events require special channel
checks to ensure messages are only send when the right events happen.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Assignment of the hci_pi(sk)->channel should be done early when binding
the HCI socket. This avoids confusion with the RAW channel that is used
for legacy access.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Only when the cookie has been assigned, then send the open and close
monitor messages. Also if the socket is bound to a device, then include
the index into the message.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Instead of keeping a version string around, use version and revision
numbers and then stringify them for use as module parameter.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Instead of manually allocating cookie information each time, use helper
functions for generating and releasing cookies.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Instead of hiding everything behind a general managment events flag,
introduce indivdual flags that allow fine control over which events are
send to a given management channel.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
This adds support for tracing all management commands and events via the
monitor interface.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
This sends new notifications to the monitor support whenever a
management channel has been opened or closed. This allows tracing of
control channels really easily.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
To further allow unique identification and tracking of control socket,
store cookie and comm information when binding the socket.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
The SOL_HCI level should be enforced when using socket options on the
HCI raw socket interface.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Similar to bt_sock_recvmsg MSG_TRUNC shall be checked using the original
flags not msg_flags.
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
The HCI_BREDR naming is confusing since it actually stands for Primary
Bluetooth Controller. Which is a term that has been used in the latest
standard. However from a legacy point of view there only really have
been Basic Rate (BR) and Enhanced Data Rate (EDR). Recent versions of
Bluetooth introduced Low Energy (LE) and made this terminology a little
bit confused since Dual Mode Controllers include BR/EDR and LE. To
simplify this the name HCI_PRIMARY stands for the Primary Controller
which can be a single mode or dual mode controller.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
If recvmsg is called with a destination buffer that is too small to
receive the contents of skb in its entirety, the return value from
recvmsg was inconsistent with common SOCK_SEQPACKET or SOCK_DGRAM
semantics.
If destination buffer provided by userspace is too small (e.g. len <
copied), then MSG_TRUNC flag is set and copied is returned. Instead, it
should return the length of the message, which is consistent with how
other datagram based sockets act. Quoting 'man recv':
"All three calls return the length of the message on successful comple‐
tion. If a message is too long to fit in the supplied buffer, excess
bytes may be discarded depending on the type of socket the message is
received from."
and
"MSG_TRUNC (since Linux 2.2)
For raw (AF_PACKET), Internet datagram (since Linux
2.4.27/2.6.8), netlink (since Linux 2.6.22), and UNIX datagram
(since Linux 3.4) sockets: return the real length of the packet
or datagram, even when it was longer than the passed buffer."
Signed-off-by: Denis Kenzior <denkenz@gmail.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Instead, allow using string formatting with send_monitor_note()
and access init_utsname().
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
To enable controller specific logging, the userspace daemon has to have
the ability to log per controller. To facilitate this support, provide
a dedicated logging channel. Messages in this channel will be included
in the monitor queue and with that also forwarded to monitoring tools
along with the actual hardware traces.
All messages from the logging channel are timestamped and with that
allow an easy correlation between userspace messages and hardware
events. This will increase the ability to debug problems faster.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
The monitor channel can be used to send generic system notes as text
strings for debugging purposes. This adds the system note monitor code
and uses it for including kernel and subsystem version into traces.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
The HCI sockets code has still some old casting coding style. Fix this
to match with the rest of the code.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
When HCI commands are injected via the raw socket, the core was not
including the decoded opcode value. So ensure that it is actually set.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
We can reduce the size of the hci_ctrl struct by converting
'bool req_start' to 'u8 req_flags' and making the two function
pointers a union (since only one is ever set at a time).
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
The new hci_skb_pkt_* wrappers only help if they are used consistently
in the Bluetooth subsystem. So first convert the core packet handling.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
The SKB context buffer for HCI request is really not just for requests,
information in their are preserved for the whole HCI layer. So it makes
more sense to actually rename it into bt_cb()->hci and also call it then
struct hci_ctrl.
In addition that allows moving the decoded opcode for outgoing packets
into that struct. So far it was just consuming valuable space from the
main shared items. And opcode are not valid for L2CAP packets.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
There are two checks that are still using (MSG_OOB) instead of just
MSG_OOB and so lets just fix them.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Before the vendor specific setup stage is triggered call back into the
core to trigger an internal notification event. That event is used to
send an index update to the monitor interface. With that specific event
it is possible to update userspace with manufacturer information before
any HCI command has been executed. This is useful for early stage
debugging of vendor specific initialization sequences.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
When using the HCI_CHANNEL_RAW, restrict the packet types to valid ones
from the Bluetooth specification.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
The HCI_VENDOR_PKT quirk was needed for BPA-100/105 devices that send
these messages. Now that there is support for proper diagnostic channel
this quirk is no longer needed.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Introduce hci_recv_diag function for HCI drivers to allow sending vendor
specific diagnostic messages into the Bluetooth core stack. The messages
are not processed, but they are forwarded to the monitor channel and can
be retrieved by user space diagnostic tools.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
The Bluetooth public device address might change during controller setup
and it makes it a lot simpler for monitoring tools if they just get told
what the new address is. In addition include the manufacturer / company
information of the controller. That allows for easy vendor specific HCI
command and event handling.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
When the core starts or shuts down the actual HCI transport, send a new
monitor event that indicates that this is happening. These new events
correspond to HCI_DEV_OPEN and HCI_DEV_CLOSE events.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
The stack internal events that are exposed to userspace should be
limited to HCI_DEV_REG, HCI_DEV_UNREG, HCI_DEV_UP and HCI_DEV_DOWN.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
With 9380f9eacf the order of unsetting
the HCI_USER_CHANNEL flag of the HCI device was reverted to ensure
the device is first closed before making it available again.
Due to hci_dev_close checking for HCI_USER_CHANNEL being set on the
device it was never really closed and was kept opened. We're now
calling hci_dev_do_close directly to make sure the device is correctly
closed and we keep the correct order to unset the flag on our device
object.
Signed-off-by: Simon Fels <simon.fels@canonical.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
During the initial setup stage of a controller, the low-level transport
is actually active. This means that HCI_UP is true. To avoid toggling
the transport off and back on again for normal operation the kernel
holds a grace period with HCI_AUTO_OFF that will turn the low-level
transport off in case no user is present.
The idea of the grace period is important to avoid having to initialize
all of the controller twice. So legacy ioctl and the new management
interface knows how to clear this grace period and then start normal
operation.
For the user channel operation this grace period has not been taken into
account which results in the problem that HCI_UP and HCI_AUTO_OFF are
set and the kernel will return EBUSY. However from a system point of
view the controller is ready to be grabbed by either the ioctl, the
management interface or the user channel.
This patch brings the user channel to the same level as the other two
entries for operating a controller.
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Cc: stable@vger.kernel.org
The hci close method needs to know if we are in user channel context.
Only add the index to mgmt once close is performed.
Signed-off-by: Loic Poulain <loic.poulain@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
In preparation for changing how struct net is refcounted
on kernel sockets pass the knowledge that we are creating
a kernel socket from sock_create_kern through to sk_alloc.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In order to shrink the size of bt_skb_cb, this patch moves the HCI
request related variables into their own req_ctrl struct. Additionall
the L2CAP and HCI request structs are placed inside the same union since
they will never be used at the same time for the same skb.
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
The mgmt.c file should be reserved purely for HCI_CHANNEL_CONTROL. The
mgmt_control() function in it is already completely generic and has a
single user in hci_sock.c. This patch moves the function there and
renames it a bit more appropriately to hci_mgmt_cmd() (as it's a command
dispatcher).
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Linus Torvalds