Pull vfs pile (part one) from Al Viro:
"Assorted stuff - cleaning namei.c up a bit, fixing ->d_name/->d_parent
locking violations, etc.
The most visible changes here are death of FS_REVAL_DOT (replaced with
"has ->d_weak_revalidate()") and a new helper getting from struct file
to inode. Some bits of preparation to xattr method interface changes.
Misc patches by various people sent this cycle *and* ocfs2 fixes from
several cycles ago that should've been upstream right then.
PS: the next vfs pile will be xattr stuff."
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (46 commits)
saner proc_get_inode() calling conventions
proc: avoid extra pde_put() in proc_fill_super()
fs: change return values from -EACCES to -EPERM
fs/exec.c: make bprm_mm_init() static
ocfs2/dlm: use GFP_ATOMIC inside a spin_lock
ocfs2: fix possible use-after-free with AIO
ocfs2: Fix oops in ocfs2_fast_symlink_readpage() code path
get_empty_filp()/alloc_file() leave both ->f_pos and ->f_version zero
target: writev() on single-element vector is pointless
export kernel_write(), convert open-coded instances
fs: encode_fh: return FILEID_INVALID if invalid fid_type
kill f_vfsmnt
vfs: kill FS_REVAL_DOT by adding a d_weak_revalidate dentry op
nfsd: handle vfs_getattr errors in acl protocol
switch vfs_getattr() to struct path
default SET_PERSONALITY() in linux/elf.h
ceph: prepopulate inodes only when request is aborted
d_hash_and_lookup(): export, switch open-coded instances
9p: switch v9fs_set_create_acl() to inode+fid, do it before d_instantiate()
9p: split dropping the acls from v9fs_set_create_acl()
...
sparse complains about these functions:
usbip/stub_dev.c:529:5: warning: symbol 'stub_pre_reset' was not declared. Should it be static?
usbip/stub_dev.c:535:5: warning: symbol 'stub_post_reset' was not declared. Should it be static?
-> add static keyword to silence the warning and make sparse happy.
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
alloc failures already get standardized OOM
messages and a dump_stack.
For the affected mallocs around these OOM messages:
Converted kzallocs with multiplies to kcalloc.
Converted kmallocs with multiplies to kmalloc_array.
Converted a kmalloc/strlen/strncpy to kstrdup.
Moved a spin_lock below a removed OOM message and
removed a now unnecessary spin_unlock.
Neatened alignment and whitespace.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
On the client side, we have a virtual hcd driver, there actually no
hardware interrupts, so we do not need worry about race conditions
caused by irq with spinlock held. Turning off irq is not good for system
performance after all. Just replace them with a non interrupt safe
version.
Signed-off-by: Harvey Yang <harvey.huawei.yang@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The function 'usbip_event_add()' may be called in interrupt context on
the stub side:
'stub_complete'->'stub_enqueue_ret_unlink'->'usbip_event_add'.
In this function it tries to get the lock 'ud->lock', so we should
disable irq when we get this lock in process context.
Signed-off-by: Harvey Yang <harvey.huawei.yang@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add missing angle bracket before and after the URL.
Signed-off-by: Masanari Iida <standby24x7@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If mkdir() of VHCI_STATE_PATH fails because the directory
already exists, that's not an error. This patch fixes
annoying "record connection" errors that would typically
come up on attach.
Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com>
Acked-by: David Chang <dchang@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This comes handy when hacking with different cross
compilers, some of which may or may not have
_FORTIFY_SOURCE turned on by default. This patch
allows us to turn _FORTIFY_SOURCE on by specifying
--with-fortify option at configuration time (or
to turn it off by specifying --without-fortify).
If nothing is specified, default compiler behavior
is assumed.
Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
USBIP daemon relies on functions available in glib2 library
to spawn handler code for incoming connection. This makes the
whole program dependent on glib2 library, which is a GNOME library
that on systems that don't have GNOME results in pulling more
dependency, only to be able to run a relatively trivial
socket-based program.
While this may not seem to be a problem on full-blown desktops
that already have the necessary libraries, it is a big issue
on small embedded systems (think USB hub with an Ethernet port)
that only have bare essentials in their file systems.
This patch eliminates glib2 dependency by reworking the
code to use lower level system calls to dispatch connection
handler. Instead of using glib2-style event loop and dispatching
mechanism, just do a ppoll(2) system call in our own loop and call
accept(2) followed by fork(2) on the socket that has incoming
connection. Stevens' books taught us that more than twenty
years ago. No need for anything smarter in a simple server, such
as usbipd.
Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There are a bunch of automatically generated files that git
should not care about.
Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
File doc/usbip_bind_driver.8 does not exist any more but it is
listed in dist_man_MANS. This breaks the build of the userspace.
Remove the file from the list.
Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com>
Acked-by: David Chang <dchang@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
CONFIG_HOTPLUG is going away as an option so __devexit_p is no longer
needed.
Signed-off-by: Bill Pemberton <wfp5p@virginia.edu>
Cc: Matt Mooney <mfm@muteddisk.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
with the last detached port state file remaining, usbip reports error on
attaching. So clean up the state files on detaching.
Signed-off-by: Harvey Yang <harvey.huawei.yang@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If probe returns with error, the kthread is still alive even when all
usbip modules unloaded. So do cleanups in error handler.
Signed-off-by: harvey.yang <harvey.huawei.yang@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
HC_STATE_RUNNING is already set by the usb core.
Signed-off-by: Bernard Blackham <b-linuxgit@largestprime.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
usbip takes a reference on a struct file which is passed in via
sysfs. Previously, this reference was never cleaned up, although
the socket it referred to was.
This patch drops the corresponding reference (found with the
socket's ->file backpointer) instead of just closing the socket.
Signed-off-by: Bernard Blackham <b-linuxgit@largestprime.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Replaced pointer arithmetic by using array indexing, and changed
function return type for usbip_alloc_iso_desc_pdu from 'void*' to
'struct usbip_iso_packet_descriptor'.
Signed-off-by: Bart Westgeest <bart@elbrys.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Here is the big USB pull request for 3.7-rc1
There are lots of gadget driver changes (including copying a bunch of
files into the drivers/staging/ccg/ directory so that the other gadget
drivers can be fixed up properly without breaking that driver), and we
remove the old obsolete ub.c driver from the tree. There are also the
usual XHCI set of updates, and other various driver changes and updates.
We also are trying hard to remove the old dbg() macro, but the final
bits of that removal will be coming in through the networking tree
before we can delete it for good.
All of these patches have been in the linux-next tree.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.19 (GNU/Linux)
iEYEABECAAYFAlBp3+AACgkQMUfUDdst+ym5vwCfe93FyJyXn/RDkGz7iBemvWFd
vrwAoIxjaOa4/yWZWcgrWc5bP4aO3ssc
=jYDr
-----END PGP SIGNATURE-----
Merge tag 'usb-3.6' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb
Pull USB changes from Greg Kroah-Hartman:
"Here is the big USB pull request for 3.7-rc1
There are lots of gadget driver changes (including copying a bunch of
files into the drivers/staging/ccg/ directory so that the other gadget
drivers can be fixed up properly without breaking that driver), and we
remove the old obsolete ub.c driver from the tree.
There are also the usual XHCI set of updates, and other various driver
changes and updates. We also are trying hard to remove the old dbg()
macro, but the final bits of that removal will be coming in through
the networking tree before we can delete it for good.
All of these patches have been in the linux-next tree.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>"
Fix up several annoying - but fairly mindless - conflicts due to the
termios structure having moved into the tty device, and often clashing
with dbg -> dev_dbg conversion.
* tag 'usb-3.6' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb: (339 commits)
USB: ezusb: move ezusb.c from drivers/usb/serial to drivers/usb/misc
USB: uas: fix gcc warning
USB: uas: fix locking
USB: Fix race condition when removing host controllers
USB: uas: add locking
USB: uas: fix abort
USB: uas: remove aborted field, replace with status bit.
USB: uas: fix task management
USB: uas: keep track of command urbs
xhci: Intel Panther Point BEI quirk.
powerpc/usb: remove checking PHY_CLK_VALID for UTMI PHY
USB: ftdi_sio: add TIAO USB Multi-Protocol Adapter (TUMPA) support
Revert "usb : Add sysfs files to control port power."
USB: serial: remove vizzini driver
usb: host: xhci: Fix Null pointer dereferencing with 71c731a for non-x86 systems
Increase XHCI suspend timeout to 16ms
USB: ohci-at91: fix null pointer in ohci_hcd_at91_overcurrent_irq
USB: sierra_ms: don't keep unused variable
fsl/usb: Add support for USB controller version 2.4
USB: qcaux: add Pantech vendor class match
...
As discussed at the kernel summit this year, CONFIG_EXPERIMENTAL means
nothing, so let's get rid of it.
Cc: Kees Cook <keescook@chromium.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The usb_device structure contains an array of usb_device "children".
This array is only valid if the usb_device is a hub, so it makes no
sense to store it there. Instead, store the usb_device child
in its parent usb_port structure.
Since usb_port is an internal USB core structure, add a new function to
get the USB device child, usb_hub_find_child(). Add a new macro,
usb_hub_get_each_child(), to iterate over all the children attached to a
particular USB hub.
Remove the printing the USB children array pointer from the usb-ip
driver, since it's really not necessary.
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Lan Tianyu <tianyu.lan@intel.com>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
USB autosuspend suspends vhci_hcd. In this process hcd_bus_suspend gets
executed which puts vhci_hcd in suspend state and calls vhci_hub_status.
vhci_hub_status function checks hub state and if it is in suspend state,
usb_hcd_resume_root_hub gets executed which resumes hub and if hub is
idle, again autosuspend puts it in suspend state and this goes on.
vhci_hub_status should resume hub only when hub port is in suspend state
and hub port status has changed.
Signed-off-by: navin patidar <navinp@cdac.in>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Almost all of usbip assumes that the_controller->lock is acquired
before vdev->priv_lock. The exception is in
vhci_device_unlink_cleanup(), where locks are acquired in the
reverse order. This leads to occasional deadlocks.
Fixing this is a bit fiddly, as the_controller->lock can't be held
when calling usb_hcd_unlink_urb_from_ep() in the middle of the list
traversal. As I can't rule out concurrent callers to this function
(perhaps it is safe?), the code here becomes slightly uglier - when
locks are dropped in the middle so the list may have emptied itself
(not even list_for_each_entry_safe is safe here).
Signed-off-by: Bernard Blackham <b-linuxgit@largestprime.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This brings them back up to date after:
commit e9837bbb3e
Author: matt mooney <mfm@muteddisk.com>
Date: Thu May 26 06:17:11 2011 -0700
staging: usbip: userspace tools v1.0.0
I couldn't find a current equivalent for
client:# usbip --port
Perhaps that functionality has been deprecated due to better logging.
It seems like libsrc/usbip_common.h's usbip_status_string is no longer
used.
Signed-off-by: W. Trevor King <wking@tremily.us>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When `--with-tcp-wrappers` is passed to `configure`, the previous code
always reset LIBS to $saved_LIBS, regardless of whether libwrap was
found or not. The current code makes the `--with-tcp-wrappers` case
look more like the default case, and it only resets LIBS if libwrap
was not found.
Signed-off-by: W. Trevor King <wking@tremily.us>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Now usbip_common.c's pr_fmt is the only thing setup by CONFIG_USBIP_DEBUG
that you can't subsequently alter using this parameter.
Signed-off-by: W. Trevor King <wking@tremily.us>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Changed setting the status bits to a byte-oriented approach. Number of
written status bits is now based on VHCI_NPORT, instead of casting
status buffer to a long. This fixes a stack corruption bug on 64-bit
based architectures, and potential byte order / endianness related
issues.
In addition updated function comments, and log statements.
Signed-off-by: Bart Westgeest <bart@elbrys.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The header files should be included in generated tarballs for archives.
Signed-off-by: Natanael Copa <ncopa@alpinelinux.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This doesn't happen with the usbip virtual hci module, but another
module wanting to interface with this user space code could cause a
seg-fault by sending data without newlines.
Signed-off-by: Christopher Harvey <charvey@matrox.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
stub_shutdown_connection() and vhci_shutdown_connection() use
task_is_dead() before kthread_stop(). This buys nothing and wrong.
kthread_stop() is fine even if this thread is dead. However, if it
is dead nothing protects this task_struct, we shouldn't touch this
memory.
Change the code to do the necessary get_task_struct/put_task_struct.
This patch assumes that
- xxx_shutdown_connection() is always called, so we can't
leak the task_struct.
- kthread_stop_put() can't be called twice on the same task.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Cc: Tobias Klauser <tklauser@distanz.ch>
Cc: Matt Mooney <mfm@muteddisk.com>,
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This was done to resolve some merge issues with the following files that
had changed in both branches:
drivers/staging/rtl8712/rtl871x_sta_mgt.c
drivers/staging/tidspbridge/rmgr/drv_interface.c
drivers/staging/zcache/zcache-main.c
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The function usbip_pad_iso never returns anything but 0 (success).
Signed-off-by: Bart Westgeest <bart@elbrys.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The stub_probe function can be called as soon as the stub_driver is
registered. This can lead to the busid_table_lock being used before it
is initialized. Moved calling the init_busid_table function (which
initalizes this spinlock) to be called earlier in the init function to
prevent this from happening.
Signed-off-by: Bart Westgeest <bart@elbrys.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Removed remnants of old function definitions which have since been
removed in earlier cleanup work.
Signed-off-by: Bart Westgeest <bart@elbrys.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The usbip_xmit function supported sending and receiving data, however
the sending part of the function was never used/executed. Renamed the
function to usbip_recv, and removed the unused code.
Signed-off-by: Bart Westgeest <bart@elbrys.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Interrupts must be disabled prior to calling usb_hcd_unlink_urb_from_ep.
If interrupts are not disabled, it can potentially lead to a deadlock.
The deadlock is readily reproduceable on a slower (ARM based) device
such as the TI Pandaboard.
Signed-off-by: Bart Westgeest <bart@elbrys.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This 2nd batch of implicit module.h users only appeared when we
removed the unnecessary module.h from include/linux/miscdevice.h
[The 1st batch is already present in Greg's staging tree.]
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Lots of drivers have in the past expected the presence of basic things
like THIS_MODULE and EXPORT_SYMBOL. With the header cleanup, they wont
have these. Call out the include explicitly.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Linus Torvalds