| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
This applies to read/write/close/poll/pipe -> unistd_posix.h
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Peter Stuge)
v1.0.6 release
Darwin: support multiple calls to libusb_init
Credit to Orin Eman for finding this bug.
Darwin: use logging functions
Use usbi_warn, usbi_err, and usbi_info instead of _usbi_log.
Darwin: fix memory leak in process_device
Credit to Mike Colagrosso for finding this bug.
Add internal abstraction for POSIX Threads
This prepares for a Windows backend without dependency on pthreads-w32.
pthread_* is renamed to usbi_* and PTHREAD_* to USBI_*.
A usbi_mutex_static_t and usbi_mutex_static_lock() and _unlock() are
introduced for statically initialized mutexes, since they may be
implemented using other types when pthreads mutexes aren't used.
Move -pthread from libusb/Makefile.am to host-specific AM_CFLAGS in
configure.ac. AM_CFLAGS is already used in Makefile.am.
(Numerous merge conflicts resolved by Michael Plante)
misc fixes to pthreads abstraction
* windows_compat.h should not be directly included (use libusbi.h)
* windows_usb.c still had some references to pthread_mutex_[un]lock
Add libusb_strerror() to get short error message in English from enum
(Merge conflicts resolved by Michael Plante)
Better cleanup on errors, stricter types and some good casts
Avoid various memory leaks in error code paths and remove warnings. Also
add usbi_cond_destroy in os/threads_posix.h because it's used for cleanup
now.
(Merge conflicts resolved by Michael Plante)
Rename all interface parameters to usb_interface or interface_number
(Merge conflicts resolved by Michael Plante)
Add type parameter to the list_for_each_entry() and _safe() macros
(Merge conflicts resolved by Michael Plante)
VA_ARGS workaround for logging with MSVC6
(Merge conflicts resolved by Michael Plante)
Fix context memory leak in libusb_init()
* Now holds default_context_lock for duration of libusb_init
* Doesn't allocate it if not needed
Fix the last MSVC /W3 warning
removal of pthreads from MSVC and mingw
* removed pthread*.lib from linking in msvc6 (2005/8 not yet edited)
* added threads_windows.[ch] to all project files in msvc6
* added usbi_cond_signal to both threading versions, unused
* added native windows threading code, now used
removed redundant time.h includes
removed pthread-win32 informational solution files
updated MSVC 2005/2008 for thread abstraction
fixed DDK build
fixed cygwin's use of POSIX threads
fixed dpfp/dpfp_threaded for MinGW compatibility
config_msvc.h -> msvc/config.h
pthread abstraction for DLL project files
fixed bad reference to thread_windows in 2005 DLL vcproj
cygwin ifdef is not required in threads_windows.h - removed
fixed residuals from previous patches
moved thread source detection into autotool scripts
|
| |
|
| |
|
|
|
|
|
|
|
| |
_libusb_####)
- fixed sscanf_s undefined for non MS compilers
- fixed cygiwn warnings in windows_usb.c
|
|
|
|
|
|
| |
- lsusb/xusb sources + ddk_build call
- missing libusb_strerror in .def
- removed "handle_timeouts_locked unused" warning in in io.c for MinGW
|
|
|
|
| |
usbi_io_exit
|
| |
|
|
|
|
| |
(latter according to linux manpage, which says glibc does not have this problem)
|
|
|
|
|
| |
- apply the same trick as usbi_handle_disconnect() to handle_timeouts(), so that usbi_handle_transfer_cancellation() can now be called from the backend without locking after a timeout.
- remove the unneeded handle_timeouts_locked()
|
|
|
|
| |
libusb_get_device_list change introduced with r52. Added definition on ssize_t in libusb.h for MSVC platforms.
|
| |
|
| |
|
| |
|
|
|
|
|
| |
- "poll we can believe in" (requires superficial modifications to core files)
- move the "interface" fix into windows_compat.h
|
| |
|
|
|
|
|
| |
The internal timing seems to be working, this will be a better test of
it before we make this timeout unlimited.
|
|
|
|
| |
Require glibc-2.9 for the working timerfd support.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Use a new file descriptor from the timerfd system calls to handle
timeouts. On supported systems, this means that there is less hassle
figuring out when the poll() timeout should be, since
libusb_get_next_timeout() will always return 0 and the timeout events will
be triggered as regular activity on the file descriptor set.
Add API function libusb_pollfds_handle_timeouts() to detect whether
you're on a platform with the timing headache, and flesh out the
surrounding documentation.
|
|
|
|
|
|
|
|
|
|
|
|
| |
At least on Linux, there were some possible races that could occur if
a transfer is cancelled from one thread while another thread is handling
an event for that transfer, or for if a transfer completes while it is
still being submitted from another thread, etc.
On the global level, transfers could be submitted and cancelled at the
same time.
Fix those issues with transfer-level locks.
|
|
|
|
|
|
|
|
|
| |
As pointed out by Dennis Muhlestein, libusb_get_max_packet_size()
doesn't really do what the documentation might suggest because it
does not consider the number of transaction opportunities per
microframe.
Add a new function to do what is useful for isochronous I/O.
|
|
|
|
| |
...because the OS does this for us.
|
|
|
|
|
|
| |
This will differ on Linux and Darwin, at least.
[dsd: minor style tweaks]
|
|
|
|
|
|
| |
When we receive data on the internal control pipe, we need to correctly
decrement the number of ready file descriptors before passing on the
remaining work to the OS implementation.
|
|
|
|
|
|
|
|
|
|
| |
This ensures that tpriv->urbs and tpriv->iso_urbs are always set to NULL
whenever a transfer is not submitted. In this way, submit_*_transfer()
and cancel_*_transfer() can error check to ensure that the transfer is
in the correct state to be either submitted or canceled, preventing
potential memory leaks or double frees.
Signed-off-by: David Moore <dcm@acm.org>
|
|
|
|
|
|
| |
Documentation brushed up, and I realised that another function is
needed for tight event handling loops -- they must be able to check
if an open/close operation is trying to interrupt it.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Ludovic Rousseau found that crashes often occur if you close a device
while another thread is doing event handling.
Fix this by adding an internal control pipe, which the close routines
use to interrupt the event handler and obtain the event handling lock,
ensuring that no other thread is handling events while the device is
closed. After the close completes, it signals all the event handlers
to start up again using the usual mechanism.
Also modified libusb_open() to do a similar thing, so that event
handlers are interrupted in order to realise that a new poll fd has
appeared.
|
| |
|
| |
|
|
|
|
|
|
| |
This fixes a problem pointed out by Lou, where resubmitting a transfer
that previously timed out never timed out again, as if the timeout had
been set to 0.
|
| |
|
| |
|
|
|
|
|
| |
Suggested by David Zeuthen. This allows multiple libraries in the same
process to independently use libusb without interfering.
|
|
|
|
|
|
| |
This reverts commit 2d3a1111caff40ebb87983c861ff548cdc9e5946.
This was based on the assumption that dying threads would automatically
release mutexes, which is not the case.
|
| |
|
|
|
|
|
| |
Pointed out by Richard Röjfors, otherwise we end up busy-looping with
a poll() timeout of 0.
|
|
|
|
|
|
|
|
|
|
|
|
| |
Ludovic Rousseau pointed out that libusb_unlock_events() is not called
when a thread gets terminated with a signal, meaning that event waiters
will not be woken up in this case.
Add a temporary hack to libusb_event_handler_active() so that at least
the other threads will realise on the next iteration of their event
handling loop.
The real fix will likely involve reworking most of this.
|
|
|
|
|
| |
This avoids a race between the user checking for active event handler
and then blocking on the condition variable
|
|
|
|
|
|
| |
Now offers a mechanism to wait for events while another thread is
doing the event handling. Complicates things for MT async apps, but
then again it's a bit of a tricky combination to start with.
|
|
|
|
|
|
|
| |
This patch closes a critical memory leak in handle_events. The fds
variable is malloced but never freed. When I'm calling
handle_events with a timeout of 0, my system runs out of memory after a
few seconds.
|
|
|
|
| |
Hopefully comprehensive enough for people to get started.
|
|
|
|
|
| |
This involved moving from select() to poll() because there is no way to
distinguish usbfs's POLLERR condition with select().
|
|
|
|
|
|
|
|
| |
Document behaviour where host-endian and bus-endian (little) should be
used respectively.
Also remove packed attribute from libusb_control_setup as all fields
are naturally aligned.
|
| |
|