Introduce a new API function sr_session_stopped_callback_set()
which can be used to receive notification when a session stops
running. This allows applications to integrate libsigrok event
processing with their own main loop, instead of blocking in
sr_session_run().
The resource API provides a generic means for accessing resources
that are bundled with sigrok, such as device firmware files. Since
the manner of resource bundling is platform-dependent, users of
libsigrok may override the functions used to open, close and read
a resource. The default implementation accesses resources as files
located in one of the XDG data directories or a directory defined
at compile time.
Use in-memory buffers instead of temporary files. This avoids
the need for low-level I/O on the FD returned by g_mkstemp().
Refactor the code accordingly. Also plug a number of leaks and
tighten the error checking.
Introduce the sr_file_get_size() utility function to retrieve the
size of an open FILE stream. This is based on fseeko() followed by
ftello(), which are POSIX functions but quite portable in practice.
Since these calls operate on FILE streams instead of filenames, the
issue of filename encoding no longer arises.
Since Autoconf places some important feature flags only into the
configuration header, it is necessary to include it globally to
guarantee a consistent build.
Disallow polling for input/error and output-ready events at the
same time, and ensure only a single FD event source is installed.
Also, do not leak if the FD event source is removed by means
other than calling serial_source_remove().
On MinGW, two implementations of printf() are available: either
the Microsoft native one or a standard-conforming replacement from
gnulib. Since we build in C99 mode, headers such as <inttypes.h>
already select the standard-conforming variant. However, MinGW's
GCC does not seem to know about this and assumes MS-style format
syntax by default, which triggers a lot of wrong warnings.
Thus, on MinGW, explicitly decorate sr_log() with the gnu_printf
format flavor attribute. Also use GLib's printf replacements in
the logging implementation to make sure we link to a conforming
printf on any platform, independently of the compiler flags.
This gets rid of the mistaken -Wformat warnings for sr_log(), but
does not cover functions such as g_strdup_printf() which do not
explicitly specify the gnu_printf flavor in the format attribute.
This can be overcome by adding "-D__printf__=__gnu_printf__" to
CPPFLAGS, but it would be inappropriate for libsigrok to define
this on its own.
Get rid of the specicialized sr_err(), sr_warn(), etc. functions.
Instead, define the logging helper macros in terms of sr_log(),
and remove the sr_log() helper macro so that no function is hidden
by a macro anymore.
Decorate sr_log() with G_GNUC_PRINTF to detect varargs errors. This
unearthed a gazillion warnings all over the place which will have
to be fixed.
Also convert the helper macros to ISO C99 __VA_ARGS__ style instead
of relying on a GNU C extension. Paste the log prefix directly into
the format string to make this work.
Replace the custom session main loop with the GLib main loop.
This is phase one of the port, which leaves the session and
driver APIs unchanged while replacing the internals.
Introduce new internal session API for changing the set of polled
file descriptors for an already installed event source. Use the
new API to apply changes to the USB poll FDs when requested to do
so by libusb. Doing so is necessary to make the generic USB code
work on Windows.
This is another attempt at getting the mess that is libusb event
handling on Windows under control. Until libusb makes its HANDLEs
available for polling, we have no choice but to block while waiting
for libusb events. Since we do not want to force drivers to deal
with multi-threading issues, that means we have to block in the
session main loop.
Fortunately, it turns out that our drivers aren't using multiple
event sources, so it is actually possible to block the main loop
without disrupting too much. This also gets rid of the USB thread
on Windows. Thankfully, libusb does not seem to care that we are
now calling libusb_handle_events_timeout_completed() twice per
iteration: first a blocking call (with timeout) in the callback
wrapper, followed by the non-blocking call in the driver-supplied
callback.
Turns out that having one event source per libusb poll FD is
a bad idea. There is only a single callback for all poll FDs,
and libusb expects to be called only once per poll iteration,
no matter how many FDs triggered.
Also, they should all share the same timeout, which should get
reset on events from any polled FD. The new timeout handling made
this problem apparent, as it caused the callback to be invoked
multiple times on timeouts, once for each separate event source.
In order to fix this, change the implementation to allow for an
arbitrary number of poll FDs per event source. This number is
zero for timer FDs, one for normal I/O sources, and one or more
for libusb sources (Unix only).
Also, on Windows, do not get an additional timeout from libusb
in the event loop. This is only appropriate when polling the
libusb FDs directly, which we aren't doing on Windows.
Handle I/O sources and timer ("dummy") sources within the same
polling loop, so that both may be used together. Slightly change
the API to improve consistency: a timeout value of -1 now disables
the timeout, and 0 makes the source always time out immediately.
The "dummy" sources already behaved that way, although it wasn't
documented as such.
Make sure that I/O events are processed preferentially: Skip any
timeout callbacks if an I/O event occurred within the same poll
iteration. This applies to both timer/idle sources and timeouts
of I/O sources.
Do not create dummy GPollFDs for timer/idle sources. Instead,
split the sources array into an I/O section and a timer section,
and create corresponding GPollFDs only for the I/O section. Use
GArray to simplify the handling of the dynamic arrays.
Keep track of when source timeouts are due and properly compare
against accumulated elapsed time between invocations. This prevents
sources with short timeouts from blocking other sources with longer
timeouts indefinitely.
This requires sr_hw_cleanup_all() and sanity_check_all_drivers()
to also take a context.
The (runtime) generation of the driver list now happens in sr_init()
and sr_driver_list() always returns that pre-generated list. This fixes
a segfault when (correctly) invoking multiple sr_init() and sr_exit()
calls with different contexts (caught by the unit tests).
This fixes bug #565.
This is a partial fix for bug #343, which lead to a large amount of
handles being created, and eventually to a frontend "hang".
It's not yet a "full" fix as some issues are still observable,
but it successfully improves the situation on Windows to the extent
that frontend hangs due to large amounts of handles no longer seem
to happen.
Thanks to Boris Gjenero <boris.gjenero@gmail.com> for the debugging
efforts, testing, and updating of this patch!
Additionally, this seems to also fix a "SysClk LWLA hanging" bug
and apparently not receiving any samples during an acquisition
(tested on an LWLA1034).
This closes bug #328.
It fills the fields with reasonable default values that should suit
most of the drivers. Drivers are obviously free to override the fields
they want after initializing.
Daniel Elstner
Uwe Hermann
Bert Vermeulen
Soeren Apel
Aurelien Jacobs
Martin Ling
Joel Holdsworth
Uffe Jakobsen