Decreases CPU load especially for embedded machines.
Requires Linux >= 4.6 and libusb >= 1.0.21. If this
is not the case or the allocation fails, we will
fall back to buffers allocated in userspace.
Signed-off-by: Steve Markgraf <steve@steve-m.de>
Since a typo in rtlsdr_set_gpio_output() was fixed,
FC0012 tuners were not detected anymore, as the reset pin
is actually 4, not 5.
Thanks to David Basden et al for reporting the bug.
Signed-off-by: Steve Markgraf <steve@steve-m.de>
So far we had 32 * 256KB which was a bit overkill,
15 are more than enough.
15 was chosen instead of 16 because at least on Linux
there seems to be a system-wide limit of 63 transfers
(when they are 256KB large), so 4 dongles can be used
on a single machine without lowering the default transfer
number.
Signed-off-by: Steve Markgraf <steve@steve-m.de>
Otherwise the new transfer status does not propagate
into the xfer->status and we try to cancel all transfers
twice.
Also replace the ifdefs with a single macro for backwards
compatibility.
Signed-off-by: Steve Markgraf <steve@steve-m.de>
after the latest sublib changes we got:
src/Makefile.am:31: compiling `convenience/convenience.c' in
subdir requires `AM_PROG_CC_C_O' in `configure.ac'
Signed-off-by: Steve Markgraf <steve@steve-m.de>
libusb < 1.0.9 doesn't have libusb_handle_events_timeout_completed(),
but libusb <= 1.0.8 doesn't have version.h, so we need to check
for the function.
The cmake-code was borrowed from UHD, which also checks
for libusb_error_name(), we add that as well since it might come
handy later on.
Signed-off-by: Steve Markgraf <steve@steve-m.de>
Now we're using libusb_handle_events_timeout_completed()
instead of the deprecated libusb_handle_events_timeout(),
and notify our async state machine when rtlsdr_cancel_async()
is called. This speeds up exiting from any program that uses the
device asynchronously.
Signed-off-by: Steve Markgraf <steve@steve-m.de>
Thanks to Joris van Rantwijk for finding what seems to be
a hardware limitation/bug (bit 28 of the rsamp register being
forced to the value of bit 27).
Signed-off-by: Steve Markgraf <steve@steve-m.de>
Since this feature also disconnects running instances
of librtlsdr that claimed the device, it is now disabled
by default and can only be enabled at compile time.
To enable it when building with cmake:
cmake . -DDETACH_KERNEL_DRIVER=ON
To enable it when building with automake:
./configure --enable-driver-detach
Signed-off-by: Steve Markgraf <steve@steve-m.de>
libusb reacts differently to a device loss during runtime,
and sporadic errors can happen when starting a second
instance interfacing another dongle, thus we need to wait
for all transfers to fail before assuming the dongle has
been removed.
Signed-off-by: Steve Markgraf <steve@steve-m.de>
Since the R820T is a power hog and gets quite hot,
this makes sense, especially when being battery-
powered.
Signed-off-by: Steve Markgraf <steve@steve-m.de>
This improves the situation on hosts with flaky host
controllers like the Raspberry Pi and the WR703N.
Signed-off-by: Steve Markgraf <steve@steve-m.de>
This allows to open a device by a name (serial number string)
that has been programmed with rtl_eeprom -s.
Signed-off-by: Steve Markgraf <steve@steve-m.de>
Otherwise the tuner would use the uncorrected xtal value
after enabling and disabling the direct sampling mode
Signed-off-by: Steve Markgraf <steve@steve-m.de>
We now use Hz instead of KHz for calculating the PLL
parameters, and use the actual crystal frequency in Hz,
which allows to correct the frequency error in ppm.
Signed-off-by: Steve Markgraf <steve@steve-m.de>
This allows tuning to 27 MHz for example (civil band).
Best results are achieved when setting the gain to
very low values.
Signed-off-by: Steve Markgraf <steve@steve-m.de>
This fixes the issue of the FC0013 locking up at frequencies
between 928.0 and 950.0 MHz, which happened because
the numerator of the fraction exceeded its 5 bit limit.
For the tuner to behave normally again, the dongle needed
to be replugged.
For the FC0013 this now results in a small gap between
948.6 MHz and 950.0 MHz, where no valid PLL values
are existant.
For the FC0012 tuning is aborted when the maximum
frequency has been reached (948.6 MHz).
Signed-off-by: Steve Markgraf <steve@steve-m.de>
This commit adds the function rtlsdr_set_direct_sampling()
which can be used to enable/disable a mode where the
RTL2832 acts as a direct sampling receiver.
This mode disables the tuner, and by attaching a long
wire, or better, a 50Ω to 200Ω transformer and a lowpass-
filter to the In-phase ADC input (pin 1 or 2 of the RTL2832,
whereas pin 1 is at the molded dot) it is possible to listen to
shortwave radio stations. The coupling capacitors can be
left in place, but for better results they should be removed.
Tuning in this mode is done with the DDC, and since the
ADC samples with 28.8 MHz, tuning is possible from 0 to
28.8 MHz.
Signed-off-by: Steve Markgraf <steve@steve-m.de>
This takes simply too much time (almost 3 seconds)
and doesn't really seem to make any difference for our
use case.
Signed-off-by: Steve Markgraf <steve@steve-m.de>
This could lead to a misdetection of the tuner on some
platforms (E4K instead of FC0013 on the raspberry pi).
Thanks to Christoph Gommel for reporting this issue.
Signed-off-by: Steve Markgraf <steve@steve-m.de>
Changing the license of the header as well,
which had been forgotten in commit
6aec27c6d0
- "tuner_e4k: relicense driver under GPLv2+"
Signed-off-by: Steve Markgraf <steve@steve-m.de>
Steve Markgraf