Remove the paragraph about writing to the Free Software Foundation's
mailing address. The FSF has changed addresses in the past, and may do
so again. In 2021 this is not useful, let's rather have a bit less
boilerplate at the start of source files.
Change-Id: I8ba71ab9ccde4ba25151ecbeb2a323f706b57d43
Before this patch, reconnecting to osmo-trx and attempting to configure it for
another band is not going to work without restarting the process.
The new variable is added in order to still allow POWEROFF followed by a
POWERON without need to reconfigure the device. In that case, previous
configuration is kept.
Change-Id: I43e5e1e4dcb36be605c6bd25dd6a5f3649e244e7
There's another related logging line also at INFO level in the caller
path in Transceiver.cpp, but it only prints if detectBurst() failed.
Let's print this one as INFO too, which proved to be a good logging
level. This way user also notices gain is too high despite osmo-trx is
still able to decode bursts.
Change-Id: Ieca4f19ae1099a430e9b838f8b6780b1c61a87a9
downsampleBurst() and the Resampler below it clearly only support or are
confgiured for 1<->4 setup currently.
Change-Id: Iebaff7a34bd24e56627f148182859918accbfa82
Gain setting without a band was apparently led to a very low output
level, thanks to defog for pointing this out.
Change-Id: I8b59d38dd7b0781776c9e61226185879541fdc53
Related: OS#3342
This is only useful if the rf path delays the signal by at least one
frame, and therefore a fairly experimental command that might be removed
or reworked in the future and should not be relied upon.
Change-Id: I29964acffad5bad4f5bcad7f631e435a72979c46
In Transceiver::addRadioVector() we scale the I/Q samples by scaling
the output voltage of the DAC. A relative factor/divisor/ration in
the voltage domain cannot be used 1:1 in the power domain.
There exist two similar formulas:
a) X_dB = 10 * log10(X_lin / X_ref)
b) Y_db = 20 * log10(Y_lin / Y_ref)
both of them are correct, and according to [1]:
a) If you convert a quantity X that relates to power or energy,
=> the factor is 10.
b) If you convert a quantity Y that relates to amplitude,
=> the factor is 20.
Therefore we should be using 20 instead of 10. This change makes
osmo-trx apply per-lchan attenuation values correctly. Otherwise
it would double the values indicated in TRXD messages.
[1] https://dspillustrations.com/pages/posts/misc/decibel-conversion-factor-10-or-factor-20.html
Change-Id: I98bc00bd25df4913d45e55eb008d715aca76fc7c
Related: SYS#4918
The memory leak was reported by ASan:
Direct leak of 32 byte(s) in 1 object(s) allocated from:
#0 0x7f23b488e459 in __interceptor_malloc /build/gcc/src/gcc/libsanitizer/asan/asan_malloc_linux.cpp:145
#1 0x558e83e39e3c in ChannelizerBase::initFilters() /osmo-trx/Transceiver52M/ChannelizerBase.cpp:84
#2 0x558e83e3a8a0 in ChannelizerBase::init() /osmo-trx/Transceiver52M/ChannelizerBase.cpp:188
#3 0x558e83e2d263 in RadioInterfaceMulti::init(int) /osmo-trx/Transceiver52M/radioInterfaceMulti.cpp:197
#4 0x558e83de76d2 in makeRadioInterface(trx_ctx*, RadioDevice*, int) /osmo-trx/Transceiver52M/osmo-trx.cpp:115
#5 0x558e83dea663 in trx_start /osmo-trx/Transceiver52M/osmo-trx.cpp:600
#6 0x558e83dead6f in main /osmo-trx/Transceiver52M/osmo-trx.cpp:695
#7 0x7f23b2576151 in __libc_start_main (/usr/lib/libc.so.6+0x28151)
Change-Id: Ibc4c7edeb9bba517db08fce152d863e6cc0c7bbb
The leak was reported by ASan.
Direct leak of 48 byte(s) in 1 object(s) allocated from:
#0 0x7fd9c9c29f41 in operator new(unsigned long) /build/gcc/src/gcc/libsanitizer/asan/asan_new_delete.cpp:99
#1 0x55bd63ae2364 in RadioInterfaceMulti::init(int) /git/osmo-trx/Transceiver52M/radioInterfaceMulti.cpp:209
#2 0x55bd63a9c6d2 in makeRadioInterface(trx_ctx*, RadioDevice*, int) /git/osmo-trx/Transceiver52M/osmo-trx.cpp:115
#3 0x55bd63a9f663 in trx_start /git/osmo-trx/Transceiver52M/osmo-trx.cpp:600
#4 0x55bd63a9fd6f in main /git/osmo-trx/Transceiver52M/osmo-trx.cpp:695
#5 0x7fd9c7910151 in __libc_start_main (/usr/lib/libc.so.6+0x28151)
Change-Id: Ia4f9d4e47caa86ada98054763573e652d281992c
Previous code relied on abort() switching sigaction to SIG_FDL +
retriggering SIGABRT in case the signal handler returns, which would
then generate the coredump + terminate the process.
However, if a SIGABRT is received from somewhere else (kill -SIGABRT),
then the process would print the talloc report and continue running,
which is not desired.
Change-Id: I3a3ff56cb2d740a33731ecfdf76aa32606872883
Fixes: OS#4865
Otherwise these logging lines end up in the automatically generated
XML VTY reference (stdout), so this breaks further XML processing.
Change-Id: I8e0fd728d406e2452c9c0ddad5bce5f6b17fab42
Related: SYS#4937, SYS#4910
Coverity warns us that a non-static class members:
- mRxLowerLoopThread, and
- mTxLowerLoopThread,
are not initialized in this constructor nor in any functions that
it calls. I don't think it's critical, because we do initialize
them in Transceiver::start(), but let's make them nullptr.
Change-Id: If9e06aa7965f17383ab6599c15945e8ce2703bbf
Fixes: CID#214952
Neither VTY nor CTRL API is used in device specific code, excluding
the 'uhd' where osmo_cpu_sched_vty_apply_localthread() is called.
Change-Id: I568b443da4b96c005734d749faa22b9c7440f951
Prior to this patch, osmo-trx relied totally on proper VTY configuration
being set in "rssi-offset" together with the RxGain set through TRXC in
order to provide correct Uplink RSSI measurements to bts-trx.
With this patch, RSSI is now by default calculated (in LMS and UHD
backends) based on the currently set RxGain, by providing empirically
discovered values. Still, for backward compatibility, the old
"rssi-offset" command will overwrite completely the per-default
calculated rssi offset.
A new optional parameter "relative" is added at the end of the
"rssi-offset" VTY command to flag the value as relative to the newly
per-default calculated value. This way specific setups (like adding a
LNA / RF fronted) can still be expressed while still keeping the
automatic per-default offset.
Related: OS#4468
Change-Id: I8ef78fd20c22c60d61bfb18d80a4a36df4fd6c20
The commandline option --vty-ref-xml is needed to enable automatic
generation of the VTY reference manual.
Change-Id: I34dd36183e013ab005f39b235c4ab561590befb7
Related: SYS#4937, OS#1601
This patch fixes MS failing to even see the network, and only RACHs of 1
zeroed byte being seen in GSMTAP.
The issue seems to only appear on some specific machines; others have
been running fine for weeks without this memset being an issue.
Fixes: 7a52e42ee0
Change-Id: I98ad885a5d71e7775973a4d881c0f1cd665ea711
In the multi-ARFCN mode, if the Tx/Rx antenna names are explicitly
set in 'chan N' sections of the configuration file:
trx
...
multi-arfcn disable
chan 0
tx-path TX/RX
rx-path RX2
chan 1
tx-path TX/RX
rx-path RX2
chan 2
tx-path TX/RX
rx-path RX2
osmo-trx would crash, because radioDevice::set_antennas() would
attempt to configure antenna names for all N physical channels,
while USRP devices usually have 2 or even 1 available.
The easiest approach is to remove both 'tx-path'/'rx-path' from
all 'chan N' sections excluding 'chan 0', so it would work fine.
This makes sense, because in the multi-ARFCN mode we actually
use only one physical channel.
However, let's still make sure that explicit configuration of the
Tx/Rx antenna names would not crash osmo-trx and skip N > 0 in
radioDevice::set_antennas().
Change-Id: I09f316f181cbbc2214e8913b73f7c1fcea4e8c05
Related: OS#4636
Oliver Smith
Vadim Yanitskiy
Pau Espin
Eric Wild
Harald Welte
Vadim Yanitskiy
Sylvain Munaut
Philipp Maier