| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|
|
|
|
|
|
|
|
| |
I have a crash with an older libosmocore and the sysmobts. I
am not sure how we can have a release for a channel where the
lapdm_channel has not gone through lapdm_channel_init. But to
prevent the crash we can promise lapdm_channel_exit to be idempotent.
It has been idempotent since f5a079f739c57d8be7c59149fd45475c402a45fc
|
|
|
|
|
|
| |
Add a #define for the check and use this in this path as well.
Fixes: Coverity CID 1103092
|
|
|
|
| |
Fixes: Coverity CID 1111545
|
|
|
|
|
|
|
| |
We could consider creating a CHECK MACRO that will print a message
in case of a tx failure.
Fixes: Coverity CID 1111544
|
|
|
|
| |
Fixes: Coverity CID 1107253
|
| |
|
|
|
|
|
|
|
|
|
| |
When a RESET is received on the same link with a different NSVCI from
a BSS on a dynamically created NS connection do not patch the nsvc
object but create a new one instead. Thus the NSVCI is never modified
at a nsvc object after the NS-VC has been established.
Sponsored-by: On-Waves ehf
|
|
|
|
|
|
|
|
|
| |
This modifies the NS stack's behavior to accept RESET and RESET_ACK
NSEI changes for NS-VC dynamically created by RESET messages from
BSSes. This feature is not used for NS-VC configured via VTY or NS-VC
to a SGSN.
Sponsored-by: On-Waves ehf
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This changes the implementations for the reception of RESET and
RESET_ACK to be compatible with 3GPP TS 08.16, 7.3.1:
- Just send a RESET_ACK with correct values back to the SGSN when a
RESET with an invalid NSVCI or NSEI has been received.
- Check RESET_ACK for matching NSEI and NSVCI.
- Ignore unexpected RESET_ACKs.
In addition, use RESET_ACK from a BSS to update the BSS source
address based on the NSVCI to be tolerant with changing UDP source
addresses/ports.
Sponsored-by: On-Waves ehf
|
|
|
|
|
|
|
|
|
|
|
|
| |
Add a test handling proper and mismatching RESETs. The latter may
occur, when ports change within a NS-VC without the SGSN getting
notice.
This tests for the behavior of the NS protocol stack for RESET and
RESET_ACK messages which changing/invalid NSEI and NSVCI like it is
being described in 3GPP TS 08.16, 7.3.1.
Sponsored-by: On-Waves ehf
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This fixes a SEGV error that happens the next time the statistics
are updated.
Addresses:
Program terminated with signal 11, Segmentation fault.
#0 0xb7711fa5 in rate_ctr_group_intv (grp=<optimized out>) at
rate_ctr.c:107
107 for (i = 0; i < grp->desc->num_ctr; i++) {
#0 0xb7711fa5 in rate_ctr_group_intv (grp=<optimized out>) at
rate_ctr.c:107
#1 rate_ctr_timer_cb (data=0x0) at rate_ctr.c:129
#2 0xb770ec59 in osmo_timers_update () at timer.c:243
#3 0xb770ef7a in osmo_select_main (polling=0) at select.c:133
#4 0x08049987 in main (argc=3, argv=0xbfba8084) at
gb_proxy_main.c:306
Sponsored-by: On-Waves ehf
|
|
|
|
|
|
|
| |
Use 'config-ns' instead of plain 'ns' and append a blank after the
'#' to make the vty python test script work.
Sponsored-by: On-Waves ehf
|
|
|
|
|
|
|
|
|
| |
This adds a 'nsvc nsvci <0-65535> (block|unblock|reset)' vty
command. It selects the NS-VC based on the nsvci instead of using the
first with a matching NSEI, like it is done when the 'nsei' keyword
is used instead.
Sponsored-by: On-Waves ehf
|
|
|
|
|
| |
* compile is generated by newer automake
* ignore the test binary
|
| |
|
|
|
|
|
|
|
| |
The state matching condition is inverted. This is corrected by this
fix.
Sponsored-by: On-Waves ehf
|
|
|
|
|
|
|
| |
This adds tests for the gprs_ns_sendmsg() function. For this it
merges back functions from the gbproxy test.
Sponsored-by: On-Waves ehf
|
|
|
|
| |
New interfaces and ABI incompatible changes in the GB library
|
|
|
|
|
|
| |
Before the assigned value (0xFF) was truncated, reg->text[0] is of
type char. A corresponding test for the same value in openbsc could
only fail.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Under special circumstances (see below) receiving a NS-RESET leads to
duplicated NS-VC entries.
This happens when the source port of a NS-VC changes to a new one
that has already been used by another NS-VC.
This patch changes gprs_ns_rx_reset() to check for this case and to
use the existing NS-VC object. The NS-VC object that was associated
with the source address before is detached from this source but kept
in the NS-VC list so that it can be reattached when a correspondent
NS-RESET is received later on. Meanwhile it will have a cleared link
layer address which will not match a real link info.
A new counter NS_CTR_REPLACED is incremented each time when the NS-VC
object is replacing another one. A new signal S_NS_REPLACED is added
which gets dispatched in this case, too.
Another new counter NS_CTR_NSEI_CHG is incremented each time when the
NSEI of a NS-VC object (with fixed NSVCI) changes.
Ticket: OW#874
Sponsored-by: On-Waves ehf
|
|
|
|
|
|
|
|
| |
Adds the functions gprs_ns_ll_copy() and gprs_ns_ll_clear(). Renames
gprs_ns_format_peer() to gprs_ns_ll_str(). All of these functions
uniformly access the link layer part within the NS-VC objects.
Sponsored-by: On-Waves ehf
|
|
|
|
|
|
|
|
|
|
|
|
| |
Currently this first NS-VC with a matching NSEI is always used to
send a UNITDATA message via gprs_ns_sendmsg(). If the NS-VC found is
either dead or blocked, an error is returned.
This patch changes to code to skip blocked or dead NS-VCs while
searching and return the error if no usable NS-VC is found. This
makes it possible to have several NS-VCs per connection group.
Sponsored-by: On-Waves ehf
|
|
|
|
|
|
|
| |
Register an osmo signal handler to print a short notice about every
SS_L_NS signal that is generated while processing the tests.
Sponsored-by: On-Waves ehf
|
|
|
|
|
|
|
|
|
| |
The testcase didn't work on Ubuntu 12.04 because vty_create will
directly call vty_event (e.g. not through the plt). This means
that the approach to override vty_event in the testcase failed.
Use the signal interface of libosmocore and make the testcase
use it. The signals can be generally useful as well.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Currently when a NS-RESET is recevied over a link that has not yet
been associated with a NS-VC, the NSEI is used to find an existing
NS-VC. If one is found, the reset procedure is initiated.
This behaviour is not conformant with 3GPP TS 08.16 (see chapter
4.2.3) which allows to use several NS-VC between two endpoints in
parallel.
The patch changes the implementation to use the NSVCI instead
of the NSEI to search for an existing NS-VC object.
Ticket: OW#874
Sponsored-by: On-Waves ehf
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
According to 3GPP TS 08.16, 7.3 "Reset procedure" the entity
receiving a NS-RESET PDU responds with a NS-RESET-ACK and 'then'
starts the test procedure which essentially means, that a NS-ALIVE
gets sent and a timer is started.
Currently the NS-ALIVE is sent before the NS-RESET-ACK.
This patch fixes the implementation by reversing the order in which
these messages are sent.
Sponsored-by: On-Waves ehf
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch refactors gprs_ns_rcvmsg() by moving the parts relevant to
the NS messages into the new functions gprs_ns_vc_create() (nsvc
object creation) and gprs_ns_process_msg() (main NS automaton). These
do not contain code that directly depends on the link layer (they
call other functions that still do). This reduces the gprs_ns_rcvmsg()
function to calling these two functions and optionally setting up the
link layer specific fields of the nsvc.
Sponsored-by: On-Waves ehf
|
|
|
|
|
|
| |
This tests the connection establishment by directly calling
gprs_ns_rcvmsg() and printing the resulting messages and the
NS-VC list.
|
|
|
|
|
|
| |
This fixes warnings that are raised by missing includes, missing casts,
missing return statements, using printf %lu with uint64_t, and unused
symbols.
|
|
|
|
|
|
|
|
|
|
| |
This explicitely allows aliasing of this struct to avoid the warning
shown below. Note, that this warning isn't show when
'-fno-strict-aliasing' is used.
Addresses:
gb/gprs_ns_test.c:85:54: warning: dereferencing type-punned pointer will
break strict-aliasing rules [-Wstrict-aliasing]
|
|
|
|
|
|
| |
It started to behave weird on Debian Testing (GCC 4.8), I compiled it
with address sanitizer support and set a breakpoint in
__asan_report_error to get a backtrace.
|
|
|
|
|
|
|
|
|
| |
gb/bssgp_fc_test.c: In function ‘fc_out_cb’:
gb/bssgp_fc_test.c:46:1: warning: no return statement in function returning non-void [-Wreturn-type]
gb/bssgp_fc_test.c: In function ‘fc_in’:
gb/bssgp_fc_test.c:56:1: warning: no return statement in function returning non-void [-Wreturn-type]
gb/bssgp_fc_test.c: In function ‘test_fc’:
gb/bssgp_fc_test.c:79:3: warning: implicit declaration of function ‘usleep’ [-Wimplicit-function-declaration]
|
|
|
|
|
|
|
|
| |
This information element has been added to the MM Information
message in GSM24.008. This patch adds it to the tlv_definition
to keep the TLV parser from breaking.
Ticket: OW#978
|
|
|
|
|
|
|
| |
This patch removes an assertion of node > CONFIG_NODE and changes
the function to handle all nodes properly.
For the sake of completeness, the generic 'exit' command
implementation is extended to work properly with all nodes, too.
|
| |
|
|
|
|
|
| |
In this case the last_node variable may hold values that are not
in enum node_type, so int is used instead.
|
|
|
|
|
|
|
|
|
|
|
| |
This adds the vty_install_default() function that is basically the
install_default() function plus the registration of the commands
'exit' and 'end'. The latter is only provided in subnodes of
ENABLED_NODE and CONFIG_NONE.
The VTY test program is extended to check these commands.
Ticket: OW#952
|
|
|
|
|
|
| |
Put all 'exit' logic into vty_go_parent() and replace the
implementations of 'exit' and 'end' by generic ones that use
vty_go_parent().
|
|
|
|
|
| |
The function is_config() returns 0 for CONFIG_NODE. Since that node
is a config node, the function is renamed to resolve this.
|
|
|
|
|
| |
Fix the build after I changed the order of header files in the
SMS test.
|
|\ |
|
|/
|
|
|
|
| |
The number of digits is the number of used octets times two (two
digits per octet). The result has been successfully dissected by
wireshark. It has not been tested with real phones.
|
|
|
|
|
|
|
|
| |
This reverts commit bf991bc0f3006eddf45d494ee7e476a75dbaf36e and
fixes the bug instead. It's just sufficient to remove the defined()
test, since identifiers that aren't macros are just considered as
0 when used with #if. (see cpp info page, section 4.2.2 'If'
in cpp-4.6.info)
|
|
|
|
|
|
|
| |
The code started to use #if defined(...) but the value was always
for GCC/Clang. The only difference was that for older compilers
the value of the definition was 0. Conditionally define these
macros.
|
|
|
|
|
| |
Separate feature checking from implementation. Supported are clang and
gcc.
|
|
|
|
|
|
|
|
|
|
| |
Fix type of argument 'argv'.
Addresses:
conv/conv_test.c:358:5: error: second parameter of 'main' (argument
array) must
be of type 'char **'
int main(int argc, char argv[])
|
|
|
|
|
|
|
|
|
|
| |
Add a new core/defs.h file for definitions that can be used from
within header files without including prototypes and extern
declarations. It's primarily meant for macro definitions and basic
typedefs.
Move the definition of OSMO_DEPRECATED there and use compiler version
info to avoid compiler errors.
|
|
|
|
|
| |
This lets make check fail earlier (in average) to shorten the
debugging cycle time.
|
|
|
|
|
| |
This is a workaround to make the FreeBSD8.4 and Debian6.0 target
build again. The deprecated attribute doesn't have arguments.
|