From: Pablo Neira Ayuso <pablo(a)gnumonks.org>
This patch fixes a segfault if we or one BTS start a TCP connection
and close it before any IPAC_MSGT_ID_RESP message is sent.
---
openbsc/src/ipaccess/ipaccess-proxy.c | 5 +++++
1 files changed, 5 insertions(+), 0 deletions(-)
diff --git a/openbsc/src/ipaccess/ipaccess-proxy.c b/openbsc/src/ipaccess/ipaccess-proxy.c
index 56fe160..d98eb54 100644
--- a/openbsc/src/ipaccess/ipaccess-proxy.c
+++ b/openbsc/src/ipaccess/ipaccess-proxy.c
@@ -806,6 +806,11 @@ static void handle_dead_socket(struct bsc_fd *bfd)
switch (bfd->priv_nr & 0xff) {
case OML_FROM_BTS: /* incoming OML data from BTS, forward to BSC OML */
+ /* The BTS started a connection with us but we got no
+ * IPAC_MSGT_ID_RESP message yet, in that scenario we did not
+ * allocate the ipa_bts_conn structure. */
+ if (ipbc == NULL)
+ break;
ipbc->oml_conn = NULL;
bsc_conn = ipbc->bsc_oml_conn;
/* close the connection to the BSC */
--
1.7.2.3
hello
i tried to switch back one of my bs-11 from multidrop to star ("./bs11_config bport0-star" after disconnecting) but i always get a NACK. whats wrong there?
T.
hello,
as you know, my 2nd bs-11(bts1) in multidrop konfig is bootstrapped as bts/trx(1/1) although only trx0 is used on bts1, and the 2nd bs-11 isnt seen by any handy. the debugp message says "abis_nm.c: 518 OC=Radio-Carrier(02) INST=(00,01,ff) STATE CHG: TRX 1 does not exist unknown object class" of course does trx1 not exist, since there are only trx0s, bts0 uses trx0 and bts1 as well (its trx0, of course)
what does that mean?
T.
Hi all!
We haven't had a functioning commit log mailinglist since mid-2009, when the
revision control system of openbsc was converted from svn to git.
Especially today, with the large number of separate git repositories on
http://cgit.osmocom.org/, it is very easy to loose track of what other
developers are doing.
The new mailing list 'osmocom-commitlog' (renamed from openbsc-commits) will
receive updates from all repositories on git.osmocom.org.
Feel free to subscribe at
http://lists.osmocom.org/mailman/listinfo/osmocom-commitlog
Cheers,
Harald
--
- Harald Welte <laforge(a)gnumonks.org> http://laforge.gnumonks.org/
============================================================================
"Privacy in residential applications is a desirable marketing option."
(ETSI EN 300 175-7 Ch. A6)
Hi all!
We're sorry to report that there are some problems with our git repositories at
the moment, resulting in 'early EOF / index-pack failed' messages when users
are attempting to clone one of our repositories.
The problem has been thoroughly analyzed, but even after many hours of
analysis, no solution has been found yet. Your best chance is to keep re-trying
the clone, in 5-10% of the cases it will succeed.
More technical details are available at this posting to the git mailing list:
http://article.gmane.org/gmane.comp.version-control.git/169909
Sorry for the inconvenience.
--
- Harald Welte <laforge(a)gnumonks.org> http://laforge.gnumonks.org/
============================================================================
"Privacy in residential applications is a desirable marketing option."
(ETSI EN 300 175-7 Ch. A6)
hello
my 2 bs-11 are now recognised (bootstrapped) by the program, but
somethings strange: bts 0, the pc-side bts, is recognised as
bts/trx(0/0), but the 2nd one is bts/trx(1/1). strange, because on the
2nd bs-11, trx 0 is activated, not 1.
the screenshot says this too (hope its been attached)
whats wrong here?
T.
--
Wer Rechtschreibfehler findet, darf sie behalten!
Hi Pablo,
On Wed, Mar 23, 2011 at 01:03:42AM +0100, Pablo Neira Ayuso wrote:
> openbsc$ git pull
> remote: Counting objects: 197, done.
> remote: Compressing objects: 100% (79/79), done.
> remote: Total 164 (delta 132), reused 104 (delta 85)
> fatal: The remote end hung up unexpectedly
> fatal: early EOF
> fatal: index-pack failed
>
> Could you check what's wrong with this?
I've wasted about half a day on this yesterday, without any solution.
It seems like it is a race condition in the git server. If you retry your pull
something like 5-10 times, it will succeed once.
I've tried with latest git.git on the server. I've tried strace and ltrace
on git-daemon, which only got me as far as the daemon trying an fstat() after a
close() on fd 1, which will obviosuly fail with -EBADF.
I'm really at a loss here...
Regards,
Harald
--
- Harald Welte <laforge(a)gnumonks.org> http://laforge.gnumonks.org/
============================================================================
"Privacy in residential applications is a desirable marketing option."
(ETSI EN 300 175-7 Ch. A6)