Commit Graph

296 Commits

Author SHA1 Message Date
aCaB
14294b605d bb#2400 2011-01-19 12:17:34 +01:00
Tomasz Kojm
3c91587c8c fix typo 2010-12-16 14:47:42 +01:00
Tomasz Kojm
2e136dcd0d freshclam, clamd: handle EINTR during waitpid() (bb#2422) 2010-12-16 14:16:11 +01:00
Tomasz Kojm
b898103236 clamd: don't report file hashes to clients (bb#2409) 2010-12-09 13:50:42 +01:00
Török Edvin
a347ecc9c1 Fix RLIMIT_DATA (bb #1941).
Set limit only when >2GB (not when equal), and set it to 2GB, not 2MB.
Based on patch from Andreas Longwitz.
2010-11-23 13:14:58 +02:00
aCaB
bbce04496c fix win32 build 2010-11-04 16:42:08 +01:00
Tomasz Kojm
bb08962d85 fix previous commit 2010-11-03 14:30:52 +01:00
Tomasz Kojm
49f8de2207 clamd: add new commands DETSTATS and DETSTATSCLEAR (part of bb#2312) 2010-11-03 13:41:06 +01:00
Török Edvin
769f37a6f6 Default off, you can turn on via 'DevLiblog'.
This also replaces the cli_*_stats variants with a callback for stats,
so that clamd can call the cl_*_callback variants instead, and pass the filename
as context.
2010-11-02 12:26:33 +02:00
Tomasz Kojm
6bee45b345 clamd: add new option OLE2BlockMacros (requested by Mike) 2010-10-29 19:04:23 +02:00
Tomasz Kojm
bd433b5ad3 clamd: be more verbose about config errors (bb#2252) 2010-10-18 20:02:53 +02:00
aCaB
f238189223 bb#2286 2010-10-14 16:06:43 +02:00
Török Edvin
8b94e259c6 Only enable RLIMIT_DATA warning on *BSD. 2010-09-22 10:38:57 +03:00
Török Edvin
9011cf1ef5 Limit RLIMIT_DATA to 2GB on 32-bit processes (bb #1941).
With RLIMIT_DATA > 2GB a 32-bit FreeBSD 7.3 allows ~120M memory to be allocated
via mmap before failing.
With RLIMIT_DATA set to 2GB it allows ~890M memory.
So check the RLIMIT_DATA and set it just below 2G.

You can also just do 'ulimit -d 2097151' to get the effects of this patch.
2010-08-31 15:51:07 +03:00
Török Edvin
e67790c658 Add clamd handling. 2010-07-29 14:07:00 +03:00
Török Edvin
cdb3cbe089 Treat MULTISCAN as CONTSCAN on non-directories (bb #1869).
This is part 1. Part 2 will be for clamdscan to send IDSESSION for
-m <file1> <file2>, but that is for 0.97.
2010-07-26 12:57:34 +03:00
Tomasz Kojm
edbba730b3 clamd: add ExtendedDetectionInfo (bb#1228, #1626) 2010-07-06 19:46:55 +02:00
Tomasz Kojm
b4b1a32734 clamd/clamd.c: be more verbose when daemonize() fails (bb#2062) 2010-06-15 16:14:07 +02:00
Török Edvin
cbb9db1941 Fix some error path leaks (bb #1990).
This doesn't actually fix bb #1990 (which is about high memory usage on RHEL6),
but fixes some leaks found while investigating that bug.
2010-05-28 18:47:21 +03:00
Tomasz Kojm
bca0b67921 tools: make sure the libclamav version is high enough (bb#2013) 2010-05-10 17:05:16 +02:00
aCaB
3d7547cf6a add collect hashes options to clamd and clamscan 2010-05-07 23:33:26 +02:00
Török Edvin
909407e70b Regenerate build system after clamconf change. 2010-04-19 17:14:57 +03:00
Török Edvin
e4a0f2c94f fix compiler warnings (bb #1872, bb #1934, bb #1935) 2010-04-13 16:19:47 +03:00
Török Edvin
d46d32b723 Fix typo
(thanks Tomasz).
2010-03-24 19:14:54 +02:00
aCaB
4fa9557e8c bb#1892 2010-03-23 00:09:00 +01:00
Török Edvin
b63681a52b Introduce BytecodeTimeout. 2010-03-22 17:16:07 +02:00
aCaB
ced85d637c bb#1889 2010-03-21 11:44:07 +01:00
aCaB
59a61e9c92 fix typo #2 2010-03-20 22:52:35 +01:00
aCaB
f4aef52ce2 fix typo 2010-03-20 22:51:16 +01:00
aCaB
f642d53c7f win32: use non tls (s)rand(), allow nonblocking sockets 2010-03-20 21:08:13 +01:00
Török Edvin
606fd6d0f1 Allow "None" bytecode security setting only in --enable-debug mode. 2010-03-10 14:30:31 +02:00
aCaB
49e5f6580e bb#1851 2010-03-06 04:00:14 +01:00
aCaB
88ce192ef5 bb#1812 2010-02-06 16:03:59 +01:00
aCaB
e394c51305 bb#1789 - part two 2010-02-04 22:01:05 +01:00
aCaB
cf5fc21912 fchxxx won't work on a socket 2010-02-04 17:56:10 +01:00
aCaB
ab56ca922e bb#1789 - part one 2010-02-04 17:32:42 +01:00
Török Edvin
1514794c49 make STATS a non-bulk command (bb #1732).
This is needed to allow clamdtop to watch in realtime what is happening in
clamd, otherwise STATS would always lag behind 50 other multiscan commands.
2010-02-01 14:38:36 +02:00
Török Edvin
6a5ec1f967 Don't allow bulk requests (multiscan, idsession) to fill more than 50% of the queue. (bb #1732)
This makes clamd responsive to simple (version,contscan,etc.) requests
even during multiscan.
Previously these would get stuck behind a ~100 item queue, and the 1:4 ratio
of executing these commands wasn't working, since the commands weren't in the queue
in the first place.
2010-02-01 14:07:48 +02:00
aCaB
3d7d59c104 clamd-win32 - works! 2010-01-30 04:20:22 +01:00
aCaB
288b945f57 clamd-win32 - where is my mind? 2010-01-30 03:17:52 +01:00
aCaB
80f4bd69c2 fix accept 2010-01-29 20:10:45 +01:00
aCaB
4155214d9a fix accept 2010-01-29 19:34:02 +01:00
aCaB
5eca45b5e7 clamd-win32 - WIP#3 2010-01-29 18:57:50 +01:00
aCaB
1c4d9a22fa clamd-win32: WIP#2 2010-01-29 14:44:55 +01:00
aCaB
3b074c7878 clamd-win32 - WIP 2010-01-28 23:54:20 +01:00
Török Edvin
e2fbea32c1 fix distcheck warnings. 2010-01-27 15:56:13 +02:00
Török Edvin
c0afc51729 Don't use C++ comments in C code. 2010-01-27 12:07:08 +02:00
Török Edvin
27b59717b8 previous commit fixes bb #1770, not #1700! 2010-01-27 11:59:51 +02:00
Török Edvin
80301d0cd6 Fix clamd hang when number of multiscan commands == MaxThreads (bb #1700).
Each multiscan command needs 1 control thread, and 1 (possibly shared) scanner
thread.
If the number of multiscan control threads would be equal to MaxThreads no
progress could be made, since each one would be waiting for 1 free scanner
thread (but none can be spawned since maxthreads was already reached).
As long as there is 1 thread free for scanning, we can have N-1 multiscan
commands active, with N MaxThreads.
2010-01-27 11:56:10 +02:00
Török Edvin
cb9a7b448d Improve error message when no ancillary data is received (bb #1700).
On Linux show a message that SELinux/AppArmor might be blocking the file
descriptor.
2010-01-27 10:54:35 +02:00