Commit Graph

4063 Commits (97f51c10111621c020d268e092aead07046ff314)
 

Author SHA1 Message Date
Victor Julien aa449d51ca Stream: use per thread ssn pool
Use per thread pools to store and retrieve SSN's from. Uses PoolThread
API.

Remove max-sessions setting. Pools are set to unlimited, but TCP memcap
limits the amount of sessions.

The prealloc_session settings now applies to each thread, so lowered the
default from 32k to 2k.
12 years ago
Victor Julien b6af6cb241 pool: add error msgs and improve memory layout 12 years ago
Victor Julien 5b9ef94f34 pool: add api for per thread pools
This API is a wrapper around the regular pools where the thread pools
are arrays of locks+pools.
12 years ago
Victor Julien 016d03bdaf pool: add error msgs and improve memory layout 12 years ago
Victor Julien 46af6b7e0f Add a per threadvars thread local thread id, that starts at 0 and increments for each thread. 12 years ago
Victor Julien b3b554c269 Coverity 1038959: DNS mpm might use initialized variable 12 years ago
Anoop Saldanha fba95e9125 Remove mpm ctxs in the wrong direction.
A lot of http mpm ctxs have now been removed as a result of this.
12 years ago
Anoop Saldanha 3c2ddf04c1 Update mpm init ctx to not accept the final cuda_rc_module argument.
It was a part of our older architecture and is no longer used.
12 years ago
Victor Julien 33818c0272 DNS: fix CUDA build 12 years ago
Victor Julien be7e6cdd7a DNS: fix warning when debug is not enabled 12 years ago
Victor Julien 16edbfd4ea DNS: disable logging by default 12 years ago
Victor Julien 571b8ac186 DNS: add support for per TX decoder events. 12 years ago
Victor Julien 9dc04d9fab app layer: add support for per TX decoder events 12 years ago
Victor Julien 72e35efbc6 Reset app layer events when we start inspecting a new TX 12 years ago
Victor Julien 28a6c1d9f8 DNS: add test for app layer event match 12 years ago
Victor Julien 6645620c03 Merge SIG_FLAG_MPM_HTTP and SIG_FLAG_MPM_DNS into SIG_FLAG_MPM_APPLAYER, do the same for the _NEG variant. 12 years ago
Victor Julien 43ba5a677e DNS: enable mpm/fast_pattern support for dns_query 12 years ago
Victor Julien 4817e1305f DNS: add /F modifier to pcre to inspect DNS query name 12 years ago
Victor Julien 4f20f72f4d DNS: add event rules file 12 years ago
Victor Julien e567e12230 DNS: add unittests for UDP and TCP for dns_query keyword 12 years ago
Victor Julien f10dd603ff DNS: adding dns_request content modifier 12 years ago
Victor Julien 6674f4892c DNS: add per tx internal id
Add per TX id. Rename transaction_cnt to transaction_max (id) and increment it on tx creation.
12 years ago
Victor Julien 59780ca770 Hacks to enable alert dns even though we have dnstcp and dnsudp parsers. Needs proper solution later. 12 years ago
Victor Julien 8e01cba85d DNS TCP and UDP parser and DNS response logger 12 years ago
Eric Leblond 4521de2dfd Use PACKET_* macro instead of UPDATE
Setting the ACTION_DROP flag can be done via PACKET_DROP instead
of using PACKET_UPDATE_ACTION.
12 years ago
Eric Leblond c0c59fbd17 decode: factorize macro code
PACKET_* are now wrapper to the newly introduced PACKET_SET_ACTION
macro.
12 years ago
Eric Leblond 3f107fa130 decode: Packet action start with PACKET
Rename all Packet action macro to have them prefixed by PACKET.
12 years ago
Anoop Saldanha 3304c91c91 Don't let geoip match on pseudo packets. 12 years ago
Anoop Saldanha 1fb4aae993 Coverity 1038523: Fix using cuda buffer slice that has been returned to the pool. 12 years ago
Victor Julien 51cdd464a6 stream: detect keep-alive and keep-alive ACK 12 years ago
Victor Julien 03c3ff5632 stream: fix typo in function name 12 years ago
Eric Leblond 4c6595f437 Coverity 1038106: fix FP out-of-bond access
A cast during the reading of a configuration variable was invalid
because a 16 bit integer was cast to a 32 bit integer. The called
function is only setting the pointer value to 1 or 0 so there is
no real issue there.
12 years ago
Eric Leblond 7df156ef50 Coverity: 1038139 suppress sanity check
The sanity check was really useless as the NULL value is checked in
the code flow.
12 years ago
Eric Leblond 75cd1f6096 Coverity 1038515: check function return
This is more cosmetic than useful but it is cleaner anyway.
12 years ago
Anoop Saldanha b4e750068f Cuda make distcheck fix for cuda-ptxdump.h 12 years ago
Anoop Saldanha 7d46d59cdd Coverity 1038522: fix memset inside cuda code. Wrong size specified to memset. 12 years ago
Victor Julien d6fcd07a31 Coverity 1038085: remove 'default' statement in SCErrorToString. This way a warning will be given if an error is defined w/o updating this function. 12 years ago
Victor Julien e2444f0ed5 Coverity 1038092 & 1038093: remove dead code 12 years ago
Victor Julien 0902c7f3aa Coverity 1038518: fix wrong error check 12 years ago
Victor Julien db1dad8cc6 Coverity 1038124: memory leak on 'seq' keyword parsing failure 12 years ago
Victor Julien b2e962da03 Coverity 1038123: memory leak on 'flowint' keyword parsing failure 12 years ago
Victor Julien 5c5b2f98dd Coverity 1038116 & 1038117: memory leaks on 'app-layer-event' keyword parsing failure 12 years ago
Victor Julien 778851626c Coverity 1038115: memory leak on 'ack' keyword parsing failure 12 years ago
Victor Julien 98dbf3e62c Coverity 1038113: possibly out of bounds read 12 years ago
Anoop Saldanha 602c91ed41 Minor cosmetic changes to the cuda code.
Moved a couple of functions to more cuda relevant files;
Re-structured some data types.
12 years ago
Anoop Saldanha c9f076def3 Modified CudaBufferCullCompletedSlices.
Allow readers specify max size of data they want to read.
12 years ago
Anoop Saldanha 70cb4d30eb Add a usleep to CudaBuffer culling process. Would lead to a situation where the thread wouldn't care to yield to others." 12 years ago
Anoop Saldanha 17c763f855 Version 1 of AC Cuda. 12 years ago
Anoop Saldanha 2de59fc235 Version 1 of CudaBuffer API. Introduced to buffer data to the gpu.
This version allows async writes to a buffer by threads.  Allows only
sequential reads though.
12 years ago
Anoop Saldanha 557cab3dc9 We call packet and stream mpm as late as possible now. Won't affect the working of the engine.
The rationale behind this is, if we have pkt buffered to the gpu, we'd want
to delay processing the pkt as much as possible.
12 years ago