January 27, 2003
View 1540fad105

parser changed to make body to point to body (w/o LWS), have a len encompassing the complete hf (name.s..CRLF) and not to put zero- terminators after header field names and cseq

Jiri Kuthan authored on 27/01/2003 21:19:48
January 21, 2003
View ff9799524a

preliminary support for FIFO/t_uac added -- compiles, but not completed yet

Jiri Kuthan authored on 21/01/2003 13:53:48
January 19, 2003
View 4d93f1459c

bunch of changes related to ability to process original request in reply_route: - textops: append_urihf(p,s) appends a hf with uri between p and s - core: supporting functions for duplicating a lump list from shmem to pkgmem (which allows further processing with pkgmem-oriented actions) - tm: reply_routes now creates a temporary, mostly pkg-mem-ed replica of shmem-ed request for additional processing

Jiri Kuthan authored on 19/01/2003 01:37:45
January 6, 2003
View 8ba9c56733

bug_fix: on_negative_reply now sets msg id and transaction context (previously, it didn't do so, which resulted in a lookup that consumed CPU and even worse, refcounted -- transaction that entered reply_route was never released)

Jiri Kuthan authored on 06/01/2003 12:48:56
View eaf10d3693

t_unsafe_reply deprecated

Jiri Kuthan authored on 06/01/2003 03:55:41
View 0afe6b05bd

on_negative_reply faked msg now copied from shmem msg (as opposed to zero-ing) -- more "read-only" actions (exec in particular) will work from reply_route; caution, rw actions may append some pkg stuff to msg, which will possibly be never released (shmem is released in a single block)

Jiri Kuthan authored on 06/01/2003 03:04:47
December 3, 2002
View 2c60a4c3d0

bug_fix: ambiguous use of TOTAG* defines fixed (report courtesy of Bogdan)

Jiri Kuthan authored on 03/12/2002 20:24:27
October 21, 2002
View 17045ca6d9

to_tags now generated all the time from TM

Jiri Kuthan authored on 21/10/2002 21:24:43
October 18, 2002
View 2c4c8a7c2b

change: to-tags no more compared in ACK lookups -- that eliminates an unlikely race conditions (reading transactional to-tag on ACK lookup whereas writing a new to tag from 200/OK) and helps UACs who screw up to-tags

Jiri Kuthan authored on 18/10/2002 02:32:14
September 25, 2002
View 4cb713e086

changed log level of the message "no resolved dst yet"

Jiri Kuthan authored on 25/09/2002 21:36:28
View c0f9ab1f8e

bug_fixed: statistics (they were previously updated in an unprotected way; hash entry locks were mistakenly used to protect access to global stats structures; now each process maintains its own stats, and fifo_stats sums all of them)

Jiri Kuthan authored on 25/09/2002 20:23:42
September 24, 2002
View e74ab20c21

- bug_fix: missed calls reporting fixed (was bound to a wrong callback previously); - t_reply.c little a bit privatized

Jiri Kuthan authored on 24/09/2002 01:15:38
September 19, 2002
View 04813d1439

fixed (blindly, hopefully correctly) signed/unsigned incompatibilites

Jiri Kuthan authored on 19/09/2002 21:43:21
View 7dd0b34273

GPLization banner introduced to *.[hc] files

Jiri Kuthan authored on 19/09/2002 12:23:52
September 18, 2002
View b286525e77

bug_fix: after return from on_negative_reply, UAS reply code is checked first prior to proceeding to introducing new branches

Jiri Kuthan authored on 18/09/2002 21:41:25