summaryrefslogtreecommitdiff
path: root/src
AgeCommit message (Collapse)Author
2005-06-27Added lots of "(void)" casts to standard function calls.Philip Hazel
2005-06-27Fix previous attempt to remove "memory" from debug "all".Philip Hazel
2005-06-27Add checks and local definitions for all T_xxx macros, not just a fewPhilip Hazel
special ones.
2005-06-27Removed support for Linux-libc5. It must be well obsolete, and the codePhilip Hazel
in the os-type script was giving problems when libc.so lives in lib64, like on x86_64 Fedora Core.
2005-06-27Add dependencies on spf.h, srs.h, dk.h to spf.o, srs.o, and dk.o in thePhilip Hazel
makefile, as these are not listed in the general $(HDRS) list.
2005-06-27Remove "memory" from -d+all; it's voluminous and rarely useful.Philip Hazel
2005-06-24Fix order of results in spf_result_id_list table conforming to lib version 1.2.xTom Kistner
2005-06-24Fixed putting a NUL char in a libdomainkeys bufferTom Kistner
2005-06-23Fix crash when callout to an address routed by a verify_only router withPhilip Hazel
no transport.
2005-06-22Insert a lot of missing (void) casts.Philip Hazel
2005-06-22Put debug statements on either side of EXIM_DBOPEN() so as to pin downPhilip Hazel
crashes while opening hints files.
2005-06-22Added the long-awaited ${if match_ip condition.Philip Hazel
2005-06-21Implement daemon_startup_{retries,sleep} to control startup retrying.Philip Hazel
2005-06-20Compiling with DomainKeys caused the test suite to show up some bugs:Philip Hazel
(1) The filter variables $n0-$n9 and $sn0-$sn9 were broken. (2) The table of control names was missing some conditional names, leading to the wrong name in some error messages.
2005-06-20Add a whole pile of casts to get rid of compiler warnings in Tom's DKPhilip Hazel
code. I hadn't compiled it before, but a user complained...
2005-06-20Return "100" instead of NULL from prvs function if system clock is wayPhilip Hazel
out.
2005-06-20Fix bug in my conversion of Tom's snprintf to string_vformat. Also makePhilip Hazel
local functions static.
2005-06-20Update Tru64 (OSF1) configuration after some email reports.Philip Hazel
2005-06-17Allow the message "All relevant MX records point to non-existent hosts"Philip Hazel
to be included in bounce messages, as it doesn't seem to be a security risk.
2005-06-17Added macros for time_t as for off_t and used them to rework the code ofPhilip Hazel
Tom's prvs patch to avoid the hardwired use of "%lld" and "long long". Replaced the call to snprintf() with a call to string_vformat().
2005-06-17Tidy up Exim's PCRE Makefile to generate tidier output like the otherPhilip Hazel
Exim Makefiles.
2005-06-17Michael Haardt's patch to update Sieve to RFC3028bis.Philip Hazel
2005-06-17Verify recipient with use_postmaster and random was losingPhilip Hazel
use_postmaster after the random test.
2005-06-17Applied Michael Haardt's HP-UX patch.Philip Hazel
2005-06-17Put in some casts for signed/unsigned character strings that Tom forgotPhilip Hazel
in expand.c. :-)
2005-06-16added "prvs" and "prvscheck" expansion itemsTom Kistner
2005-06-16(1) Applied Matthew Newton's exicyclog patch. (2) Another wish.Philip Hazel
2005-06-16Check the length of off_t at build time and use %ld or %lld to printPhilip Hazel
off_t values as appropriate. Assume that %lld is available if the size of off_t is greater than 4. This involved some modifications to string_vformat() in order to handle %lld.
2005-06-15Fix typo in previous commit.Tony Finch
2005-06-15Some systems need an explicit -ldl in order to support EXPAND_DLFUNCTony Finch
2005-06-15Installed PCRE 6.0 sources, which involved adding a number of files andPhilip Hazel
renaming some others.
2005-06-14Auto_thaw no longer applies to bounce messages.Philip Hazel
2005-06-14Applied Alex Kiernan's patch for the API change in BDB 4.3.Philip Hazel
2005-06-13Fix a typo in an error message for ${extract reported byTony Finch
Jeremy Harris <jgh@wizmail.org> in exim-users Message-ID: <42AC48EB.90406@wizmail.org>
2005-06-10Another ratelimit tweak. I didn't anticipate people using per_cmdTony Finch
ratelimits more than once in the same ACL (e.g. to increase the delay for higher rates) so the code didn't do the right thing in that situation. There's now a per_cmd cache which is reset at the start of each ACL run.
2005-06-10A minor code clean-up and a comment fix for the CSA patch arisingTony Finch
from a bug report by John Horne. The lack of initialization of the found variable in acl_verify_csa() was not a bug but I've fixed it because it's risky. Bug ID 26
2005-06-10Revert accidental bogus checkin, sorryTom Kistner
2005-06-10Fix possible endless loop due to dying spamdTom Kistner
2005-06-07Upgrade appendfile so that it is capable of handling quotas that arePhilip Hazel
greater than 2G.
2005-06-07Fix ${def: bug which treated "0" as false. Also diagnose syntax error ofPhilip Hazel
unexpected characters after the variable name.
2005-06-07Installed exipick 20050604.1 as supplied by John Jetmore.Philip Hazel
2005-06-06Added Chris Webb contributionTom Kistner
2005-06-06Merged patched from Alex Miller: 1) Improve aveserver failure handling 2) ↵Tom Kistner
Write proper mbox delimiter in .eml files. 3) add X-Envelope-Sender and X-Envelope-To headers to .eml file
2005-06-03Added -xls and the ability to specify output files (Frank Heydlauf).Steve Campbell
Use FileHandles for outputing results. Allow any combination of xls, txt, and html output. Fixed display of large numbers with -nvr option Fixed merging of reports with empty tables. Added the -include_original_destination flag Removed tabs and trailing whitespace.
2005-05-31ratelimit fix: omit the limit parameter from the database lookupTony Finch
key, so that minor configuration tweaks do not cause Exim to forget clients' stored rates. The smoothing period and options affect the meaning of the stored state, so are still included in the key.
2005-05-31Minor tweak to a comment.Philip Hazel
2005-05-31Add "fullpostmaster" to check <postmaster> without a domain, if thePhilip Hazel
domained version fails.
2005-05-25Add simple SPF lookup method (armored in EXPERIMENTAL_SPF)Tom Kistner
2005-05-25Fix possible NULL secretTom Kistner
2005-05-25Change the result of the ratelimit ACL condition when clients are sendingTony Finch
exactly at the limit, in order to improve the behaviour of certain edge cases. The result was FAIL (under limit) and it's now OK (over limit).