]> git.ipfire.org Git - thirdparty/squid.git/log
thirdparty/squid.git
26 years agoavoid NPR if helper is NULL
wessels [Sun, 18 Oct 1998 14:29:09 +0000 (14:29 +0000)] 
avoid NPR if helper is NULL

26 years agouse EBIT_CLR,EBIT_TEST for comm_select fd_mask bits
wessels [Sun, 18 Oct 1998 14:26:40 +0000 (14:26 +0000)] 
use EBIT_CLR,EBIT_TEST for comm_select fd_mask bits

26 years agoread queued bytes on linux before closing a client socket to prevent
wessels [Sun, 18 Oct 1998 14:10:05 +0000 (14:10 +0000)] 
read queued bytes on linux before closing a client socket to prevent
RST packets going to the browser

26 years agoshutdown/reconfigure bugs
wessels [Sun, 18 Oct 1998 14:08:58 +0000 (14:08 +0000)] 
shutdown/reconfigure bugs

26 years ago- added missing case counters to refreshCheck()
rousskov [Sun, 18 Oct 1998 07:30:09 +0000 (07:30 +0000)] 
- added missing case counters to refreshCheck()
request_reload2ims_stale, request_reload_stale,
min_age_override_exp_fresh, min_age_override_lmt_fresh
- display only non-zero stats
- hack to double check that all cases are counted
  ("total" line will not display 100% if not)

26 years agouse array syntax instead of pointer syntax
wessels [Sat, 17 Oct 1998 23:03:42 +0000 (23:03 +0000)] 
use array syntax instead of pointer syntax

26 years agochange clock skew debug threshold to 2 hours
wessels [Sat, 17 Oct 1998 11:38:45 +0000 (11:38 +0000)] 
change clock skew debug threshold to 2 hours

26 years agoreport clock skew
wessels [Sat, 17 Oct 1998 11:37:48 +0000 (11:37 +0000)] 
report clock skew

26 years agoupdate
wessels [Sat, 17 Oct 1998 11:08:46 +0000 (11:08 +0000)] 
update

26 years agoclear HAVE_GNUMALLOC_H for --enable-dlmalloc
wessels [Sat, 17 Oct 1998 11:08:16 +0000 (11:08 +0000)] 
clear HAVE_GNUMALLOC_H for --enable-dlmalloc

26 years ago- updated current Cache Digests version to "4" (required version remains "3")
rousskov [Sat, 17 Oct 1998 10:53:40 +0000 (10:53 +0000)] 
- updated current Cache Digests version to "4" (required version remains "3")
  the change will help to track down caches that use old refresh algorithm

26 years ago- fixed typos in refreshCountsStats
rousskov [Sat, 17 Oct 1998 10:41:48 +0000 (10:41 +0000)] 
- fixed typos in refreshCountsStats

26 years ago- eliminated refreshWhen() which was out-of-sync with refreshCheck()
rousskov [Sat, 17 Oct 1998 10:34:08 +0000 (10:34 +0000)] 
- eliminated refreshWhen() which was out-of-sync with refreshCheck()
  potentially causing under-utilized cache digests
- maintain refreshCheck statistics on per-protocol basis so we
  can tell why ICP or Digests return too many misses, etc.

26 years agofixed helper restart bugs
wessels [Sat, 17 Oct 1998 02:02:45 +0000 (02:02 +0000)] 
fixed helper restart bugs

26 years agorearrange debugging
wessels [Sat, 17 Oct 1998 01:20:17 +0000 (01:20 +0000)] 
rearrange debugging

26 years agowe used to CLOSE persistent connections if the ENTRY_BAD_LENGTH bit
wessels [Sat, 17 Oct 1998 01:19:29 +0000 (01:19 +0000)] 
we used to CLOSE persistent connections if the ENTRY_BAD_LENGTH bit
was set (because we didn't get exactly content-length bytes).  But
when we read TOO MUCH we can still handle this gracefully.  Instead of
closing on != condition, now we close on < condition.

26 years agofix debugging message
wessels [Sat, 17 Oct 1998 01:18:16 +0000 (01:18 +0000)] 
fix debugging message

26 years agoclientKeepaliveNextRequest() used to assume (assert) that conn->chr
wessels [Fri, 16 Oct 1998 05:44:37 +0000 (05:44 +0000)] 
clientKeepaliveNextRequest() used to assume (assert) that conn->chr
was either NULL, or if conn->chr was set, then conn->chr->entry must
also be set.  But the cilent request could be blocked on ACL or
redirector operations, in which case conn->chr->entry is NULL.

26 years agosupport 301: 302: redirector hacks
wessels [Fri, 16 Oct 1998 05:40:04 +0000 (05:40 +0000)] 
support 301: 302: redirector hacks

26 years agolittle StatHist cleanup and debugging
wessels [Thu, 15 Oct 1998 23:34:32 +0000 (23:34 +0000)] 
little StatHist cleanup and debugging

26 years agoupdate
wessels [Thu, 15 Oct 1998 06:47:03 +0000 (06:47 +0000)] 
update

26 years agoallow whitespace in request URI's.
wessels [Thu, 15 Oct 1998 03:11:55 +0000 (03:11 +0000)] 
allow whitespace in request URI's.

26 years agotypo
wessels [Thu, 15 Oct 1998 02:14:04 +0000 (02:14 +0000)] 
typo

26 years agotypos and whitespace
wessels [Thu, 15 Oct 1998 02:12:14 +0000 (02:12 +0000)] 
typos and whitespace

26 years agoupdate
wessels [Wed, 14 Oct 1998 22:30:27 +0000 (22:30 +0000)] 
update

26 years agoFMR bug
wessels [Wed, 14 Oct 1998 22:09:30 +0000 (22:09 +0000)] 
FMR bug

26 years agocheck for working inet_ntoa()
wessels [Wed, 14 Oct 1998 09:04:21 +0000 (09:04 +0000)] 
check for working inet_ntoa()

26 years agofix potential buffer overrun and negative array index bugs
wessels [Wed, 14 Oct 1998 08:40:25 +0000 (08:40 +0000)] 
fix potential buffer overrun and negative array index bugs

26 years agowrong version in welcome comment
wessels [Wed, 14 Oct 1998 05:42:03 +0000 (05:42 +0000)] 
wrong version in welcome comment

26 years agoFrom: Q <q@fan.net.au>
wessels [Wed, 14 Oct 1998 05:39:08 +0000 (05:39 +0000)] 
From:    Q <q@fan.net.au>
Here is a patch to make squid2-p1 work properly when it's both an
ipf-transparent proxy and a local http-accelerator at the same time. It
also closes a potential DoS window.

26 years agosome compilers cant grok variable names same as type names
wessels [Wed, 14 Oct 1998 05:33:29 +0000 (05:33 +0000)] 
some compilers cant grok variable names same as type names

26 years agotypo
wessels [Wed, 14 Oct 1998 05:15:01 +0000 (05:15 +0000)] 
typo

26 years agoupdate
wessels [Wed, 14 Oct 1998 02:39:01 +0000 (02:39 +0000)] 
update

26 years ago - Changed storeClientCopy2() so that it keeps sending the remainder
wessels [Wed, 14 Oct 1998 02:38:42 +0000 (02:38 +0000)] 
    - Changed storeClientCopy2() so that it keeps sending the remainder
          of a STORE_ABORTED request, instead of cutting off the client as
          soon as the object becomes aborted.

26 years agoreplaced errorCon() with call to fwdFail() for httpTimeout.
wessels [Wed, 14 Oct 1998 02:37:19 +0000 (02:37 +0000)] 
replaced errorCon() with call to fwdFail() for httpTimeout.

26 years agomake Unused ICP version warning debug level 1
wessels [Wed, 14 Oct 1998 02:35:55 +0000 (02:35 +0000)] 
make Unused ICP version warning debug level 1

26 years agoupdate
wessels [Tue, 13 Oct 1998 23:40:36 +0000 (23:40 +0000)] 
update

26 years agoneed to release objects on swapout errors when we close the swapout FD
wessels [Tue, 13 Oct 1998 10:46:30 +0000 (10:46 +0000)] 
need to release objects on swapout errors when we close the swapout FD

26 years agoupdate
wessels [Tue, 13 Oct 1998 04:02:40 +0000 (04:02 +0000)] 
update

26 years agogetAnyParent -- last resort for when all parents down
wessels [Tue, 13 Oct 1998 03:40:57 +0000 (03:40 +0000)] 
getAnyParent -- last resort for when all parents down

26 years ago2.0 branch
wessels [Tue, 13 Oct 1998 03:39:06 +0000 (03:39 +0000)] 
2.0 branch

26 years agoupdate
wessels [Tue, 13 Oct 1998 03:25:11 +0000 (03:25 +0000)] 
update

26 years agoSGI fixes for rusage stuff
wessels [Tue, 13 Oct 1998 02:26:41 +0000 (02:26 +0000)] 
SGI fixes for rusage stuff

26 years ago2.0 branch merge
wessels [Mon, 12 Oct 1998 05:19:47 +0000 (05:19 +0000)] 
2.0 branch merge

26 years agoupdate
wessels [Sat, 10 Oct 1998 20:58:46 +0000 (20:58 +0000)] 
update

26 years agoMoved many common dns/redirect/authenticate functions into helper.c
wessels [Sat, 10 Oct 1998 20:57:36 +0000 (20:57 +0000)] 
Moved many common dns/redirect/authenticate functions into helper.c

26 years ago2.0 branch merge
wessels [Fri, 9 Oct 1998 23:52:54 +0000 (23:52 +0000)] 
2.0 branch merge

26 years agoChanged the policy of storeReleaseRequest() which used to require
wessels [Fri, 9 Oct 1998 23:46:35 +0000 (23:46 +0000)] 
Changed the policy of storeReleaseRequest() which used to require
the entry to be locked.  This was to prevent RELEASE_REQUEST entries
from getting stranded.

But there were some places where we manually set the RELEASE_REQUEST
bits for maybe-unlocked objects.  This sucks because we also need
to clear ENTRY_CACHABLE if we set RELEASE_REQUEST.

Now anyone can call storeReleaseRequest().

26 years ago2_0 branch merge
wessels [Fri, 9 Oct 1998 02:10:09 +0000 (02:10 +0000)] 
2_0 branch merge

26 years agofix for unsinged size_t
wessels [Thu, 8 Oct 1998 09:17:30 +0000 (09:17 +0000)] 
fix for unsinged size_t

26 years ago2.0 branch merge
wessels [Thu, 8 Oct 1998 08:40:01 +0000 (08:40 +0000)] 
2.0 branch merge

26 years agoupdate
wessels [Thu, 8 Oct 1998 01:34:51 +0000 (01:34 +0000)] 
update

26 years agoadded drand48
wessels [Thu, 8 Oct 1998 01:34:36 +0000 (01:34 +0000)] 
added drand48

26 years agoadding
wessels [Thu, 8 Oct 1998 01:34:01 +0000 (01:34 +0000)] 
adding

26 years agoadd 5/60 minute byte/request hit ratios to info page
wessels [Wed, 7 Oct 1998 00:32:41 +0000 (00:32 +0000)] 
add 5/60 minute byte/request hit ratios to info page

26 years agoDEC CC -O2 only works on OSF4 and later?
wessels [Sat, 3 Oct 1998 10:06:26 +0000 (10:06 +0000)] 
DEC CC -O2 only works on OSF4 and later?

26 years agomerge SQUID_2_0 branch
wessels [Sat, 3 Oct 1998 09:56:47 +0000 (09:56 +0000)] 
merge SQUID_2_0 branch

26 years agounused var
wessels [Sat, 3 Oct 1998 09:39:46 +0000 (09:39 +0000)] 
unused var

26 years agoadd -O2 for DEC CC
wessels [Sat, 3 Oct 1998 09:36:53 +0000 (09:36 +0000)] 
add -O2 for DEC CC

26 years agouse fd_mask vs int
wessels [Fri, 2 Oct 1998 11:02:29 +0000 (11:02 +0000)] 
use fd_mask vs int
try a faster loop for commDeferRead() checks

26 years agocheck for type fd_mask
wessels [Fri, 2 Oct 1998 10:36:22 +0000 (10:36 +0000)] 
check for type fd_mask

26 years agoupdate
wessels [Fri, 2 Oct 1998 04:58:09 +0000 (04:58 +0000)] 
update

26 years ago2.1.PRE1
wessels [Fri, 2 Oct 1998 04:57:11 +0000 (04:57 +0000)] 
2.1.PRE1

26 years agoavoid scanning loops in comm_select(); from Jeff Mogul patches
wessels [Fri, 2 Oct 1998 04:28:26 +0000 (04:28 +0000)] 
avoid scanning loops in comm_select(); from Jeff Mogul patches

26 years agomake delayPoolsUpdate() an event
wessels [Wed, 30 Sep 1998 10:34:06 +0000 (10:34 +0000)] 
make delayPoolsUpdate() an event

26 years agoAdded offline mode
wessels [Wed, 30 Sep 1998 08:53:17 +0000 (08:53 +0000)] 
Added offline mode

26 years ago- better comments for httpHdrRangeWillBeComplex + HIT case
rousskov [Wed, 30 Sep 1998 05:53:27 +0000 (05:53 +0000)] 
- better comments for httpHdrRangeWillBeComplex + HIT case

26 years agoAdded SNMP changes
glenn [Wed, 30 Sep 1998 03:17:33 +0000 (03:17 +0000)] 
Added SNMP changes

26 years agorearrange stuff so snprintf() doesn't call vsnprintf() and we don't
wessels [Tue, 29 Sep 1998 22:47:53 +0000 (22:47 +0000)] 
rearrange stuff so snprintf() doesn't call vsnprintf() and we don't
get prototype warnings

26 years agogindent
wessels [Tue, 29 Sep 1998 22:33:39 +0000 (22:33 +0000)] 
gindent

26 years agodebug levels
wessels [Tue, 29 Sep 1998 22:32:50 +0000 (22:32 +0000)] 
debug levels

26 years agoneighborType() needs to be global
wessels [Tue, 29 Sep 1998 07:32:58 +0000 (07:32 +0000)] 
neighborType() needs to be global

26 years agonetdbGetClosestParent() needs to check the peer is really a parent
wessels [Sat, 26 Sep 1998 00:31:27 +0000 (00:31 +0000)] 
netdbGetClosestParent() needs to check the peer is really a parent

26 years agoincrease debug
wessels [Fri, 25 Sep 1998 12:21:56 +0000 (12:21 +0000)] 
increase debug

26 years agogindent
wessels [Fri, 25 Sep 1998 12:21:30 +0000 (12:21 +0000)] 
gindent

26 years ago2.0.PRE1
wessels [Fri, 25 Sep 1998 12:03:41 +0000 (12:03 +0000)] 
2.0.PRE1

26 years agoupdate
wessels [Fri, 25 Sep 1998 12:02:57 +0000 (12:02 +0000)] 
update

26 years agoupdate
wessels [Fri, 25 Sep 1998 11:58:15 +0000 (11:58 +0000)] 
update

26 years agoprobable off-by-one bug (Joe Ramey)
wessels [Fri, 25 Sep 1998 11:35:31 +0000 (11:35 +0000)] 
probable off-by-one bug (Joe Ramey)

26 years agoAdded Connection/Meter
glenn [Fri, 25 Sep 1998 04:41:06 +0000 (04:41 +0000)] 
Added Connection/Meter

26 years ago- synced with Range support changes
rousskov [Fri, 25 Sep 1998 02:47:30 +0000 (02:47 +0000)] 
- synced with Range support changes
- added X-Forwarded-For TODO entry

26 years ago- added httpHdrRangeWillBeComplex
rousskov [Fri, 25 Sep 1998 02:42:09 +0000 (02:42 +0000)] 
- added httpHdrRangeWillBeComplex

26 years ago- changed the way Range requests are handled:
rousskov [Fri, 25 Sep 1998 02:41:19 +0000 (02:41 +0000)] 
- changed the way Range requests are handled:
- do not "advertise" our ability to process ranges at all
- on hits, handle simple ranges and forward complex ones
- on misses, fetch the whole document for simple ranges
  and forward range request for complex ranges
  The change is supposed to decrease the number of cases when clients
  such as Adobe acrobat reader get confused when we send a "200" response
  instead of "206" (because we cannot handle complex ranges, even for hits)
  Note: Support for complex ranges requires storage of partial objects.

26 years ago- minor fix
rousskov [Fri, 25 Sep 1998 02:35:11 +0000 (02:35 +0000)] 
- minor fix

26 years ago- fixed(?) cbdata handling by the peer_digest module
rousskov [Fri, 25 Sep 1998 02:29:13 +0000 (02:29 +0000)] 
- fixed(?) cbdata handling by the peer_digest module
  there should be less coredumps on reconfigure now
  however, the digest structure might leak when peer is gone -- TBF

26 years agoupdate
wessels [Fri, 25 Sep 1998 02:26:54 +0000 (02:26 +0000)] 
update

26 years ago- added support for undocumented Request-Range header
rousskov [Fri, 25 Sep 1998 02:20:20 +0000 (02:20 +0000)] 
- added support for undocumented Request-Range header
  the header is used by Netscape (with Adobe plugin working)
  Request-Range is assumed to match Range header so only
  the former is actually used when both are provided.

26 years ago- fixed a bug when one of the merged ranges could get lost
rousskov [Fri, 25 Sep 1998 02:16:24 +0000 (02:16 +0000)] 
- fixed a bug when one of the merged ranges could get lost
- commented out merging of overlapping ranges
  because it might break efficieancy of some clients
- added httpHdrRangeWillBeComplex()

26 years agoRemoved static reference to squid used global variable for name instead
glenn [Thu, 24 Sep 1998 05:15:53 +0000 (05:15 +0000)] 
Removed static reference to squid used global variable for name instead

26 years ago- added missing "bytes" to Content-Range reply header
rousskov [Thu, 24 Sep 1998 03:52:20 +0000 (03:52 +0000)] 
- added missing "bytes" to Content-Range reply header

26 years agoMoved the squid OID and added and removed a number of MIB objects
glenn [Thu, 24 Sep 1998 03:31:29 +0000 (03:31 +0000)] 
Moved the squid OID and added and removed a number of MIB objects

26 years agoRemoved SNMP code.
glenn [Thu, 24 Sep 1998 03:29:29 +0000 (03:29 +0000)] 
Removed SNMP code.

26 years agoAdded and removed code so only returns basic stats
glenn [Thu, 24 Sep 1998 03:29:07 +0000 (03:29 +0000)] 
Added and removed code so only returns basic stats

26 years agoMinor modifications, clean up the static situation
glenn [Thu, 24 Sep 1998 03:26:06 +0000 (03:26 +0000)] 
Minor modifications, clean up the static situation

26 years agoModified to remove and add some groups.
glenn [Thu, 24 Sep 1998 03:25:11 +0000 (03:25 +0000)] 
Modified to remove and add some groups.

26 years agoModified to no longer return the whole cache, just statistics
glenn [Thu, 24 Sep 1998 03:24:37 +0000 (03:24 +0000)] 
Modified to no longer return the whole cache, just statistics

26 years agoAdded code to return basic statistics
glenn [Thu, 24 Sep 1998 03:23:33 +0000 (03:23 +0000)] 
Added code to return basic statistics

26 years agoMinor modifications, removing statics.
glenn [Thu, 24 Sep 1998 03:22:50 +0000 (03:22 +0000)] 
Minor modifications, removing statics.

26 years agoModified to use SMIv2
glenn [Thu, 24 Sep 1998 03:21:08 +0000 (03:21 +0000)] 
Modified to use SMIv2

26 years agoget the order right for quick_abort
wessels [Thu, 24 Sep 1998 03:21:05 +0000 (03:21 +0000)] 
get the order right for quick_abort