Author: Henrik Nordstrom <henrik@henriknordstrom.net>
http_port allow-direct option to allow direct forwarding in accelerator mode
normally direct forwarding is disabled in accelerator mode unless overridden
by always_direct, to avoid unintentional security loops. But there is setups
where it makes sense to not have this restriction as this has effects on
peer selection as well.
Alex Rousskov [Sat, 11 Jul 2009 05:39:44 +0000 (23:39 -0600)]
Limit X-Forwarded-For growth.
X-Forwarded-For growth leads to String size limit assertions and probably
other problems.
We now replace huge XFF values with a string "error", warn the admin the
first 100 times, and hope that something will stop the loop (if it is a
loop). TODO: we should probably deny requests with huge XFF.
To make growth-associated problems visible during forwarding loops, the
loop breaking code must be disabled (no Via) or not applicable (direct
forwarding) and request_header_max_size has to be raised or disabled.
The X-Forwarded-For header value may also grow too large for reasons
unrelated to forwarding loops.
This change also prevents most cases of pointless computation of the
original X-Forwarded-For value list. That computation can be quite
expensive.
Shuffle cachemgr.cgi.8 into tools/ with cachemgr.cgi code
To perform this shuffle the SUBSTITUTE make code pushed into a sub-include
at doc/manuals/Substitute.am for shared use.
Also clean up the manual titles to reflect the binary they apply to
rather than the generic Squid project release name.
The use of generic @SYSCONFDIR@ and specific filenames allows the
translation strings to be more specific, readable for translators and
reduces Makefile complexity at once. squid.conf etc are widely mentioned
and used enough that allowing for alternative names at this point seems
meaningless.
Attempts to translate, and installs the error pages content if present.
Both srcdir and builddir are scanned used since the file location varies
depending on whether a source bundle or VCS is used to install.
Alex Rousskov [Thu, 2 Jul 2009 16:36:36 +0000 (10:36 -0600)]
Truncate too-long HTTP response bodies to match their Content-Length header.
Sometimes a broken server sends more than Content-Length bytes in the
response. For example, a 302 redirect message with "Content-Length: 0" header
may include an HTML body. Squid used to send "everything" it read to the
client, even if it read more than the Content-Length bytes. That may have
helped in some cases, but we should be more conservative when dealing
with broken servers to combat message smuggling attacks and other bad
side-effects for clients.
We now do not forward more than the advertised content length and declare the
connection with a broken server non-persistent.
Chunked responses (that HTTP/1.0 Squid should not receive and that must not
have a Content-Length header) are not truncated because RFC 2616 says we
MUST ignore their Content-Length header.
TODO: Do not cache the truncated entry and purge the cached version, if any.
Alex Rousskov [Thu, 2 Jul 2009 15:24:24 +0000 (09:24 -0600)]
Break forwarding loops for "transparent" or "intercept" http_ports.
Squid detected forwarding loops in most configurations, but broke
them (using a customizable HTTP_FORBIDDEN response) only when working as
an accelerator. Squid now breaks loops when working as a transparent
proxy as well.
A persistent loop is going to be broken anyway, when the Via and
X-Forwarded-For headers exceed header size limit, but that wastes a lot of
resources and may also crash misconfigured Squids.
TODO: Consider breaking all loops, regardless of the http_port options.
TODO: Consider adding a specific and/or configurable error page for this case
instead of using hard-coded ACCESS_DENIED.
Amos Jeffries [Sun, 28 Jun 2009 08:28:20 +0000 (20:28 +1200)]
Bug 422: RFC 2616 Date header requirements
The Date: header appears to already be implemented on all generated
pages and ICAP processed pages.
This tests and enforces Date: on all other outgoing replies as required.
I'm not certain this is the right place, it appears to be post-caching.
The RFC indicates the Date: should be enforced pre-caching. But was
unable to find a place of input cloning/processing after initial parse.
The storeEntry timestamp is used to estimate correct receiving date.
Amos Jeffries [Sun, 28 Jun 2009 08:11:27 +0000 (20:11 +1200)]
Simpler test for PURGE method
Uses a tri-state setting on enable_purge and acl parsing to
detect PURGE method addition/removal instead of a complicated ACL
creation test post-configure.
This removes the annoying false errors about temp ACL and some minor
speed up in all actions that parse squid.conf.
Alex Rousskov [Sat, 27 Jun 2009 01:14:58 +0000 (19:14 -0600)]
Author: Philip Allison <philip.allison@smoothwall.net>
Bug 2614 fix: Potential loss of adapted body data from eCAP adapters
It was possible for Squid to stop reading buffered adapted body data before it
has all been sent to the browser.
Squid treated a call to noteAbContentDone by an adapter as a signal to stop
consuming and sending adapted body data. The correct behaviour is to use
noteAbContentDone to record the fact that the adapter has stopped producing
new adapted body data, but continue to consume and send data until all
buffered ab content is consumed and sent (i.e., abContent returns an empty
Area).
Amos Jeffries [Fri, 26 Jun 2009 13:32:27 +0000 (01:32 +1200)]
Language Updates: Add aliases from live traffic info
Taking a scan of the last 98 days traffic and locating the country-code
Accept-Language headers used in that traffic to refer to the existing
languages gives a subset we can alias to further improve the coverage.
Amos Jeffries [Fri, 26 Jun 2009 07:02:45 +0000 (19:02 +1200)]
Bug 2674: Remove limit on HTTP headers read.
Headers may be accumulated over more than one read. It does not make
sense to limit the internal copy of the accumulated read buffer to 64KB.
Reverts the internal read buffer to MemBuf defaults. This may cause
issues where headers are of unbounded size. But those are expected to be
caught by the header parser.
Amos Jeffries [Sun, 21 Jun 2009 06:32:16 +0000 (18:32 +1200)]
Fix alias linker dist/install
make requires ';' after a SHELL command apparently.
Make alias-link.sh handle case where the DESTDIR is non-existent.
This occurs on some distro packaging systems (ie using langpack as a
separate package may not install errors).
Amos Jeffries [Sat, 20 Jun 2009 06:50:37 +0000 (18:50 +1200)]
Correct Licensing Credits
Several of the licenses mentioned in the CREDITS file are not relevant
to Squid-3.1 code any more. Several license disclaimers were found to be
missing.
Thanks to the Debian Project for identifying these incorrect entries.
Amos Jeffries [Wed, 17 Jun 2009 14:20:11 +0000 (02:20 +1200)]
Language alias linker/installer/upgrade scripts
alias-link.sh
This is a script set designed to be called via make/Makefile and setup
language codes for those languages which it would be impractical to
bundle duplicate translated files for.
Relies on local environment tools to be detected by automake.
make install
- now also calls generation of aliases after existing install.
Provided in file aliases.
make upgrade
- cleans out legacy files from pre-3.1 and replaces with symlinks
to the new upgraded language codes.
Provided in file alias-upgrade.
NP: this is a destructive process and must be manually run.
Bundle aliasing scripts and Makefile to use them with the langpack.
Amos Jeffries [Sun, 14 Jun 2009 06:18:11 +0000 (18:18 +1200)]
Update .pot and rename errpages.pot
The old name dictionary.pot was picked semi-randomly based on my old
experience in other translation arenas. It does not fit with the l18n
requirements of Squid very well. This renaming adjusts to make things
a bit clearer.
Amos Jeffries [Fri, 12 Jun 2009 08:46:29 +0000 (20:46 +1200)]
Bug 2395: FTP auth errors not displayed
Round 2 for this bug. Now we handle missing auth as an expected result
rather than a failure. FTP operations are now well tested and this patch
does not affect code shared with other components.
Side-effect is that browser authentication popups now appear when the
FTP server needs authentication. This has been a long missed event.
The root cause of the issue is not found so other subsequent errors in
FTP sub-protocol still silently lost due to the same issue.
Amos Jeffries [Fri, 5 Jun 2009 23:13:17 +0000 (11:13 +1200)]
Make debug_options rotate=N option back-compatible.
This ties older config which do not set rotate=N to still use the value
of logfile_rotate. If set rotate=N overrides logfile_rotate and can be
any value 0 or up.