It seems like glibc's getaddrinfo is entirely happy to accept 70000 as
a valid port, so use something that's not a valid service name so make
check still works.


git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3293 d4fa192b-c00b-0410-8231-f00ffab90ce4

later.  Use EOF-encoding for sessions where the request is lower
HTTP protocol versions.



git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3292 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3291 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3290 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3289 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3288 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3287 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3286 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3285 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3284 d4fa192b-c00b-0410-8231-f00ffab90ce4

unexpected result codes.



git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3283 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3282 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3281 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3280 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3279 d4fa192b-c00b-0410-8231-f00ffab90ce4

IPv4 & IPv6 address in /etc/hosts' localhost entry.

Use "127.0.0.1" intead.

Send me a telegram when IPv6 matters.



git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3278 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3277 d4fa192b-c00b-0410-8231-f00ffab90ce4

tests a given number of times:

	varnishtest -n 100 tests/b*.vtc




git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3276 d4fa192b-c00b-0410-8231-f00ffab90ce4

Suggested by:	 "chen xiaoyong"



git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3275 d4fa192b-c00b-0410-8231-f00ffab90ce4

- Synced rpm changelog from fedora. There was some strange items that
  had to be sorted


git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3274 d4fa192b-c00b-0410-8231-f00ffab90ce4

* Wed Oct 08 2008 Ingvar Hagelund <ingvar@linpro.no> - 2.0-0.11.rc1
- 2.0-rc1 released. New upstream sources
- Added a patch for test a00008, from r3269
- Removed condrestart in postscript at upgrade. We don't want that.



git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3273 d4fa192b-c00b-0410-8231-f00ffab90ce4

(or might not) be the root cause of #346:

The delay function would access the toplevel log element from client/server
contexts with no logging.

Fix by passing the log element to the subcommands.

This race would only affect testcases using delay or sema inside client
or server threads.

Revert test a00008.vtc to use only one sema, using two just moved the
race condition.

This also has the side effect of moving "sema" related messages to
the tread that carries out the "sema" operation.

Various other cleanups now that I had my fingers in the code anyway:
memory leaks plugged, stylistic nits etc.



git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3272 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3271 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3270 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3269 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3268 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3266 d4fa192b-c00b-0410-8231-f00ffab90ce4

Partially addresses #280


git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3265 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3264 d4fa192b-c00b-0410-8231-f00ffab90ce4

error was unimplemented in vcl_fetch.  Implement it and add a test
case.


git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3263 d4fa192b-c00b-0410-8231-f00ffab90ce4

Fixes #339



git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3262 d4fa192b-c00b-0410-8231-f00ffab90ce4

Hopefully, this will mitigate pile-ups somewhat and prevent us from
running out of file descriptors, at least as quickly.

See #330.


git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3261 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3260 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3259 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3258 d4fa192b-c00b-0410-8231-f00ffab90ce4

If the error references the first token and there is no preceding
newline we would core dump.



git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3257 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3256 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3255 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3254 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/trunk/varnish-cache@3253 d4fa192b-c00b-0410-8231-f00ffab90ce4