git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5443 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5442 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5441 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5438 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5437 d4fa192b-c00b-0410-8231-f00ffab90ce4

Add a distcheck hook that checks the version field of the RH spec
file, ensuring that we don't accidentially release a version without
at least making sure that number is correct.



git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5436 d4fa192b-c00b-0410-8231-f00ffab90ce4

The DNS director allows Varnish to pick backend based on the Host
header provided by the client, and how it resolves in DNS. A suffix
can be added to make it "internal" (see vcl(7)).
    
This includes commits 5062, 5063, 5064, 5099, 5172, 5173, 5174, 5175



git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5434 d4fa192b-c00b-0410-8231-f00ffab90ce4

Fix up the documentation build system, hopefully making "make
distcheck" work correctly again.

The irrelevant bits of 5427 has been elided.



git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5433 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5430 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5424 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5423 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5400 d4fa192b-c00b-0410-8231-f00ffab90ce4

This makes it compile against 2.1.4 again


git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5396 d4fa192b-c00b-0410-8231-f00ffab90ce4

This is not a merge since this is fixed by a rewrite in trunk


git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5383 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5372 d4fa192b-c00b-0410-8231-f00ffab90ce4

In general the "be liberal in what you accept" is a good thing, but
I am not going to be liberal about anything that could cause us to
return wrong content.

The truly consistent thing to do would be to error the fetch, but that
is a bit on the harsh-side for this specific cornercase, so I have
added a test to ignore the first character of the vary header, if it
is a colon.

Fixes:  #763

5335 is:

Add regression test for #763




git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5369 d4fa192b-c00b-0410-8231-f00ffab90ce4

After considerable thinking, I have come to the conclusion that this
assert is bogus.

There is a window from our lookup until we lock the refcount where
a different thread conceiveably could loose the last reference to
the objhdr, before we get the lock to increase it.

The situation the assert was meant to protect, where we insert 
a new objhdr, is handled in the "oh == noh" clause earlier, were
the weaker "> 0" check is used (rather than "== 1") to cater for
the exact same race.

Fixes: #761
Fixes: #783



git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5368 d4fa192b-c00b-0410-8231-f00ffab90ce4

Fixes #629


git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5367 d4fa192b-c00b-0410-8231-f00ffab90ce4

Fixes #670
Fixes #671



git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5366 d4fa192b-c00b-0410-8231-f00ffab90ce4

If we fail to enqueue a session because all threadpools are filled
we drop the session before we have rendered the clients address.

Report IP and port as "-" rather than passing NULL pointers, which
is not nice under any, and core dump solaris under some circumstances.

Fixes:  #598



git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5365 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5364 d4fa192b-c00b-0410-8231-f00ffab90ce4

Don't try to trim a storage segment we filled completely.

Fixes: #776



git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5363 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5362 d4fa192b-c00b-0410-8231-f00ffab90ce4

Merge r5211: Set the int status when setting the txt header status in VRT_l_[obj|resp|beresp]_status.

Fixes: #769



git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5361 d4fa192b-c00b-0410-8231-f00ffab90ce4

Fixes: #768



git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5360 d4fa192b-c00b-0410-8231-f00ffab90ce4

Not quite sure this is the correct fix, but at least try to access
a default.vcl file we know will exist.


git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5359 d4fa192b-c00b-0410-8231-f00ffab90ce4

Merge r5198: fleshed out the directors a bit more. unsure about if the hash/client dir take the retries parameter


git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5358 d4fa192b-c00b-0410-8231-f00ffab90ce4

The effect of this is that the root object in the binheap could be
wrong and thus hold the expiry thread hostage, even though other
objects were ripe for expiry.

This would show up as inflated obj/objcore/objhdr counts, but have no
other deleritous effects.

Detected, conclusively, by:	sky



git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5356 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5352 d4fa192b-c00b-0410-8231-f00ffab90ce4

Thanks to DocWilco for spotting this.



git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5350 d4fa192b-c00b-0410-8231-f00ffab90ce4

closes #708



git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5349 d4fa192b-c00b-0410-8231-f00ffab90ce4

close #775



git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5348 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5347 d4fa192b-c00b-0410-8231-f00ffab90ce4

Always include <sys/param.h> on FreeBSD so we are sure we have a
definition for __FreeBSD_version.




git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5346 d4fa192b-c00b-0410-8231-f00ffab90ce4

Fixes: #760



git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5345 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5344 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5343 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5342 d4fa192b-c00b-0410-8231-f00ffab90ce4

Add explicit log flushing when backend connections close to keep log
chronology. This is to avoid log lines for the same (reused) FD coming
before the closing connection when load is high.

Fixes #739



git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5341 d4fa192b-c00b-0410-8231-f00ffab90ce4

git-svn-id: http://www.varnish-cache.org/svn/branches/2.1@5340 d4fa192b-c00b-0410-8231-f00ffab90ce4