- 05 Dec, 2016 4 commits
-
-
Martin Blix Grydeland authored
When doing range delivery, make the delivery processing finish early when all of the bytes of the requested range has been delivered. This to avoid waiting around for a slow fetch to finish before handling the next request on the connection. Fixes: #2035
-
Martin Blix Grydeland authored
A delivery processor may want to finish the delivery early. Change the return value from VDP_bytes and the delivery processors into 3 categories. 0 means continue, negative is error and positive is finished. The latching of errors (to help ESI delivery continue while still recording the error) will latch the lowest non-zero value observed. VDP_gunzip is fixed to not unconditionally return errors. Also add some text to explain the behaviour of VDP_bytes and its delivery processors.
-
Poul-Henning Kamp authored
-
Nils Goroll authored
Merges #2152
-
- 04 Dec, 2016 1 commit
-
-
Federico G. Schwindt authored
In particular VSL_Error() and VSM_Error() return a string without the newline now.
-
- 02 Dec, 2016 6 commits
-
-
Poul-Henning Kamp authored
builtin echo doesn't understand -n ? Really ?
-
Nils Goroll authored
-
Federico G. Schwindt authored
This is a hard dep iff the svg files need to be updated.
-
Nils Goroll authored
-
Nils Goroll authored
-
Nils Goroll authored
-
- 01 Dec, 2016 1 commit
-
-
Dridi Boukelmoune authored
The `held == 0` branch in the Lck__AssertHeld function is *never* taken and `cache.h` instructs to only use the macro. Being generated from calling code, the panic message becomes more informative. Panic messages were up until now not very helpful: Assert error in Lck__Assert(), cache/cache_lck.c line 175: Condition(ilck->held) not true.
-
- 30 Nov, 2016 8 commits
-
-
Federico G. Schwindt authored
-
Federico G. Schwindt authored
E.g. and 0 tests failed, 7 tests skipped, 558 tests passed This could use some tlc but will do it for now.
-
Nils Goroll authored
VCC was silently accepting enum values valid for other arguments following in the argument list as well as other identifiers in the argument spec (for example "ENUM"). Consequently, wrong enum values in VCL were not detected at VCC time and passed to vmod functions, which, in the best case, would detect the error (and, if following the varnish good practice, panic on a failed assertion). This is another forgotten case since the enum list was changed to be terminated by \1 in a78efad8: vcc_expr would just loop over the \1 terminator up to the final \0 terminator at the end of the argument spec.
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
It's a loose comparison of socket addresses, not including the port number. But not loose enough to compare IPv4 addresses to v4-mapped IPv6 addresses. Fixes #2142
-
Dridi Boukelmoune authored
Adapted from an original patch from @michbsd: on top of normalizing the host header we now have a regular expressions that only match the right domains and potential sub-domains. Closes #2140
-
Federico G. Schwindt authored
-
Federico G. Schwindt authored
-
- 29 Nov, 2016 3 commits
-
-
Dridi Boukelmoune authored
-
Nils Goroll authored
-
Poul-Henning Kamp authored
-
- 28 Nov, 2016 8 commits
-
-
Poul-Henning Kamp authored
This should stabilize c00049
-
Nils Goroll authored
Previously, unless HAVE_TCP_KEEP was defined equally for varnishd and an API client (vmod), the latter used a wrong struct declaration and, consequently, struct params members were accessed at a wrong location. This could have adverse effects from reading bogus values to overwriting wrong parameters or other memory. For consistency, we keep uninplemented parameters also in the cli and rst output with an appropriate description. Setting them over the cli is not an error, but has no effect.
-
Dridi Boukelmoune authored
-
Federico G. Schwindt authored
-
Federico G. Schwindt authored
-
Federico G. Schwindt authored
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
is no longer around.
-
- 27 Nov, 2016 2 commits
-
-
Federico G. Schwindt authored
-
Federico G. Schwindt authored
Fixes #2148.
-
- 26 Nov, 2016 7 commits
-
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
correct course of action is.
-
Dridi Boukelmoune authored
Spotted by VTEST
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
They live as private macros in varnish.m4, which is now included in the Varnish build during the autogen phase. This way it is now possible to dogfood on macros we distribute downstream. Considering how simple they are, they could be made public macros for the convenience of VMOD or VUT writers. We don't need to have the LIBS-saving-dance in configure.ac for every single dependency.
-
Poul-Henning Kamp authored
-