- 20 Jun, 2018 40 commits
-
-
Poul-Henning Kamp authored
Fixes #2685
-
Poul-Henning Kamp authored
Fixes #2647
-
Dridi Boukelmoune authored
-
Nils Goroll authored
-
Nils Goroll authored
-
Dag Haavi Finstad authored
-
Dag Haavi Finstad authored
The h/2 request body VFP would drop data when the input buffer was too small to fit the data in the received frame. With this fix we have the VFP code call us again with a fresh buffer when we run out. Fixes: #2679
-
Nils Goroll authored
Ref #2683
-
Nils Goroll authored
HSH_Unbusy() calls BAN_NewObjCore() not holding the objhead lock, so the ban lurker may race and grab the ban mtx just after the new oc has been inserted, but the busy flag not yet cleared. While it would be correct to call BAN_NewObjCore() with the objhead mtx held, doing so would increase the pressure on the combined ban & objhead mtx. If the ban lurker encounters a busy object, we know that there must be an unbusy in progress and it would be wiser to rather back off in favor of the it. Fixes #2681
-
Federico G. Schwindt authored
The check is added to the builtin logic for now. Fixes #2631.
-
Nils Goroll authored
Conflicts: doc/changes.rst
-
Poul-Henning Kamp authored
-
Nils Goroll authored
or should we use "%" PRIu64 in the format ? reported by Travis via @fgsch
-
Nils Goroll authored
The blind caching is OBE for six years now, see 344a709c
-
Poul-Henning Kamp authored
the schema intended for JSON output. Conflicts: bin/varnishd/cache/cache_backend.c bin/varnishd/cache/cache_backend_probe.c bin/varnishd/cache/cache_director.c bin/varnishtest/tests/d00005.vtc bin/varnishtest/tests/v00014.vtc Only help -j is fixed, the backend changes are left out.
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
-
Nils Goroll authored
to avoid micro-managing memory Follow up cf20e04e
-
Nils Goroll authored
-
Dag Haavi Finstad authored
-
Nils Goroll authored
required after the constification in 8f83087f I have no idea why the umem authors did not make this argument a const, they call strncpy on it
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
-
Nils Goroll authored
A show stopper used to be that the top request's workspace was reserved for delivery during ESI processing, but since beeaa19c the path is clear.
-
Nils Goroll authored
spotted by @lkarsten
-
Nils Goroll authored
Ref #2672
-
Nils Goroll authored
-
Nils Goroll authored
-
Nils Goroll authored
-
Nils Goroll authored
-
Federico G. Schwindt authored
After recent failures discovered by travis-ci it should be safe to have this back on.
-
Federico G. Schwindt authored
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
rely on h2_errcheck() calls anyway.
-
Dag Haavi Finstad authored
-
Poul-Henning Kamp authored
-
Dridi Boukelmoune authored
It complains that &pfh_vut is never NULL...
-
Dridi Boukelmoune authored
For out-of-tree code that may set up multiple VUTs in a single process, we don't want the VUT that initialized a PID file to remove it if it finishes before other VUTs. A global pseudo-VUT is introduced to keep track of the PID file and the error callback in order to use them in the atexit(3) callback. Original test case by Geoff, slightly modified. Closes #2651
-
Dridi Boukelmoune authored
This can happen if a signal is caught after `VUT_Fini` was called, but at this point we should no longer manipulate the VUT being dismantled. Original test case by Geoff, slightly modified. Closes #2650
-