- 25 Nov, 2017 1 commit
-
-
Poul-Henning Kamp authored
-
- 23 Nov, 2017 1 commit
-
-
Dag Haavi Finstad authored
This test case should not rely on first_byte_timeout/between_bytes_timeout.
-
- 21 Nov, 2017 1 commit
-
-
Nils Goroll authored
-
- 20 Nov, 2017 5 commits
-
-
Nils Goroll authored
This issue is hidden by the fix in the previous commits, so we cannot test for it directly any more. Applying this patch to fa388439 exposes the "Constructor not found". Fixes #2498
-
Poul-Henning Kamp authored
-
Nils Goroll authored
When we are looking for ! SYM_NONE and find ! SYM_NONE, do not return the symbol we found. The semantics of (kind != SYM_NONE && sym->kind != kind) are inclusive because the next conditional is not true for this case. Ref: 9d81a71a Fixes: #2497
-
Dag Haavi Finstad authored
The second time around, we force a fresh connection. The VCL user may choose to do 'return (retry);' in vcl_backend_error{} if further attempts are deemed warranted. Fixes: #2135
-
Dag Haavi Finstad authored
Passing a nonzero value will force a fresh backend connection.
-
- 19 Nov, 2017 5 commits
-
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
that we now actually *do* get the "child panicked" exit code.
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
-
- 18 Nov, 2017 8 commits
-
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
it uses a 'double' which only has approx 52 bits of mantissa, which is marginal already for PetaBytes. Test various edge-cases.
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
-
- 17 Nov, 2017 1 commit
-
-
Pål Hermunn Johansen authored
The counter cache_hit_grace counts the number of grace hits. To be precise, it counts the number of times lookup returns an expired object, but vcl_hit is called and decides to return(deliver). Every time cache_hit_grace is incremented, cache_hit is also incremented (so this commit does not change the cache_hit counter).
-
- 16 Nov, 2017 6 commits
-
-
Dag Haavi Finstad authored
-
Martin Blix Grydeland authored
I'm guessing this is due to rounding. All test cases involving file stevedore has a minimum 10m file in the test, was silly to attempt a smaller one in this test. Fixes: #2496
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
of sourcefile names over the project.
-
- 15 Nov, 2017 8 commits
-
-
Nils Goroll authored
Fixes #2448
-
Nils Goroll authored
for example, mgt_shm_atexit could race in two processes, one in VSMW_Destroy and the other in system("rm -rf") Closes #2484
-
Nils Goroll authored
... now that I've found the layer 8 bug Thank you to @bsdphk Ref #2484
-
Nils Goroll authored
I had left in superfluous stuff from bug analysis. Ref: #2319
-
Nils Goroll authored
Creating "backend synthetic" content was impossible whenever a fetch had already started, unless storage was assigned (again) explicitly from VCL. Fixes #2494
-
Nils Goroll authored
The oa_present member acts as a filter at the level above the stevedore to check set attributes efficiently. We missed to clear it when recreating freeing an abandoned backend fetch object in vcl_backend_error. Fixes #2319
-
Poul-Henning Kamp authored
sanely, in this case by failing VCL initialization. Fixes #2036
-
Poul-Henning Kamp authored
order uncertainty.
-
- 14 Nov, 2017 4 commits
-
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
-
Poul-Henning Kamp authored
(Replaces cache_backend_cfg.c::backend_find)
-