- 21 May, 2020 1 commit
-
-
Nils Goroll authored
and, in particular, matching client.ip against a vcl Ref #3334
-
- 19 May, 2020 5 commits
-
-
Martin Tzvetanov Grigorov authored
There is a problem with fakeroot-sysv on Ubuntu 20.04 aarch64: Install Build-Depends packages... + mk-build-deps --install debian/control + yes semop(1): encountered an error: Function not implemented Error in the build process: exit status 1
-
Dridi Boukelmoune authored
That is, in the context of varnish.m4, because this changes existing behavior in a way that breaks backward compatibility. Partial revert of 134d0633.
-
Dridi Boukelmoune authored
-
Nils Goroll authored
This is to support generated vcc files with out-of-tree vmod builds.
-
Nils Goroll authored
our python sources use spaces
-
- 18 May, 2020 1 commit
-
-
Dridi Boukelmoune authored
-
- 15 May, 2020 3 commits
-
-
Martin Tzvetanov Grigorov authored
-
Martin Tzvetanov Grigorov authored
-
Martin Tzvetanov Grigorov authored
In order to build aarch64 packages it relies on the machine executors feature based on QEMU. Since the same scripts were needed to build packages for different architectures they were extracted in standalone shell scripts. Alpine Linux packages are now built in addition to RPM and Deb packages but considering how long emulation takes for aarch64 packages, packaging was moved to a nightly workflow. Some changes happened in the pkg-varnish-cache repository to accomodate the CircleCI reorganization, mainly around dependency management. More information in the CircleCI README file. Closes #3313 Reviewed-by:
Guillaume Quintard <guillaume@varnish-software.com> Signed-off-by:
Dridi Boukelmoune <dridi.boukelmoune@gmail.com>
-
- 14 May, 2020 3 commits
-
-
Jordan Christiansen authored
-
Jordan Christiansen authored
-
Jordan Christiansen authored
-
- 13 May, 2020 4 commits
-
-
Nils Goroll authored
Ref #3311
-
Nils Goroll authored
See comments in #3311, but done slightly differently
-
Poul-Henning Kamp authored
This is an alternative implementation of #2680
-
Martin Blix Grydeland authored
-
- 12 May, 2020 4 commits
-
-
Nils Goroll authored
Ref #3312
-
Nils Goroll authored
Closes #3312 (probably)
-
Dridi Boukelmoune authored
-
Dridi Boukelmoune authored
-
- 11 May, 2020 1 commit
-
-
Dag Haavi Finstad authored
Fixes: #3266
-
- 10 May, 2020 3 commits
-
-
Federico G. Schwindt authored
-
Federico G. Schwindt authored
-
Federico G. Schwindt authored
-
- 07 May, 2020 2 commits
-
-
Nils Goroll authored
vcc_compile.h is not generated EXTRA_DIST is not necessarily generated
-
Nils Goroll authored
- missing dependencies in each of the both places of invocation - vrt_obj.h is generated, so it cannot be a dependency at the same time notice with out-of-tree parallel build
-
- 05 May, 2020 7 commits
-
-
Martin Blix Grydeland authored
-
Martin Blix Grydeland authored
Copy paste the text from beresp.status to resp.status about the special handling of status codes.
-
Martin Blix Grydeland authored
-
Martin Blix Grydeland authored
Some status codes are illegal and will cause VRT_fail() when executed as normal set instructions in VCL. But this test is bypassed when status is set as a side effect of a `return (synth(code))` statement. This patch applies the same rules as when executing a set-instruction to the return(synth()) handling. Fixes second part of: #3301
-
Martin Blix Grydeland authored
The compound statement doing return() in VCL may have failed for some reason, for example running out of workspace while putting together the reason field. The fail handling would be overwritten by the forced handling at the end of the generated code for a return statement. This patch adds a test for the failure condition before forcing the handling to the result of the return statement.
-
Martin Blix Grydeland authored
In VCL we allow status codes larger then 999, and do modulo 1000 when sending it to the client. But the code to apply the modulo was missing in H/2. This issue was uncovered and reported by @Francois-v3. Fixes first case of: #3301
-
Dridi Boukelmoune authored
-
- 04 May, 2020 2 commits
-
-
Guillaume Quintard authored
-
Nils Goroll authored
-
- 02 May, 2020 1 commit
-
-
Dridi Boukelmoune authored
It's not a gadget for production use.
-
- 30 Apr, 2020 1 commit
-
-
Guillaume Quintard authored
-
- 28 Apr, 2020 2 commits
-
-
Jordan Christiansen authored
-
Nils Goroll authored
When making a connection a "real" backend (VBE), we checked the probed health state and did not take into account the administrative health state as set with `varnishadm backend.set_health ... {healthy,sick}`. Our documentation was not particularly explicit on this aspect either, but the administrative states `sick` and `healthy` made no sense if `auto` semantics was implied always. Also, the semantics were implicitly documented for `backend.list`. Implementation note: The relevant change is to call `VRT_Healthy()`, which does check the administrative health, in place of checking `(struct backend *)->sick` in `vbe_dir_getfd()`. As a `VRT_CTX` is required by `VRT_Healthy()`, we change the arguments of `vbe_dir_getfd()` accordingly: The busyobj can now be taken from the ctx, but the worker argument differs for pipe mode vs. fetch, so we preserve it as an explicit argument. A test for overriding a probed backend as healthy has been added to c00048.vtc, which requires a second probe to hit server s1 and fail. This is timing sensitive, so I hope that the backend probe interval of 5 seconds is long enough for all our test environments. If not, we probably need to make it longer or add another vtc. Fixes #3299
-