Commit graph

686 commits

Author SHA1 Message Date
Matthew Holt
c8da8ca673
Update readme 2020-05-07 13:01:33 -06:00
Jose Donizetti
43fba378d6
docs: Fix command.Func documentation (#3371) 2020-05-07 09:31:58 -06:00
Matthew Holt
cd9317e5df
httpcaddyfile: Fix route ordering bug
https://caddy.community/t/cant-get-simple-alias-to-work/7911/8?u=matt

This removes an optimization where we amortized path matcher decoding.
The decoded matchers were index by... position... which obviously
changes during sorting. Duh.

Anyway, sorting is sliiightly slower now but the Caddyfile is not
really CPU-sensitive, so this is fine.
2020-05-06 19:41:37 -06:00
Matthew Holt
8dbc5f70a5
Update dependencies and get rid of placeholder hacks in CA code
With the latest commit on smallstep/certificates, placeholders in config
are no longer needed.
2020-05-06 16:02:21 -06:00
Francis Lavoie
07c6076ea0
ci: Add release tagged event triggers to sister repos (#3321) 2020-05-06 16:42:55 -04:00
Matthew Holt
28ab0bfb13
core: Support loading modules from [][]json.RawMessage fields 2020-05-06 13:18:56 -06:00
Matthew Holt
1c17e6c6bb
reverseproxy: Allow using TLS for port 80 upstreams (see #3361)
An upstream like https://localhost:80 is still forbidden, but an addr of
localhost:80 can be used while explicitly enabling TLS as an override;
we just don't allow the implicit behavior to be ambiguous.
2020-05-06 12:37:44 -06:00
Karol Będkowski
b814c0af9c
tls/client auth: verify first certificates in client request (#3344)
When client certificate is enabled Caddy check only last certificate from
request. When this cert is not in list of trusted leaf certificates,
connection is rejected. According to RFC TLS1.x the sender's certificate
must come first in the list.  Each following certificate must directly
certify the one preceding it.

This patch fix this problem - first certificate is checked instead of last.
2020-05-06 10:07:13 -06:00
Dave Henderson
9e5d9e2530
ci: Add linux-armv5 builds (#3356)
Signed-off-by: Dave Henderson <dhenderson@gmail.com>
2020-05-05 19:13:56 -06:00
Dave Henderson
9408dacc27
Fixing goreleaser syntax error (#3355)
Signed-off-by: Dave Henderson <dhenderson@gmail.com>
2020-05-06 00:50:30 +00:00
Daniel Black
12cfc19487
ci: add s390x and ppc64le builds for linux (#3325) 2020-05-05 12:41:08 -06:00
Francis Lavoie
afecd90a6c
reverseproxy: Add tls_server_name option to Caddyfile (#3322) 2020-05-05 12:39:39 -06:00
Matt Holt
2f59467ac3
httpcaddyfile: Only append TLS conn policy if it's non-empty (#3319)
This can lead to nicer, smaller JSON output for Caddyfiles like this:

	a {
		tls internal
	}
	b {
		tls foo@bar.com
	}

i.e. where the tls directive only configures automation policies, and
is merely meant to enable TLS on a server block (if it wasn't implied).
This helps keeps implicit config implicit.

Needs a little more testing to ensure it doesn't break anything
important.
2020-05-05 12:37:52 -06:00
Matt Holt
184e8e9f71
pki: Embedded ACME server (#3198)
* pki: Initial commit of embedded ACME server (#3021)

* reverseproxy: Support auto-managed TLS client certificates (#3021)

* A little cleanup after today's review session
2020-05-05 12:35:32 -06:00
Matt Holt
1e8c9764df
file_server: Accept files args in one-liner of Caddyfile matcher (#3298)
Previously, matching by trying files other than the actual path of the
URI was:

    file {
        try_files <files...>
    }

Now, the same can be done in one line:

    file <files...>

As before, an empty file matcher:

    file

still matches if the request URI exists as a file in the site root.
2020-05-05 12:34:58 -06:00
Matt Holt
41c7bd27b4
httpserver: Add experimental H2C support (#3289)
* reverse_proxy: Initial attempt at H2C transport/client support (#3218)

I have not tested this yet

* Experimentally enabling H2C server support (closes #3227)

See also #3218

I have not tested this

* reverseproxy: Clean up H2C transport a bit

* caddyhttp: Update godoc for h2c server; clarify experimental status

* caddyhttp: Fix trailers when recording responses (fixes #3236)

* caddyhttp: Tweak h2c config settings and docs
2020-05-05 12:33:21 -06:00
Francis Lavoie
96d6d277a4
caddyconfig: Don't start comments in middle of tokens (#3267)
* caddyconfig: Only parse # as start of comment if preceded by space

* caddyconfig: Simplify # logic using len(val), add a test
2020-05-05 12:32:12 -06:00
Francis Lavoie
26e559662d
httpcaddyfile: Support single-line matchers (#3263)
* httpcaddyfile: Support single-line matchers

* httpcaddyfile: Add single-line matcher test

* httpcaddyfile: Add a matcher syntax adapt test
2020-05-05 12:29:21 -06:00
Matt Holt
52305618df
caddyfile: Support backticks as quotes (closes #2591) (#3242) 2020-05-05 12:27:49 -06:00
Mohammed Al Sahaf
e051e119d1
ci: add tests on s390x and ppc64le (#3328)
* ci: add tests on s390x and ppc64le

* ci: use Travis as CI for ppc64le and s390x

* ci: cache Go builds on Travis

* ci: avoid Travis duplicate builds
2020-05-02 17:24:54 -06:00
Matthew Holt
8e42661060
caddytls: Finish upgrading to libdns DNS providers for ACME challenges
Until we finish the migration to the new acme library, we have to bring
the solver type in-house. It's small and temporary.
2020-05-02 17:23:36 -06:00
Matthew Holt
86a4f2c9f4
caddytls: Fix namespace tls.dns -> dns.providers
Coulda sworn I did this already but I think I messed up my git commands
2020-05-02 16:28:10 -06:00
Matthew Holt
a507a5bbc7 reverseproxy: Remove circuitbreaker module (see #3331)
Moving to https://github.com/caddyserver/circuitbreaker

Nobody was using it anyway -- it works well, but something got fumbled
in a refactoring *months* ago. Turns out that we forgot the interface
guards AND botched a method name (my bad) - Ok() should have been OK().
So it would always have thrown a runtime panic if it tried to be loaded.
The module itself works well, but obviously nobody used it because
nobody reported the error. Fixing this while we move it to the new repo.

Removing this removes the last Bazaar/Launchpad dependency (I think).
2020-05-01 19:47:46 -06:00
Mark Sargent
d0770dbbb3
expose caddytest timeouts (#3329) 2020-05-02 10:24:35 +12:00
Matthew Holt
a77bd1d887
httpcaddyfile: Update tls parsing for DNS providers 2020-05-01 10:41:08 -06:00
Matthew Holt
bca610fbde
httpcaddyfile: Minor fixes to parsing storage options 2020-05-01 09:34:32 -06:00
Matthew Holt
1fa8c185a8
go.mod: Remove DNSProviderMaker interface; update to lego 3.6 2020-04-30 18:17:39 -06:00
Matthew Holt
a1796c2f14
caddytls: Adjust DNS challenge structure; clarify some docs 2020-04-30 16:15:20 -06:00
Matthew Holt
f931c26f68
caddyhttp: Better duration logging
Also un-nest all the error handling, that was unnecessary indentation
2020-04-28 15:38:45 -06:00
Matt Holt
10db57027d
caddyhttp: General improvements to access logging (#3301)
* httpcaddyfile: Exclude access logs written to files from default log

Even though any logs can just be ignored, most users don't seem to like
configuring an access log to go to a file only to have it doubly appear
in the default log.

Related to:
- #3294
- https://caddy.community/t/v2-logging-format/7642/4?u=matt
- https://caddy.community/t/caddyfile-questions/7651/3?u=matt

* caddyhttp: General improvements to access log controls (fixes #3310)

* caddyhttp: Move log config nil check higher

* Rename LoggerName -> DefaultLoggerName
2020-04-28 08:32:04 -06:00
Matthew Holt
c11d0e47a3
cmd: Clean up, simplify reverse proxy command; fix some edge cases
Now we take advantage of the address parsing capabilities of the HTTP
caddyfile.
2020-04-27 15:53:38 -06:00
Matthew Holt
9770ce7c9f Minor comment fix 2020-04-27 14:49:27 -06:00
Francis Lavoie
5ae1a5617c
caddyhttp: Add split_path to file matcher (used by php_fastcgi) (#3302)
* matcher: Add `split_path` option to file matcher; used in php_fastcgi

* matcher: Skip try_files split if not the final part of the filename

* matcher: Add MatchFile tests

* matcher: Clarify SplitPath godoc
2020-04-27 14:46:46 -06:00
Matthew Holt
83c85c53f5
caddyhttp: Fix listener overlap detection on Linux
Sigh, apparently Linux is incapable of distinguishing host interfaces
in socket addresses, even though it works fine on Mac. I suppose we just
have to assume that any listeners with the same port are the same
address, completely ignoring the host interface on Linux... oh well.
2020-04-26 22:28:49 -06:00
Francis Lavoie
768383a610
ci: Enable GoReleaser .deb support (#3309)
* ci: Enable GoReleaser .deb support

* ci: Test .deb build

* ci: Fix typo

* ci: Turn off snapshot (breaks due to go mod edit)

* ci: Force the tag to rc3 for now

* ci: Let's try to publish the .debs

* ci: Attempt to enable build cache, rebuild after fixed line endings

* ci: Fix yml dupe ID issue, add caddy-api.service

* ci: Split cache keys between files so they're separate

* ci: Fix bindir

* ci: Update the script files

* ci: Retrigger

* ci: Push to gemfury

* ci: Use loop, fix bad env var

* ci: Retrigger

* ci: Try to force blank password?

* ci: Check if the token is actually present

* ci: Cleanup, remove debugging stuff

* ci: Remove useless comment
2020-04-26 20:20:14 -06:00
Mark Sargent
570d84f7d3
refactored caddytest helpers (#3285)
* refactored caddytest helpers
* added cookie jar support. Added support for more http verbs
2020-04-27 13:23:46 +12:00
Christoffer Andersson
a6761153cb
Fix misspelling in onDemandAskRequest error (#3308) 2020-04-25 10:34:56 -06:00
Matthew Holt
02845bc9fd
docs: Improve template documentation slightly; use const, not literal 2020-04-24 21:05:09 -06:00
Matthew Holt
97ed9e111d
httpcaddyfile: Add nil check to prevent panic, fix validation logic
Panic would happen if an automation policy was specified in a singular
server block that had no hostnames in its address. Definitely an edge
case.

Fixed a bug related to checking for server blocks with a host-less key
that tried to make an automation policy. Previously if you had only two
server blocks like ":443" and another one at ":80", the one at ":443"
could not create a TLS automation policy because it thought it would
interfere with TLS automation for the block at ":80", but obviously that
key doesn't enable TLS because it is on the HTTP port. So now we are a
little smarter and count only non-HTTP-empty-hostname keys.

Also fixed a bug so that a key like "https://:1234" is sure to have TLS
enabled by giving it a TLS connection policy. (Relaxed conditions
slightly; the previous conditions were too strict, requiring there to be
a TLS conn policy already or a default SNI to be non-empty.)

Also clarified a comment thanks to feedback from @Mohammed90
2020-04-24 20:57:51 -06:00
Matthew Holt
100d19e3af
dangit, of course I would bork my git commit 2020-04-24 17:48:33 -06:00
Matthew Holt
ebf07f853b
caddyhttp: Fix auto redirects for catch-all HTTPS sites
Prior logic was not setting up redirects for the case when domain names
are not known, but the server still clearly has TLS enabled.
2020-04-24 17:36:52 -06:00
Matthew Holt
1b061815b2
reverseproxy: Don't forget to provision embedded headers handler
https://caddy.community/t/set-cookie-manipulation-in-reverse-proxy/7666?u=matt
2020-04-22 19:57:06 -06:00
Matthew Holt
026937fab5
caddyhttp: Fix trailers when recording responses (fixes #3236) 2020-04-22 11:10:13 -06:00
Matthew Holt
295604d6df
httpcaddyfile: Why was this code repeated?? 2020-04-22 09:20:39 -06:00
Francis Lavoie
bacf50a59e
caddyhttp: Fix common_log format's user ID placeholder (#3300) 2020-04-22 09:05:26 -06:00
westwin
da8686c4b9
reverseproxy: always set req.URL.Host with upstream (#3297) 2020-04-21 20:34:00 -06:00
Matthew Holt
e3a8f72f1c
docs: Minor improvements 2020-04-21 19:30:04 -06:00
Mohammed Al Sahaf
bae4f15fad
ci: fuzz: remove the fuzzer of the Caddyfile parser (#3288) 2020-04-20 15:21:19 -06:00
Francis Lavoie
0798459e44
readme: Fix broken links (#3283)
Credit to @kanagawa41 for spotting these!

Fixes #3282
2020-04-19 17:37:03 -06:00
Matthew Holt
f980170909
doc: Improve comment 2020-04-17 12:03:57 -06:00