Fix null pointer dereference #57

… that netproc failing doesn't cause short reads and exits.

… on Linux.

https://github.com/pozdnychev -- thanks!

…g but looks up addresses as used by netproc.

Prune a lot of unused variables.

…in the challenge directory on another system. Don't use this.

…, and have the netproc wait til they have: there's no point in talking to Let's Encrypt if these services haven't started.

… bail earlier.

Continue building in revocation facility.

have files made in the challengedir at all.

…sible to simply run letskencrypt as a cronjob without worrying about overloading the ACME server.

a -portable and OpenBSD version of the software with a guarded config.h inclusion.

ongoing re-structure into the main and -portable branch.

…ble.

…portable.

Remove wrong-headed attempt at setproctitle.

bit more readable.

invoke the JSON functions locally, within a doXXXX function; second, don't read into the JSON parser, but into an intermediary buffer (allowing us to dump it on error); third, move some fetch bodies from the main netproc() function into their own functions; and lastly, store the CA nonce agency as a variable (we'll use this later when using other servers).

…ving one SAN entry per domain. However, apparently this is not allowed; instead we now have a single SAN extension entry with the full list.

of file (epipe), and 1 on success. This addresses all the callers.

…www repo. Strip down the README.md file to only what's necessary.

Also, the console can get spammed by multiple procs writing into stderr.

…port their error.

those on the certificate, if found. NOTE: what if there's no SAN entry at all?

nil-terminated (and valgrind suggests they aren't), so make sure that they are always nil-terminated.

…d by https://github.com/snimmagadda/http.

… itself. Move the maximum number of queried servers into extern.h.

…sproc comm identifiers.

and sreq() functions with those using http.h. There is still a fair amount of superfluous debugging going on, and the system isn't "optimum" regarding memory at all.

versions of the API.

…t is the same.

put the head and body buffer pointers into http_get.

… messages, use the new http_get members instead of calling functions. "Clean-up."

inet pledge, which was a holdover from libcurl.

the whitepath doesn't exist (yet) for the former, and the challenge process has unknown files in a known path.

carry around all sorts of extra information we don't use.

fully pledged need the root user for chroot.

…d by Remco--thanks!

@kaworu

former from a patch by @kaworu in pull/2 -- thanks!

@kaworu

I'm not touching the memory (right now) of this mystery function. This raised by @kaworu in pull/2 -- thanks!

…, and makes the key extraction be generic. It will need more interoperability with the signing process to actually work.

Remco--thank you!

restrictions, as the certificate creation is opaque. This is from a patch by Remco--thanks!

domain to all paths. This makes it easier to use in systems where one's invoking letskencrypt multiple times. Also add some simple validation of the domain names to prevent them from (1) trampling the directory structure and (2) being bogus in general.

…---thanks!

patch by Remco---thanks!

…co---thanks!

Remco and as kristapsdz#5

…s was prompted by kristapsdz#7

usage (current) by acctproc and (pending) keyproc.

…d in kristapsdz/acme-client-portable#3 ) and finally change the error code to be "2" if nothing changed on the disc, otherwise 0 on success (the certificates updated) and 1 on failure.

…ystems have the -m flag), and use the return codes in the example.

if it doesn't exist. This allows using -nN even after first creating the files. From kristapsdz#8

…e this directory, so let them provide their own.

to netproc and so on. This reverts a prior change that was incomplete.

Suggested by kristapsdz#12 and kristapsdz#9

…hed) in kristapsdz#15 and by ``pasta'' over e-mail -- thanks!

be manage the challenge itself. Compile-tested but needs work. Also put all asprintf calls into doasprintf, which manages the null-ification on the failure track.

…enge process is going to do anything in the webroot.

shipped as acme-client in OpenBSD.

inclusion into OpenBSD as acme-client and further... suggested... by Josh Aas of Let's Encrypt.

…hanks!

…OpenBSD.

USING THE DEFAULT PATHS, YOU NEED TO RENAME!

…t been overriden. For example, if providing -m foo.com without a -f, this would create the foo.com directory within /etc/acme.

…ne each time the connections are made. This needs to happen before pledging and/or jailing. Inspired by OpenBSD's downstream commit, but modified to prevent a static value being used for the configuration. While here, merge OpenBSD churn for variables.

…mprove the -portable logic.

…stead of merely forking into them.

…Also document some other functions.

latter causes us to reach out over the file-system for locale data with some operating systems, which trips up our sandbox.

kristapsdz/acme-client-portable#12 that pushes the check for file pre-existence with -n and -N only into the main process, stripping out the arguments before passing them to the child. This fixes a race condition.

allows certain variables to be overriden. This allows us to provide -nN and have them properly be nullified by the child processes.

…back- port that to the older API of libressl.

…ortable#13 . This creates a triplet exported to the operator: challenge type, domain, and print.

.

…s can be noted approriately. From kristapsdz#29 -- thanks!

…s will make this optional.)

…p to date version of libressl out there.

…versions as old as that.

…nginx and Apache (more to come). Inspired by kristapsdz#32 --- thanks!

Ooops! I hope I don't see any more of those!

Commits on Jul 10, 2016

Add some example usage.

kristaps committed Jul 10, 2016

Configuration menu

View commit details

Copy full SHA for bd3de43

Browse repository at this point

Copy the full SHA

bd3de43 View commit details

Browse the repository at this point in the history

Commits on Dec 5, 2016

Force umask for creation of challenge file.

kristaps committed Dec 5, 2016

Configuration menu

View commit details

Copy full SHA for 6c44c54

Browse repository at this point

Copy the full SHA

6c44c54 View commit details

Browse the repository at this point in the history

Commits on Feb 14, 2017

Be more clear.

kristaps committed Feb 14, 2017

Configuration menu

View commit details

Copy full SHA for d1ba902

Browse repository at this point

Copy the full SHA

d1ba902 View commit details

Browse the repository at this point in the history

Fix null pointer dereference #57

Are you sure you want to change the base?

Fix null pointer dereference #57

Commits on May 15, 2016

Commits on May 16, 2016

Commits on May 17, 2016

Commits on May 18, 2016

Commits on May 19, 2016

Commits on May 20, 2016

Commits on May 22, 2016

Commits on May 24, 2016

Commits on May 25, 2016

Commits on May 26, 2016

Commits on Jun 1, 2016

Commits on Jun 2, 2016

Commits on Jun 3, 2016

Commits on Jun 25, 2016

Commits on Jun 28, 2016

Commits on Jul 2, 2016

Commits on Jul 10, 2016

Commits on Jul 12, 2016

Commits on Jul 13, 2016

Commits on Jul 16, 2016

Commits on Jul 28, 2016

Commits on Aug 6, 2016

Commits on Aug 19, 2016

Commits on Aug 22, 2016

Commits on Aug 31, 2016

Commits on Sep 1, 2016

Commits on Sep 2, 2016

Commits on Sep 3, 2016

Commits on Oct 29, 2016

Commits on Oct 30, 2016

Commits on Oct 31, 2016

Commits on Nov 1, 2016

Commits on Nov 2, 2016

Commits on Nov 9, 2016

Commits on Nov 24, 2016

Commits on Nov 25, 2016

Commits on Dec 2, 2016

Commits on Dec 5, 2016

Commits on Jan 28, 2017

Commits on Jan 30, 2017

Commits on Feb 1, 2017

Commits on Feb 12, 2017

Commits on Feb 14, 2017

Commits on Mar 12, 2017

Commits on Jul 1, 2017

Commits on Jul 2, 2017

Commits on Sep 8, 2018