Currently dogfooding this patch at gemini.sgregoratto.me. To test,
run the following command and look for the "OCSP response" header:
openssl s_client -connect "gemini.sgregoratto.me:1965" -status
This adds a barebone dumping of the parsed configuration. It is not
complete, but I'm interested in dumping the full path to `cert' and
`key' in order to write some scripts that can inspect the
configuration, extract the certificates and renew them when expired
automatically.
It's not easy to parse gmid configuration otherwise because the syntax
is flexible and users can use macros. Instead, the idea is to run
gmid and let it dump the configuration once it's been parsed in a
static and predictable format.
Now is possible to parse gmid configuration with, say, awk or perl.
It's not intuitive to print
open ... for domain xyz
it doesn't convey that the open failed.
now it appends the error string, at least the user can understand that
something went wrong.
reported by cage on irc, thanks!
FastCGI is designed to multiplex requests over a single connection, so
ideally the server can open only one connection per worker to the
FastCGI application and that's that.
Doing this kind of multiplexing makes the code harder to follow and
easier to break/leak etc on the gmid side however. OpenBSD' httpd
seems to open one connection per client, so why can't we too?
One connection per request is still way better (lighter) than using
CGI, and we can avoid all the pitfalls of the multiplexing (keeping
track of "live ids", properly shut down etc...)
There's no difference, but bzero(3) says
STANDARDS
The bzero() function conforms to the X/Open System Interfaces option of
the IEEE Std 1003.1-2004 (“POSIX.1”) specification. It was removed from
the standard in IEEE Std 1003.1-2008 (“POSIX.1”), which recommends using
memset(3) instead.
so here we are.
Initialize the logger as soon as possible and log by default to
stderr. With this, some (common?) errors are printed early instead of
ending up in syslog.
# NB: this is in configless mode
% ./gmid -p 80
[2021-07-07 11:05:57] bind: Address already in use
% ./gmid -p 81
[2021-07-07 11:13:53] bind: Permission denied
%
we need to delete the events associated with the backends, otherwise
the server process won't ever quit.
Here, we add a pending counter to every backend and shut down
immediately if they aren't handling any client; otherwise we try to
close them as soon as possible (i.e. when they close the connection to
the last connected client.)
GMID_VERSION follows the CGI/FastCGI style, i.e. project_name/version.
Define GMID_STRING with a more "human" variant "project_name version",
and reuse that in the --help and --version codepath.
the logger process now can receive a file descriptor to write logs
to. At the moment the logic is simple, if it receives a file it logs
there, otherwise it logs to syslog. This will allow to log on custom
log files.
it's not technically required, since a couple of lines below we free
whole host struct, and we don't have code that may use
h->{env,aliases} afterwards, but it's nice not to have invalid
pointers around. it may bite in the future.
Not production-ready yet, but it's a start.
This adds a third ``backend'' for gmid: until now there it served
local files or CGI scripts, now FastCGI applications too.
FastCGI is meant to be an improvement over CGI: instead of exec'ing a
script for every request, it allows to open a single connection to an
``application'' and send the requests/receive the responses over that
socket using a simple binary protocol.
At the moment gmid supports three different methods of opening a
fastcgi connection:
- local unix sockets, with: fastcgi "/path/to/sock"
- network sockets, with: fastcgi tcp "host" [port]
port defaults to 9000 and can be either a string or a number
- subprocess, with: fastcgi spawn "/path/to/program"
the fastcgi protocol is done over the executed program stdin
of these, the last is only for testing and may be removed in the
future.
P.S.: the fastcgi rule is per-location of course :)
this fixes a bug introduced with the prefork mechanics: every server
process shared the same socket, and this would cause a race condition
when multiple server processes asked for a script cgi being executed.
This gives each server process its own socket to talk to the executor,
so the race cannot happen.
Include gmid.h as first header in every file, as it then includes
config.h (that defines _GNU_SOURCE for instance).
Fix also a warning about unsigned vs signed const char pointers in
openssl.
I got bitten by the scope visibility rules. After the end of the
block, the path variable is no longer valid, and in fact later
load_vhosts fails to open that (because the buffer gets invalidated)
the new logging code was crashing if the client didn't support SNI or
if required an unknown vhost: this because we short-circuit in
handle_handshake to an error, so c->iri isn't populated yet (we don't
even read the request).
fixes#1
err/warn is not available on some systems (unfortunately!) and in any
case don't play well with our daemon mode (that closes std{in,out,err}).
Use our fatal that is daemon-aware.
cgi.c wasn't really needed; it better to group all the server related
functions together, cgi or not. Now gmid.c contains only startup and
utility code.
this way, we can sandbox the listener with seccomp (todo) or capsicum
(already done) and still have CGI scripts. When we want to exec, we
tell the executor what to do, the executor executes the scripts and
send the fd backt to the listener.
* gmid.c (main): changed behaviour: daemon off by default
(main): changed -c in -C (cert option)
(main): changed -k in -K (key option, for consistency with -C)
(main): added -c to load a configuration
(main): certs, key and doc (-C -K and -d) doesn't have a default value anymore
(handle_handshake): add vhosts support
This alter the current state machine by adding S_HANDSHAKE as the
initial state. There, we ensure we did the handshake and we check
SNI. ATM we simply continue in S_OPEN, but later we can add virtual
host checks there, and skip to S_INITIALIZING with an error state if
the client is accessing a wrong host.
Stephen Gregoratto
Omar Polo