1
0
Fork 0
mirror of https://github.com/warmcat/libwebsockets.git synced 2025-03-16 00:00:07 +01:00
Commit graph

285 commits

Author SHA1 Message Date
Andy Green
3073655759 ss: zero on destroy 2021-08-09 17:31:16 +01:00
Andy Green
48541efcc2 mqtt: handle NULL mqtt publish metadata 2021-07-08 10:16:06 +01:00
Andy Green
ee510154a0 fi: wsi: fault to fake peer hangup after a delay range
Example usage, client

$ valgrind ./bin/lws-minimal-secure-streams-stress-client -c 4 --budget 40

proxy

$ valgrind ./lws-minimal-secure-streams-proxy --fault-injection "ss=mintest/wsi/timedclose(20%),ss=mintest/wsi/timedclose_ms(0..100)"
2021-07-04 10:29:54 +01:00
Andy Green
fabe78d222 Wextra
Add -Wextra (with -Wno-unused-parameter) to unix builds in addition to
-Wall -Werror.

This can successfully build everything in Sai without warnings / errors.
2021-07-04 10:29:54 +01:00
Andy Green
5432e8ae7b sspc: add concurrency stress
Add a way to perform sspc tests with increased concurrency and sequential
tests
2021-07-04 10:29:52 +01:00
Andy Green
5d8cf03221 logs: log contexts 2021-07-01 05:20:53 +01:00
Andy Green
8e76634ed3 logs: introduce log_cx 2021-07-01 05:20:53 +01:00
Andy Green
66afeaa5bc dns: handle EAI_NONAME as fatal
This causes the blocking dns lookup to treat EAI_NONAME as immediately
fatal, this is usually caused by an assertive NXDOMAIN from the DNS server
or similar.

Not being able to reach the server should continue to retry.

In order to make the problem visible, it reports the situation using
CLIENT_CONNECTION_ERROR, even though it is still inside the outer client
creation call.
2021-06-28 08:52:20 +01:00
Andy Green
9e8eb28c72 sul: improve debugging 2021-06-26 13:03:13 +01:00
Andy Green
a095e6c60b ss: sspc: enforce not allowed to call lws_ss[pc]_destroy from callbacks
For both ss and sspc, enforce at runtime that user code cannot call
lws_ss[pc]_destroy on a handle from a callback.

The error indicates the remedy (return DESTROY_ME) and asserts.
2021-06-23 16:51:59 +01:00
Andy Green
5ea9194378 ss: check destroy null policy 2021-06-22 17:44:40 +01:00
Andy Green
7135b3113b ss: check serialize null policy 2021-06-22 16:54:04 +01:00
Andy Green
c859b60854 ss: avoid null ss policy on req tx len 2021-06-22 16:47:59 +01:00
Andy Green
9bd8819045 ss: avoid null ss policy on req tx 2021-06-22 16:47:11 +01:00
Haopeng Li
cb5b1f720b ss: h1: log dereferences NULL 2021-06-22 15:54:22 +01:00
Jed Lu
2abf4115a8 ss: direct protocol string 2021-06-11 14:44:10 +01:00
Haopeng Li
b3bff76a73 strexp: handle NULL better 2021-06-07 15:11:12 +01:00
Andy Green
890023f821 ss: h2: handle zero length COLON_PATH for metadata 2021-06-04 15:31:49 +01:00
Andy Green
6c5636889d sspc: handle nonexistent metadata cleanly 2021-06-04 15:05:36 +01:00
Andy Green
436bb70919 coverity: h->cwsi must be valid if we are handling rx on it 2021-06-04 14:17:13 +01:00
Jed Lu
5b1a260881 ss: sspc: handle destroy for client_connect and request_tx 2021-05-28 09:41:39 +01:00
Chunho Lee
c5484c0232 ss-mqtt: Add support for MQTT UNSUBSCRIBE and DISCONNECT
Send UNSUBSCRIBE and DISCONNECT when LWSSSSRET_DISCONNECT_ME is
returned.
2021-05-27 11:37:44 +01:00
Andy Green
c70c4bac89 coverity: ss: handle NULL omd 2021-05-26 14:07:02 +01:00
Andy Green
632ffa1eca ss: deal with missing handle metadata 2021-05-25 14:57:57 +01:00
Jin Wang
2197c2eafc sspc: handle DESTROY_ME at proxy link close cb
It's open to user state callback to respond to the DISCONNECT we issue on
the proxy link closure by escalating it to a DESTROY_ME... we have to deal
with that if it comes.
2021-05-20 15:24:42 +01:00
Andy Green
a3765d118c ss: h2: leave the ss bound to its affiliated wsi until close clears it
In sai, on Xenial (only...) noticed that the wsi is still bound to the ss
handle, and can reference it even after the ss has been destroyed on
ss-testsfail sometimes.

Leave the handle knowing its wsi and able to detach it later during close.
2021-05-07 11:12:51 +01:00
Andy Green
88c824b89d ss: avs: let multipart handle content-type
We added the neccessary return check on set_metadata which is good, but it
doesn't take into account that for LWA, we will normally use the multipart
stuff, that provides its own conflicting content-type.

Currently the lwa streamtype policy doesn't even provide ctype metadata
type, let's just comment it out.
2021-05-07 10:15:05 +01:00
Andy Green
2dba500d4f h2: just log nwsi close and pass thru
User reports problems with the close / retry flow not happening if we don't
pass thru the nwsi close... it may be happening before the sid1 migration.

Just log it and don't end the handling before the passthru.  Logging it
because there was a reason for the change to not passing it through...
2021-04-21 16:05:24 +01:00
Andy Green
891f28b43e ss: proxy: fix conn deref on onward
Trying to use the opaque pointer in the handle to point to the conn isn't
going to work when we need it to point to the ss handle.

Move it to have its on place in the handle.
2021-04-21 10:02:00 +01:00
Andy Green
68d9f3a7f2 ss: wire up EVENT_WAIT_CANCELLED
Currently the lws_cancel_service() api only manifests itself at lws level.
This adds a state LWSSSCS_EVENT_WAIT_CANCELLED that is broadcast to all
SS in the event loop getting the cancel service api call, and allows
SS-level user code to pick up handling events from other threads.

There's a new example minimal-secure-streams-threads which shows the
pattern for other threads to communicate with and trigger the event in the
lws service thread.
2021-04-15 19:10:54 +01:00
Andy Green
5b6a89f79d ss: add test for OS trust store mdoe
It's already the case that leaving off the "tls_trust_store" member of the
streamtype definition in the policy causes the streamtype to validate its
tls connections via the OS trust store, usually a bundle OpenSSL has been
configured to load at init automagically, but also literally the OS trust
store in windows case.

Add tests to confirm that.
2021-04-15 19:10:54 +01:00
Chunho Lee
0276f3e635 ss: mqtt: Avoid CONNECTING to DISCONNECTED transition
Check previous states with 'ss_dangling_connected'.
If it did not visit CONNECTED, transit to UNREACHABLE
instead of DISCONNECTED.
2021-04-15 17:34:01 +01:00
Andy Green
402e97a679 ss: metrics: http: just report at wsi close
Defer recording the ss metrics histogram until wsi close, so it has a
chance to collect all the tags that apply.

Defer dumping metrics until the FINALIZE phase of context destroy, so we
had a chance to get any metrics recorded.
2021-04-08 06:47:54 +01:00
Haopeng
2a7ded310f ss: metrics: add result tag reflecting ss txn disposition
For http ss, add a tag result=SS_ACK_REMOTE or SS_NACK_REMOTE
2021-04-08 06:37:35 +01:00
Andy Green
42dc817d8f ss: proxy: get rx flow control working
This fixes the proxy rx flow by adding an lws_dsh helper to hide the
off-by-one in the "kind" array (kind 0 is reserved for tracking the
unallocated dsh blocks).

For testing, it adds a --blob option on minimal-secure-streams[-client]
which uses a streamtype "bulkproxflow" from here

https://warmcat.com/policy/minimal-proxy-v4.2-v2.json

		"bulkproxflow": {
			"endpoint": "warmcat.com",
			"port": 443,
			"protocol": "h1",
			"http_method": "GET",
			"http_url": "blob.bin",
			"proxy_buflen": 32768,
			"proxy_buflen_rxflow_on_above": 24576,
			"proxy_buflen_rxflow_off_below": 8192,
			"tls": true,
			"retry": "default",
			"tls_trust_store": "le_via_dst"
		}

This downloads a 51MB blob of random data with the SHA256sum

ed5720c16830810e5829dfb9b66c96b2e24efc4f93aa5e38c7ff4150d31cfbbf

The minimal-secure-streams --blob example client delays the download by
50ms every 10KiB it sees to force rx flow usage at the proxy.

It downloads the whole thing and checks the SHA256 is as expected.

Logs about rxflow status are available at LLL_INFO log level.
2021-04-07 15:54:26 +01:00
Andy Green
fdc1e1e9a5 sspc: states for proxy failure 2021-04-07 14:48:52 +01:00
Andy Green
877fcc3e4d ss: policy: atoll needed for 32-bit machines
On 32-bit Linux compilers, long int == int == 32-bit.  So even atol() cannot
handle ints above 0x7fffffff and clips any it finds at that.

There's only one instance in policy-json.c, use atoll() cast to uint64_t
to allow values up to 64-bit INT_MAX even on 32-bit machines.
2021-04-07 09:52:04 +01:00
Andy Green
3ad0a3d223 ss: server: allow bind to existing vhost
If the larger application is defining vhosts using lejp-conf JSON, it's
often more convenient to describe the vhost for ss server binding to
that.

If the server policy endpoint (usually used to describe the server
interface bind) begins with '!', take the remainder of the endpoint
string as the name of a preexisting vhost to bind ss server to at
creation-time.
2021-04-05 10:55:04 +01:00
Andy Green
d291c02a23 ss: sspc: add conmon performance telemetry
This provides a way to get ahold of LWS_WITH_CONMON telemetry from Secure
Streams, it works the same with direct onward connections or via the proxy.

You can mark streamtypes with a "perf": true policy attribute... this
causes the onward connections on those streamtypes to collect information
about the connection performance, and the unsorted DNS results.

Streams with that policy attribute receive extra data in their rx callback,
with the LWSSS_FLAG_PERF_JSON flag set on it, containing JSON describing the
performance of the onward connection taken from CONMON data, in a JSON
representation.  Streams without the "perf" attribute set never receive
this extra rx.

The received JSON is based on the CONMON struct info and looks like

{"peer":"46.105.127.147","dns_us":596,"sockconn_us":31382,"tls_us":28180,"txn_resp_us:23015,"dns":["2001:41d0:2:ee93::1","46.105.127.147"]}

A new minimal example minimal-secure-streams-perf is added that collects
this data on an HTTP GET from warmcat.com, and is built with a -client
version as well if LWS_WITH_SECURE_STREAMS_PROXY_API is set, that operates
via the ss proxy and produces the same result at the client.
2021-04-05 10:55:04 +01:00
Andy Green
51490ae6e6 Fault injection
add lws_xos: xoshiro256 PRNG
2021-04-05 10:55:04 +01:00
Andy Green
e4d381eadd metrics: increase type of us_schedule to uint64_t 2021-04-02 16:50:49 +01:00
Jed Lu
69f2ad9ced ss: allow that request_tx may get told DESTROY_ME 2021-04-01 09:09:10 +01:00
Chunho Lee
ff76604006 ss: mqtt: Wrap ACK_REMOTE to ACK_REMOTE transition with LWS_ROLE_MQTT
Wrap ACK_REMOTE to ACK_REMOTE transition with
"#if defined(LWS_ROLE_MQTT) /.../ #endif".
2021-03-31 09:39:12 +01:00
Andy Green
0c98d94481 ss: mqtt: allow ACK_REMOTE to ACK_REMOTE transition
MQTT doesn't follow http type transaction flow, so it's
quite possible to have n PUBLISH getting acked in a row.
2021-03-30 08:10:52 +01:00
Chunho Lee
25ae9facc9 mqtt: topic validation for different mqtt servers
AWS IoT enforces limits topic level and length. If 'aws_iot' is set
on the policy, the topic limits will be enforced for AWS IoT.
2021-03-30 07:38:37 +01:00
Sakthi Kannan
f3531ef673 mqtt: wildcard topic and topic to 256 chars
Adding supports to MQTT wildcard support, topic to 256 chars,
incorrect topic validation.
2021-03-30 07:38:37 +01:00
Sakthi Kannan
a088b72696 mqtt: Setting the CONNECTED state only when SUBACK is received
Setting the CONNECTED state only when SUBACK is received if the stream has
defined a subscription topic. This is to avoid SS from sending out SUBSCRIBE
right after CONNACK, even when the connection is not valid.
2021-03-30 07:38:32 +01:00
Andy Green
eda5c1b60a sai: openbsd: fix sigv4 2021-03-29 19:36:26 +01:00
Andy Green
1000a52df3 logs: prioritize WITH_NO_LOGS even in release
Really not having any logs makes it difficult to know what is really
happening, but if that's you're thing this will align debug and release
modes to just have ERR and USER if you give WITH_NO_LOGS
2021-03-15 12:56:15 +00:00
Andy Green
ad5047a687 sspc: make sure we cancel any retry 2021-03-14 19:50:08 +00:00