This should be a NOP for h2 support and only affects internal
apis. But it lets us reuse the working and reliable h2 mux
arrangements directly in other protocols later, and share code
so building for h2 + new protocols can take advantage of common
mux child handling struct and code.
Break out common mux handling struct into its own type.
Convert all uses of members that used to be in wsi->h2 to wsi->mux
Audit all references to the members and break out generic helpers
for anything that is useful for other mux-capable protocols to
reuse wsi->mux related features.
Saw this on travis selftests during context destroy
==18895== Invalid read of size 8
==18895== at 0x415909: __lws_vhost_destroy2 (vhost.c:1063)
==18895== by 0x40E65B: lws_context_destroy2 (context.c:929)
==18895== by 0x40EBE5: lws_context_destroy (context.c:1128)
==18895== by 0x40CC41: main (minimal-http-client-post.c:267)
==18895== Address 0x6168688 is 728 bytes inside a block of size 792 free'd
==18895== at 0x4C2BDEC: free (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==18895== by 0x45B29E: _realloc (alloc.c:120)
==18895== by 0x45B2D6: lws_realloc (alloc.c:130)
==18895== by 0x415ED7: __lws_vhost_destroy2 (vhost.c:1204)
==18895== by 0x419164: lws_vhost_unbind_wsi (wsi.c:82)
==18895== by 0x41236B: __lws_free_wsi (close.c:154)
==18895== by 0x4134CF: __lws_close_free_wsi_final (close.c:650)
==18895== by 0x4133BA: __lws_close_free_wsi (close.c:610)
==18895== by 0x413528: lws_close_free_wsi (close.c:660)
==18895== by 0x4158C7: __lws_vhost_destroy2 (vhost.c:1053)
==18895== by 0x40E65B: lws_context_destroy2 (context.c:929)
==18895== by 0x40EBE5: lws_context_destroy (context.c:1128)
Removing the last wsi from the vhost we started to destroy finalized the
vhost destruction, which is aimed at libuv async close cleanup. But if
we already entered __lws_vhost_destroy2, we will definitely destroy the vhost
ourselves at the end of that function already. So defeat the wsi close
triggering it.
If a client connects to a SSL server and the server sends handshake
alert (e.g. no matching ciphers) SSL_connect() fails, but because
SSL_ERROR_SSL return value is not handled, it's not considered a
failure. SSL_want_read() will return 1 and the client will happily wait
for more data from the server. Now if the server closes connection after
sending handshake alert, POLLIN event will be triggered,
lws_tls_client_connect() called again, but SSL_connect() will fail
without calling read(), so the client will end up consuming 100% CPU
because POLLIN will be triggered repeatedly.
Similar error handling is used in lws_tls_server_accept() and the
condition checks for SSL_ERROR_SSL. Using the same condition in
lws_tls_client_connect() fixes the problem.
Tested with OpenSSL 1.0.2k.
On some platforms, it's possible that logging flow may reset errno. In the case where
we try to log errno on those platforms and afterwards try to query it, we will get a
nasty surprise that the logged errno is destroyed by the time we come to test it.
In the two cases of this in the tree at the moment, sample errno into a temp and
log and test the temp.
Thanks to Sakthi Ramabadran for finding this.
Now the generic lws_system blobs can cover client certs + key, let's
add support for applying one of the blob sets to a specific client
connection (rather than doing it via the vhost).
Remove the auth lws_system stuff and redo it using generic blobs
with separate namespaces. Support pointing to already-in-memory
blobs without using heap as well as multi-fragment appened blobs
eg, parsed out of JSON chunk by chunk and chained in heap.
Support auth the new way, along with client cert + key in DER
namespaces.
It looks to semmle like the int size can be bigger than the char loop var.
But the size is the size of the IPv4 or IPv6 address, so it cannot make
a problem.
Client connection items for protocols other than http ones
will never get into an ah. Allow use of the values from the
client stash allocation instead if present.
This teaches http client stuff how to handle 303 redirects... these
can happen after POST where the server side wants you to come back with
a GET to the Location: mentioned.
lws client will follow the redirect and force GET, this works for both
h1 and h2. Client protocol handler has to act differently if it finds
it is connecting for the initial POST or the subsequent GET, it can
find out which by checking a new api lws_http_is_redirected_to_get(wsi)
which returns nonzero if in GET mode.
Minimal example for server form-post has a new --303 switch to enable
this behaviour there and the client post example has additions to
check lws_http_is_redirected_to_get().
Resetting the ah and waiting a bit is the right strategy at the end of
http/1.1 client transaction. But it's wrong for h2... drop the ah
instead if it's the end of a client transaction on h2.
