Commit Graph

1278 Commits

Author SHA1 Message Date
Tatsuhiro Tsujikawa 4ceb9fd767 Cancel stream if flow control window gets overflow
Previously, we did terminate session, but it did not actually cancel
streams.  Now we just cancel streams in this case.
2016-01-23 00:24:12 +09:00
Tatsuhiro Tsujikawa c7de58d865 Validate :authroity, host, and :scheme value more strictly 2016-01-16 15:11:48 +09:00
Tatsuhiro Tsujikawa 15ea11d5db Fix bug that HTTP upgrade was not working 2016-01-12 21:05:05 +09:00
Tatsuhiro Tsujikawa afc6f556d1 Update doc 2016-01-11 17:47:46 +09:00
Tatsuhiro Tsujikawa e14da859b6 Merge branch 'master' into simple-extensions 2016-01-11 16:39:35 +09:00
Tatsuhiro Tsujikawa bb6f842b37 Check request/response submission error based side of session
Disallow request from server, and response from client respectively.
When the violation is detected, return NGHTTP2_ERR_PROTO from
nghttp2_submit_request, nghttp2_submit_response,
nghttp2_submit_headers.

We also did some refactoring, and now self-dependency detection is
placed where it is only required.
2016-01-11 16:18:39 +09:00
Tatsuhiro Tsujikawa 9cea986819 Strict outgoing idle stream detection
Previously, we use session->next_stream_id to detect that given stream
ID was idle or not.  But this was suboptimal, since it was updated
when stream ID was assigned, and it did not necessarily mean that it
actually has been sent to the peer.  Now we introduced
session->sent_stream_id, which only updated when HEADERS/PUSH_PROMISE
has sent.  Using sent_stream_id instead of next_stream_id tightens
idle stream detection, and misbehaved peer which sends frame with
stream ID that has not been generated.

This commit also overhauls test code which involves opening streams.
Now we have some wrapper functions for nghttp2_session_open_stream()
which also take care of updating next_stream_id and
last_recv_stream_id.  They are crucial for some tests.
2016-01-11 00:31:52 +09:00
Tatsuhiro Tsujikawa 0caefe20ef Merge branch 'master' into simple-extensions 2016-01-09 19:08:28 +09:00
kumagi f3a37b2ef1 fix typos: heder->header alreay->already reponse->response 2016-01-05 22:19:33 +09:00
Tatsuhiro Tsujikawa 8716dd05d4 Return error from nghttp2_submit_{headers,request} when self dependency is made
Return NGHTTP2_ERR_INVALID_ARGUMENT from nghttp2_submit_headers() if
given stream ID and pri_spec->stream_id are the same (thus trying to
depend on itself).

Also return NGHTTP2_ERR_INVALID_ARGUMENT from nghttp2_submit_request()
and nghttp2_submit_headers() with stream_id == 1, when new stream ID
equals to pri_spec->stream_id.

Previously, these cases are not checked, and just sent to peer.
2015-12-25 21:37:18 +09:00
Tatsuhiro Tsujikawa 894c1bd02e Create idle stream on client side
With the presence of idle stream related API (e.g.,
nghttp2_create_idle_stream()), it is more predictable for client to
create idle streams with its dependency to another idle stream.
Previously, we didn't create complete parent idle stream in this case.
Now we create idle streams as we do on server side.
2015-12-25 00:10:37 +09:00
Tatsuhiro Tsujikawa 92a56d034f Fix bug that idle/closed stream may be destroyed while it is referenced 2015-12-23 16:38:30 +09:00
Tatsuhiro Tsujikawa ca4a40b8e0 Don't schedule response HEADERS with priority tree
Previously we scheduled the transmission of response HEADERS using
priority tree in the belief that it allows more better utilization of
bandwidth for prioritized streams.  But to reduce the overhead of
reconstruction of priority queue when connection level flow control
window is depleted, we just don't check priority tree in this case.
This means that response HEADERS frames are not sent even though they
are not flow controlled.  This could waste bandwidth.  To improve this
situation, we stop scheduling response HEADERS with priority tree for
now.  Now they are just sent in the order they submitted.  The
response body DATA continued to be scheduled with priority tree as
before.
2015-12-21 21:33:58 +09:00
Tatsuhiro Tsujikawa cb73ba948d Simplify HEADERS handling; handle push response in one function 2015-12-20 23:47:16 +09:00
Tatsuhiro Tsujikawa e957147249 Make obvious implementation error connection error 2015-12-20 23:29:24 +09:00
Tatsuhiro Tsujikawa 010726a875 Simplify error handling in nghttp2_session_on_headers_received
return session_inflate_handle_invalid_stream(...) case is for streams
for INITIAL state, but this is rare case.  In general, we'd like to
reduce RST_STREAM transmission, and it is suffice to ignore this frame
for now.
2015-12-20 23:20:14 +09:00
Tatsuhiro Tsujikawa 19146211d7 Update doc 2015-12-20 13:20:21 +09:00
Tatsuhiro Tsujikawa 9f0083309e Fix strange use of session_detect_idle_stream 2015-12-20 13:19:29 +09:00
Tatsuhiro Tsujikawa 80f7abb565 Fix crash caused by the regression in 9f8fc7b2bb 2015-12-17 22:25:28 +09:00
Tatsuhiro Tsujikawa 9f8fc7b2bb Strict error handling for frames which are not allowed after closed (remote)
This makes h2spec strict mode a bit happier.  We still one failing
test with h2spec -S (strict mode).
2015-12-17 21:31:20 +09:00
Tatsuhiro Tsujikawa 9cfda0c070 Update doc 2015-12-17 18:04:16 +09:00
Tatsuhiro Tsujikawa 228d92244a Optimize nghttp2_pq swap 2015-12-15 22:07:51 +09:00
Tatsuhiro Tsujikawa 68c5deea5a Clarify the condition when opening new stream from remote is allowed 2015-12-11 21:23:49 +09:00
Tatsuhiro Tsujikawa f14ac74316 At least check stream ID is valid when PUSH_PROMISE is received in goaway mode 2015-12-10 23:54:54 +09:00
Tatsuhiro Tsujikawa 0c70ff5658 Simplify 2015-12-10 23:18:02 +09:00
Tatsuhiro Tsujikawa 7cc2d22ab5 Fix compile error with gcc 2015-12-08 23:33:26 +09:00
Tatsuhiro Tsujikawa abcdbf0039 Take into account remainder due to integer division when calculating cycle 2015-12-08 23:15:55 +09:00
Tatsuhiro Tsujikawa 4bcc14fc67 Reschedule stream when only weight is changed
Previously, we only updated stream's weight field when only weight was
changed by PRIORITY frame.  If stream is queued, it would be better to
actually reschedule it based on new weight.  This could be especially
useful if weight is increased.
2015-12-07 23:14:22 +09:00
Tatsuhiro Tsujikawa 72f815d535 Update descendant_last_cycle on nghttp2_stream_next_outbound_item
Previously, we updated descendant_last_cycle in
nghttp2_stream_reschedule, which is called after non-zero DATA frame.
But this was not optimal since we still had old descendant_last_cycle,
and new stream was scheduled based on it.  Now descendant_last_cycle
is updated in nghttp2_stream_next_outbound_item, which is called when
stream with highest priority is selected from queue.  And new stream
is scheduled based on it.  This commit also removes 0-reset of
descendant_last_cycle and cycle in nghttp2_stream_reschedule.  This
could help making them lower, so that they are not overflow.  But
there is a pattern that it doesn't work, and we are not sure they are
really useful at this moment.
2015-12-07 22:48:50 +09:00
Tatsuhiro Tsujikawa 9c84f60ba0 Merge branch 'master' into simple-extensions 2015-12-04 23:48:40 +09:00
Tatsuhiro Tsujikawa a151a44caf Set max number of outgoing concurrent streams to 100 by default
Instead of using nonsensical large value for max outgoing concurrent
streams, use more sensible value, 100.
2015-12-03 23:45:03 +09:00
Tatsuhiro Tsujikawa 478a423bcf Reduce nghttp2_stream size 2015-12-03 22:53:02 +09:00
Tatsuhiro Tsujikawa 93d8636fb0 Keep incoming streams only at server side
We should only keep incoming closed streams because we only keep at
most max concurrent streams, which only applied to incoming streams.
2015-12-03 22:48:41 +09:00
Tatsuhiro Tsujikawa 2288ee8060 Create stream object for pushed resource during nghttp2_submit_push_promise()
Previously, stream object for pushed resource was not created during
nghttp2_submit_push_promise().  It was created just before
nghttp2_before_frame_send_callback was called for that PUSH_PROMISE
frame.  This means that application could not call
nghttp2_submit_response for the pushed resource before
nghttp2_before_frame_send_callback was called.  This could be solved
by callback chaining, but for web server with back pressure from
backend stream, it is a bit unnecessarily hard to use.

This commit changes nghttp2_submit_push_promise() behaviour so that
stream object is created during that call.  It makes application call
nghttp2_submit_response right after successful
nghttp2_submit_push_promise call.
2015-12-02 21:16:30 +09:00
Tatsuhiro Tsujikawa 2754d9e2bb Update doc 2015-11-28 15:24:13 +09:00
Tatsuhiro Tsujikawa 12b2e0a2b3 Add nghttp2_session_create_idle_stream() API
See GH-436
2015-11-28 15:23:44 +09:00
Tatsuhiro Tsujikawa aacac613af Assert dep_stream is non-null to shut up scan-build 2015-11-28 00:50:29 +09:00
Tatsuhiro Tsujikawa a70445e122 Retain stream in reserved state on client side
Application may use it using nghttp2_stream_* functions, and traverse
its dependency.
2015-11-27 22:54:55 +09:00
Tatsuhiro Tsujikawa f23e34fa3c Handle response in nghttp2_on_begin_frame_callback
Previously, nghttp2_session_end_request_headers_received assumes
stream is still writable (in other words, local endpoint has not sent
END_STREAM).  But this assumption is false, because application can
send response in nghttp2_on_begin_frame_callback.  Probably, this
assumption was made before the callback was introduced.  This commit
addresses this issue.  Since all
nghttp2_session_end_*_headers_received functions are identical, we
refactored them into one function.
2015-11-27 22:50:13 +09:00
Tatsuhiro Tsujikawa 3d1d54e2ce Remove dead code 2015-11-27 21:13:44 +09:00
Tatsuhiro Tsujikawa faad041868 Use seq to break a tie for stream weight
Because of the nature of heap based priority queue, and our compare
function, streams with the same weight and same parent are handled in
the reverse order they pushed to the queue.  To allow stream pushed
earlier to be served first, secondary key "seq" is introduced to break
a tie.  "seq" is monotonically increased integer per parent stream,
and it is assigned to stream when it is pused to the queue, and gets
incremented.
2015-11-25 22:26:56 +09:00
Tatsuhiro Tsujikawa b08d5b1975 Explicitly treat stream_id 0 as error in nghttp2_session_change_stream_priority 2015-11-24 22:34:36 +09:00
Tatsuhiro Tsujikawa b53b1381b7 Fix bug that nghttp2_session_find_stream(session, 0) returned NULL
Previously, nghttp2_session_find_stream(session, 0) returned NULL
despite the fact that documentation said that it should return root
stream.  Now it is corrected, and it returns root stream as
documented.
2015-11-24 22:30:12 +09:00
Tatsuhiro Tsujikawa 8f970dec0e Update doc 2015-11-23 21:05:25 +09:00
Tatsuhiro Tsujikawa aa317c89ea Add API to change stream priority without sending PRIORITY frame
The added API is nghttp2_session_change_stream_priority().  This
provides the same functionality to re-prioritize stream when PRIORITY
frame.  is received, but we do it without PRIORITY frame.  This could
be useful for server to change pushed stream's priority silently.
2015-11-21 18:32:42 +09:00
Tatsuhiro Tsujikawa 63a50b1ccf Add nghttp2_session_check_server_session() API
This is very simple API, and it returns nonzero if session is
initialized as server.
2015-11-21 15:07:55 +09:00
Tatsuhiro Tsujikawa d7b0768ab8 Fix bug that dep_stream->sum_dep_weight was not updated 2015-11-20 21:24:54 +09:00
Tatsuhiro Tsujikawa 4db3828fa6 Reset last_writelen to 0 when stream is removed from tree
When stream is removed from tree, stream is either closed, or its
remote flow control window is depleted.  In the latter case, we
schedule this stream as fast as possible if its remote window gets
positive, since it did not sent anything in its turn.  To achieve
this, reset last_writelen to 0 when stream is removed from tree.
2015-11-20 00:19:17 +09:00
Tatsuhiro Tsujikawa 3bbb05f59b Optimize the case when only weight is changed 2015-11-20 00:19:02 +09:00
Tatsuhiro Tsujikawa 83cc2511e3 Remove flags parameter from nghttp2_pack_extension_callback
It has no usecase at the moment.  It is most likely that applications
know the flags when it submitted extension frame, no need to modify it
later.  Possibly feature bloat.
2015-11-17 21:29:21 +09:00
Tatsuhiro Tsujikawa b95a3c4b28 Consider to use CANCEL error code when closing streams with GOAWAY
For clients, CANCEL is more appropriate for both incoming/outgoing
streams.  For servers, CANCEL is appropriate for its pushed stream
(outgoing), but REFUSED_STREAM is more appropriate for incoming
stream.
2015-11-16 22:53:31 +09:00
Tatsuhiro Tsujikawa b918f9650a Don't send push response if GOAWAY has been received 2015-11-16 22:47:12 +09:00
Tatsuhiro Tsujikawa 7463493259 Use error code CANCEL to reset pushed reserved stream from remote 2015-11-16 22:20:27 +09:00
Tatsuhiro Tsujikawa eb4e402aae Merge branch 'master' into simple-extensions 2015-11-14 22:31:52 +09:00
Tatsuhiro Tsujikawa c6ef1c02b9 Switch to clang-format-3.6 2015-11-13 00:53:29 +09:00
Tatsuhiro Tsujikawa 269a100081 Add nghttp2_session_upgrade2(), deprecate nghttp2_session_upgrade()
To validate actual response body length against the value declared in
content-length response header field, we first check request method.
If request method is HEAD, respose body must be 0 regardless of the
value in content-length.  nghttp2_session_upgrade() has no parameter
to indicate the request method is HEAD, so we failed to validate
response body if HEAD is used with HTTP Upgrade.  New
nghttp2_session_upgrade2() accepts new parameter to indicate that
request method is HEAD or not to fix this issue.  Although, this issue
affects client side only, we deprecate nghttp2_session_upgrade() in
favor of nghttp2_session_upgrade2() for both client and server side.
2015-11-07 16:13:06 +09:00
Tatsuhiro Tsujikawa 5e7e479c6c Workaround HTTP upgrade with HEAD request
By default, we check the length of response body matches
content-length.  For HEAD request, this is not necessarily true, so we
sniff request method, and if it is HEAD, make sure that response body
length is 0.  But this does not work for HTTP Upgrade, since
nghttp2_session_upgrade() has no parameter to tell the request method
was HEAD.  This commit disables this response body length validation
for the stream upgraded by HTTP Upgrade.  We will add new version of
nghttp2_session_upgrade with the parameter to pass the request method
information so that we can handle this situation properly.
2015-11-07 10:56:40 +09:00
Tatsuhiro Tsujikawa dfbbb08124 Silence warning with scan-build 2015-11-06 20:07:40 +09:00
Tatsuhiro Tsujikawa 43b230685f Introduce NGHTTP2_NV_FLAG_NO_COPY_NAME and NGHTTP2_NV_FLAG_NO_COPY_VALUE 2015-11-05 22:48:30 +09:00
Tatsuhiro Tsujikawa 58d636abdb Revert 918f8cca36
For first scheduling, we might ignore weight.  This is OK since weight
is just a distribution of resource, rather than strict priority
ordering.
2015-11-02 23:38:29 +09:00
Tatsuhiro Tsujikawa be0d0e2995 Rename nghttp2_session_request_allowed as nghttp2_session_check_request_allowed 2015-10-30 22:48:27 +09:00
Tatsuhiro Tsujikawa 5d611d2e24 Merge branch 'master' into simple-extensions 2015-10-29 23:24:34 +09:00
Tatsuhiro Tsujikawa 41002a77d4 Remove bogus debug log 2015-10-29 23:01:02 +09:00
Tatsuhiro Tsujikawa 1c37abc0dd Add nghttp2_session_request_allowed() API function
This function is useful for the client application to know that there
is a chance that request can be sent.  If this function returns 0,
there is zero chance to make a request.

This commit also set error_code passed to
nghttp2_on_stream_close_callback to NGHTTP2_REFUSED_STREAM if request
is not sent.
2015-10-29 22:53:30 +09:00
Alexis La Goutte b4f363ef53 Fix Value stored to 'rv' is never read found by Clang Analyzer 2015-10-26 15:22:13 +01:00
Tatsuhiro Tsujikawa 5d3d377c03 Add missing NGHTTP2_EXTERN to public functions 2015-10-25 16:51:50 +09:00
Tatsuhiro Tsujikawa 4960583637 Increase NGHTTP2_MAX_OBQ_FLOOD_ITEM to avoid false positives 2015-10-25 16:23:29 +09:00
Tatsuhiro Tsujikawa 25bf567cd7 Don't always expect dynamic table size update
The encoder is not required to send dynamic table size update if the
table size is not changed from the previous value after accepting new
maximum value.
2015-10-24 17:49:51 +09:00
Tatsuhiro Tsujikawa 5bc3dfa1cd Shrink to the minimum table size seen in local SETTINGS 2015-10-24 17:49:50 +09:00
Tatsuhiro Tsujikawa ce74a30990 Use -fvisibility=hidden for internal functions
This will improve performance since we can avoid indirect call of
internal functions.  The downside is we now require libnghttp2 static
library to run unit tests.
2015-10-23 00:08:15 +09:00
Tatsuhiro Tsujikawa 46d7c30c17 Update doc 2015-10-21 00:18:27 +09:00
Tatsuhiro Tsujikawa 837e716306 Fix compile error with gcc 2015-10-15 00:30:42 +09:00
Tatsuhiro Tsujikawa 061a557839 Add nghttp2_option_set_user_recv_extension_type to opt-in incoming extension type 2015-10-15 00:17:07 +09:00
Tatsuhiro Tsujikawa 3785cf07ba Add simple HTTP/2 extension framework
Application can utilize this framework to send/receive user defined
extension frames.  These frames are expected not to change existing
protocol behaviour.
2015-10-11 17:46:23 +09:00
Tatsuhiro Tsujikawa 3f4b6f2b1c Clean up 2015-10-11 17:29:29 +09:00
Tatsuhiro Tsujikawa ba08948307 Merge branch 'master' of https://github.com/alagoutte/nghttp2 into alagoutte-master 2015-10-04 22:02:47 +09:00
Alexis La Goutte 81addc77b4 Add (and fix) -Wredundant-decls 2015-10-04 14:48:32 +02:00
Tatsuhiro Tsujikawa b1de6ca390 Update doc 2015-10-04 10:36:09 +09:00
Tatsuhiro Tsujikawa 45d4c9dece Add new error code NGHTTP2_ERR_PAUSE to send_data_callback
If application returns NGHTTP2_ERR_PAUSE from send_data_callback, it
means application processed all data, but wants to make
nghttp2_session_mem_send or nghttp2_session_send return immediately.
This is useful if application writes to fixed sized buffers, and there
is no room to write more data.
2015-10-03 17:16:42 +09:00
Tatsuhiro Tsujikawa cea76226b1 Avoid excessive WINDOW_UPDATE queuing 2015-10-01 01:19:57 +09:00
Tatsuhiro Tsujikawa d22ced77c0 Return fatal error if flooding is detected to close session immediately
This change adds new return error code from nghttp2_session_mem_recv
and nghttp2_session_recv functions, namely NGHTTP2_ERR_FLOODED.  It is
fatal error, and is returned when flooding was detected.
2015-09-30 22:19:03 +09:00
Tatsuhiro Tsujikawa 0cb8c82125 Detect flooding and tear down session 2015-09-30 00:44:08 +09:00
Tatsuhiro Tsujikawa ad5d73738c Return type of nghttp2_submit_trailer is int 2015-09-29 00:06:20 +09:00
Tatsuhiro Tsujikawa ef3e2c7764 Don't send WINDOW_UPDATE with 0 increment 2015-09-27 00:22:33 +09:00
Tatsuhiro Tsujikawa b5427d9651 Remove unused nghttp2_memdup 2015-09-26 22:44:08 +09:00
Tatsuhiro Tsujikawa 2d5b42693d Don't expect NULL terminated name/value in nghttp2_hd_deflate_hd
If it is called through libnghttp2 internally, name/value pairs are
all NULL-terminated.  But it is one of public API, and we cannot
expect that applications always make NULL-terminated string for
name/value pairs.
2015-09-26 22:41:05 +09:00
Tatsuhiro Tsujikawa 637288359e Ignore 0 length header partial block if it is not final 2015-09-25 22:53:35 +09:00
Tatsuhiro Tsujikawa 93ba75b602 Fix bug that headers in CONTINUATION were ignored after HEADERS with padding 2015-09-25 22:28:03 +09:00
Tatsuhiro Tsujikawa 1f72648b0e Fix compile error with --enable-debug and --enable-werror 2015-09-25 22:19:18 +09:00
Tatsuhiro Tsujikawa d804780c25 Fix compile error with Android NDK and --enable-werror 2015-09-25 01:45:28 +09:00
Tatsuhiro Tsujikawa 913b95336b Fix warning 'this decimal constant is unsigned only in ISO C90' 2015-09-25 00:03:38 +09:00
Tatsuhiro Tsujikawa 170f2a144a Fix warning with gcc 2015-09-23 18:26:02 +09:00
Tatsuhiro Tsujikawa 28fe3e7e89 More warning flags for --enable-werror 2015-09-23 16:49:45 +09:00
Tatsuhiro Tsujikawa 230b1f927f Use hash table for dynamic table lookup 2015-09-19 21:29:40 +09:00
Tatsuhiro Tsujikawa eacd6eeed2 Check header block limit after new stream is opened 2015-09-16 00:40:31 +09:00
Tatsuhiro Tsujikawa 991352d201 Just assign stream->cycle to descendant_last_cycle 2015-09-14 22:20:08 +09:00
Tatsuhiro Tsujikawa 1bbef4be74 Avoid usage of typeof and replace __builtin_offsetof with offsetof 2015-09-10 21:27:14 +09:00
Tatsuhiro Tsujikawa 918f8cca36 Honor stream->weight even if stream->last_writelen is 0 2015-09-09 21:20:59 +09:00
Tatsuhiro Tsujikawa 22b39f0337 Revert "last_write_len must not be 0, which makes weight useless"
This reverts commit e260006182.
2015-09-09 21:16:53 +09:00
Tatsuhiro Tsujikawa e260006182 last_write_len must not be 0, which makes weight useless 2015-09-08 02:27:59 +09:00