I was observing rare crashes on my mac, where I was using SecureTransport.
This would imply both the client (the test) and the server were working
in h2c mode. But this is against the test's logic - the first request
will be HTTP/1.1 (and upgrading protocol) and I wanted to send 3 HTTP/2
requests instead.
Fixes: QTBUG-77476
Change-Id: I048ca242e2096ca36dd112277807d1fee530150c
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
And send it in our 'SETTINGS' frame. Add an auto-test
for this and (as a bonus) - fix a bug accidentally
introduced by the previous change.
Task-number: QTBUG-77412
Change-Id: I4277ff47e8d8d3b6b8666fbcd7dc73c827f349c0
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Similar to TLS configuration that we can use on QNetworkRequest,
we can configure different options in our HTTP/2 handling by
providing QNetworkAccessManager with h2 configuration. Previously,
it was only possible internally in our auto-test - a hack with
QObject's properties and a private class. Now it's time to provide
a public API for this.
[ChangeLog][QtNetwork][QNetworkRequest] Add an ability to configure HTTP/2 protocol
Change-Id: I80266a74f6dcdfabb7fc05ed1dce17897bcda886
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
some assumptions were incorrect: our test server immediately sends
its SETTINGS frame, as a result we have to reply with client preface +
SETTINGS(ACK). So QVERIFY(!prefaceOK) was wrong from the beginning and
was only passing by pure luck.
Change-Id: Ie43f0d4ac41deb0e5339badaae6149a9b2f9d9b3
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
One of the tests above was unsetting a variable that enforces
the use of a temporary keychain. We have to set it back, otherwise
the test is failing. What surprises me though - why I had this
problem only locally and not on CI? Apparently, SecureTransport
is not covered by our configurations ...
Change-Id: I0ff1e3e304632869391ed61213c245b949d8c778
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
QNetworkAccessManager::connectToHostEncrypted()/connectToHost()
creates 'fake' requests with pseudo-schemes 'preconnect-https'/
'preconnect-http'. QHttp2ProtocolHandler should handle this
requests in a special way - reporting them immediately as
finished (so that QNAM emits finished as it does in case of
HTTP/1.1) and not trying to send anything.
We also have to properly cache the connection - 'https' or
'http' scheme is too generic - it allows (unfortunately)
mixing H2/HTTP/1.1 in a single connection in case an attribute
was missing on a request, which is wrong.
h2c is more complicated, since it needs a real request
to negotiate the protocol switch to H2, with the current
QNetworkHttpConnection(Channel)'s design it's not possible
without large changes (aka regressions and new bugs introduced).
Auto-test extended.
Fixes: QTBUG-77082
Change-Id: I03467673a620c89784c2d36521020dc9d08aced7
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
1. Use per-case QNAM objects
2. In a slots (connected to QNetworkReplies) - if an error detected - stop
the event loop (no reason to continue waiting) and then do normal
QVERIFY/QCOMPARE things.
3. In tests, check QTest::currentTestFailed after the event loop returned -
if an error was detected by a slot, no need to continue with QCOMPARE/QVERIFY
in the test itself.
Task-number: QTBUG-77053
Change-Id: I3827a629a2749becd3dc6eee7fd6994d96441e65
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
fix the bloody test for good - the idea to have a shared QNAM (shared
by test cases in this test) was somewhat wrong to start with.
Fixes: QTBUG-77053
Change-Id: I5755e96ec988e2dd546f527f3f902fc43914b0b7
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
to test different h2 modes: "h2c" (clear text with protocol upgrade),
"h2" (encrypted, negotiating via ALPN extension), "h2-direct" (encrypted,
no ALPN), "h2c-direct" (plain text, no protocol upgrade). This patch-set
is an amendment to the recent fix in the protocol handler where we were
crashing in "h2c-direct" mode.
Change-Id: I3ff5ed1396a59b72b59a95f927d404ccd202d0b8
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
1. Fix erroneous logic, which was triggered in 'h2' mode (non-TLS connection)
- after the initial protocol upgrade/POST request was handled, the server
(on Windows specifically) was erroneously handling upcoming DATA frames by replying
with another redirect response.
2. Make the test less heavy by sending 1 MB of Qt::Uninitialize instead of 10 MB
- theoretically this could cause a timeout before the redirected request finished
successfully.
Task-number: QTBUG-73873
Change-Id: I961e0a5f50252988edd46d0e73baf96ee22eef3f
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
[ChangeLog][QtNetwork][SSL] The Schannel backend now supports ALPN and
thus HTTP/2.
Change-Id: I1819a936ec3c9e0118b9dad12681f791262d4db2
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Some tests were fixed and others were skipped/blacklisted.
Task-number: QTBUG-63152
Change-Id: Ica7df555f8d152ee589865911130525101d4b941
Reviewed-by: Liang Qi <liang.qi@qt.io>
Unlike QNAM, our toy http2 server sends payload as one big chunk as soon as
it fits in the receive window's size. Internally, 'frame writer' splits this
payload into many DATA frames of the appropriate size (imposed either by the
default value or the one from the client's SETTINGS frame). If some test fails,
we can end up with a server waiting for the writer to send all the DATA frames
though it is not needed anymore - there is nobody to receive them after a failure.
This patch moves such a loop into the test server instead and stops the loop early
if needed.
Change-Id: Iea2dcd718d8f83386fd16004807f6447bf999435
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
QTestEventLoop (conveniently so) takes care of timeouts thus no
external QTimer/handling logic needed at all.
Change-Id: Id65ea928daec1e7d9380107e63916896f19d3d14
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
1. If a request was redirected or some error was encountered, we
try to reset the uploading byte-device.
2. Disconnecting from the byte-device is not enough, since we have a
queued connection, _q_uploadDataReadyRead() gets called even if
byte-device was deleted and thus sender() can return null -
we have to check this condition.
3. Update auto-test with a case where our server immediately
replies with a redirect status code.
Task-number: QTBUG-67469
Task-number: QTBUG-66913
Change-Id: I9b364cf3dee1717940ddbe50cba37c3398cc9c95
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
The vast majority is actually switched to QRandomGenerator::bounded(),
which gives a mostly uniform distribution over the [0, bound)
range. There are very few floating point cases left, as many of those
that did use floating point did not need to, after all. (I did leave
some that were too ugly for me to understand)
This commit also found a couple of calls to rand() instead of qrand().
This commit does not include changes to SSL code that continues to use
qrand() (job for someone else):
src/network/ssl/qsslkey_qt.cpp
src/network/ssl/qsslsocket_mac.cpp
tests/auto/network/ssl/qsslsocket/tst_qsslsocket.cpp
Change-Id: Icd0e0d4b27cb4e5eb892fffd14b5285d43f4afbf
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
1. Recently we have updated our receive window size to a larger value.
Unfortunately, this also results in auto-test pumping through
more data, which probably takes more time on CI.
At the moment we do not have any public API on QNAM's level to
customize HTTP/2 parameters (aka 5.10/FF and so on). So we use the fact
that QNAM is QObject and we can set a property on it. This property
is our Http2::ProtocolParameters object that allows us to configure:
- HPACK parameters (in 5.10 - noop)
- session receive window size
- different SETTINGS as described by RFC 7540, 6.5.2.
2. Undocumented environment variable to set ENABLE_PUSH is not needed
anymore.
3. In 5.11 Http2::ProtocolParameter will become a public API
and we'll introduce a new setter in QNAM.
Change-Id: If08fd5e09e7c0b61cf9700b426b60b5837b6b2e6
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
We were using the default ones, provided by RFC7540. It appears they are way
too restrictive and conservative: when downloading something relatively big,
a stream keeps spending the whole session/its own 'recv' windows and thus
we have to constantly send WINDOW_UPDATE frames. This significantly slows
down our HTTP/2 implementation, making it orders of magnitude slower than
HTTP/1.1. To fix this:
- We send SETTINGS_INITIAL_WINDOW_SIZE in the first SETTINGS frame
to inform our peer that per-stream WINDOW is bigger than 64Kb
- We increase the session's receive window size.
Task-number: QTBUG-63722
Change-Id: I31312fcfd5f0fc0aee6aaa5d3562cc7d1b931adc
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Remaining uses of Q_DECL_OVERRIDE are in:
src/corelib/global/qcompilerdetection.h
src/corelib/global/qglobal.cpp
doc/global/qt-cpp-defines.qdocconf
(definition and documentation of Q_DECL_OVERRIDE)
tests/manual/qcursor/qcursorhighdpi/main.cpp
(a test executable compilable both under Qt4 and Qt5)
Change-Id: Ib9b05d829add69e98a86238274b6a1fcb19b49ba
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Ville Voutilainen <ville.voutilainen@qt.io>
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Without TLS (and thus ALPN/NPN negotiation) HTTP/2 requires
a protocol upgrade procedure, as described in RFC 7540, 3.2.
We start as HTTP/1.1 (and thus we create QHttpProtocolHandler first),
augmenting the headers we send with 'Upgrade: h2c'. In case
we receive HTTP/1.1 response with status code 101 ('Switching
Protocols'), we continue as HTTP/2 session, creating QHttp2ProtocolHandler
and pretending the first request we sent was HTTP/2 request
on a real HTTP/2 stream. If the first response is something different
from 101, we continue as HTTP/1.1. This change also required
auto-test update: our toy-server now has to respond to
the initial HTTP/1.1 request on a platform without ALPN/NPN.
As a bonus a subtle flakyness in 'goaway' auto-test went
away (well, it was fixed).
[ChangeLog][QtNetwork][HTTP/2] In case of clear text HTTP/2 we
now initiate a required protocol upgrade procedure instead of
'H2Direct' connection.
Task-number: QTBUG-61397
Change-Id: I573fa304fdaf661490159037dc47775d97c8ea5b
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
tst_qvariant.cpp(80): warning C4309: 'initializing': truncation of constant value
tst_qvariant.cpp(4635): warning C4309: 'initializing': truncation of constant value
tst_qbytearray.cpp(1438): warning C4267: 'argument': conversion from 'size_t' to 'uint', possible loss of data
tst_qbytearray.cpp(1440): warning C4267: 'argument': conversion from 'size_t' to 'uint', possible loss of data
http2srv.cpp(64): warning C4018: '<=': signed/unsigned mismatch
tst_qinputdialog.cpp(352): warning C4804: '<=': unsafe use of type 'bool' in operation
Change-Id: Id012d88b7b20c5c9f128f2ef53753cc1d479f358
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Previously we were always setting SpdyWasUsedAttribute for SPDY/HTTP/2/HTTP/1.1
(true/false) which is confusing. Now if HTTP2AllowedAttribute was set to true on
a request, we set HTTP2WasUsedAttribute. Otherwise, as we did before, we're setting
SpdyWasUsedAttribute.
Change-Id: I0c44cfb5469fef0c12719baa951197ee2accee4a
Reviewed-by: Markus Goetz (Woboq GmbH) <markus@woboq.com>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
After it started to fail (somehow it's only OpenSUSE 42.1) again and again
and after a quick re-evaluation it appears the logic testing SETTINGS|ACK
is incorrect. We (client side) start by sending the preface and then
continue to send our request(s). The other side (server) starts from sending its
SETTINGS frame. These settings must be ACKed, but apparently it can happen,
that server receives a requests and sends a reply before it receives SETTINGS|ACK,
resulting in replyFinished (replyFinishedWithError) signal and event loop stopping.
As a result - QVERIFY(serverGotSettingsACK) fails.
Task-number: QTBUG-58758
Change-Id: I8184cf459b2b88f70c646171e0115c184237fad1
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
- Fix the case when we erroneously handled stream ID == 0 in a GOAWAY frame as
an invalid stream ID.
- _q_receivedReply: convert do{}while() loop into to while(){} to prevent
it from handling any frames after GOAWAY frame received and all active frame
finished.
- sendRequest - if we received GOAWAY, also clear spdyRequests in the connection
channel, otherwise it keeps re-trying to send requests!
- Http network connection channel never resets a protocolHandler in _q_encrypted/
_q_connected, which is BAD for HTTP/2, since HTTP/2 has unique per-connection
compression context and must be reset - now we recreate the protocol handler in
_q_encrypted or _q_connected (https/http).
- Update autotest.
Task-number: QTBUG-57600
Change-Id: Ib864ce52287bab23334ff43a83ba4b0b7cb52c60
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
HTTP/2 allows a server to pre-emptively send (or "push") responses (along
with corresponding "promised" requests) to a client in association with a
previous client-initiated request. This can be useful when the server
knows the client will need to have those responses available in order
to fully process the response to the original request.
Server push is semantically equivalent to a server responding to a request;
however, in this case, that request is also sent by the server, as a
PUSH_PROMISE frame.
The PUSH_PROMISE frame includes a header block that contains a complete set
of request header fields that the server attributes to the request.
After sending the PUSH_PROMISE frame, the server can begin delivering the
pushed response as a response on a server-initiated stream that uses the
promised stream identifier.
This patch:
- fixes the HPACK decompression of PUSH_PROMISE frames;
- allows a user to enable PUSH_PROMISE;
- processes and caches pushed data for promised streams;
- updates auto-test - emulates a simple PUSH_PROMISE
scenario.
Change-Id: Ic4850863a5e3895320baac3871a723fc091b4aca
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Source (.cpp and .h) files should not be executable.
Change-Id: I021d8733185d73d071fcaf3df7e529862a490b63
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@theqtcompany.com>
Introduce new entity: class Http2::Frame with accessors like
payloadSize/type/flags/streamID etc. (they actually read
/interpret raw bytes from a frame's buffer) instead of
duplicating this functionality in reader/writer classes.
Delete defaulted members and remove explicitly defined
move ctors/operators (not needed actually).
Update auto-test ('HTTP/2 server') to use these new classes.
Change-Id: Ie3516efbd095704e212142eef9e792323678ccfa
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Since headersFrame is a reference to a vector's element, clearing this
vector before accessing headersFrame.flags is not a good idea, must be
done later.
Change-Id: I80eee0761ac1cad580e979be9371ec7588a694ac
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Recently enabled cleartext fails to build with QT_NO_SSL - fix
test and QNAM.
Change-Id: I467edab8e4eb5113715ad2d3b3022e0d8c027de8
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
HTTP/2 does not require TLS connection, it can work in a cleartext mode.
Plus at the moment only OpenSSL backend allows HTTP/2 negotiation
via ALPN/NPN (and none of our CI configurations with OpenSSL supports
these extensions, rendering HTTP/2 auto-test useless). This patch
implements cleartext HTTP/2 ('h2c') in 'direct' mode - this is
allowed if a client has a prior knowledge that HTTP/2 is supported by
a server.
Change-Id: I4978775e9732c40bc77f549b83bb4a5d1761887e
Reviewed-by: Alex Trotsenko <alex1973tr@gmail.com>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Add autotest for QHttp2ProtocolHandler. This patch contains a very simplistic
"in-process HTTP2 server" for testing the protocol's basic logic/flow control/error
handling and emulating possible scenarios.
Task-number: QTBUG-50956
Change-Id: Ie02d3329c5182277a3c7c84f1bae8d02308e945d
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Friedemann Kleint
Qt Forward Merge Bot
Timur Pocheptsov
Liang Qi
Mårten Nordheim
Sami Nurmenniemi
Thiago Macieira
Kevin Funk
Timur Pocheptsov
Edward Welbourne
Timur Pocheptsov