pull in changes from qtcreator #12
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
So far, the only option is whether to use the default proxy or not. More will follow. Change-Id: Icd1ec407a0269cc8e7577901853d3873fd03015c Reviewed-by: Tobias Hunger <tobias.hunger@digia.com>
This is the "direct-tcpip" port forwarding specified in RFC 4254. Change-Id: I1ffa2e923b4479c7211b1b4304e66895b565fb64 Reviewed-by: hjk <qthjk@ovi.com> Reviewed-by: Christian Kandeler <christian.kandeler@nokia.com>
Change-Id: I0b104942d97238aff8ac21e0e047a17079a53181 Reviewed-by: Fawzi Mohamed <fawzi.mohamed@digia.com>
Otherwise we might get false positives in the check in SshConnectionManager. Change-Id: Ia54be53ce294c75fbfad2e6475c9d489d2639504 Reviewed-by: Tobias Hunger <tobias.hunger@digia.com>
While we want to conform to the specs, there are systems out there today whose SSH servers send non-conforming identifications strings. We now enable API clients to switch the respecive checks off, and we do so ourselves in the RemoteLinux plugin, since the only known problems are with OpenSSH servers. Change-Id: I9e6f9076f2dc7435a0bde7016f99cfb2fcb30a9c Reviewed-by: Tobias Hunger <tobias.hunger@digia.com>
These appear when compiling in release mode. Change-Id: I76ee3b1b8d728fd839d713ee4f914b6965851b99 Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
There is now at least one Linux distribution (openSUSE 12.3) that
disables the "password" authentication method in its default
sshd_config, while others allow it, but disable "keyboard-interactive".
This patch tackles the problem as follows:
1) Implement RFC 4256 ("keyboard-interactive") and make this method
available in the API.
2) In addition, the API offers to try both password-based methods
one after the other, until one has succeeded or all have failed.
3) Dialogs continue to offer just the choice between "Password" and
"Key", as to not confuse users. Internally, "Password" uses the feature
described in 2).
Task-number: QTCREATORBUG-9568
Change-Id: Ic81bd5d2dc4b1332ea1a8be938c19811c21a9087
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Christian Kandeler <christian.kandeler@digia.com>
* adds missing fields dmp1, dmq1 and iqmp to RSA structure * verified with openssl 1.0.1 Change-Id: I2891cd04303f7b44bbe24a844b2d74669c07f521 Reviewed-by: Christian Kandeler <christian.kandeler@digia.com>
If there are still open channels in a connection being released, that is not necessarily a mistake: The channels could already be in the closing state, and it would be unreasonable for us to expect API clients to always wait for confirmation before releasing the connection, as this can complicate code quite a bit, turning a synchronous operation into an asynchronous one. Change-Id: If8c604c9ee1294728e7947c8d5c8130d6e704b49 Reviewed-by: Tobias Hunger <tobias.hunger@digia.com>
Botan requires to be initialized before first use. We do that on-demand, but currently we miss one of the two places where this is necessary. The problem only manifests itself with multithreading, because Botan itself also does an on-demand initialization, but does not guard against concurrent access when doing so. Task-number: QTCREATORBUG-9980 Change-Id: Id65c428e3d8842c8a6f16942b11d52412239d3c9 Reviewed-by: David Kaspar <dkaspar@blackberry.com> Reviewed-by: Christian Kandeler <christian.kandeler@digia.com>
Provide a QProcess-like abstraction that can be used
to implement processes running locally or on a remote
device. Objects of a concrete class implementing the functionality
are created by IDevice objects.
Current implementations are:
- Local execution (QProcess-based), provided via the DesktopDevice.
- Remote execution via SSH.
- A specialized case of the former for remote Linux systems (provided by
LinuxDevice).
The latter is already being used in a number of places. As a result, lots of
code dealing with details such as setting the remote environment could be
moved to a central location. These things are no longer the concern of whoever
is wishing to run a remote process.
Change-Id: I919260ee6e77a020ca47226a4a534e7b8398106f
Reviewed-by: hjk <hjk121@nokiamail.com>
Change-Id: Id1541f83f431171dbdd94d5dd48f93e1c2cdf6fb Reviewed-by: Christian Kandeler <christian.kandeler@digia.com>
Calling fromLocal8Bit() for literals is wrong, since that depends ont the user's environment. We use latin1 strings exclusively in Qt. Change-Id: I0cd0986787ea630425773672d3a892fbd0f0a77a Reviewed-by: hjk <hjk121@nokiamail.com>
So far we ignored crashes and unexpected exits of the remote SFTP service under the assumption that the SSH server would catch these itself and act accordingly. This is not the case, however: OpenSSH, for instance, does not even realize if its sftp-server binary is not present at all. As a result, Qt Creator waits indefinitely for an SFTP operation to finish. Now we emit an error and close the offending channel. Task-number: QTCREATORBUG-10339 Change-Id: I132ed4a0098434a4cfce6056b964bd6363951fd7 Reviewed-by: Volker Vogelhuber <wiendl@web.de> Reviewed-by: Christian Kandeler <christian.kandeler@digia.com>
Change-Id: Ib608bb49e26781aef1914085a5d801fcdcd5eb56 Reviewed-by: Leena Miettinen <riitta-leena.miettinen@digia.com>
Change-Id: I3a1803cc57050c6fb90c5d143ffb2d4e98765be3 Reviewed-by: Christian Kandeler <christian.kandeler@digia.com>
Change-Id: I2f90c8ae7b5e968b9de882833f8661ab540a9232 Reviewed-by: hjk <hjk121@nokiamail.com>
The statement in the RFC was misunderstood. Task-number: QTCREATORBUG-12884 Change-Id: Icd533362ebf95d963cacc8d11da03360e742f070 Reviewed-by: Christian Kandeler <christian.kandeler@digia.com>
Channel exit is completely normal if we closed the channel, so this must not cause an error message. Change-Id: I57f5165b339a52a25118ad5e357f41334dadebc4 Reviewed-by: Christian Kandeler <christian.kandeler@digia.com>
New OpenSSH versions do not support the (required) CBC modes out of the box anymore, so let's add some CTR ones. Task-number: QTCREATORBUG-13340 Change-Id: Ia3e38be3aab95be258e64396283736d246c8b93b Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com> Reviewed-by: Christian Kandeler <christian.kandeler@digia.com>
This can be triggered rather quickly by clicking "Run" for a short- lived application on a QNX device. The state in this situation is QSsh::Internal::AbstractSshChannel::SessionRequested. Change-Id: I4447b85a4e8d6ba63e30f3783d29bb0c816e0dff Reviewed-by: Christian Kandeler <christian.kandeler@theqtcompany.com>
- It is not an error if the server replies to one of our earlier requests while we are in closing state. This can happen if we close the channel shortly after sending the other request. - Also change the message itself, which could be interpreted as "packet corrupt" when we actually meant "packet unexpected". Change-Id: I735c67b2a9b41af0c5e0b8d229369d94ec37277c Reviewed-by: hjk <hjk121@nokiamail.com>
... and connect it using Qt5-style connects. Change-Id: Ic7f36949b38d4773f5ac0f04853abf93bebcf467 Reviewed-by: Christian Kandeler <christian.kandeler@theqtcompany.com>
If closeChannel() is called in between our channel open request to the
server and the corresponding reply, we cannot forward the close request
to the server, as we don't have its channel id yet. So wait until we do.
Our failure to correctly handle this sequence of events was the root cause
for the following user-visible errors:
- A (since-removed) Q_ASSERT in ~SshRemoteProcess() was hit.
- The server closed the connection because we referred to an invalid
channel ("Received ieof for nonexistent channel -1").
Commits 26920307f0 and 3027bcc952 are also related to this issue.
Change-Id: I4994d85f5b21a72682f75389cdf8769738bd6768
Reviewed-by: hjk <hjk121@nokiamail.com>
Change-Id: Ic537fd123478a4f8fe4e54736d444fa64fd62d9c Reviewed-by: hjk <hjk121@nokiamail.com>
Change-Id: I5f10bd801bb5cf43e58193c41e62d9ea2f9cb645 Task-number: QTCREATORBUG-13339 Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
We keep connections open to avoid re-doing the authentication procedure when the user runs or deploys a remote app repeatedly. However, the connections should not be around indefinitely; we therefore now close them after about five minutes of inactivity. Change-Id: Ie677eb133b618d34c08528da2f2bc221472316b3 Reviewed-by: hjk <hjk121@nokiamail.com> Reviewed-by: Christian Kandeler <christian.kandeler@theqtcompany.com>
Probably just a thinko in the original code. Change-Id: Ie089737a3501c5aede9a2234c10e4f1c1611d952 Reviewed-by: Tobias Hunger <tobias.hunger@theqtcompany.com>
Change-Id: I9cdfa51dadf1014ee21a37ed97f4c543eb5c8fbb Reviewed-by: Christian Kandeler <christian.kandeler@theqtcompany.com>
Change-Id: I84834f37dd15108ed8c5fbf5353bcabc4e564f70 Reviewed-by: hjk <hjk@theqtcompany.com> Reviewed-by: Christian Kandeler <christian.kandeler@theqtcompany.com>
The Botan library ships a pkg-config file, which helps finding where the library has been installed. This commit does not remove the previously hardcoded paths, in order to keep compatibility for existing users of the library.
|
+1, this fixes many bugs with Qt5 |
We claimed to support the diffie-hellman-group14-sha1 protocol, but forgot to actually handle it. Change-Id: I3193997d7270328700486a15a6281ebf8528ed56 Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
…n non-strict mode
upstream commit:
commit 1c6c64df2058cb8f2a06fce1a4ffa5476a80b762
Author: Christian Kandeler <christian.kandeler@qt.io>
Date: Tue Oct 16 17:38:12 2018 +0200
SSH: Turn some manual tests into autotests
Change-Id: Id098709ff74dfc31a26ef42d559ce35d0120c0c3
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
There is no minimum requirement for the per-channel maximum packet size. Change-Id: Iced6e3bb1b77f7642d39d4a113ab4236af6b247b Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
There are other "channel open" requests than the one for forwarded TCP/ IP. Factor out the generic parts. Change-Id: I593a9a075d87d57850697459ab0814c07b75c4d4 Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
Change-Id: Ia7b15e784cb098bc7c6c6be2748d772192187e97 Reviewed-by: hjk <hjk@qt.io> Reviewed-by: Christian Kandeler <christian.kandeler@qt.io>
This format has been around for a while, but with release 7.8 has become the default for all keys generated by ssh-keygen. Change-Id: I0d8f666996832de49aad4f6bd246a5ed820df6cd Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
The SshX11InfoRetriever object must not be allocated on the stack, and the failure signal was potentially emitted too early for the test to catch it. Change-Id: Iac53546deee183c8f02bafdcc11a7910f3e392c0 Reviewed-by: Christian Stenger <christian.stenger@qt.io>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
lots of fixes, and adds support for e. g. elliptic curve stuff.