summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* ssh-agent: Make public key parsing even robusterwip/dueno/openssh-parseDaiki Ueno2018-07-141-0/+15
| | | | | | | This amends commit f3f3cc70 to take into account of the fact that the key type is prefixed to the decoded blob. Suggested by Mantas Mikulėnas in: https://bugzilla.gnome.org/show_bug.cgi?id=795699
* Merge branch 'wip/dueno/gitlab-ci' into 'master'Daiki Ueno2018-06-122-2/+85
|\ | | | | | | | | build: Enable gitlab-ci See merge request GNOME/gnome-keyring!1
| * build: Enable gitlab-ciDaiki Ueno2018-06-121-0/+83
| |
| * pkcs11: Don't null terminate PKCS #11 string fieldsDaiki Ueno2018-06-121-2/+2
|/
* doap: add <description>Daiki Ueno2018-05-251-0/+10
|
* doap: add myself as maintainerDaiki Ueno2018-05-251-0/+8
|
* Revert "ssh-agent: Set may_block to TRUE while waiting for ssh-agent process"Daiki Ueno2018-05-131-1/+1
| | | | This reverts commit fb0d66553753bdc0d700cb5c0bb2803d0690e9ff.
* ssh-agent: Set may_block to TRUE while waiting for ssh-agent processDaiki Ueno2018-05-071-1/+1
| | | | | Potentially fix the busy loop reported in: https://bugzilla.gnome.org/show_bug.cgi?id=794848
* ssh-agent: Don't treat packet write error as fatalDaiki Ueno2018-05-071-2/+5
|
* ssh-agent: Make public key parsing robusterDaiki Ueno2018-05-073-11/+24
| | | | | | | | | | | | | | | | | Previously, _gkd_ssh_agent_parse_public_key() accepted OpenSSH v1 keys, because the second component of the key line looks like a valid base64 blob: 2048 65537 2444136... This patch checks that the component is really base64 encoded, by checking the length is a multiple of 4. Note that this solution is not perfect, as there could be a key with a public exponent whose decimal length is multiple of 4. More thorough approach would be to call ssh-keygen -l on each public key. https://bugzilla.gnome.org/show_bug.cgi?id=795699
* Update Romanian translationDaniel Șerbănescu2018-04-181-203/+202
| | | | (cherry picked from commit d8b6de0c65c4206bc47942963c285d4ee76cf0c6)
* Updated Slovenian translationMatej Urbančič2018-03-271-253/+241
|
* login: Use password from login keyring once for the same interactionDaiki Ueno2018-03-241-7/+13
| | | | https://bugzilla.gnome.org/show_bug.cgi?id=794631
* egg: Port cosmetic fixes to egg-secure-memory.c from libsecretDaiki Ueno2018-03-231-4/+4
|
* login: Use the same label as before when storing passwordDaiki Ueno2018-03-202-1/+6
| | | | https://bugzilla.gnome.org/show_bug.cgi?id=794500
* ssh-agent: Don't be too verbose on password promptDaiki Ueno2018-03-201-3/+1
| | | | | | | This partially reverts the change in 869b5c6d, so as not to display duplicate words on the password prompt. https://bugzilla.gnome.org/show_bug.cgi?id=794500
* build: Suppress compiler warnings with -Wdiscarded-qualifiersDaiki Ueno2018-03-201-2/+2
|
* ssh-agent: Make EOF handling robusterDaiki Ueno2018-03-193-5/+12
| | | | https://bugzilla.gnome.org/show_bug.cgi?id=794369
* ssh-agent: Allow opening multiple connections to inferior ssh-agentDaiki Ueno2018-03-196-46/+51
| | | | | | | | | Previously, it keeps only one connection to the inferior ssh-agent process. That prevented simultaneous access to gnome-keyring's ssh-agent service. With this patch, it always opens a new connection to the inferior ssh-agent process when a new client connects. https://bugzilla.gnome.org/show_bug.cgi?id=794369
* ssh-agent: Use the same parameters for accessing login keyringDaiki Ueno2018-03-191-2/+3
| | | | | | | | | | | | | When looking up a secret in the login keyring, do not supply any schema in the criteria, while using "org.freedesktop.Secret.Generic" as schema when storing it. This is for backward compatibility with gnome-keyring 2.29, which used "org.gnome.keyring.EncryptionKey" as schema. In addtion, use the same label for the newly stored passwords as before. https://bugzilla.gnome.org/show_bug.cgi?id=794368
* login: Allow different sets of secret attributes for lookup/storingDaiki Ueno2018-03-191-5/+23
| | | | https://bugzilla.gnome.org/show_bug.cgi?id=794368
* ssh-agent: Propagate stderr to journal when spawning ssh-addDaiki Ueno2018-03-194-3/+51
| | | | | | | | | | | ssh-add fails in certain occasions, such as when the file permissions of private key is not unsafe. To help diagnostics, propagate the stderr output from the command to journal. As the ssh commands send error message with trailing CR for each line, we need to scrub it so as not to confuse journald. https://bugzilla.gnome.org/show_bug.cgi?id=794361
* Release 3.28.0.13.28.0.1Daiki Ueno2018-03-132-1/+4
|
* build: Fix linking of test modules with "-z defs"Daiki Ueno2018-03-132-2/+2
| | | | | | Split out mock-interaction.c from libegg.la to libegg-test.la. https://bugzilla.gnome.org/show_bug.cgi?id=794274
* Release 3.28.03.28.0Daiki Ueno2018-03-122-1/+4
|
* build: Fix tap-driver to parse TAP directiveDaiki Ueno2018-03-051-2/+5
|
* Release 3.27.923.27.92Daiki Ueno2018-03-052-1/+6
|
* Update POTFILES.inPiotr Drąg2018-03-041-0/+1
|
* build: Improve ssh-agent command detectionDaiki Ueno2018-03-041-9/+11
| | | | | | | With this patch, configure skips the checks for the commands when --disable-ssh-agent; otherwise, error out if any of those commands ias unavailable. In either case, set SSH_AGENT and SSH_ADD macros so that ssh-agent code compile.
* ssh-agent: Use stock ssh-agentStef Walter2018-03-0429-3928/+3194
| | | | | | | | | | | | | | | | | This patch removes our own implementation of ssh-agent and switches to using the ssh-agent program provided by OpenSSH. We can't simply drop the ssh-agent functionality from gnome-keyring, as it enables the following: * Automatic loading and unlocking of keys * Prompting in the UI Instead we wrap the ssh-agent program as a subprocess and augment the protocol as we need. Co-authored-by: Daiki Ueno <dueno@src.gnome.org> https://bugzilla.gnome.org/show_bug.cgi?id=775981
* configure: Require gcr 3.27.90Daiki Ueno2018-03-041-1/+1
| | | | https://bugzilla.gnome.org/show_bug.cgi?id=775981
* ssh-agent: Add GkdSshAgentInteraction APIDaiki Ueno2018-03-045-1/+446
| | | | | | | This could be used by the new implementation of ssh-agent, through the GcrSshAskpass and GkdLoginInteraction. https://bugzilla.gnome.org/show_bug.cgi?id=775981
* egg: Import mock-interaction.[ch] from gcrDaiki Ueno2018-03-043-1/+141
| | | | https://bugzilla.gnome.org/show_bug.cgi?id=775981
* login: Add GkdLoginInteraction APIDaiki Ueno2018-03-045-0/+494
| | | | | | | This decorates underlying GTlsInteraction with a capability of retrieving/storing passwords in the login keyring. https://bugzilla.gnome.org/show_bug.cgi?id=775981
* login: Add non-varargs version of gkd_login_*_password() functionsDaiki Ueno2018-03-042-32/+104
| | | | | | | Those functions could be used in GkdLoginInteraction where the fields are stored in a table. https://bugzilla.gnome.org/show_bug.cgi?id=775981
* configure: Require GLib 2.44Daiki Ueno2018-03-041-3/+3
| | | | https://bugzilla.gnome.org/show_bug.cgi?id=775981
* gkm-timer: Use monotonic time exclusivelyDaiki Ueno2018-03-021-7/+3
| | | | | | | | Previously, it used g_get_real_time(), g_get_current_time(), and g_get_monotonic_time() in a mixed manner. To measure the elapsed time, it would be sufficient to use g_get_monotonic_time() only. https://bugzilla.gnome.org/show_bug.cgi?id=791932
* pkcs11: Don't install p11-kit module configurationRay Strode2018-03-014-16/+0
| | | | | | | | | It doesn't work well in multithreaded environments, and it overlaps with better maintained options like SoftHSM. To avoid any confusion, stop installing the p11-kit configuration for that module so that it is not registered by default. https://bugzilla.gnome.org/show_bug.cgi?id=791401
* Fix Spanish translation headerPiotr Drąg2018-02-231-1/+1
|
* Update Scottish Gaelic translationGNOME Translation Robot2018-02-231-219/+268
| | | | (cherry picked from commit 66ca4f41faa3df5b19f67a1d8be1982515ecc903)
* egg: Fix null termination behavior of egg_secure_strndup()Daiki Ueno2018-02-212-0/+26
| | | | | Even if the given string is shorter than n, the result should be null terminated. This matches the behavior of strndup().
* login: Fix memleak when secret store is not foundDaiki Ueno2018-02-191-0/+1
|
* login: Don't call g_object_unref() on NULLDaiki Ueno2018-02-171-1/+1
|
* login: Don't assume error if there is no matching tokenDaiki Ueno2018-02-171-1/+4
| | | | | gck_modules_token_for_uri() could return NULL without setting error if there is no matching token.
* secret-store/test*: Cleanup temporary directoryDaiki Ueno2018-02-151-0/+1
|
* daemon: Set NoDisplay=true for autostart filesJeremy Bicha2018-01-282-0/+2
| | | | | | | | | | This hides them from Ubuntu's Startup Applications app https://bugzilla.gnome.org/show_bug.cgi?id=777513 The SSH service is intentionally shown since there is no other easy GUI way currently to tell gnome-keyring to let something else handle ssh-agent.
* Release 3.27.43.27.4Daiki Ueno2018-01-082-1/+6
|
* main: Remove unnecessary HAVE_LOCALE_H checksTing-Wei Lan2018-01-072-4/+0
| | | | https://bugzilla.gnome.org/show_bug.cgi?id=792278
* build: Fix issues on desktop file generationDaiki Ueno2017-12-132-3/+4
| | | | | | This fixes a couple issues in commit 8818e939: - intermidiate files are left after "make distcheck" - to incorporate translations, it didn't properly use a suffix rule
* ssh-agent: Support SHA2 extension for RSA signaturesJakub Jelen2017-11-294-64/+129
| | | | | | | | | | | | | | * Includes the test adjustments IETF draft: https://tools.ietf.org/html/draft-ietf-curdle-rsa-sha2-12 This will be most probably needed for OpenSSH 7.7p1: https://bugzilla.mindrot.org/show_bug.cgi?id=2799 Signed-off-by: Jakub Jelen <jjelen@redhat.com> https://bugzilla.gnome.org/show_bug.cgi?id=790910
View Lorry Controller Status