- Jan 24, 2025
-
-
Change-Id: I42b10e4025d2b37c2c377e874bc5abb1005b6e6c
-
- Jan 23, 2025
-
- Jan 22, 2025
-
-
ovari authored
Change-Id: I10597a207dcca24ead8e936680c60af7eb36bc03
-
- Jan 17, 2025
-
-
Adrien Béraud authored
Change-Id: I0c093ea4e1834866fd08abb30d51d5043fe1f455
-
- Jan 16, 2025
-
-
François-Simon Fauteux-Chapleau authored
This patch removes the custom retry limit and timeout in the NatPmp::readResponse function; they were not necessary given that libnatpmp already implements this functionality. This change means that readResponse can no longer return the NATPMP_TRYAGAIN error code. The fact that it could previously seems to be what enabled the bug described in the following issue: jami-client-qt#1898 More precisely, it seems that the readResponse function was repeatedly returning NATPMP_TRYAGAIN (even though the mapping requests had actually succeeded), which isn't a "fatal" error (as determined by the NatPmp::isErrorFatal function) and therefore didn't cause the error counter to be incremented and the IGD to be invalidated. GitLab: #43 Change-Id: I1374ba1b40e867399dab0f2dc4f6f75424f4a6b6
-
- Jan 15, 2025
-
-
François-Simon Fauteux-Chapleau authored
This patch fixes a bug where FAILED mappings with the "auto-update" option enabled are left in the local list, even after a new mapping has been requested. This can cause the list to grow substantially over time and was partly responsible for the bug described in the following issue: jami-client-qt#1898 GitLab: #45 Change-Id: I009ff791ae90d10f64d46dede3b2cbfc9c1b7160
-
- Jan 06, 2025
-
-
François-Simon Fauteux-Chapleau authored
GitLab: #632 Change-Id: I54a9e41f1cd026183a0fa3aa0891e49e99785517
-
- Jan 05, 2025
-
-
ovari authored
authentification → authentication chunck → chunk DTLS GnuTLS IPv4 IPv6 MiM attack → MITM attack MTU non-reliable → unreliable non-valid → invalid OCSP OK Renogotiate → Renegotiate retry → reattempt shuting → shutting tried → attempt try → attempt URI x509 → X.509 Change-Id: I5ab67f1fb10b24aa8de79217cba219137de6a88b
-
- Dec 22, 2024
-
-
ovari authored
Change-Id: Id59bce284de0a8a683c62d5eb4388fe228f4d950
-
- Dec 19, 2024
-
-
ovari authored
I/O queue ID intermidiate → intermediate IP address IPv4 IPv6 not initialized → Uninitialized STUN too much → Too many Change-Id: Ice56bef0cf363ce676f516ed32f475c9cd248b15
-
- Dec 15, 2024
-
-
ovari authored
… Change-Id: I8dfaa1fab2dd712bc09d9689b9eaf657b72a2c5e
-
- Dec 13, 2024
-
-
Adrien Béraud authored
Gitlab: #631 Change-Id: I17b5e5edd7bf4d7a7ce26d766146d796d41c8889
-
- Dec 04, 2024
-
-
François-Simon Fauteux-Chapleau authored
The UPnPContext class currently has hardcoded limits on the number of mappings that need to be "available" at any given time. This patch adds the ability to modify these limits and sets them to zero in the upnpctrl tool. This changes the behavior of the tool in two ways: 1) setting the minimum limit to zero means that no mappings are opened when upnpctrl starts; 2) setting the maximum limit to zero means that the "close" command will always cause the relevant mapping to be closed immediately instead of (potentially) being put back in the list of available mappings. Change-Id: Idcf4262ce80a8b39f19ce7d3a201b40b3ea56911
-
- Dec 02, 2024
-
-
ovari authored
... → … wish → want Change-Id: Iedfd8fcf6cdb56819b80a954d511f83eb2b8e906
-
- Nov 27, 2024
-
-
ovari authored
ssh → SSH stun server → STUN server try → attempt turn server → TURN server UPNP → UPnP Change-Id: Ia31a3baae908850a7e99d50128447f2eb9113df1
-
won't → unable to Change-Id: I1ce5af3ac16ed364df62d20338507b714a47b913 GitLab: jami-client-qt#1730
-
ovari authored
platform → platforms Change-Id: I99a79c2073e4c9a593fd264f75e22e4e1df1afcc GitLab: jami-client-qt#1730
-
- Nov 26, 2024
-
-
ovari authored
{can't, won't} → unable to plateform → platform Change-Id: I7723ac5ed881572201fb8367f1e5c84c52330888 GitLab: jami-client-qt#1730
-
Andreas Traczyk authored
Change-Id: I3acce7150d5f21bfdaf6174060e8671086c4f56e
-
- Nov 24, 2024
-
-
ovari authored
3-character ellipsis ... → 1-character ellipsis … Change-Id: If14c822aaf7c32a50a5367c26cc2deededebfd90 GitLab: jami-client-qt#1730
-
- Nov 18, 2024
-
-
Change-Id: Ied3cc750a1dca7dd14f3ef75ffc8729c09bf7087
-
- Nov 11, 2024
-
-
Adrien Béraud authored
* fix: use the same mutex in callback and to wait * fix: notify with mutex locked * fix: avoid accessing this from callback, preventing potential use-after-free * allows to stop waiting for upnp mappings immediately when the transport is stopped * perform single allocation of shared state, and free the memory of mutex and cv after mapping completes Change-Id: I573937d2e8a90b31501a237607bbc7f52942bcea
-
- Oct 25, 2024
-
-
Adrien Béraud authored
Change-Id: Ic7c5bf97c9ac159519eea2c265787cf47d378ce7
-
- Oct 24, 2024
-
-
Louis Maillard authored
On Ubuntu/debian, it have happenned that some build have inconsistent paths like /usr/usr/lib. To avoid this, hard-coded most of the paths like we are doing in Fedora. Change-Id: I8c1a3567490e02ac11ec57315bc3a7cc38d02ea0
-
Louis Maillard authored
Cleaned all the dead code in debian/ubuntu post install script. Change-Id: Idb23b5974e544d704ee9feb753990bf185575ec5
-
Louis Maillard authored
Building dhtnet for ubuntu 24.10 require the addition of libzstd for compilation to work (linker require it). Change-Id: I2168a84ee9fde3dc21925f000d5f136ef76fb207
-
Add dhtnet.spec for RPM build and fedora_40,39 dockerfile for env build. Fix and changes was required in Cmake instruction in order to build: - Dependencies can now be built in static mode even with dynamic library - Remove occurence of INSTALL_DIR path in dhtnet.pc and dnc.service but only when option TRIM_PREFIX_PATH is turned On. Fix and replace previously-reverted review 28703 Replace change id (I2f2161d7c65241a35c4acbcc006ba27bfcc1f176) Change-Id: I5a80317d83791743ba6ee443fae7f89156e8013b
-
Adrien Béraud authored
Change-Id: Ifef17404e31f7478d310179255092d0e17743b16
-
Adrien Béraud authored
Only process the callback if it was not already done by the tls shutdown callback. Change-Id: Iee9b479100268c4815f40f7b5d51b9db7dbb22cd
-
- Oct 23, 2024
-
-
Avoid re-using the same ID from the previous connection attempt, which prevents various issues, including the new connectionInfo being destroyed just after the retry by the async shutdown callback from the previous connection. Change-Id: I9c9017f7d0b3107dd8311b5066ba558d8c87b053
-
* mtx -> mutex * cbIds_ -> pendingCbs_ Change-Id: Iea5c1ed55efe197024f25adda4860e599bc94892
-
- Oct 22, 2024
-
-
Upload generated artefacts on nigtly folder and keep the last 10 build in Jenkins for now. Avoid using the exclude method used before for artifact and switch to creating a publish/ folder as it's easier for rsync. Change-Id: I311a1ef457bb96b8106263e21d5e9abc2a5f43e3
-
Adrien Béraud authored
And remove unneeded check added in previous commit Change-Id: I1b1f946fde02a09c9695b1f092c041fbd9278e06
-
- Oct 21, 2024
-
-
Adrien Béraud authored
Change-Id: I3bad2a7673e9faf860a4a34544c51ef2df2d387c
-
- Oct 18, 2024
-
-
François-Simon Fauteux-Chapleau authored
The change made in commit 41d80670 to prevent a mapping's NotifyCallback from being called more than once unnecessarily didn't correctly handle the case where the callback is null. As a result, the callback would sometimes not be called at all in situations where it should have been. https://git.jami.net/savoirfairelinux/jami-client-ios/-/issues/410 Change-Id: Icce78210b8be873be244e93e430d8c6123619db9
-
- Oct 17, 2024
-
-
This patch modifies the Mapping class to prevent a mapping's NotifyCallback from being called more than once if the mapping's state hasn't changed (which can cause bugs, cf. the GitLab issue linked below). The patch also fixes a race condition in the reserveMapping function where a mapping reserved by one thread could wrongly appear to still be available from the perspective of a different thread. https://git.jami.net/savoirfairelinux/jami-client-ios/-/issues/410 Change-Id: I3784d36ccbb2f278ecc1c2573ac6f36b125f58c1
-
- Oct 14, 2024
-
-
François-Simon Fauteux-Chapleau authored
GitLab: #630 Change-Id: If4aece97b8c2e63d63981a36bafb35523c01d3a6
-
- Oct 11, 2024
-
-
Louis Maillard authored
Theses changes include using underscore instead of dash, to include the full version name for ubuntu (24 => 24.04) and a refactor of Jenkins to reduce duplications. Change-Id: I45eada9d5078e7e973527141c743fe1736e87a1b
-
Change-Id: I7154f3adeec0dd09bcfe42fb8adc30891a24bb2f
-
Louis Maillard authored
There was missing documentation on how to use the tools as an DevOps or SysAdmin. It now explain where to download, how to setup client and server, as well as basics about security in the tool. Change-Id: I7271d2dc37aa65459de16d0cf8a5cd41628e13e8
-