Branch: refs/heads/master Home: https://github.com/bluez/bluez Commit: ddec3358fe5cd991c4da89b35dbfff0058b0434a https://github.com/bluez/bluez/commit/ddec3358fe5cd991c4da89b35dbfff0058b0434a Author: Luiz Augusto von Dentz <luiz.von.dentz@xxxxxxxxx> Date: 2025-07-15 (Tue, 15 Jul 2025) Changed paths: M doc/tester.config Log Message: ----------- tester.config: Add missing drivers This adds missing config options to enable all drivers which is required in order to catch build problem with github CI workflow. Commit: 009cfb5188f8287819cb73ae4910dc55b659e41c https://github.com/bluez/bluez/commit/009cfb5188f8287819cb73ae4910dc55b659e41c Author: Christian Eggers <ceggers@xxxxxxx> Date: 2025-07-15 (Tue, 15 Jul 2025) Changed paths: M mesh/net-keys.c Log Message: ----------- mesh: net-keys: more uses of BEACON_LEN_SNB and BEACON_LEN_MPB Commit: 646cbe922792f1c299703c111a41c0c959a9d045 https://github.com/bluez/bluez/commit/646cbe922792f1c299703c111a41c0c959a9d045 Author: Christian Eggers <ceggers@xxxxxxx> Date: 2025-07-15 (Tue, 15 Jul 2025) Changed paths: M mesh/mesh-defs.h M mesh/mesh-io-generic.c M mesh/mesh-io-mgmt.c M mesh/mesh-io-unit.c M mesh/net-keys.c M mesh/net.c Log Message: ----------- mesh: introduce MESH_AD_MAX_LEN and MESH_NET_MAX_PDU_LEN Use symbolic names rather than magic numbers. Remove unneeded extra pointer in send_seg(). Commit: cd651d8e21d785399086e1e6136267e407c1e863 https://github.com/bluez/bluez/commit/cd651d8e21d785399086e1e6136267e407c1e863 Author: Christian Eggers <ceggers@xxxxxxx> Date: 2025-07-15 (Tue, 15 Jul 2025) Changed paths: M mesh/manager.c M mesh/mesh-defs.h M mesh/mesh-io-generic.c M mesh/mesh-io-mgmt.c M mesh/mesh-io.c M mesh/mesh.c M mesh/net-keys.c M mesh/net.c M mesh/pb-adv.c M mesh/prov-acceptor.c M mesh/prov-initiator.c Log Message: ----------- mesh: replace MESH internal defines by shared ones Replace BT_AD_MESH_* by MESH_AD_TYPE_*. Both sets of definition have been added almost at the time, so maybe it was a 'race condition'. Commit: b692d72b15b3d732a482d342db1d5934fc84a8b8 https://github.com/bluez/bluez/commit/b692d72b15b3d732a482d342db1d5934fc84a8b8 Author: Christian Eggers <ceggers@xxxxxxx> Date: 2025-07-15 (Tue, 15 Jul 2025) Changed paths: M mesh/net.c Log Message: ----------- mesh: net: constify tx path Although the first byte of network packets has the same value for all 4 'send' functions, it feels much more natural to assign this byte at the location(s) where the packet is assembled, rather than where it is sent. This improves the readability because send_msg_pkt() isn't called with a partially uninitialized buffer anymore. Commit: 0e586df2d0aecd181714191b2c62c83657d380fd https://github.com/bluez/bluez/commit/0e586df2d0aecd181714191b2c62c83657d380fd Author: Christian Eggers <ceggers@xxxxxxx> Date: 2025-07-15 (Tue, 15 Jul 2025) Changed paths: M mesh/net.c M mesh/net.h Log Message: ----------- mesh: net: remove unused stuff 'struct mesh_key_set' and 'struct net_decode' arent't used anymore since commit 994932b740c7 ("mesh: Refactor friend.c and net.c for central key DB"). 'mesh_status_func_t' isn't used anymore since commit c4bf0626fb62 ("mesh: Replace storage_save_config with mesh_config_save_config"). Commit: cb4b20c71fd3c426902b853cb70e8ffdfea7e85f https://github.com/bluez/bluez/commit/cb4b20c71fd3c426902b853cb70e8ffdfea7e85f Author: Christian Eggers <ceggers@xxxxxxx> Date: 2025-07-15 (Tue, 15 Jul 2025) Changed paths: M mesh/net.c Log Message: ----------- mesh: net: update comment term 'master credentials' has been replaced by 'flooding credentials' in commit 09f87c80f1d5 ("mesh: Inclusive language changes") Commit: 9a32ff5aa7ed1d744f0d70403fe7851d8e0ad7c0 https://github.com/bluez/bluez/commit/9a32ff5aa7ed1d744f0d70403fe7851d8e0ad7c0 Author: Christian Eggers <ceggers@xxxxxxx> Date: 2025-07-15 (Tue, 15 Jul 2025) Changed paths: M mesh/crypto.c M mesh/crypto.h M mesh/net-keys.c M mesh/prov-initiator.c M unit/test-mesh-crypto.c Log Message: ----------- mesh: crypto: mesh_crypto_aes_ccm_encrypt(): remove unused parameter The 'out_mic' pointer isn't written by this function and all callers pass a NULL pointer for this. It's obviously not required (and would not work), so lets remove it. Commit: 159101c7bc38da3b2ca2f791ccca4cedb8da8a0c https://github.com/bluez/bluez/commit/159101c7bc38da3b2ca2f791ccca4cedb8da8a0c Author: Christian Eggers <ceggers@xxxxxxx> Date: 2025-07-15 (Tue, 15 Jul 2025) Changed paths: M mesh/crypto.c Log Message: ----------- mesh: crypto: simplify mesh_crypto_packet_parse() - NULL pointer checks are not required (no caller passes NULL pointers) - reuse result of mesh_crypto_network_header_parse() rather than determining value of 'CTL' again. Commit: 2b0a6fa0840715bbffe5b7a2195795575ce14600 https://github.com/bluez/bluez/commit/2b0a6fa0840715bbffe5b7a2195795575ce14600 Author: Christian Eggers <ceggers@xxxxxxx> Date: 2025-07-15 (Tue, 15 Jul 2025) Changed paths: M mesh/crypto.c M mesh/net-keys.c Log Message: ----------- mesh: fix corrupted relay packets Currently, all relayed packets are broken! At the point when net_key_encrypt() is called from net_rx() for encrypting a relay packet, the packet size must include the NetMIC field. But the length of this field has already been removed during decryption of the incoming packet (by decrypt_net_pkt()), although mesh_crypto_packet_decrypt() has correctly reset the NetMIC field to zeroes. Move stripping of the NetMIC field length from decrypt_net_pkt() to mesh_crypto_packet_parse(), so that the field length is only stripped from the payload message (but keeping the field length for the network PDU). Additionally add extra length checks during parsing. Commit: 8c9977b02169621f4643106c18de4de9824485a9 https://github.com/bluez/bluez/commit/8c9977b02169621f4643106c18de4de9824485a9 Author: Ye He <ye.he@xxxxxxxxxxx> Date: 2025-07-15 (Tue, 15 Jul 2025) Changed paths: M src/adapter.c Log Message: ----------- adapter: Fix RemoveDevice timeout when device already disconnected When attempting to use RemoveDevice to delete a BIS source device that was synchronized by the BIS sink scan delegator, the kernel marks the device as disconnected due to PA(period adv) sync termination. However, BlueZ is not notified of this disconnection and still proceeds to send MGMT Disconnect command. The kernel responds with MGMT_STATUS_DISCONNECTED, which BlueZ does not currently handle as a successful case. As a result, the RemoveDevice call never completes and no D-Bus reply is returned. Fixes: https://github.com/bluez/bluez/issues/1421 Compare: https://github.com/bluez/bluez/compare/b0a1386f98c2...8c9977b02169 To unsubscribe from these emails, change your notification settings at https://github.com/bluez/bluez/settings/notifications
