{"payload":{"feedbackUrl":"https://github.com/orgs/community/discussions/53140","repo":{"id":230270531,"defaultBranch":"stable","name":"wireguard-linux","ownerLogin":"WireGuard","currentUserCanPush":false,"isFork":false,"isEmpty":false,"createdAt":"2019-12-26T13:39:04.000Z","ownerAvatar":"https://avatars.githubusercontent.com/u/13991055?v=4","public":true,"private":false,"isOrgOwned":true},"refInfo":{"name":"","listCacheKey":"v0:1657757598.784256","currentOid":""},"activityList":{"items":[{"before":"492303379c1a9c59f8b4cbf82e375f8dd30e6191","after":"d24b03535e5eb82e025219c2f632b485409c898f","ref":"refs/heads/stable","pushedAt":"2024-03-22T18:13:58.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"nfc: nci: Fix uninit-value in nci_dev_up and nci_ntf_packet\n\nsyzbot reported the following uninit-value access issue [1][2]:\n\nnci_rx_work() parses and processes received packet. When the payload\nlength is zero, each message type handler reads uninitialized payload\nand KMSAN detects this issue. The receipt of a packet with a zero-size\npayload is considered unexpected, and therefore, such packets should be\nsilently discarded.\n\nThis patch resolved this issue by checking payload size before calling\neach message type handler codes.\n\nFixes: 6a2968aaf50c (\"NFC: basic NCI protocol implementation\")\nReported-and-tested-by: syzbot+7ea9413ea6749baf5574@syzkaller.appspotmail.com\nReported-and-tested-by: syzbot+29b5ca705d2e0f4a44d2@syzkaller.appspotmail.com\nCloses: https://syzkaller.appspot.com/bug?extid=7ea9413ea6749baf5574 [1]\nCloses: https://syzkaller.appspot.com/bug?extid=29b5ca705d2e0f4a44d2 [2]\nSigned-off-by: Ryosuke Yasuoka \nReviewed-by: Jeremy Cline \nReviewed-by: Krzysztof Kozlowski \nSigned-off-by: David S. Miller ","shortMessageHtmlLink":"nfc: nci: Fix uninit-value in nci_dev_up and nci_ntf_packet"}},{"before":"e30cef001da259e8df354b813015d0e5acc08740","after":"d24b03535e5eb82e025219c2f632b485409c898f","ref":"refs/heads/davem/net","pushedAt":"2024-03-22T18:13:58.000Z","pushType":"push","commitsCount":7601,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"nfc: nci: Fix uninit-value in nci_dev_up and nci_ntf_packet\n\nsyzbot reported the following uninit-value access issue [1][2]:\n\nnci_rx_work() parses and processes received packet. When the payload\nlength is zero, each message type handler reads uninitialized payload\nand KMSAN detects this issue. The receipt of a packet with a zero-size\npayload is considered unexpected, and therefore, such packets should be\nsilently discarded.\n\nThis patch resolved this issue by checking payload size before calling\neach message type handler codes.\n\nFixes: 6a2968aaf50c (\"NFC: basic NCI protocol implementation\")\nReported-and-tested-by: syzbot+7ea9413ea6749baf5574@syzkaller.appspotmail.com\nReported-and-tested-by: syzbot+29b5ca705d2e0f4a44d2@syzkaller.appspotmail.com\nCloses: https://syzkaller.appspot.com/bug?extid=7ea9413ea6749baf5574 [1]\nCloses: https://syzkaller.appspot.com/bug?extid=29b5ca705d2e0f4a44d2 [2]\nSigned-off-by: Ryosuke Yasuoka \nReviewed-by: Jeremy Cline \nReviewed-by: Krzysztof Kozlowski \nSigned-off-by: David S. Miller ","shortMessageHtmlLink":"nfc: nci: Fix uninit-value in nci_dev_up and nci_ntf_packet"}},{"before":"aee4de74523d555025d308940d9db5f8d7d2bc28","after":"492303379c1a9c59f8b4cbf82e375f8dd30e6191","ref":"refs/heads/stable","pushedAt":"2024-03-14T22:38:43.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"wireguard: selftests: set RISCV_ISA_FALLBACK on riscv{32,64}\n\nThis option is needed to continue booting with QEMU. Recent changes that\nmade this optional meant that it gets unset in the test harness, and so\nWireGuard CI has been broken. Fix this by simply setting this option.\n\nCc: stable@vger.kernel.org\nFixes: 496ea826d1e1 (\"RISC-V: provide Kconfig & commandline options to control parsing \"riscv,isa\"\")\nSigned-off-by: Jason A. Donenfeld ","shortMessageHtmlLink":"wireguard: selftests: set RISCV_ISA_FALLBACK on riscv{32,64}"}},{"before":"d7d75124965aee23e5e4421d78376545cf070b0a","after":"e30cef001da259e8df354b813015d0e5acc08740","ref":"refs/heads/davem/net","pushedAt":"2024-03-14T22:38:43.000Z","pushType":"push","commitsCount":8,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"net: txgbe: fix clk_name exceed MAX_DEV_ID limits\n\ntxgbe register clk which name is i2c_designware.pci_dev_id(),\nclk_name will be stored in clk_lookup_alloc. If PCIe bus number\nis larger than 0x39, clk_name size will be larger than 20 bytes.\nIt exceeds clk_lookup_alloc MAX_DEV_ID limits. So the driver\nshortened clk_name.\n\nFixes: b63f20485e43 (\"net: txgbe: Register fixed rate clock\")\nSigned-off-by: Duanqiang Wen \nReviewed-by: Michal Kubiak \nLink: https://lore.kernel.org/r/20240313080634.459523-1-duanqiangwen@net-swift.com\nSigned-off-by: Paolo Abeni ","shortMessageHtmlLink":"net: txgbe: fix clk_name exceed MAX_DEV_ID limits"}},{"before":"32511f43ea9bba810bd68a6c0f63b9ee7c895057","after":"aee4de74523d555025d308940d9db5f8d7d2bc28","ref":"refs/heads/stable","pushedAt":"2024-03-14T04:29:11.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"wireguard: selftests: set RISCV_ISA_FALLBACK on riscv{32,64}\n\nThis option is needed to continue booting with QEMU. Recent changes that\nmade this optional meant that it gets unset in the test harness, and so\nWireGuard CI has been broken. Fix this by simply setting this option.\n\nCc: stable@vger.kernel.org\nFixes: 496ea826d1e1 (\"RISC-V: provide Kconfig & commandline options to control parsing \"riscv,isa\"\")\nSigned-off-by: Jason A. Donenfeld ","shortMessageHtmlLink":"wireguard: selftests: set RISCV_ISA_FALLBACK on riscv{32,64}"}},{"before":"0915bdea02e3fb4303d23a73edd56c3925d45be8","after":"32511f43ea9bba810bd68a6c0f63b9ee7c895057","ref":"refs/heads/stable","pushedAt":"2024-03-14T03:47:31.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"wireguard: netlink: access device through ctx instead of peer\n\nThe previous commit fixed a bug that led to a NULL peer->device being\ndereferenced. It's actually easier and faster performance-wise to\ninstead get the device from ctx->wg. This semantically makes more sense\ntoo, since ctx->wg->peer_allowedips.seq is compared with\nctx->allowedips_seq, basing them both in ctx. This also acts as a\ndefence in depth provision against freed peers.\n\nCc: stable@vger.kernel.org\nFixes: e7096c131e51 (\"net: WireGuard secure network tunnel\")\nSigned-off-by: Jason A. Donenfeld ","shortMessageHtmlLink":"wireguard: netlink: access device through ctx instead of peer"}},{"before":"3436e03d9d6682fd5a3e52c698488e879a2d07f0","after":"0915bdea02e3fb4303d23a73edd56c3925d45be8","ref":"refs/heads/stable","pushedAt":"2024-03-14T03:28:33.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"wireguard: device: remove generic .ndo_get_stats64\n\nCommit 3e2f544dd8a33 (\"net: get stats64 if device if driver is\nconfigured\") moved the callback to dev_get_tstats64() to net core, so,\nunless the driver is doing some custom stats collection, it does not\nneed to set .ndo_get_stats64.\n\nSince this driver is now relying in NETDEV_PCPU_STAT_TSTATS, then, it\ndoesn't need to set the dev_get_tstats64() generic .ndo_get_stats64\nfunction pointer.\n\nSigned-off-by: Breno Leitao \nReviewed-by: Simon Horman \nSigned-off-by: Jason A. Donenfeld ","shortMessageHtmlLink":"wireguard: device: remove generic .ndo_get_stats64"}},{"before":"1ce2654d87e2fb91fea83b288bd9b2641045e42a","after":"d7d75124965aee23e5e4421d78376545cf070b0a","ref":"refs/heads/davem/net","pushedAt":"2024-03-14T03:28:33.000Z","pushType":"push","commitsCount":7180,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"devlink: Fix devlink parallel commands processing\n\nCommit 870c7ad4a52b (\"devlink: protect devlink->dev by the instance\nlock\") added devlink instance locking inside a loop that iterates over\nall the registered devlink instances on the machine in the pre-doit\nphase. This can lead to serialization of devlink commands over\ndifferent devlink instances.\n\nFor example: While the first devlink instance is executing firmware\nflash, all commands to other devlink instances on the machine are\nforced to wait until the first devlink finishes.\n\nTherefore, in the pre-doit phase, take the devlink instance lock only\nfor the devlink instance the command is targeting. Devlink layer is\ntaking a reference on the devlink instance, ensuring the devlink->dev\npointer is valid. This reference taking was introduced by commit\na380687200e0 (\"devlink: take device reference for devlink object\").\nWithout this commit, it would not be safe to access devlink->dev\nlockless.\n\nFixes: 870c7ad4a52b (\"devlink: protect devlink->dev by the instance lock\")\nSigned-off-by: Shay Drory \nReviewed-by: Jiri Pirko \nSigned-off-by: David S. Miller ","shortMessageHtmlLink":"devlink: Fix devlink parallel commands processing"}},{"before":"a2437565dc9110d38e00e2ba38f03947a6fc4c10","after":"3436e03d9d6682fd5a3e52c698488e879a2d07f0","ref":"refs/heads/stable","pushedAt":"2024-03-14T03:24:37.000Z","pushType":"push","commitsCount":0,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"}},{"before":"8e762c9ec4ef4cc26e8d8ca6d27031b4d3cc16e1","after":"a2437565dc9110d38e00e2ba38f03947a6fc4c10","ref":"refs/heads/stable","pushedAt":"2024-03-14T02:22:23.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"wireguard: netlink: check for dangling peer via is_dead instead of empty list\n\nIf all peers are removed via wg_peer_remove_all(), rather than setting\npeer_list to empty, the peer is added to a temporary list with a head on\nthe stack of wg_peer_remove_all(). If a netlink dump is resumed and the\ncursored peer is one that has been removed via wg_peer_remove_all(), it\nwill iterate from that peer and then attempt to dump freed peers.\n\nFix this by instead checking peer->is_dead, which was explictly created\nfor this purpose. Also move up the device_update_lock lockdep assertion,\nsince reading is_dead relies on that.\n\nIt can be reproduced by a small script like:\n\n echo \"Setting config...\"\n ip link add dev wg0 type wireguard\n wg setconf wg0 /big-config\n (\n while true; do\n echo \"Showing config...\"\n wg showconf wg0 > /dev/null\n done\n ) &\n sleep 4\n wg setconf wg0 <(printf \"[Peer]\\nPublicKey=$(wg genkey)\\n\")\n\nResulting in:\n\n BUG: KASAN: slab-use-after-free in __lock_acquire+0x182a/0x1b20\n Read of size 8 at addr ffff88811956ec70 by task wg/59\n CPU: 2 PID: 59 Comm: wg Not tainted 6.8.0-rc2-debug+ #5\n Call Trace:\n \n dump_stack_lvl+0x47/0x70\n print_address_description.constprop.0+0x2c/0x380\n print_report+0xab/0x250\n kasan_report+0xba/0xf0\n __lock_acquire+0x182a/0x1b20\n lock_acquire+0x191/0x4b0\n down_read+0x80/0x440\n get_peer+0x140/0xcb0\n wg_get_device_dump+0x471/0x1130\n\nCc: stable@vger.kernel.org\nFixes: e7096c131e51 (\"net: WireGuard secure network tunnel\")\nReported-by: Lillian Berry \nReported-by: Jann Horn \nSigned-off-by: Jason A. Donenfeld ","shortMessageHtmlLink":"wireguard: netlink: check for dangling peer via is_dead instead of em…"}},{"before":"06ba660ab3082d77acb685dd15fd0bb616058039","after":"8e762c9ec4ef4cc26e8d8ca6d27031b4d3cc16e1","ref":"refs/heads/stable","pushedAt":"2024-03-13T04:09:23.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"wireguard: netlink: check for dangling peer via is_dead instead of empty list\n\nIf all peers are removed via wg_peer_remove_all(), rather than setting\npeer_list to empty, the peer is added to a temporary list with a head on\nthe stack of wg_peer_remove_all(). If a netlink dump is resumed and the\ncursored peer is one that has been removed via wg_peer_remove_all(), it\nwill iterate from that peer and then attempt to dump freed peers.\n\nFix this by instead checking peer->is_dead, which was explictly created\nfor this purpose. Also move up the device_update_lock lockdep assertion,\nsince reading is_dead relies on that.\n\nCc: stable@vger.kernel.org\nFixes: e7096c131e51 (\"net: WireGuard secure network tunnel\")\nReported-by: Jann Horn \nSigned-off-by: Jason A. Donenfeld ","shortMessageHtmlLink":"wireguard: netlink: check for dangling peer via is_dead instead of em…"}},{"before":"9098f7c91085ce5514a29cb4aac7aee29a563451","after":"06ba660ab3082d77acb685dd15fd0bb616058039","ref":"refs/heads/stable","pushedAt":"2024-03-13T03:59:24.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"wireguard: netlink: check for dangling peer via is_dead instead of empty list\n\nIf all peers are removed via wg_peer_remove_all(), rather than setting\npeer_list to empty, the peer is added to a temporary list with a head on\nthe stack of wg_peer_remove_all(). If a netlink dump is resumed and the\ncursored peer is one that has been removed via wg_peer_remove_all(), it\nwill iterate from that peer and then attempt to dump freed peers.\n\nFix this by instead checking peer->is_dead, which was explictly created\nfor this purpose.\n\nCc: stable@vger.kernel.org\nFixes: e7096c131e51 (\"net: WireGuard secure network tunnel\")\nReported-by: Jann Horn \nSigned-off-by: Jason A. Donenfeld ","shortMessageHtmlLink":"wireguard: netlink: check for dangling peer via is_dead instead of em…"}},{"before":"1ce2654d87e2fb91fea83b288bd9b2641045e42a","after":"9098f7c91085ce5514a29cb4aac7aee29a563451","ref":"refs/heads/stable","pushedAt":"2024-02-06T19:16:44.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"wireguard: receive: annotate data-race around receiving_counter.counter\n\nSyzkaller with KCSAN identified a data-race issue when accessing\nkeypair->receiving_counter.counter. Use READ_ONCE() and WRITE_ONCE()\nannotations to mark the data race as intentional.\n\n BUG: KCSAN: data-race in wg_packet_decrypt_worker / wg_packet_rx_poll\n\n write to 0xffff888107765888 of 8 bytes by interrupt on cpu 0:\n counter_validate drivers/net/wireguard/receive.c:321 [inline]\n wg_packet_rx_poll+0x3ac/0xf00 drivers/net/wireguard/receive.c:461\n __napi_poll+0x60/0x3b0 net/core/dev.c:6536\n napi_poll net/core/dev.c:6605 [inline]\n net_rx_action+0x32b/0x750 net/core/dev.c:6738\n __do_softirq+0xc4/0x279 kernel/softirq.c:553\n do_softirq+0x5e/0x90 kernel/softirq.c:454\n __local_bh_enable_ip+0x64/0x70 kernel/softirq.c:381\n __raw_spin_unlock_bh include/linux/spinlock_api_smp.h:167 [inline]\n _raw_spin_unlock_bh+0x36/0x40 kernel/locking/spinlock.c:210\n spin_unlock_bh include/linux/spinlock.h:396 [inline]\n ptr_ring_consume_bh include/linux/ptr_ring.h:367 [inline]\n wg_packet_decrypt_worker+0x6c5/0x700 drivers/net/wireguard/receive.c:499\n process_one_work kernel/workqueue.c:2633 [inline]\n ...\n\n read to 0xffff888107765888 of 8 bytes by task 3196 on cpu 1:\n decrypt_packet drivers/net/wireguard/receive.c:252 [inline]\n wg_packet_decrypt_worker+0x220/0x700 drivers/net/wireguard/receive.c:501\n process_one_work kernel/workqueue.c:2633 [inline]\n process_scheduled_works+0x5b8/0xa30 kernel/workqueue.c:2706\n worker_thread+0x525/0x730 kernel/workqueue.c:2787\n ...\n\nFixes: a9e90d9931f3 (\"wireguard: noise: separate receive counter from send counter\")\nReported-by: syzbot+d1de830e4ecdaac83d89@syzkaller.appspotmail.com\nSigned-off-by: Nikita Zhandarovich \nSigned-off-by: Jason A. Donenfeld ","shortMessageHtmlLink":"wireguard: receive: annotate data-race around receiving_counter.counter"}},{"before":"fa41884c1c6deb6774135390e5813a97184903e0","after":"1ce2654d87e2fb91fea83b288bd9b2641045e42a","ref":"refs/heads/stable","pushedAt":"2024-02-06T19:13:03.000Z","pushType":"push","commitsCount":10000,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"net: stmmac: xgmac: fix a typo of register name in DPP safety handling\n\nDDPP is copied from Synopsys Data book:\n\nDDPP: Disable Data path Parity Protection.\n When it is 0x0, Data path Parity Protection is enabled.\n When it is 0x1, Data path Parity Protection is disabled.\n\nThe macro name should be XGMAC_DPP_DISABLE.\n\nFixes: 46eba193d04f (\"net: stmmac: xgmac: fix handling of DPP safety error for DMA channels\")\nSigned-off-by: Furong Xu <0x1207@gmail.com>\nReviewed-by: Serge Semin \nLink: https://lore.kernel.org/r/20240203053133.1129236-1-0x1207@gmail.com\nSigned-off-by: Paolo Abeni ","shortMessageHtmlLink":"net: stmmac: xgmac: fix a typo of register name in DPP safety handling"}},{"before":"fa41884c1c6deb6774135390e5813a97184903e0","after":"1ce2654d87e2fb91fea83b288bd9b2641045e42a","ref":"refs/heads/davem/net","pushedAt":"2024-02-06T19:13:03.000Z","pushType":"push","commitsCount":10000,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"net: stmmac: xgmac: fix a typo of register name in DPP safety handling\n\nDDPP is copied from Synopsys Data book:\n\nDDPP: Disable Data path Parity Protection.\n When it is 0x0, Data path Parity Protection is enabled.\n When it is 0x1, Data path Parity Protection is disabled.\n\nThe macro name should be XGMAC_DPP_DISABLE.\n\nFixes: 46eba193d04f (\"net: stmmac: xgmac: fix handling of DPP safety error for DMA channels\")\nSigned-off-by: Furong Xu <0x1207@gmail.com>\nReviewed-by: Serge Semin \nLink: https://lore.kernel.org/r/20240203053133.1129236-1-0x1207@gmail.com\nSigned-off-by: Paolo Abeni ","shortMessageHtmlLink":"net: stmmac: xgmac: fix a typo of register name in DPP safety handling"}},{"before":"fcd4c76a3c15cd7f4a4ab49bec1f1176388ea2c5","after":"fa41884c1c6deb6774135390e5813a97184903e0","ref":"refs/heads/stable","pushedAt":"2023-08-07T19:31:03.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"Merge branch 'wireguard-fixes-for-6-5-rc6'\n\nJason A. Donenfeld says:\n\n====================\nwireguard fixes for 6.5-rc6\n\nJust one patch this time, somewhat late in the cycle:\n\n1) Fix an off-by-one calculation for the maximum node depth size in the\n allowedips trie data structure, and also adjust the self-tests to hit\n this case so it doesn't regress again in the future.\n====================\n\nLink: https://lore.kernel.org/r/20230807132146.2191597-1-Jason@zx2c4.com\nSigned-off-by: Jakub Kicinski ","shortMessageHtmlLink":"Merge branch 'wireguard-fixes-for-6-5-rc6'"}},{"before":"1733d0be68ab1b89358a3b0471ef425fd61de7c5","after":"fa41884c1c6deb6774135390e5813a97184903e0","ref":"refs/heads/davem/net","pushedAt":"2023-08-07T19:31:03.000Z","pushType":"push","commitsCount":19,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"Merge branch 'wireguard-fixes-for-6-5-rc6'\n\nJason A. Donenfeld says:\n\n====================\nwireguard fixes for 6.5-rc6\n\nJust one patch this time, somewhat late in the cycle:\n\n1) Fix an off-by-one calculation for the maximum node depth size in the\n allowedips trie data structure, and also adjust the self-tests to hit\n this case so it doesn't regress again in the future.\n====================\n\nLink: https://lore.kernel.org/r/20230807132146.2191597-1-Jason@zx2c4.com\nSigned-off-by: Jakub Kicinski ","shortMessageHtmlLink":"Merge branch 'wireguard-fixes-for-6-5-rc6'"}},{"before":"bf400e83708d055bdf442577ed2f2a8eb87a06f2","after":"fcd4c76a3c15cd7f4a4ab49bec1f1176388ea2c5","ref":"refs/heads/stable","pushedAt":"2023-08-07T13:16:49.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"wireguard: allowedips: expand maximum node depth\n\nIn the allowedips self-test, nodes are inserted into the tree, but it\ngenerated an even amount of nodes, but for checking maximum node depth,\nthere is of course the root node, which makes the total number\nnecessarily odd. With two few nodes added, it never triggered the\nmaximum depth check like it should have. So, add 129 nodes instead of\n128 nodes, and do so with a more straightforward scheme, starting with\nall the bits set, and shifting over one each time. Then increase the\nmaximum depth to 129, and choose a better name for that variable to\nmake it clear that it represents depth as opposed to bits.\n\nCc: stable@vger.kernel.org\nFixes: e7096c131e51 (\"net: WireGuard secure network tunnel\")\nSigned-off-by: Jason A. Donenfeld ","shortMessageHtmlLink":"wireguard: allowedips: expand maximum node depth"}},{"before":"1733d0be68ab1b89358a3b0471ef425fd61de7c5","after":"bf400e83708d055bdf442577ed2f2a8eb87a06f2","ref":"refs/heads/stable","pushedAt":"2023-08-04T13:56:24.000Z","pushType":"push","commitsCount":1,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"wireguard: allowedips: expand maximum node depth\n\nIn the allowedips self-test, nodes are inserted into the tree, but it\nforgot the generate the all zero node. As a result, the >= 128 check\nwasn't triggered as it should have. In fact, the maximum bit depth for\nthe tree is 129 nodes deep, not 128 nodes, so adjust the self-test to\ngenerate a full tree, and expand the maximum depth to fit it.\n\nCc: stable@vger.kernel.org\nFixes: e7096c131e51 (\"net: WireGuard secure network tunnel\")\nSigned-off-by: Jason A. Donenfeld ","shortMessageHtmlLink":"wireguard: allowedips: expand maximum node depth"}},{"before":"acd9755894c96c27078b52e0bfd894e48b0b1508","after":"1733d0be68ab1b89358a3b0471ef425fd61de7c5","ref":"refs/heads/stable","pushedAt":"2023-08-04T13:31:23.000Z","pushType":"push","commitsCount":9430,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"Merge branch 'at803x-wol'\n\nLi Yang says:\n\n====================\nfix at803x wol setting\n\nv3:\n Break long lines\n Add back error checking of phy_read\n\nv4:\n Disable WoL in 1588 register for AR8031 in probe\n====================\n\nSigned-off-by: David S. Miller ","shortMessageHtmlLink":"Merge branch 'at803x-wol'"}},{"before":"acd9755894c96c27078b52e0bfd894e48b0b1508","after":"1733d0be68ab1b89358a3b0471ef425fd61de7c5","ref":"refs/heads/davem/net","pushedAt":"2023-08-04T13:31:23.000Z","pushType":"push","commitsCount":9430,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"Merge branch 'at803x-wol'\n\nLi Yang says:\n\n====================\nfix at803x wol setting\n\nv3:\n Break long lines\n Add back error checking of phy_read\n\nv4:\n Disable WoL in 1588 register for AR8031 in probe\n====================\n\nSigned-off-by: David S. Miller ","shortMessageHtmlLink":"Merge branch 'at803x-wol'"}},{"before":"d3690685c544a12cdc2ce1a9064b594d014454c6","after":"acd9755894c96c27078b52e0bfd894e48b0b1508","ref":"refs/heads/stable","pushedAt":"2023-07-03T17:01:20.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"Documentation: ABI: sysfs-class-net-qmi: pass_through contact update\n\nSwitch to the quicinc.com id.\n\nFixes: bd1af6b5fffd (\"Documentation: ABI: sysfs-class-net-qmi: document pass-through file\")\nSigned-off-by: Subash Abhinov Kasiviswanathan \nSigned-off-by: David S. Miller ","shortMessageHtmlLink":"Documentation: ABI: sysfs-class-net-qmi: pass_through contact update"}},{"before":"97791d3c6d0a501109968cae769dc6256c2829e1","after":"acd9755894c96c27078b52e0bfd894e48b0b1508","ref":"refs/heads/davem/net","pushedAt":"2023-07-03T17:01:20.000Z","pushType":"push","commitsCount":11,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"Documentation: ABI: sysfs-class-net-qmi: pass_through contact update\n\nSwitch to the quicinc.com id.\n\nFixes: bd1af6b5fffd (\"Documentation: ABI: sysfs-class-net-qmi: document pass-through file\")\nSigned-off-by: Subash Abhinov Kasiviswanathan \nSigned-off-by: David S. Miller ","shortMessageHtmlLink":"Documentation: ABI: sysfs-class-net-qmi: pass_through contact update"}},{"before":"ab87603b251134441a67385ecc9d3371be17b7a7","after":"97791d3c6d0a501109968cae769dc6256c2829e1","ref":"refs/heads/davem/net","pushedAt":"2023-07-03T01:19:04.000Z","pushType":"push","commitsCount":6295,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"Merge branch 'octeontx2-af-fixes'\n\nHariprasad Kelam says:\n\n====================\nocteontx2-af: MAC block fixes for CN10KB\n\nThis patch set contains fixes for the issues encountered in testing\nCN10KB MAC block RPM_USX.\n\nPatch1: firmware to kernel communication is not working due to wrong\n interrupt configuration. CSR addresses are corrected.\n\nPatch2: NIX to RVU PF mapping errors encountered due to wrong firmware\n config. Corrects this mapping error.\n\nPatch3: Driver is trying to access non exist cgx/lmac which is resulting\n in kernel panic. Address this issue by adding proper checks.\n\nPatch4: MAC features are not getting reset on FLR. Fix the issue by\n resetting the stale config.\n====================\n\nSigned-off-by: David S. Miller ","shortMessageHtmlLink":"Merge branch 'octeontx2-af-fixes'"}},{"before":"1311c750d075c70a380d9790c400da4e9f238819","after":"d3690685c544a12cdc2ce1a9064b594d014454c6","ref":"refs/heads/stable","pushedAt":"2023-07-03T01:19:04.000Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"wireguard: timers: move to using timer_delete_sync\n\nThe documentation says that del_timer_sync is obsolete, and code should\nuse the equivalent timer_delete_sync instead, so switch to it.\n\nSigned-off-by: Jason A. Donenfeld ","shortMessageHtmlLink":"wireguard: timers: move to using timer_delete_sync"}},{"before":"543e550938e436bf7db19521e679234c024963c1","after":"1311c750d075c70a380d9790c400da4e9f238819","ref":"refs/heads/stable","pushedAt":"2023-05-23T18:20:17.695Z","pushType":"push","commitsCount":1,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"wireguard: timers: move to using timer_delete_sync\n\nThe documentation says that del_timer_sync is obsolete, and code should\nuse the equivalent timer_delete_sync instead, so switch to it.\n\nSigned-off-by: Jason A. Donenfeld ","shortMessageHtmlLink":"wireguard: timers: move to using timer_delete_sync"}},{"before":"3ac1bf099766f1e9735883d5127148054cd5b30a","after":"543e550938e436bf7db19521e679234c024963c1","ref":"refs/heads/stable","pushedAt":"2023-05-19T19:28:29.053Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"wireguard: netlink: send staged packets when setting initial private key\n\nPackets bound for peers can queue up prior to the device private key\nbeing set. For example, if persistent keepalive is set, a packet is\nqueued up to be sent as soon as the device comes up. However, if the\nprivate key hasn't been set yet, the handshake message never sends, and\nno timer is armed to retry, since that would be pointless.\n\nBut, if a user later sets a private key, the expectation is that those\nqueued packets, such as a persistent keepalive, are actually sent. So\nadjust the configuration logic to account for this edge case, and add a\ntest case to make sure this works.\n\nMaxim noticed this with a wg-quick(8) config to the tune of:\n\n [Interface]\n PostUp = wg set %i private-key somefile\n\n [Peer]\n PublicKey = ...\n Endpoint = ...\n PersistentKeepalive = 25\n\nHere, the private key gets set after the device comes up using a PostUp\nscript, triggering the bug.\n\nFixes: e7096c131e51 (\"net: WireGuard secure network tunnel\")\nCc: stable@vger.kernel.org\nReported-by: Maxim Cournoyer \nTested-by: Maxim Cournoyer \nLink: https://lore.kernel.org/wireguard/87fs7xtqrv.fsf@gmail.com/\nSigned-off-by: Jason A. Donenfeld ","shortMessageHtmlLink":"wireguard: netlink: send staged packets when setting initial private key"}},{"before":"54d5e4329efe0d1dba8b4a58720d29493926bed0","after":"3ac1bf099766f1e9735883d5127148054cd5b30a","ref":"refs/heads/stable","pushedAt":"2023-05-18T01:14:40.097Z","pushType":"push","commitsCount":1,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"wireguard: netlink: send staged packets when setting initial private key\n\nPackets bound for peers can queue up prior to the device private key\nbeing set. For example, if persistent keepalive is set, a packet is\nqueued up to be sent as soon as the device comes up. However, if the\nprivate key hasn't been set yet, the handshake message never sends, and\nno timer is armed to retry, since that would be pointless.\n\nBut, if a user later sets a private key, the expectation is that those\nqueued packets, such as a persistent keepalive, are actually sent. So\nadjust the configuration logic to account for this edge case, and add a\ntest case to make sure this works.\n\nMaxim noticed this with a wg-quick(8) config to the tune of:\n\n [Interface]\n PostUp = wg set %i private-key somefile\n\n [Peer]\n PublicKey = ...\n Endpoint = ...\n PersistentKeepalive = 25\n\nHere, the private key gets set after the device comes up using a PostUp\nscript, triggering the bug.\n\nFixes: e7096c131e51 (\"net: WireGuard secure network tunnel\")\nCc: stable@vger.kernel.org\nReported-by: Maxim Cournoyer \nLink: https://lore.kernel.org/wireguard/87fs7xtqrv.fsf@gmail.com/\nSigned-off-by: Jason A. Donenfeld ","shortMessageHtmlLink":"wireguard: netlink: send staged packets when setting initial private key"}},{"before":"7d0b9ab9ac20b727f9c59b241a49ed5338754ae1","after":"54d5e4329efe0d1dba8b4a58720d29493926bed0","ref":"refs/heads/stable","pushedAt":"2023-05-18T00:41:52.076Z","pushType":"force_push","commitsCount":0,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"wireguard: queueing: use saner cpu selection wrapping\n\nUsing `% nr_cpumask_bits` is slow and complicated, and not totally\nrobust toward dynamic changes to CPU topologies. Rather than storing the\nnext CPU in the round-robin, just store the last one, and also return\nthat value. This simplifies the loop drastically into a much more common\npattern.\n\nFixes: e7096c131e51 (\"net: WireGuard secure network tunnel\")\nCc: stable@vger.kernel.org\nReported-by: Linus Torvalds \nSigned-off-by: Jason A. Donenfeld ","shortMessageHtmlLink":"wireguard: queueing: use saner cpu selection wrapping"}},{"before":"f038f3917baf04835ba2b7bcf2a04ac93fbf8a9c","after":"ab87603b251134441a67385ecc9d3371be17b7a7","ref":"refs/heads/davem/net","pushedAt":"2023-05-18T00:41:52.076Z","pushType":"push","commitsCount":10000,"pusher":{"login":"zx2c4-bot","name":"Bot for git.zx2c4.com","path":"/zx2c4-bot","primaryAvatarUrl":"https://avatars.githubusercontent.com/u/26463517?s=80&v=4"},"commit":{"message":"net: wwan: t7xx: Ensure init is completed before system sleep\n\nWhen the system attempts to sleep while mtk_t7xx is not ready, the driver\ncannot put the device to sleep:\n[ 12.472918] mtk_t7xx 0000:57:00.0: [PM] Exiting suspend, modem in invalid state\n[ 12.472936] mtk_t7xx 0000:57:00.0: PM: pci_pm_suspend(): t7xx_pci_pm_suspend+0x0/0x20 [mtk_t7xx] returns -14\n[ 12.473678] mtk_t7xx 0000:57:00.0: PM: dpm_run_callback(): pci_pm_suspend+0x0/0x1b0 returns -14\n[ 12.473711] mtk_t7xx 0000:57:00.0: PM: failed to suspend async: error -14\n[ 12.764776] PM: Some devices failed to suspend, or early wake event detected\n\nMediatek confirmed the device can take a rather long time to complete\nits initialization, so wait for up to 20 seconds until init is done.\n\nSigned-off-by: Kai-Heng Feng \nSigned-off-by: David S. Miller ","shortMessageHtmlLink":"net: wwan: t7xx: Ensure init is completed before system sleep"}}],"hasNextPage":true,"hasPreviousPage":false,"activityType":"all","actor":null,"timePeriod":"all","sort":"DESC","perPage":30,"cursor":"djE6ks8AAAAEHQWajQA","startCursor":null,"endCursor":null}},"title":"Activity · WireGuard/wireguard-linux"}