)]}' { "log": [ { "commit": "85c698863c15176f743fd1d1fcf39ceb9172c820", "tree": "8870e1382a730cfd492080dc4f2126424f5b27c7", "parents": [ "222a31408ab0f418cdb3a0b65d5a2008013c2e27" ], "author": { "name": "Mianhan Liu", "email": "liumh1@shanghaitech.edu.cn", "time": "Mon Sep 20 19:55:36 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Sep 20 13:09:06 2021 +0100" }, "message": "net/ipv4/tcp_minisocks.c: remove superfluous header files from tcp_minisocks.c\n\ntcp_minisocks.c hasn\u0027t use any macro or function declared in mm.h, module.h,\nslab.h, sysctl.h, workqueue.h, static_key.h and inet_common.h. Thus, these\nfiles can be removed from tcp_minisocks.c safely without affecting the\ncompilation of the net module.\n\nSigned-off-by: Mianhan Liu \u003cliumh1@shanghaitech.edu.cn\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "222a31408ab0f418cdb3a0b65d5a2008013c2e27", "tree": "994135ee5bad326def1642317a451c9f5f1755f9", "parents": [ "ffa66f15e4506dcfd48d144be27870f13438dcfc" ], "author": { "name": "Mianhan Liu", "email": "liumh1@shanghaitech.edu.cn", "time": "Mon Sep 20 19:34:16 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Sep 20 13:09:06 2021 +0100" }, "message": "net/ipv4/tcp_fastopen.c: remove superfluous header files from tcp_fastopen.c\n\ntcp_fastopen.c hasn\u0027t use any macro or function declared in crypto.h, err.h,\ninit.h, list.h, rculist.h and inetpeer.h. Thus, these files can be removed\nfrom tcp_fastopen.c safely without affecting the compilation of the net module.\n\nSigned-off-by: Mianhan Liu \u003cliumh1@shanghaitech.edu.cn\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "ffa66f15e4506dcfd48d144be27870f13438dcfc", "tree": "41b2ae8de444275737e3f9abdd6166e2a24a783e", "parents": [ "998ac358019e491217e752bc6dcbb3afb2a6fa3e" ], "author": { "name": "Mianhan Liu", "email": "liumh1@shanghaitech.edu.cn", "time": "Mon Sep 20 19:31:37 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Sep 20 13:09:06 2021 +0100" }, "message": "net/ipv4/route.c: remove superfluous header files from route.c\n\nroute.c hasn\u0027t use any macro or function declared in uaccess.h, types.h,\nstring.h, sockios.h, times.h, protocol.h, arp.h and l3mdev.h. Thus, these\nfiles can be removed from route.c safely without affecting the compilation\nof the net module.\n\nSigned-off-by: Mianhan Liu \u003cliumh1@shanghaitech.edu.cn\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "998ac358019e491217e752bc6dcbb3afb2a6fa3e", "tree": "0a0c57c35563fd536e121d6dd2890f094add0064", "parents": [ "13f356f5dc9de17181fa255a379fde29eea060d4" ], "author": { "name": "Aleksander Jan Bajkowski", "email": "olek2@wp.pl", "time": "Sun Sep 19 20:24:28 2021 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Sep 20 10:07:52 2021 +0100" }, "message": "net: lantiq: add support for jumbo frames\n\nAdd support for jumbo frames. Full support for jumbo frames requires\nchanges in the DSA switch driver (lantiq_gswip.c).\n\nTested on BT Hone Hub 5A.\n\nSigned-off-by: Aleksander Jan Bajkowski \u003colek2@wp.pl\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "13f356f5dc9de17181fa255a379fde29eea060d4", "tree": "f7c57b33a1304c698db558a8a6a641fff44ddc7d", "parents": [ "d44fd8604a4ab92119adb35f05fd87612af722b5", "607d574aba6e2b3adb5cd5cff31194bd31a8048e" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Sep 20 10:04:01 2021 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Sep 20 10:04:01 2021 +0100" }, "message": "Merge branch \u0027wwan-iosm-fw-flashing\u0027\n\nM Chetan Kumar says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nnet: wwan: iosm: fw flashing \u0026 cd collection\n\nThis patch series brings-in support for M.2 7560 Device firmware flashing \u0026\ncoredump collection using devlink.\n- Driver Registers with Devlink framework.\n- Register devlink params callback for configuring device params\n required in flashing or coredump flow.\n- Implements devlink ops flash_update callback that programs modem\n firmware.\n- Creates region \u0026 snapshot required for device coredump log collection.\n\nOn early detection of device in boot rom stage. Driver registers with\nDevlink framework and establish transport channel for PSI (Primary Signed\nImage) injection. Once PSI is injected to device, the device execution\nstage details are read to determine whether device is in flash or\nexception mode. The collected information is reported to devlink user\nspace application \u0026 based on this informationi, application proceeds with\neither modem firmware flashing or coredump collection.\n\nRefer to iosm devlink documentation for details on Devlink Params, flash\nupdate and coredump collection command usage.\n\nNote: Patches are interdependent. Need to apply complete patch series for\ncompilation.\n\nChanges since v1:\n * Break down single patch into multiple patches.\n * IOSM Driver devlink documentation.\n * Fixes NULL parameter deference in ipc_devlink_flash_update() reported\n by smatch static checker.\n * Fixes memory leak in ipc_devlink_create_region().\n * Use kmemdup instead of kzalloc and memcpy in ipc_flash_boot_psi().\n * Fixes linux-net build error.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "607d574aba6e2b3adb5cd5cff31194bd31a8048e", "tree": "f7c57b33a1304c698db558a8a6a641fff44ddc7d", "parents": [ "64302024bce5d52594b80e5e6188abc85ff7cb0e" ], "author": { "name": "M Chetan Kumar", "email": "m.chetan.kumar@linux.intel.com", "time": "Sun Sep 19 22:58:35 2021 +0530" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Sep 20 10:03:37 2021 +0100" }, "message": "net: wwan: iosm: fw flashing \u0026 cd collection infrastructure changes\n\nIOSM Makefile \u0026 WWAN Kconfig changes to support fw flashing \u0026 cd\ncollection module compliation.\n\nSigned-off-by: M Chetan Kumar \u003cm.chetan.kumar@linux.intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "64302024bce5d52594b80e5e6188abc85ff7cb0e", "tree": "5ff40995ecf0a186bf5e5daa172aa04df8f7d00d", "parents": [ "8d9be06341816e5fb7e29b2fd44b3ffe8dd3263a" ], "author": { "name": "M Chetan Kumar", "email": "m.chetan.kumar@linux.intel.com", "time": "Sun Sep 19 22:58:18 2021 +0530" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Sep 20 10:03:37 2021 +0100" }, "message": "net: wwan: iosm: devlink fw flashing \u0026 cd collection documentation\n\nDocuments devlink params, fw update \u0026 cd collection commands\nand its usage.\n\nSigned-off-by: M Chetan Kumar \u003cm.chetan.kumar@linux.intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "8d9be06341816e5fb7e29b2fd44b3ffe8dd3263a", "tree": "3b72d8e3a5853305b937c42eecd96b5cca7b45d3", "parents": [ "09e7b002ff67342364af735f7bbf13b0be1fcdfc" ], "author": { "name": "M Chetan Kumar", "email": "m.chetan.kumar@linux.intel.com", "time": "Sun Sep 19 22:57:56 2021 +0530" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Sep 20 10:03:37 2021 +0100" }, "message": "net: wwan: iosm: transport layer support for fw flashing/cd\n\nImplements transport layer protocol for fw flashing/coredump\ncollection.\n\nSigned-off-by: M Chetan Kumar \u003cm.chetan.kumar@linux.intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "09e7b002ff67342364af735f7bbf13b0be1fcdfc", "tree": "5ff4322c73b362af093f400d8943b1cbc4fa5744", "parents": [ "b55734745568234146c83fa52b67580288b382ec" ], "author": { "name": "M Chetan Kumar", "email": "m.chetan.kumar@linux.intel.com", "time": "Sun Sep 19 22:57:27 2021 +0530" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Sep 20 10:03:37 2021 +0100" }, "message": "net: wwan: iosm: coredump collection support\n\nImplements protocol for coredump collection.\n\nSigned-off-by: M Chetan Kumar \u003cm.chetan.kumar@linux.intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "b55734745568234146c83fa52b67580288b382ec", "tree": "02e0fa19fc842c1aef9d9b0ef5ca02f535cc191b", "parents": [ "4dcd183fbd67b105decc8be262311937730ccdbf" ], "author": { "name": "M Chetan Kumar", "email": "m.chetan.kumar@linux.intel.com", "time": "Sun Sep 19 22:57:05 2021 +0530" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Sep 20 10:03:37 2021 +0100" }, "message": "net: wwan: iosm: fw flashing support\n\nImplements protocol for fw flashing and PSI injection for\ncoredump collection.\n\nSigned-off-by: M Chetan Kumar \u003cm.chetan.kumar@linux.intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4dcd183fbd67b105decc8be262311937730ccdbf", "tree": "7211a52f210325ca7a24a17f805cb3770639e511", "parents": [ "d44fd8604a4ab92119adb35f05fd87612af722b5" ], "author": { "name": "M Chetan Kumar", "email": "m.chetan.kumar@linux.intel.com", "time": "Sun Sep 19 22:56:18 2021 +0530" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Sep 20 10:03:37 2021 +0100" }, "message": "net: wwan: iosm: devlink registration\n\nRegister with devlink framework and implment callbacks required\nfor fw flashing and coredump collection.\n\nSigned-off-by: M Chetan Kumar \u003cm.chetan.kumar@linux.intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "d44fd8604a4ab92119adb35f05fd87612af722b5", "tree": "d3229a3afd8540d6a9107bd236f3df471af222b6", "parents": [ "15b9df4ece17d084f14eb0ca1cf05f2ad497e425" ], "author": { "name": "Ansuel Smith", "email": "ansuelsmth@gmail.com", "time": "Sun Sep 19 18:28:17 2021 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Sep 20 09:51:21 2021 +0100" }, "message": "net: phy: at803x: fix spacing and improve name for 83xx phy\n\nFix spacing and improve name for 83xx phy following other phy in the\nsame driver.\n\nSigned-off-by: Ansuel Smith \u003cansuelsmth@gmail.com\u003e\nReviewed-by: Andrew Lunn \u003candrew@lunn.ch\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "15b9df4ece17d084f14eb0ca1cf05f2ad497e425", "tree": "402aaa34c521ea67d1fe7075c561aa2e3b101478", "parents": [ "b4df02b562f4aa14ff6811f30e1b4d2159585c59" ], "author": { "name": "Ansuel Smith", "email": "ansuelsmth@gmail.com", "time": "Sun Sep 19 18:28:16 2021 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Sep 20 09:51:21 2021 +0100" }, "message": "net: phy: at803x: add resume/suspend function to qca83xx phy\n\nAdd resume/suspend function to qca83xx internal phy.\nWe can\u0027t use the at803x generic function as the documentation lacks of\nany support for WoL regs.\n\nSigned-off-by: Ansuel Smith \u003cansuelsmth@gmail.com\u003e\nReviewed-by: Andrew Lunn \u003candrew@lunn.ch\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "b4df02b562f4aa14ff6811f30e1b4d2159585c59", "tree": "9410d7506e517aa176cec6e4c1cc655239a7e1eb", "parents": [ "a520794b063b9567799533c4fec230d860f5c9e5" ], "author": { "name": "Ansuel Smith", "email": "ansuelsmth@gmail.com", "time": "Sun Sep 19 18:28:15 2021 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Sep 20 09:51:21 2021 +0100" }, "message": "net: phy: at803x: add support for qca 8327 A variant internal phy\n\nFor qca8327 internal phy there are 2 different switch variant with 2\ndifferent phy id. Add this missing variant so the internal phy can be\ncorrectly identified and fixed.\n\nSigned-off-by: Ansuel Smith \u003cansuelsmth@gmail.com\u003e\nReviewed-by: Andrew Lunn \u003candrew@lunn.ch\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "a520794b063b9567799533c4fec230d860f5c9e5", "tree": "652045543f6c2fb6c19894c177867904d68f11fe", "parents": [ "14e94f9445a9e91d460f5d4b519f8892c3fb14bb" ], "author": { "name": "Tony Lu", "email": "tony.ly@linux.alibaba.com", "time": "Fri Sep 17 16:40:06 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Sep 20 09:42:51 2021 +0100" }, "message": "virtio_net: introduce TX timeout watchdog\n\nThis implements ndo_tx_timeout handler and put this into stats. When\nthere is something wrong to send out packets, we could notice tx timeout\nevents and total timeout counter.\n\nWe have suffered send timeout issues due to the backends hung. With this,\nwe can find the details, and collect the counters by monitor systems.\n\nSigned-off-by: Tony Lu \u003ctony.ly@linux.alibaba.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "14e94f9445a9e91d460f5d4b519f8892c3fb14bb", "tree": "886a27a5f7c7ace15b6f18dfbde78b736eb6f486", "parents": [ "f7116fb4608500ad93eca03e0ad158b75fa85d0b" ], "author": { "name": "Hariprasad Kelam", "email": "hkelam@marvell.com", "time": "Fri Sep 17 18:40:24 2021 +0530" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Sep 19 14:05:50 2021 +0100" }, "message": "octeontx2-af: verify CQ context updates\n\nAs per HW errata AQ modification to CQ could be discarded on heavy\ntraffic. This patch implements workaround for the same after each\nCQ write by AQ check whether the requested fields (except those\nwhich HW can update eg: avg_level) are properly updated or not.\n\nIf CQ context is not updated then perform AQ write again.\n\nSigned-off-by: Hariprasad Kelam \u003chkelam@marvell.com\u003e\nSigned-off-by: Sunil Goutham \u003csgoutham@marvell.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "f7116fb4608500ad93eca03e0ad158b75fa85d0b", "tree": "5565bd03953c0c7cb1dfa581d438d25f19e0039d", "parents": [ "cbcca2e3961eac736566ac13ef0d0bf6f0b764ec" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Fri Sep 17 06:55:06 2021 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Sep 19 13:26:01 2021 +0100" }, "message": "net: sched: move and reuse mq_change_real_num_tx()\n\nThe code for handling active queue changes is identical\nbetween mq and mqprio, reuse it.\n\nSuggested-by: Cong Wang \u003ccong.wang@bytedance.com\u003e\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "cbcca2e3961eac736566ac13ef0d0bf6f0b764ec", "tree": "a1638941e4d1d3acc7698f0e3ed5d1325ae1c77b", "parents": [ "4fc29989835ab7379f807854b5f7338b752b9f1a" ], "author": { "name": "Russell King (Oracle)", "email": "rmk+kernel@armlinux.org.uk", "time": "Fri Sep 17 14:36:31 2021 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Sep 19 13:23:47 2021 +0100" }, "message": "net: phylink: don\u0027t call netif_carrier_off() with NULL netdev\n\nDan Carpenter points out that we have a code path that permits a NULL\nnetdev pointer to be passed to netif_carrier_off(), which will cause\na kernel oops. In any case, we need to set pl-\u003eold_link_state to false\nto have the desired effect when there is no netdev present.\n\nFixes: f97493657c63 (\"net: phylink: add suspend/resume support\")\nReported-by: Dan Carpenter \u003cdan.carpenter@oracle.com\u003e\nSigned-off-by: Russell King (Oracle) \u003crmk+kernel@armlinux.org.uk\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4fc29989835ab7379f807854b5f7338b752b9f1a", "tree": "129d9b2d493ca913acc2b20e15f1c7c9cf5aa119", "parents": [ "9ce4e3d6d85600153e61c4b8c3c346bd84d84470" ], "author": { "name": "Yajun Deng", "email": "yajun.deng@linux.dev", "time": "Sat Sep 18 14:36:07 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Sep 19 12:56:02 2021 +0100" }, "message": "net: rtnetlink: convert rcu_assign_pointer to RCU_INIT_POINTER\n\nIt no need barrier when assigning a NULL value to an RCU protected\npointer. So use RCU_INIT_POINTER() instead for more fast.\n\nSigned-off-by: Yajun Deng \u003cyajun.deng@linux.dev\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "9ce4e3d6d85600153e61c4b8c3c346bd84d84470", "tree": "8b5de1d26d4684fd0b20761bc9a2cb5a8cb63b4b", "parents": [ "db9c8e2b1e246fc2dc20828932949437793146cc" ], "author": { "name": "Xuan Zhuo", "email": "xuanzhuo@linux.alibaba.com", "time": "Sat Sep 18 14:06:15 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Sep 19 12:46:17 2021 +0100" }, "message": "virtio_net: use netdev_warn_once to output warn when without enough queues\n\nThis warning is output when virtnet does not have enough queues, but it\nonly needs to be printed once to inform the user of this situation. It\nis not necessary to print it every time. If the user loads xdp\nfrequently, this log appears too much.\n\nSigned-off-by: Xuan Zhuo \u003cxuanzhuo@linux.alibaba.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "db9c8e2b1e246fc2dc20828932949437793146cc", "tree": "63aa61228e6b84238b5a19642a47f90f17268eb0", "parents": [ "aed0826b0cf2e488900ab92193893e803d65c070" ], "author": { "name": "wangzhitong", "email": "wangzhitong@uniontech.com", "time": "Sat Sep 18 17:39:10 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Sep 19 12:43:56 2021 +0100" }, "message": "NET: IPV4: fix error \"do not initialise globals to 0\"\n\nthis patch fixes below Errors reported by checkpatch\n ERROR: do not initialise globals to 0\n +int cipso_v4_rbm_optfmt \u003d 0;\n\nSigned-off-by: wangzhitong \u003cwangzhitong@uniontech.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "aed0826b0cf2e488900ab92193893e803d65c070", "tree": "505d95bcbc13d94f7ad1bbbe0b31b27ac6162851", "parents": [ "9eb7b5e7cb50942f55c1346b0f8d940c07201202" ], "author": { "name": "Yajun Deng", "email": "yajun.deng@linux.dev", "time": "Sat Sep 18 17:04:10 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Sep 19 12:43:04 2021 +0100" }, "message": "net: net_namespace: Fix undefined member in key_remove_domain()\n\nThe key_domain member in struct net only exists if we define CONFIG_KEYS.\nSo we should add the define when we used key_domain.\n\nFixes: 9b242610514f (\"keys: Network namespace domain tag\")\nSigned-off-by: Yajun Deng \u003cyajun.deng@linux.dev\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "9eb7b5e7cb50942f55c1346b0f8d940c07201202", "tree": "3d32b71d12f6cbde70ce53f65cbce02f9bfc8287", "parents": [ "983e59a27b92d7e1a3432a1009522177cc1e187b" ], "author": { "name": "Russell King", "email": "rmk+kernel@armlinux.org.uk", "time": "Fri Sep 17 14:41:17 2021 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Sep 19 12:11:40 2021 +0100" }, "message": "net: dpaa2-mac: add support for more ethtool 10G link modes\n\nPhylink documentation says:\n Note that the PHY may be able to transform from one connection\n technology to another, so, eg, don\u0027t clear 1000BaseX just\n because the MAC is unable to BaseX mode. This is more about\n clearing unsupported speeds and duplex settings. The port modes\n should not be cleared; phylink_set_port_modes() will help with this.\n\nSo add the missing 10G modes.\n\nSigned-off-by: Russell King \u003crmk+kernel@armlinux.org.uk\u003e\nAcked-by: Marek Behún \u003ckabel@kernel.org\u003e\nAcked-by: Ioana Ciornei \u003cioana.ciornei@nxp.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "983e59a27b92d7e1a3432a1009522177cc1e187b", "tree": "4252af1ca2a9aefc97ba62cea0f9d78932580f08", "parents": [ "95dca2d578d2ae702b9d9e67a1facfdc918cfa27", "ce9979129a0ba700112151a83a6d4cf09c7a1158" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Sep 18 14:20:01 2021 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Sep 18 14:20:01 2021 +0100" }, "message": "Merge branch \u0027mptcp-next\u0027\n\nMat Martineau says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nmptcp: Add SOL_MPTCP getsockopt support\n\nHere\u0027s the first new MPTCP feature for the v5.16 cycle, and I\u0027ll defer\nto Florian\u0027s helpful description of the series implementing some new\nMPTCP socket options:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nThis adds the MPTCP_INFO, MPTCP_TCPINFO and MPTCP_SUBFLOW_ADDRS\nmptcp getsockopt optnames.\n\nMPTCP_INFO exposes the mptcp_info struct as an alternative to the\nexisting netlink diag interface.\n\nMPTCP_TCPINFO exposes the tcp_info struct.\nUnlike SOL_TCP/TCP_INFO, this returns one struct for each active\nsubflow.\n\nMPTCP_SUBFLOW_ADDRS allows userspace to discover the ip addresses/ports\nused by the local and remote endpoints, one for each active tcp subflow.\n\nMPTCP_TCPINFO and MPTCP_SUBFLOW_ADDRS share the same meta-header that\nneeds to be pre-filled by userspace with the size of the data structures\nit expects. This is done to allow extension of the involved structs\nlater on, without breaking backwards compatibility.\n\nThe meta-structure can also be used to discover the required space\nto obtain all information, as kernel will fill in the number of\nactive subflows even if there is not enough room for the requested info\nitself.\n\nMore information is available in the individual patches.\nLast patch adds test cases for the three optnames.\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "ce9979129a0ba700112151a83a6d4cf09c7a1158", "tree": "4252af1ca2a9aefc97ba62cea0f9d78932580f08", "parents": [ "c11c5906bc0aba62a78da69035f6b30c6da6d13b" ], "author": { "name": "Florian Westphal", "email": "fw@strlen.de", "time": "Fri Sep 17 16:33:22 2021 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Sep 18 14:20:01 2021 +0100" }, "message": "selftests: mptcp: add mptcp getsockopt test cases\n\nAdd a test program that retrieves the three info types:\n1. mptcp meta information\n2. tcp info for subflow\n3. subflow endpoint addresses\n\nFor all three rudimentary checks are added.\n\n1. Meta information checks that the logical mptcp\n sequence numbers advance as expected, based on the bytes read\n (init seq + bytes_received/sent) and the connection state\n (after close, we should exect 1 extra byte due to FIN).\n\n2. TCP info checks the number of bytes sent/received vs.\n sums of read/write syscall return values.\n\n3. Subflow endpoint addresses are checked vs. getsockname/getpeername\n result.\n\nTests for forward compatibility (0-initialisation of output-only\nfields in mptcp_subflow_data structure) are added as well.\n\nCo-developed-by: Matthieu Baerts \u003cmatthieu.baerts@tessares.net\u003e\nSigned-off-by: Matthieu Baerts \u003cmatthieu.baerts@tessares.net\u003e\nSigned-off-by: Florian Westphal \u003cfw@strlen.de\u003e\nSigned-off-by: Mat Martineau \u003cmathew.j.martineau@linux.intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "c11c5906bc0aba62a78da69035f6b30c6da6d13b", "tree": "503a6dd7e0654e69aeec170f46f3551f894a7cda", "parents": [ "06f15cee369535a383c9c82ed37a25f0a413f6f1" ], "author": { "name": "Florian Westphal", "email": "fw@strlen.de", "time": "Fri Sep 17 16:33:21 2021 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Sep 18 14:20:01 2021 +0100" }, "message": "mptcp: add MPTCP_SUBFLOW_ADDRS getsockopt support\n\nThis retrieves the address pairs of all subflows currently\nactive for a given mptcp connection.\n\nIt re-uses the same meta-header as for MPTCP_TCPINFO.\n\nA new structure is provided to hold the subflow\naddress data:\n\nstruct mptcp_subflow_addrs {\n\tunion {\n\t\t__kernel_sa_family_t sa_family;\n\t\tstruct sockaddr sa_local;\n\t\tstruct sockaddr_in sin_local;\n\t\tstruct sockaddr_in6 sin6_local;\n\t\tstruct sockaddr_storage ss_local;\n\t};\n\tunion {\n\t\tstruct sockaddr sa_remote;\n\t\tstruct sockaddr_in sin_remote;\n\t\tstruct sockaddr_in6 sin6_remote;\n\t\tstruct sockaddr_storage ss_remote;\n\t};\n};\n\nUsage of the new getsockopt is very similar to\nMPTCP_TCPINFO one.\n\nUserspace allocates a\n\u0027struct mptcp_subflow_data\u0027, followed by one or\nmore \u0027struct mptcp_subflow_addrs\u0027, then inits the\nmptcp_subflow_data structure as follows:\n\nstruct mptcp_subflow_addrs *sf_addr;\nstruct mptcp_subflow_data *addr;\nsocklen_t olen \u003d sizeof(*addr) + (8 * sizeof(*sf_addr));\n\naddr \u003d malloc(olen);\naddr-\u003esize_subflow_data \u003d sizeof(*addr);\naddr-\u003enum_subflows \u003d 0;\naddr-\u003esize_kernel \u003d 0;\naddr-\u003esize_user \u003d sizeof(struct mptcp_subflow_addrs);\n\nsf_addr \u003d (struct mptcp_subflow_addrs *)(addr + 1);\n\nand then retrieves the endpoint addresses via:\nret \u003d getsockopt(fd, SOL_MPTCP, MPTCP_SUBFLOW_ADDRS,\n\t\t addr, \u0026olen);\n\nIf the call succeeds, kernel will have added up to 8\nendpoint addresses after the \u0027mptcp_subflow_data\u0027 header.\n\nUserspace needs to re-check \u0027olen\u0027 value to detect how\nmany bytes have been filled in by the kernel.\n\nUserspace can check addr-\u003enum_subflows to discover when\nthere were more subflows that available data space.\n\nCo-developed-by: Matthieu Baerts \u003cmatthieu.baerts@tessares.net\u003e\nSigned-off-by: Matthieu Baerts \u003cmatthieu.baerts@tessares.net\u003e\nSigned-off-by: Florian Westphal \u003cfw@strlen.de\u003e\nSigned-off-by: Mat Martineau \u003cmathew.j.martineau@linux.intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "06f15cee369535a383c9c82ed37a25f0a413f6f1", "tree": "0f4ddb0681586e8d5b4c49f1b7a6de4767776a8a", "parents": [ "55c42fa7fa331f98062c32799456420930b8bf8c" ], "author": { "name": "Florian Westphal", "email": "fw@strlen.de", "time": "Fri Sep 17 16:33:20 2021 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Sep 18 14:20:01 2021 +0100" }, "message": "mptcp: add MPTCP_TCPINFO getsockopt support\n\nAllow users to retrieve TCP_INFO data of all subflows.\n\nUsers need to pre-initialize a meta header that has to be\nprepended to the data buffer that will be filled with the tcp info data.\n\nThe meta header looks like this:\n\nstruct mptcp_subflow_data {\n __u32 size_subflow_data;/* size of this structure in userspace */\n __u32 num_subflows;\t/* must be 0, set by kernel */\n __u32 size_kernel;\t/* must be 0, set by kernel */\n __u32 size_user;\t/* size of one element in data[] */\n} __attribute__((aligned(8)));\n\nsize_subflow_data has to be set to \u0027sizeof(struct mptcp_subflow_data)\u0027.\nThis allows to extend mptcp_subflow_data structure later on without\nbreaking backwards compatibility.\n\nIf the structure is extended later on, kernel knows where the\nuserspace-provided meta header ends, even if userspace uses an older\n(smaller) version of the structure.\n\nnum_subflows must be set to 0. If the getsockopt request succeeds (return\nvalue is 0), it will be updated to contain the number of active subflows\nfor the given logical connection.\n\nsize_kernel must be set to 0. If the getsockopt request is successful,\nit will contain the size of the \u0027struct tcp_info\u0027 as known by the kernel.\nThis is informational only.\n\nsize_user must be set to \u0027sizeof(struct tcp_info)\u0027.\n\nThis allows the kernel to only fill in the space reserved/expected by\nuserspace.\n\nExample:\n\nstruct my_tcp_info {\n struct mptcp_subflow_data d;\n struct tcp_info ti[2];\n};\nstruct my_tcp_info ti;\nsocklen_t olen;\n\nmemset(\u0026ti, 0, sizeof(ti));\n\nti.d.size_subflow_data \u003d sizeof(struct mptcp_subflow_data);\nti.d.size_user \u003d sizeof(struct tcp_info);\nolen \u003d sizeof(ti);\n\nret \u003d getsockopt(fd, SOL_MPTCP, MPTCP_TCPINFO, \u0026ti, \u0026olen);\nif (ret \u003c 0)\n\tdie_perror(\"getsockopt MPTCP_TCPINFO\");\n\nmptcp_subflow_data.num_subflows is populated with the number of\nsubflows that exist on the kernel side for the logical mptcp connection.\n\nThis allows userspace to re-try with a larger tcp_info array if the number\nof subflows was larger than the available space in the ti[] array.\n\nolen has to be set to the number of bytes that userspace has allocated to\nreceive the kernel data. It will be updated to contain the real number\nbytes that have been copied to by the kernel.\n\nIn the above example, if the number if subflows was 1, olen is equal to\n\u0027sizeof(struct mptcp_subflow_data) + sizeof(struct tcp_info).\nFor 2 or more subflows olen is equal to \u0027sizeof(struct my_tcp_info)\u0027.\n\nIf there was more data that could not be copied due to lack of space\nin the option buffer, userspace can detect this by checking\nmptcp_subflow_data-\u003enum_subflows.\n\nSigned-off-by: Florian Westphal \u003cfw@strlen.de\u003e\nSigned-off-by: Mat Martineau \u003cmathew.j.martineau@linux.intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "55c42fa7fa331f98062c32799456420930b8bf8c", "tree": "42ce80bd04e563a3ee02ef07819e49ed253073ab", "parents": [ "61bc6e82f92ec105505266e484d90a935e1249ac" ], "author": { "name": "Florian Westphal", "email": "fw@strlen.de", "time": "Fri Sep 17 16:33:19 2021 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Sep 18 14:20:01 2021 +0100" }, "message": "mptcp: add MPTCP_INFO getsockopt\n\nIts not compatible with multipath-tcp.org kernel one.\n\n1. The out-of-tree implementation defines a different \u0027struct mptcp_info\u0027,\n with embedded __user addresses for additional data such as\n endpoint addresses.\n\n2. Mat Martineau points out that embedded __user addresses doesn\u0027t work\nwith BPF_CGROUP_RUN_PROG_GETSOCKOPT() which assumes that copying in\noptsize bytes from optval provides all data that got copied to userspace.\n\nThis provides mptcp_info data for the given mptcp socket.\n\nUserspace sets optlen to the size of the structure it expects.\nThe kernel updates it to contain the number of bytes that it copied.\n\nThis allows to append more information to the structure later.\n\nSigned-off-by: Florian Westphal \u003cfw@strlen.de\u003e\nSigned-off-by: Mat Martineau \u003cmathew.j.martineau@linux.intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "61bc6e82f92ec105505266e484d90a935e1249ac", "tree": "1d188dec8b3ebb92979a3b89d9b7e308ad15f485", "parents": [ "95dca2d578d2ae702b9d9e67a1facfdc918cfa27" ], "author": { "name": "Florian Westphal", "email": "fw@strlen.de", "time": "Fri Sep 17 16:33:18 2021 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Sep 18 14:20:00 2021 +0100" }, "message": "mptcp: add new mptcp_fill_diag helper\n\nWill be re-used from getsockopt path.\nSince diag can be a module, we can\u0027t export the helper from diag, it\nneeds to be moved to core.\n\nSigned-off-by: Florian Westphal \u003cfw@strlen.de\u003e\nSigned-off-by: Mat Martineau \u003cmathew.j.martineau@linux.intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "95dca2d578d2ae702b9d9e67a1facfdc918cfa27", "tree": "d6f69c29026b62949b82996ab7ad7d1afdec7e55", "parents": [ "b972b54a68b2512a7528658ecd023aea108c03a5", "0f4f6d7332bbc3567ae01466812c089f133cfc83" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Sep 18 14:14:39 2021 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Sep 18 14:14:39 2021 +0100" }, "message": "Merge branch \u0027macb-MII-on-RGMII\u0027\n\nClaudiu Beznea says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nnet: macb: add support for MII on RGMII interface\n\nThis series adds support for MII mode on RGMII interface (patches 3/4,\n4/4). Along with this the series also contains minor cleanups (patches 1/3,\n2/3) on macb.h.\n\nChanges in v2:\n- added patch 4/4 to enable MII on RGMII support for SAMA7G5 MAC IPs\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "0f4f6d7332bbc3567ae01466812c089f133cfc83", "tree": "d6f69c29026b62949b82996ab7ad7d1afdec7e55", "parents": [ "1a9b5a26daf606868220f24b9783c0f37085454d" ], "author": { "name": "Claudiu Beznea", "email": "claudiu.beznea@microchip.com", "time": "Fri Sep 17 16:26:15 2021 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Sep 18 14:14:39 2021 +0100" }, "message": "net: macb: enable mii on rgmii for sama7g5\n\nBoth MAC IPs available on SAMA7G5 support MII on RGMII feature.\nEnable these by adding proper capability to proper macb_config\nobjects.\n\nSigned-off-by: Claudiu Beznea \u003cclaudiu.beznea@microchip.com\u003e\nAcked-by: Nicolas Ferre \u003cnicolas.ferre@microchip.com\u003e\nReviewed-by: Andrew Lunn \u003candrew@lunn.ch\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "1a9b5a26daf606868220f24b9783c0f37085454d", "tree": "76ecaa01745a60b9b6110fcb5d2a700d4436085e", "parents": [ "d7b3485f1c2b917328890ba28f4e7cac4b98ca1e" ], "author": { "name": "Claudiu Beznea", "email": "claudiu.beznea@microchip.com", "time": "Fri Sep 17 16:26:14 2021 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Sep 18 14:14:39 2021 +0100" }, "message": "net: macb: add support for mii on rgmii\n\nCadence IP has option to enable MII support on RGMII interface. This\ncould be selected though bit 28 of network control register. This option\nis not enabled on all the IP versions thus add a software capability to\nbe selected by the proper implementation of this IP.\n\nSigned-off-by: Claudiu Beznea \u003cclaudiu.beznea@microchip.com\u003e\nAcked-by: Nicolas Ferre \u003cnicolas.ferre@microchip.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "d7b3485f1c2b917328890ba28f4e7cac4b98ca1e", "tree": "9762ba17abd2783e2ae6a48915ac424ddb2efefb", "parents": [ "1dac0084d41297c7efc4f271e1e6430b624f2713" ], "author": { "name": "Claudiu Beznea", "email": "claudiu.beznea@microchip.com", "time": "Fri Sep 17 16:26:13 2021 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Sep 18 14:14:39 2021 +0100" }, "message": "net: macb: align for OSSMODE offset\n\nAlign for OSSMODE offset.\n\nSigned-off-by: Claudiu Beznea \u003cclaudiu.beznea@microchip.com\u003e\nAcked-by: Nicolas Ferre \u003cnicolas.ferre@microchip.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "1dac0084d41297c7efc4f271e1e6430b624f2713", "tree": "d23833fd01cc2c09e871e0ae9c2a96505f932e7b", "parents": [ "b972b54a68b2512a7528658ecd023aea108c03a5" ], "author": { "name": "Claudiu Beznea", "email": "claudiu.beznea@microchip.com", "time": "Fri Sep 17 16:26:12 2021 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Sep 18 14:14:38 2021 +0100" }, "message": "net: macb: add description for SRTSM\n\nAdd description for SRTSM bit.\n\nSigned-off-by: Claudiu Beznea \u003cclaudiu.beznea@microchip.com\u003e\nAcked-by: Nicolas Ferre \u003cnicolas.ferre@microchip.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "b972b54a68b2512a7528658ecd023aea108c03a5", "tree": "c4fcee652454d58925ab62c2632d8f794ab51f51", "parents": [ "0efcc3f201452aa42670d2da1d72858f95d0b7f7" ], "author": { "name": "Florian Fainelli", "email": "f.fainelli@gmail.com", "time": "Fri Sep 17 14:55:38 2021 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Sep 18 11:52:07 2021 +0100" }, "message": "net: bcmgenet: Patch PHY interface for dedicated PHY driver\n\nWhen we are using a dedicated PHY driver (not the Generic PHY driver)\nchances are that it is going to configure RGMII delays and do that in a\nway that is incompatible with our incorrect interpretation of the\nphy_interface value.\n\nAdd a quirk in order to reverse the PHY_INTERFACE_MODE_RGMII to the\nvalue of PHY_INTERFACE_MODE_RGMII_ID such that the MAC continues to be\nconfigured the way it used to be, but the PHY driver can account for\nadding delays. Conversely when PHY_INTERFACE_MODE_RGMII_TXID is\nspecified, return PHY_INTERFACE_MODE_RGMII_RXID to the PHY since we will\nhave enabled a TXC MAC delay (id_mode_dis\u003d0, meaning there is a delay\ninserted).\n\nThis is not considered a bug fix at this point since it only affects\nBroadcom STB platforms shipping with a Device Tree blob that is not\nupdatable in the field (quite a few devices out there) and which was\ngenerated using the scripted Device Tree environment shipped with those\nplatforms\u0027 SDK.\n\nSigned-off-by: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "0efcc3f201452aa42670d2da1d72858f95d0b7f7", "tree": "ea96dc334f606785b7b6a900e136a594bbbc4b2b", "parents": [ "af54faab84f754ebd42ecdda871f8d71940ae40b" ], "author": { "name": "Heiner Kallweit", "email": "hkallweit1@gmail.com", "time": "Thu Sep 16 23:40:37 2021 +0200" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Fri Sep 17 18:27:43 2021 -0700" }, "message": "sky2: Stop printing VPD info to debugfs\n\nSky2 is parsing the VPD and adds the parsed information to its debugfs\nfile. This isn\u0027t needed in kernel, userspace tools like lspci can be\nused to display such information nicely. Therefore remove this from\nthe driver.\n\nlspci -vv:\n\nCapabilities: [50] Vital Product Data\n\tProduct Name: Marvell Yukon 88E8070 Gigabit Ethernet Controller\n\tRead-only fields:\n\t\t[PN] Part number: Yukon 88E8070\n\t\t[EC] Engineering changes: Rev. 1.0\n\t\t[MN] Manufacture ID: Marvell\n\t\t[SN] Serial number: AbCdEfG970FD4\n\t\t[CP] Extended capability: 01 10 cc 03\n\t\t[RV] Reserved: checksum good, 9 byte(s) reserved\n\tRead/write fields:\n\t\t[RW] Read-write area: 1 byte(s) free\n\tEnd\n\nRelevant part in debugfs file:\n\n0000:01:00.0 Product Data\nMarvell Yukon 88E8070 Gigabit Ethernet Controller\n Part Number: Yukon 88E8070\n Engineering Level: Rev. 1.0\n Manufacturer: Marvell\n Serial Number: AbCdEfG970FD4\n\nSigned-off-by: Heiner Kallweit \u003chkallweit1@gmail.com\u003e\nAcked-by: Stephen Hemminger \u003cstephen@networkplumber.org\u003e\nLink: https://lore.kernel.org/r/bbaee8ab-9b2e-de04-ee7b-571e094cc5fe@gmail.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "af54faab84f754ebd42ecdda871f8d71940ae40b", "tree": "63a6465f20e891afcfc7dceecd8cc8764f20e6ae", "parents": [ "f68d08c437f98ee19a14142b9de2d7afe2032d5c", "ca21a3e5edfd47c90141724557f9d6f5000e46f3" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Fri Sep 17 12:40:20 2021 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Fri Sep 17 12:40:21 2021 -0700" }, "message": "Merge https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next\n\nAlexei Starovoitov says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\npull-request: bpf-next 2021-09-17\n\nWe\u0027ve added 63 non-merge commits during the last 12 day(s) which contain\na total of 65 files changed, 2653 insertions(+), 751 deletions(-).\n\nThe main changes are:\n\n1) Streamline internal BPF program sections handling and\n bpf_program__set_attach_target() in libbpf, from Andrii.\n\n2) Add support for new btf kind BTF_KIND_TAG, from Yonghong.\n\n3) Introduce bpf_get_branch_snapshot() to capture LBR, from Song.\n\n4) IMUL optimization for x86-64 JIT, from Jie.\n\n5) xsk selftest improvements, from Magnus.\n\n6) Introduce legacy kprobe events support in libbpf, from Rafael.\n\n7) Access hw timestamp through BPF\u0027s __sk_buff, from Vadim.\n\n* https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next: (63 commits)\n selftests/bpf: Fix a few compiler warnings\n libbpf: Constify all high-level program attach APIs\n libbpf: Schedule open_opts.attach_prog_fd deprecation since v0.7\n selftests/bpf: Switch fexit_bpf2bpf selftest to set_attach_target() API\n libbpf: Allow skipping attach_func_name in bpf_program__set_attach_target()\n libbpf: Deprecated bpf_object_open_opts.relaxed_core_relocs\n selftests/bpf: Stop using relaxed_core_relocs which has no effect\n libbpf: Use pre-setup sec_def in libbpf_find_attach_btf_id()\n bpf: Update bpf_get_smp_processor_id() documentation\n libbpf: Add sphinx code documentation comments\n selftests/bpf: Skip btf_tag test if btf_tag attribute not supported\n docs/bpf: Add documentation for BTF_KIND_TAG\n selftests/bpf: Add a test with a bpf program with btf_tag attributes\n selftests/bpf: Test BTF_KIND_TAG for deduplication\n selftests/bpf: Add BTF_KIND_TAG unit tests\n selftests/bpf: Change NAME_NTH/IS_NAME_NTH for BTF_KIND_TAG format\n selftests/bpf: Test libbpf API function btf__add_tag()\n bpftool: Add support for BTF_KIND_TAG\n libbpf: Add support for BTF_KIND_TAG\n libbpf: Rename btf_{hash,equal}_int to btf_{hash,equal}_int_tag\n ...\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nLink: https://lore.kernel.org/r/20210917173738.3397064-1-ast@kernel.org\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "f68d08c437f98ee19a14142b9de2d7afe2032d5c", "tree": "f918fb703bc0e08647d6df6ffdb2bd9eed4054e5", "parents": [ "1a575cde596c44aee04d2853d3d7067942a9612c" ], "author": { "name": "Florian Fainelli", "email": "f.fainelli@gmail.com", "time": "Fri Sep 17 11:15:50 2021 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Fri Sep 17 11:49:55 2021 -0700" }, "message": "net: phy: bcm7xxx: Add EPHY entry for 72165\n\n72165 is a 16nm process SoC with a 10/100 integrated Ethernet PHY,\ncreate a new macro and set of functions for this different process type.\n\nSigned-off-by: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nLink: https://lore.kernel.org/r/20210917181551.2836036-1-f.fainelli@gmail.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "1a575cde596c44aee04d2853d3d7067942a9612c", "tree": "a1d56173353244c978dd0dd6ce1d0292079884ef", "parents": [ "5ef8a0291513c624b568acf15eee87e905736f2c" ], "author": { "name": "Nathan Chancellor", "email": "nathan@kernel.org", "time": "Thu Sep 16 21:52:05 2021 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Fri Sep 17 10:32:08 2021 -0700" }, "message": "ptp: ocp: Avoid operator precedence warning in ptp_ocp_summary_show()\n\nClang warns twice:\n\ndrivers/ptp/ptp_ocp.c:2065:16: error: operator \u0027?:\u0027 has lower precedence\nthan \u0027\u0026\u0027; \u0027\u0026\u0027 will be evaluated first\n[-Werror,-Wbitwise-conditional-parentheses]\n on \u0026 map ? \" ON\" : \"OFF\", src);\n ~~~~~~~~ ^\ndrivers/ptp/ptp_ocp.c:2065:16: note: place parentheses around the \u0027\u0026\u0027\nexpression to silence this warning\n on \u0026 map ? \" ON\" : \"OFF\", src);\n ^\n ( )\ndrivers/ptp/ptp_ocp.c:2065:16: note: place parentheses around the \u0027?:\u0027\nexpression to evaluate it first\n on \u0026 map ? \" ON\" : \"OFF\", src);\n ^\n\non and map are both booleans so this should be a logical AND, which\nclears up the operator precedence issue.\n\nFixes: a62a56d04e63 (\"ptp: ocp: Enable 4th timestamper / PPS generator\")\nLink: https://github.com/ClangBuiltLinux/linux/issues/1457\nSuggested-by: Jonathan Lemon \u003cjonathan.lemon@gmail.com\u003e\nSigned-off-by: Nathan Chancellor \u003cnathan@kernel.org\u003e\nAcked-by: Jonathan Lemon \u003cjonathan.lemon@gmail.com\u003e\nLink: https://lore.kernel.org/r/20210917045204.1385801-1-nathan@kernel.org\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "ca21a3e5edfd47c90141724557f9d6f5000e46f3", "tree": "266f00cbcfbba465aa7f3342b988cd1bda65faa5", "parents": [ "f706f6c66c4360e41dd650bcf24a45880e75bb39" ], "author": { "name": "Yonghong Song", "email": "yhs@fb.com", "time": "Thu Sep 16 21:33:43 2021 -0700" }, "committer": { "name": "Alexei Starovoitov", "email": "ast@kernel.org", "time": "Fri Sep 17 09:10:54 2021 -0700" }, "message": "selftests/bpf: Fix a few compiler warnings\n\nWith clang building selftests/bpf, I hit a few warnings like below:\n\n .../bpf_iter.c:592:48: warning: variable \u0027expected_key_c\u0027 set but not used [-Wunused-but-set-variable]\n __u32 expected_key_a \u003d 0, expected_key_b \u003d 0, expected_key_c \u003d 0;\n ^\n\n .../bpf_iter.c:688:48: warning: variable \u0027expected_key_c\u0027 set but not used [-Wunused-but-set-variable]\n __u32 expected_key_a \u003d 0, expected_key_b \u003d 0, expected_key_c \u003d 0;\n ^\n\n .../tc_redirect.c:657:6: warning: variable \u0027target_fd\u0027 is used uninitialized whenever \u0027if\u0027 condition is true [-Wsometimes-uninitialized]\n if (!ASSERT_OK_PTR(nstoken, \"setns \" NS_FWD))\n ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n .../tc_redirect.c:743:6: note: uninitialized use occurs here\n if (target_fd \u003e\u003d 0)\n ^~~~~~~~~\n\nRemoving unused variables and initializing the previously-uninitialized variable\nto ensure these warnings are gone.\n\nSigned-off-by: Yonghong Song \u003cyhs@fb.com\u003e\nSigned-off-by: Alexei Starovoitov \u003cast@kernel.org\u003e\nLink: https://lore.kernel.org/bpf/20210917043343.3711917-1-yhs@fb.com\n" }, { "commit": "f706f6c66c4360e41dd650bcf24a45880e75bb39", "tree": "41c2d3963751330b621ff21f8e1014f3d058ba88", "parents": [ "336562752acc1a723f9a24b5b8129ae22e0478c6", "942025c9f37ee45e69eb5f39a2877afab66d9555" ], "author": { "name": "Alexei Starovoitov", "email": "ast@kernel.org", "time": "Fri Sep 17 09:04:13 2021 -0700" }, "committer": { "name": "Alexei Starovoitov", "email": "ast@kernel.org", "time": "Fri Sep 17 09:05:41 2021 -0700" }, "message": "Merge branch \u0027Improve set_attach_target() and deprecate open_opts.attach_prog_fd\u0027\n\nAndrii Nakryiko says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nThis patch set deprecates bpf_object_open_opts.attach_prog_fd (in libbpf 0.7+)\nby extending bpf_program__set_attach_target() to support some more flexible\nscenarios. Existing fexit_bpf2bpf selftest is updated accordingly to not use\ndeprecated APIs.\n\nWhile at it, also deprecate no-op relaxed_core_relocs option (they are always\n\"relaxed\").\n\nLast patch also const-ifies all high-level libbpf attach APIs, as there is no\nreason for them to assume bpf_program/bpf_map modifications.\n\nPatch #1 also removes one more unneeded use of find_sec_def(), relying on\nprog-\u003esec_def that\u0027s set during bpf_object__open() operation, simplifying\nupcoming refactoring a little bit more.\n\nAll these changes are preparatory patches before SEC() handling refactoring\nthat will come next.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: Alexei Starovoitov \u003cast@kernel.org\u003e\n" }, { "commit": "942025c9f37ee45e69eb5f39a2877afab66d9555", "tree": "41c2d3963751330b621ff21f8e1014f3d058ba88", "parents": [ "91b555d73e53879fc6d4cf82c8c0e14c00ce212d" ], "author": { "name": "Andrii Nakryiko", "email": "andrii@kernel.org", "time": "Wed Sep 15 18:58:36 2021 -0700" }, "committer": { "name": "Alexei Starovoitov", "email": "ast@kernel.org", "time": "Fri Sep 17 09:05:41 2021 -0700" }, "message": "libbpf: Constify all high-level program attach APIs\n\nAttach APIs shouldn\u0027t need to modify bpf_program/bpf_map structs, so\nchange all struct bpf_program and struct bpf_map pointers to const\npointers. This is completely backwards compatible with no functional\nchange.\n\nSigned-off-by: Andrii Nakryiko \u003candrii@kernel.org\u003e\nSigned-off-by: Alexei Starovoitov \u003cast@kernel.org\u003e\nAcked-by: Yonghong Song \u003cyhs@fb.com\u003e\nLink: https://lore.kernel.org/bpf/20210916015836.1248906-8-andrii@kernel.org\n" }, { "commit": "91b555d73e53879fc6d4cf82c8c0e14c00ce212d", "tree": "4af595a6acc6fc236991b4f3ec25949d82df53e0", "parents": [ "60aed22076b0d0ec2b7c7f9dba3ccd642520e1f3" ], "author": { "name": "Andrii Nakryiko", "email": "andrii@kernel.org", "time": "Wed Sep 15 18:58:35 2021 -0700" }, "committer": { "name": "Alexei Starovoitov", "email": "ast@kernel.org", "time": "Fri Sep 17 09:05:41 2021 -0700" }, "message": "libbpf: Schedule open_opts.attach_prog_fd deprecation since v0.7\n\nbpf_object_open_opts.attach_prog_fd makes a pretty strong assumption\nthat bpf_object contains either only single freplace BPF program or all\nof BPF programs in BPF object are freplaces intended to replace\ndifferent subprograms of the same target BPF program. This seems both\na bit confusing, too assuming, and limiting.\n\nWe\u0027ve had bpf_program__set_attach_target() API which allows more\nfine-grained control over this, on a per-program level. As such, mark\nopen_opts.attach_prog_fd as deprecated starting from v0.7, so that we\nhave one more universal way of setting freplace targets. With previous\nchange to allow NULL attach_func_name argument, and especially combined\nwith BPF skeleton, arguable bpf_program__set_attach_target() is a more\nconvenient and explicit API as well.\n\nSigned-off-by: Andrii Nakryiko \u003candrii@kernel.org\u003e\nSigned-off-by: Alexei Starovoitov \u003cast@kernel.org\u003e\nAcked-by: Yonghong Song \u003cyhs@fb.com\u003e\nLink: https://lore.kernel.org/bpf/20210916015836.1248906-7-andrii@kernel.org\n" }, { "commit": "60aed22076b0d0ec2b7c7f9dba3ccd642520e1f3", "tree": "6920a9f255245e636e0909a518e08c7df0d1ac19", "parents": [ "2d5ec1c66e25f0b4dd895a211e651a12dec2ef4f" ], "author": { "name": "Andrii Nakryiko", "email": "andrii@kernel.org", "time": "Wed Sep 15 18:58:34 2021 -0700" }, "committer": { "name": "Alexei Starovoitov", "email": "ast@kernel.org", "time": "Fri Sep 17 09:05:41 2021 -0700" }, "message": "selftests/bpf: Switch fexit_bpf2bpf selftest to set_attach_target() API\n\nSwitch fexit_bpf2bpf selftest to bpf_program__set_attach_target()\ninstead of using bpf_object_open_opts.attach_prog_fd, which is going to\nbe deprecated. These changes also demonstrate the new mode of\nset_attach_target() in which it allows NULL when the target is BPF\nprogram (attach_prog_fd !\u003d 0).\n\nSigned-off-by: Andrii Nakryiko \u003candrii@kernel.org\u003e\nSigned-off-by: Alexei Starovoitov \u003cast@kernel.org\u003e\nAcked-by: Yonghong Song \u003cyhs@fb.com\u003e\nLink: https://lore.kernel.org/bpf/20210916015836.1248906-6-andrii@kernel.org\n" }, { "commit": "2d5ec1c66e25f0b4dd895a211e651a12dec2ef4f", "tree": "2e0b4d06450323d5c9cd64e2a9afb6e189009747", "parents": [ "277641859e833549722eb82c97cbc2d55421df7c" ], "author": { "name": "Andrii Nakryiko", "email": "andrii@kernel.org", "time": "Wed Sep 15 18:58:33 2021 -0700" }, "committer": { "name": "Alexei Starovoitov", "email": "ast@kernel.org", "time": "Fri Sep 17 09:05:00 2021 -0700" }, "message": "libbpf: Allow skipping attach_func_name in bpf_program__set_attach_target()\n\nAllow to use bpf_program__set_attach_target to only set target attach\nprogram FD, while letting libbpf to use target attach function name from\nSEC() definition. This might be useful for some scenarios where\nbpf_object contains multiple related freplace BPF programs intended to\nreplace different sub-programs in target BPF program. In such case all\nprograms will have the same attach_prog_fd, but different\nattach_func_name. It\u0027s convenient to specify such target function names\ndeclaratively in SEC() definitions, but attach_prog_fd is a dynamic\nruntime setting.\n\nTo simplify such scenario, allow bpf_program__set_attach_target() to\ndelay BTF ID resolution till the BPF program load time by providing NULL\nattach_func_name. In that case the behavior will be similar to using\nbpf_object_open_opts.attach_prog_fd (which is marked deprecated since\nv0.7), but has the benefit of allowing more control by user in what is\nattached to what. Such setup allows having BPF programs attached to\ndifferent target attach_prog_fd with target functions still declaratively\nrecorded in BPF source code in SEC() definitions.\n\nSelftests changes in the next patch should make this more obvious.\n\nSigned-off-by: Andrii Nakryiko \u003candrii@kernel.org\u003e\nSigned-off-by: Alexei Starovoitov \u003cast@kernel.org\u003e\nAcked-by: Yonghong Song \u003cyhs@fb.com\u003e\nLink: https://lore.kernel.org/bpf/20210916015836.1248906-5-andrii@kernel.org\n" }, { "commit": "277641859e833549722eb82c97cbc2d55421df7c", "tree": "a2239b4f700ca4316878b00ccfb8e497e107c444", "parents": [ "23a7baaa93882c241ad3464cdeeb8ef0d1d40a12" ], "author": { "name": "Andrii Nakryiko", "email": "andrii@kernel.org", "time": "Wed Sep 15 18:58:32 2021 -0700" }, "committer": { "name": "Alexei Starovoitov", "email": "ast@kernel.org", "time": "Fri Sep 17 09:04:12 2021 -0700" }, "message": "libbpf: Deprecated bpf_object_open_opts.relaxed_core_relocs\n\nIt\u0027s relevant and hasn\u0027t been doing anything for a long while now.\nDeprecated it.\n\nSigned-off-by: Andrii Nakryiko \u003candrii@kernel.org\u003e\nSigned-off-by: Alexei Starovoitov \u003cast@kernel.org\u003e\nAcked-by: Yonghong Song \u003cyhs@fb.com\u003e\nLink: https://lore.kernel.org/bpf/20210916015836.1248906-4-andrii@kernel.org\n" }, { "commit": "23a7baaa93882c241ad3464cdeeb8ef0d1d40a12", "tree": "a7df8559269f5632795d6b073207ca058d4da871", "parents": [ "f11f86a3931b5d533aed1be1720fbd55bd63174d" ], "author": { "name": "Andrii Nakryiko", "email": "andrii@kernel.org", "time": "Wed Sep 15 18:58:31 2021 -0700" }, "committer": { "name": "Alexei Starovoitov", "email": "ast@kernel.org", "time": "Fri Sep 17 09:04:12 2021 -0700" }, "message": "selftests/bpf: Stop using relaxed_core_relocs which has no effect\n\nrelaxed_core_relocs option hasn\u0027t had any effect for a while now, stop\nspecifying it. Next patch marks it as deprecated.\n\nSigned-off-by: Andrii Nakryiko \u003candrii@kernel.org\u003e\nSigned-off-by: Alexei Starovoitov \u003cast@kernel.org\u003e\nAcked-by: Yonghong Song \u003cyhs@fb.com\u003e\nLink: https://lore.kernel.org/bpf/20210916015836.1248906-3-andrii@kernel.org\n" }, { "commit": "f11f86a3931b5d533aed1be1720fbd55bd63174d", "tree": "499439e013261aec23ca7690d766b29c3bd7e25a", "parents": [ "336562752acc1a723f9a24b5b8129ae22e0478c6" ], "author": { "name": "Andrii Nakryiko", "email": "andrii@kernel.org", "time": "Wed Sep 15 18:58:30 2021 -0700" }, "committer": { "name": "Alexei Starovoitov", "email": "ast@kernel.org", "time": "Fri Sep 17 09:04:12 2021 -0700" }, "message": "libbpf: Use pre-setup sec_def in libbpf_find_attach_btf_id()\n\nDon\u0027t perform another search for sec_def inside\nlibbpf_find_attach_btf_id(), as each recognized bpf_program already has\nprog-\u003esec_def set.\n\nAlso remove unnecessary NULL check for prog-\u003esec_name, as it can never\nbe NULL.\n\nSigned-off-by: Andrii Nakryiko \u003candrii@kernel.org\u003e\nSigned-off-by: Alexei Starovoitov \u003cast@kernel.org\u003e\nAcked-by: Yonghong Song \u003cyhs@fb.com\u003e\nLink: https://lore.kernel.org/bpf/20210916015836.1248906-2-andrii@kernel.org\n" }, { "commit": "5ef8a0291513c624b568acf15eee87e905736f2c", "tree": "88b068ad0271031c88b5ea6687521a5489838e8b", "parents": [ "6db9350a9db370bb2657a37f28cabe5ddf4939a1" ], "author": { "name": "Krzysztof Kozlowski", "email": "krzysztof.kozlowski@canonical.com", "time": "Thu Sep 16 19:05:08 2021 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 17 14:21:00 2021 +0100" }, "message": "net: microchip: encx24j600: drop unneeded MODULE_ALIAS\n\nThe MODULE_DEVICE_TABLE already creates proper alias for spi driver.\nHaving another MODULE_ALIAS causes the alias to be duplicated.\n\nSigned-off-by: Krzysztof Kozlowski \u003ckrzysztof.kozlowski@canonical.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "6db9350a9db370bb2657a37f28cabe5ddf4939a1", "tree": "a309c10c55a95e2644b572bbe92016ca85c49ca7", "parents": [ "b20b54fb00a8c7cfc91ead288a8048cf343d9e7c" ], "author": { "name": "Leon Romanovsky", "email": "leonro@nvidia.com", "time": "Thu Sep 16 13:38:33 2021 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 17 14:19:39 2021 +0100" }, "message": "devlink: Delete not-used devlink APIs\n\nDevlink core exported generously the functions calls that were used\nby netdevsim tests or not used at all.\n\nDelete such APIs with one exception - devlink_alloc_ns(). That function\nshould be spared from deleting because it is a special form of devlink_alloc()\nneeded for the netdevsim.\n\nSigned-off-by: Leon Romanovsky \u003cleonro@nvidia.com\u003e\nAcked-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\nReviewed-by: Jiri Pirko \u003cjiri@nvidia.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "b20b54fb00a8c7cfc91ead288a8048cf343d9e7c", "tree": "acd393af23c64b01df11e744ff6fdbdbe3f2189f", "parents": [ "3503e673db234be0af57b80a03d7f5ac851e471d" ], "author": { "name": "Cai Huoqing", "email": "caihuoqing@baidu.com", "time": "Thu Sep 16 15:37:36 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 17 14:02:40 2021 +0100" }, "message": "net: stmmac: dwmac-visconti: Make use of the helper function dev_err_probe()\n\nWhen possible use dev_err_probe help to properly deal with the\nPROBE_DEFER error, the benefit is that DEFER issue will be logged\nin the devices_deferred debugfs file.\nAnd using dev_err_probe() can reduce code size, and the error value\ngets printed.\n\nSigned-off-by: Cai Huoqing \u003ccaihuoqing@baidu.com\u003e\nAcked-by: Nobuhiro Iwamatsu \u003cnobuhiro1.iwamatsu@toshiba.co.jp\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "3503e673db234be0af57b80a03d7f5ac851e471d", "tree": "92eb809ecda7785dc9198930ddab277d77e53c67", "parents": [ "d853f1d3c900cc16758b12853375c1ab888fae65" ], "author": { "name": "Colin Ian King", "email": "colin.king@canonical.com", "time": "Fri Sep 17 13:03:33 2021 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 17 14:00:51 2021 +0100" }, "message": "octeontx2-af: Remove redundant initialization of variable blkaddr\n\nThe variable blkaddr is being initialized with a value that is never\nread, it is being updated later on in a for-loop. The assignment is\nredundant and can be removed.\n\nAddresses-Coverity: (\"Unused value\")\nSigned-off-by: Colin Ian King \u003ccolin.king@canonical.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "d853f1d3c900cc16758b12853375c1ab888fae65", "tree": "367a296728277f89e4f0d8c3040f2717ed82e1c3", "parents": [ "6042d4348a341957d5eddadda2137c493a97131e" ], "author": { "name": "Colin Ian King", "email": "colin.king@canonical.com", "time": "Fri Sep 17 12:57:47 2021 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 17 14:00:51 2021 +0100" }, "message": "octeontx2-af: Fix uninitialized variable val\n\nIn the case where the condition !is_rvu_otx2(rvu) is false variable\nval is not initialized and can contain a garbage value. Fix this by\ninitializing val to zero and bit-wise or\u0027ing in BIT_ULL(51) to val\nfor the true condition case of !is_rvu_otx2(rvu).\n\nAddresses-Coverity: (\"Uninitialized scalar variable\")\nFixes: 4b5a3ab17c6c (\"octeontx2-af: Hardware configuration for inline IPsec\")\nSigned-off-by: Colin Ian King \u003ccolin.king@canonical.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "6042d4348a341957d5eddadda2137c493a97131e", "tree": "30671932df4810d866990db05b25399b4f5da7aa", "parents": [ "61524e43abadffbe450f9429f05c19d54cc95496" ], "author": { "name": "Hao Chen", "email": "chenhaoa@uniontech.com", "time": "Fri Sep 17 14:16:54 2021 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 17 11:11:25 2021 +0100" }, "message": "net: e1000e: solve insmod \u0027Unknown symbol mutex_lock\u0027 error\n\nAfter I turn on the CONFIG_LOCK_STAT\u003dy, insmod e1000e.ko will report:\n[ 5.641579] e1000e: Unknown symbol mutex_lock (err -2)\n[ 90.775705] e1000e: Unknown symbol mutex_lock (err -2)\n[ 132.252339] e1000e: Unknown symbol mutex_lock (err -2)\n\nThis problem fixed after include \u003clinux/mutex.h\u003e.\n\nSigned-off-by: Hao Chen \u003cchenhaoa@uniontech.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "61524e43abadffbe450f9429f05c19d54cc95496", "tree": "2f1f5498715a43da017ab7ab5ec6001a9e2d8859", "parents": [ "8dc84dcd7f74b50f81de3dbf6f6b5b146e3a8eea" ], "author": { "name": "Cai Huoqing", "email": "caihuoqing@baidu.com", "time": "Thu Sep 16 15:37:29 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 17 09:42:29 2021 +0100" }, "message": "net: netsec: Make use of the helper function dev_err_probe()\n\nWhen possible use dev_err_probe help to properly deal with the\nPROBE_DEFER error, the benefit is that DEFER issue will be logged\nin the devices_deferred debugfs file.\nAnd using dev_err_probe() can reduce code size, and the error value\ngets printed.\n\nSigned-off-by: Cai Huoqing \u003ccaihuoqing@baidu.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "8dc84dcd7f74b50f81de3dbf6f6b5b146e3a8eea", "tree": "26b550162b1d4a1ac01f2614a1025c03fd13def4", "parents": [ "44ded7ca63f1957c55bd1cb280a717aa69c51ec6" ], "author": { "name": "Florian Fainelli", "email": "f.fainelli@gmail.com", "time": "Thu Sep 16 14:27:41 2021 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Sep 16 19:11:17 2021 -0700" }, "message": "net: phy: broadcom: Enable 10BaseT DAC early wake\n\nEnable the DAC early wake when then link operates at 10BaseT allows\npower savings in the hundreds of milli Watts by shutting down the\ntransmitter. A number of errata have been issued for various Gigabit\nPHYs and the recommendation is to enable both the early and forced DAC\nwake to be on the safe side. This needs to be done dynamically based\nupon the link state, which is why a link_change_notify callback is\nutilized.\n\nSigned-off-by: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nLink: https://lore.kernel.org/r/20210916212742.1653088-1-f.fainelli@gmail.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "44ded7ca63f1957c55bd1cb280a717aa69c51ec6", "tree": "c4cc2cad041a6e5767124330276d05406478c35d", "parents": [ "561bed688bffedd6bbdfa70af7a4f64f0a3e2140", "7d5af56418d7d01e43247a33b6fe6492ea871923" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Sep 16 19:09:54 2021 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Sep 16 19:09:54 2021 -0700" }, "message": "Merge branch \u0027net-dsa-b53-clean-up-cpu-imp-ports\u0027\n\nRafał Miłecki says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nnet: dsa: b53: Clean up CPU/IMP ports\n\nThis has been tested on:\n\n1. Luxul XBR-4500 with used CPU port 5\n[ 8.361438] b53-srab-switch 18007000.ethernet-switch: found switch: BCM53012, rev 0\n\n2. Netgear R8000 with used CPU port 8\n[ 4.453858] b53-srab-switch 18007000.ethernet-switch: found switch: BCM53012, rev 5\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nLink: https://lore.kernel.org/r/20210916120354.20338-1-zajec5@gmail.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "7d5af56418d7d01e43247a33b6fe6492ea871923", "tree": "c4cc2cad041a6e5767124330276d05406478c35d", "parents": [ "3ff26b29230c54fea2353b63124c589b61953e14" ], "author": { "name": "Rafał Miłecki", "email": "rafal@milecki.pl", "time": "Thu Sep 16 14:03:54 2021 +0200" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Sep 16 19:06:36 2021 -0700" }, "message": "net: dsa: b53: Drop unused \"cpu_port\" field\n\nIt\u0027s set but never used anymore.\n\nSigned-off-by: Rafał Miłecki \u003crafal@milecki.pl\u003e\nReviewed-by: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nTested-by: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "3ff26b29230c54fea2353b63124c589b61953e14", "tree": "427a43948573e68b38d85bedf56997b7774811c7", "parents": [ "b290c6384afabbca5ae6e2af72fb1b2bc37922be" ], "author": { "name": "Rafał Miłecki", "email": "rafal@milecki.pl", "time": "Thu Sep 16 14:03:53 2021 +0200" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Sep 16 19:06:36 2021 -0700" }, "message": "net: dsa: b53: Improve flow control setup on BCM5301x\n\nAccording to the Broadcom\u0027s reference driver flow control needs to be\nenabled for any CPU switch port (5, 7 or 8 - depending on which one is\nused). Current code makes it work only for the port 5. Use\ndsa_is_cpu_port() which solved that problem.\n\nSigned-off-by: Rafał Miłecki \u003crafal@milecki.pl\u003e\nReviewed-by: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nTested-by: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "b290c6384afabbca5ae6e2af72fb1b2bc37922be", "tree": "124c295fe76786c93363d49c56f2a965495f1d5e", "parents": [ "983d96a9116a328668601555d96736261d33170c" ], "author": { "name": "Rafał Miłecki", "email": "rafal@milecki.pl", "time": "Thu Sep 16 14:03:52 2021 +0200" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Sep 16 19:06:35 2021 -0700" }, "message": "net: dsa: b53: Drop BCM5301x workaround for a wrong CPU/IMP port\n\nOn BCM5301x port 8 requires a fixed link when used.\n\nYears ago when b53 was an OpenWrt downstream driver (with configuration\nbased on sometimes bugged NVRAM) there was a need for a fixup. In case\nof forcing fixed link for (incorrectly specified) port 5 the code had to\nactually setup port 8 link.\n\nFor upstream b53 driver with setup based on DT there is no need for that\nworkaround. In DT we have and require correct ports setup.\n\nSigned-off-by: Rafał Miłecki \u003crafal@milecki.pl\u003e\nReviewed-by: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nTested-by: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "983d96a9116a328668601555d96736261d33170c", "tree": "df543a977c3f25337b5e465134c40a2a8d9b1b85", "parents": [ "561bed688bffedd6bbdfa70af7a4f64f0a3e2140" ], "author": { "name": "Rafał Miłecki", "email": "rafal@milecki.pl", "time": "Thu Sep 16 14:03:51 2021 +0200" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Sep 16 19:06:35 2021 -0700" }, "message": "net: dsa: b53: Include all ports in \"enabled_ports\"\n\nMake \"enabled_ports\" bitfield contain all available switch ports\nincluding a CPU port. This way there is no need for fixup during\ninitialization.\n\nFor BCM53010, BCM53018 and BCM53019 include also other available ports.\n\nSigned-off-by: Rafał Miłecki \u003crafal@milecki.pl\u003e\nReviewed-by: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nTested-by: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "561bed688bffedd6bbdfa70af7a4f64f0a3e2140", "tree": "19256b7a70dcdb9a95c4a7494a0b8bb9db10ef30", "parents": [ "4b5a3ab17c6c942bd428984b6b37fe3c07f18ab3", "fc0c0548c1a2e676d3a928aaed70f2d4d254e395" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Sep 16 13:58:38 2021 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Sep 16 13:58:38 2021 -0700" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net\n\nNo conflicts!\n\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "fc0c0548c1a2e676d3a928aaed70f2d4d254e395", "tree": "6875da13286e88f0eee2b045ecde268c1dc10270", "parents": [ "3c0d2a46c0141913dc6fd126c57d0615677d946e", "ee8a9600b5391f434905c46bec7f77d34505083e" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 16 13:05:42 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 16 13:05:42 2021 -0700" }, "message": "Merge tag \u0027net-5.15-rc2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net\n\nPull networking fixes from Jakub Kicinski:\n \"Including fixes from bpf.\n\n Current release - regressions:\n\n - vhost_net: fix OoB on sendmsg() failure\n\n - mlx5: bridge, fix uninitialized variable usage\n\n - bnxt_en: fix error recovery regression\n\n Current release - new code bugs:\n\n - bpf, mm: fix lockdep warning triggered by stack_map_get_build_id_offset()\n\n Previous releases - regressions:\n\n - r6040: restore MDIO clock frequency after MAC reset\n\n - tcp: fix tp-\u003eundo_retrans accounting in tcp_sacktag_one()\n\n - dsa: flush switchdev workqueue before tearing down CPU/DSA ports\n\n Previous releases - always broken:\n\n - ptp: dp83640: don\u0027t define PAGE0, avoid compiler warning\n\n - igc: fix tunnel segmentation offloads\n\n - phylink: update SFP selected interface on advertising changes\n\n - stmmac: fix system hang caused by eee_ctrl_timer during suspend/resume\n\n - mlx5e: fix mutual exclusion between CQE compression and HW TS\n\n Misc:\n\n - bpf, cgroups: fix cgroup v2 fallback on v1/v2 mixed mode\n\n - sfc: fallback for lack of xdp tx queues\n\n - hns3: add option to turn off page pool feature\"\n\n* tag \u0027net-5.15-rc2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (67 commits)\n mlxbf_gige: clear valid_polarity upon open\n igc: fix tunnel offloading\n net/{mlx5|nfp|bnxt}: Remove unnecessary RTNL lock assert\n net: wan: wanxl: define CROSS_COMPILE_M68K\n selftests: nci: replace unsigned int with int\n net: dsa: flush switchdev workqueue before tearing down CPU/DSA ports\n Revert \"net: phy: Uniform PHY driver access\"\n net: dsa: destroy the phylink instance on any error in dsa_slave_phy_setup\n ptp: dp83640: don\u0027t define PAGE0\n bnx2x: Fix enabling network interfaces without VFs\n Revert \"Revert \"ipv4: fix memory leaks in ip_cmsg_send() callers\"\"\n tcp: fix tp-\u003eundo_retrans accounting in tcp_sacktag_one()\n net-caif: avoid user-triggerable WARN_ON(1)\n bpf, selftests: Add test case for mixed cgroup v1/v2\n bpf, selftests: Add cgroup v1 net_cls classid helpers\n bpf, cgroups: Fix cgroup v2 fallback on v1/v2 mixed mode\n bpf: Add oversize check before call kvcalloc()\n net: hns3: fix the timing issue of VF clearing interrupt sources\n net: hns3: fix the exception when query imp info\n net: hns3: disable mac in flr process\n ...\n" }, { "commit": "3c0d2a46c0141913dc6fd126c57d0615677d946e", "tree": "7176048edd93285918f8de3617c4afe9ee6a7778", "parents": [ "040b8907ccf1c78d020aca29800036565d761d73" ], "author": { "name": "Guenter Roeck", "email": "linux@roeck-us.net", "time": "Wed Sep 08 20:57:43 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 16 12:52:06 2021 -0700" }, "message": "net: 6pack: Fix tx timeout and slot time\n\ntx timeout and slot time are currently specified in units of HZ. On\nAlpha, HZ is defined as 1024. When building alpha:allmodconfig, this\nresults in the following error message.\n\n drivers/net/hamradio/6pack.c: In function \u0027sixpack_open\u0027:\n drivers/net/hamradio/6pack.c:71:41: error:\n \tunsigned conversion from \u0027int\u0027 to \u0027unsigned char\u0027\n \tchanges value from \u0027256\u0027 to \u00270\u0027\n\nIn the 6PACK protocol, tx timeout is specified in units of 10 ms and\ntransmitted over the wire:\n\n https://www.linux-ax25.org/wiki/6PACK\n\nDefining a value dependent on HZ doesn\u0027t really make sense, and\npresumably comes from the (very historical) situation where HZ was\noriginally 100.\n\nNote that the SIXP_SLOTTIME use explicitly is about 10ms granularity:\n\n mod_timer(\u0026sp-\u003etx_t, jiffies + ((when + 1) * HZ) / 100);\n\nand the SIXP_TXDELAY walue is sent as a byte over the wire.\n\nSigned-off-by: Guenter Roeck \u003clinux@roeck-us.net\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "040b8907ccf1c78d020aca29800036565d761d73", "tree": "9cb48e4e955054c6ed2b910fa04cd37ab53be5f8", "parents": [ "b60cee5bae733f49ba33840804c159a8e474cfda" ], "author": { "name": "Arnd Bergmann", "email": "arnd@arndb.de", "time": "Tue Apr 28 23:31:24 2020 +0200" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 16 12:02:33 2021 -0700" }, "message": "drm/rockchip: cdn-dp-core: Make cdn_dp_core_resume __maybe_unused\n\nWith the new static annotation, the compiler warns when the functions\nare actually unused:\n\n drivers/gpu/drm/rockchip/cdn-dp-core.c:1123:12: error: \u0027cdn_dp_resume\u0027 defined but not used [-Werror\u003dunused-function]\n 1123 | static int cdn_dp_resume(struct device *dev)\n | ^~~~~~~~~~~~~\n\nMark them __maybe_unused to suppress that warning as well.\n\n[ Not so \u0027new\u0027 static annotations any more, and I removed the part of\n the patch that added __maybe_unused to cdn_dp_suspend(), because it\u0027s\n used by the shutdown/remove code.\n\n So only the resume function ends up possibly unused if CONFIG_PM isn\u0027t\n set - Linus ]\n\nFixes: 7c49abb4c2f8 (\"drm/rockchip: cdn-dp-core: Make cdn_dp_core_suspend/resume static\")\nSigned-off-by: Arnd Bergmann \u003carnd@arndb.de\u003e\nReviewed-by: Enric Balletbo i Serra \u003cenric.balletbo@collabora.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "b60cee5bae733f49ba33840804c159a8e474cfda", "tree": "9fe3527223f575ef622b760aff8a08737fae8d00", "parents": [ "35a3f4ef0ab543daa1725b0c963eb8c05e3376f8" ], "author": { "name": "Guenter Roeck", "email": "linux@roeck-us.net", "time": "Thu Sep 09 11:47:14 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 16 11:29:27 2021 -0700" }, "message": "cpufreq: vexpress: Drop unused variable\n\narm:allmodconfig fails to build with the following error.\n\n drivers/cpufreq/vexpress-spc-cpufreq.c:454:13: error:\n\t\t\t\t\tunused variable \u0027cur_cluster\u0027\n\nRemove the unused variable.\n\nFixes: bb8c26d9387f (\"cpufreq: vexpress: Set CPUFREQ_IS_COOLING_DEV flag\")\nCc: Viresh Kumar \u003cviresh.kumar@linaro.org\u003e\nSigned-off-by: Guenter Roeck \u003clinux@roeck-us.net\u003e\nReviewed-by: Kees Cook \u003ckeescook@chromium.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "35a3f4ef0ab543daa1725b0c963eb8c05e3376f8", "tree": "b2b1e36021e5c03fe782134e2a81397799731cd1", "parents": [ "db71f8fb44956714249a526647c143bac5bb96a1" ], "author": { "name": "Guenter Roeck", "email": "linux@roeck-us.net", "time": "Wed Sep 08 22:00:33 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 16 11:27:32 2021 -0700" }, "message": "alpha: Declare virt_to_phys and virt_to_bus parameter as pointer to volatile\n\nSome drivers pass a pointer to volatile data to virt_to_bus() and\nvirt_to_phys(), and that works fine. One exception is alpha. This\nresults in a number of compile errors such as\n\n drivers/net/wan/lmc/lmc_main.c: In function \u0027lmc_softreset\u0027:\n drivers/net/wan/lmc/lmc_main.c:1782:50: error:\n\tpassing argument 1 of \u0027virt_to_bus\u0027 discards \u0027volatile\u0027\n\tqualifier from pointer target type\n\n drivers/atm/ambassador.c: In function \u0027do_loader_command\u0027:\n drivers/atm/ambassador.c:1747:58: error:\n\tpassing argument 1 of \u0027virt_to_bus\u0027 discards \u0027volatile\u0027\n\tqualifier from pointer target type\n\nDeclare the parameter of virt_to_phys and virt_to_bus as pointer to\nvolatile to fix the problem.\n\nSigned-off-by: Guenter Roeck \u003clinux@roeck-us.net\u003e\nAcked-by: Arnd Bergmann \u003carnd@arndb.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "db71f8fb44956714249a526647c143bac5bb96a1", "tree": "2e05c0d5d765025889862d1f29814d138e4ed898", "parents": [ "5fe983d3f1a5b103fc00a24f9e0408302e60c39c" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 16 11:14:47 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 16 11:14:47 2021 -0700" }, "message": "3com 3c515: make it compile on 64-bit architectures\n\nThis driver isn\u0027t enabled most places because of the ISA config\ndependency, but alpha still has it. And I think the \u0027Jensen\u0027 actually\ndid have an ISA slot.\n\nHowever, it doesn\u0027t build cleanly, because the \"Vortex bus master\" code\njust casts the skb-\u003edata pointer to \u0027int\u0027:\n\n outl((int) (skb-\u003edata), ioaddr + Wn7_MasterAddr);\n\nwhich is all kinds of broken. Even on a good old traditional PC/AT it\nwould be broken because the high bits will be random kernel address\nbits, but presumably the hardware ignores those bits. I mean, it\u0027s ISA.\nWe\u0027re talking 16MB dma limits. The \"good old days\".\n\nMake the build happy with this kind of craziness by using the proper\nisa_virt_to_bus() handling that the full bus master code uses anyway\n(the Vortex bus mastering is a limited special case).\n\nWho knows, this might even work.\n\nReported-by: Guenter Roeck \u003clinux@roeck-us.net\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "5fe983d3f1a5b103fc00a24f9e0408302e60c39c", "tree": "71a1c74332accbb43f9f69f110a62fc1b8204d49", "parents": [ "077a6ccf2588c7c893b443fad62d0f8ed342cafc", "90cc7bed1ed19f869ae7221a6b41887fe762a6a3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 16 10:32:00 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 16 10:32:00 2021 -0700" }, "message": "Merge tag \u0027for-5.15/parisc-4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux\n\nPull parisc fix from Helge Deller:\n \"Fix a build warning when using the PAGE0 pointer\"\n\n* tag \u0027for-5.15/parisc-4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux:\n parisc: Use absolute_pointer() to define PAGE0\n" }, { "commit": "077a6ccf2588c7c893b443fad62d0f8ed342cafc", "tree": "fd95a326eebca544fffd429c135b6fa67d7115ae", "parents": [ "ff1ffd71d5f0612cf194f5705c671d6b64bf5f91", "a7b68ed15d1fd72c1e451d5eb6edebee2a624b90" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 16 10:26:48 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 16 10:26:48 2021 -0700" }, "message": "Merge tag \u0027m68k-for-v5.15-tag2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/geert/linux-m68k\n\nPull m68k fixes from Geert Uytterhoeven:\n\n - Warning fixes to mitigate CONFIG_WERROR\u003dy\n\n* tag \u0027m68k-for-v5.15-tag2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/geert/linux-m68k:\n m68k: mvme: Remove overdue #warnings in RTC handling\n m68k: Double cast io functions to unsigned long\n" }, { "commit": "4b5a3ab17c6c942bd428984b6b37fe3c07f18ab3", "tree": "e516bfceff3cb94322dc2dd25751ec236f8bbbba", "parents": [ "227b9644ab16d2ecd98d593edbe15c32c0c9620a" ], "author": { "name": "Srujana Challa", "email": "schalla@marvell.com", "time": "Thu Sep 16 15:11:14 2021 +0530" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 16 14:37:38 2021 +0100" }, "message": "octeontx2-af: Hardware configuration for inline IPsec\n\nOn OcteonTX2/CN10K SoC, the admin function (AF) is the only one\nwith all priviliges to configure HW and alloc resources, PFs and\nit\u0027s VFs have to request AF via mailbox for all their needs.\nThis patch adds new mailbox messages for CPT PFs and VFs to configure\nHW resources for inline-IPsec.\n\nSigned-off-by: Subbaraya Sundeep \u003csbhatta@marvell.com\u003e\nSigned-off-by: Srujana Challa \u003cschalla@marvell.com\u003e\nSigned-off-by: Vidya Sagar Velumuri \u003cvvelumuri@marvell.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "227b9644ab16d2ecd98d593edbe15c32c0c9620a", "tree": "ff65b8bc94c89fa11eb9425f252876a84937e825", "parents": [ "d1ab2647de3272e878604720ac0af66442e8d1d3" ], "author": { "name": "Tianjia Zhang", "email": "tianjia.zhang@linux.alibaba.com", "time": "Thu Sep 16 11:37:38 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 16 14:36:26 2021 +0100" }, "message": "net/tls: support SM4 GCM/CCM algorithm\n\nThe RFC8998 specification defines the use of the ShangMi algorithm\ncipher suites in TLS 1.3, and also supports the GCM/CCM mode using\nthe SM4 algorithm.\n\nSigned-off-by: Tianjia Zhang \u003ctianjia.zhang@linux.alibaba.com\u003e\nAcked-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "d1ab2647de3272e878604720ac0af66442e8d1d3", "tree": "93eb8b63d72be96c126faa5a1257a72c05e71851", "parents": [ "63f85c401ebaa62094c373c780d4051a84608ac3" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Sep 15 14:58:23 2021 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 16 14:32:53 2021 +0100" }, "message": "Revert \"net: wwan: iosm: firmware flashing and coredump collection\"\n\nThe devlink parameters are not the right mechanism to pass\nextra parameters to device flashing. The params added are\nalso undocumented.\n\nThis reverts commit 13bb8429ca98 (\"net: wwan: iosm: firmware\nflashing and coredump collection\").\n\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "ee8a9600b5391f434905c46bec7f77d34505083e", "tree": "790bd1b127b2221a338d811193d93ff591c3ff5f", "parents": [ "40ee363c844fcb6ae0f1f5cfea68aed7e268c2f4" ], "author": { "name": "David Thompson", "email": "davthompson@nvidia.com", "time": "Wed Sep 15 14:08:48 2021 -0400" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 16 14:31:58 2021 +0100" }, "message": "mlxbf_gige: clear valid_polarity upon open\n\nThe network interface managed by the mlxbf_gige driver can\nget into a problem state where traffic does not flow.\nIn this state, the interface will be up and enabled, but\nwill stop processing received packets. This problem state\nwill happen if three specific conditions occur:\n 1) driver has received more than (N * RxRingSize) packets but\n less than (N+1 * RxRingSize) packets, where N is an odd number\n Note: the command \"ethtool -g \u003cinterface\u003e\" will display the\n current receive ring size, which currently defaults to 128\n 2) the driver\u0027s interface was disabled via \"ifconfig oob_net0 down\"\n during the window described in #1.\n 3) the driver\u0027s interface is re-enabled via \"ifconfig oob_net0 up\"\n\nThis patch ensures that the driver\u0027s \"valid_polarity\" field is\ncleared during the open() method so that it always matches the\nreceive polarity used by hardware. Without this fix, the driver\nneeds to be unloaded and reloaded to correct this problem state.\n\nFixes: f92e1869d74e (\"Add Mellanox BlueField Gigabit Ethernet driver\")\nReviewed-by: Asmaa Mnebhi \u003casmaa@nvidia.com\u003e\nSigned-off-by: David Thompson \u003cdavthompson@nvidia.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "63f85c401ebaa62094c373c780d4051a84608ac3", "tree": "263dfafd8b662c79bc1e65fbd3fd53ce876c654f", "parents": [ "52583c8d8b12f232bd1128de9f5895bcdad7d7a3" ], "author": { "name": "Hariprasad Kelam", "email": "hkelam@marvell.com", "time": "Wed Sep 15 21:29:46 2021 +0530" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 16 14:31:05 2021 +0100" }, "message": "octeontx2-pf: CN10K: Hide RPM stats over ethtool\n\nCN10K MAC block (RPM) differs in number of stats compared to Octeontx2\nMAC block (CGX). RPM supports stats for each class of PFC and error\npackets etc. It would be difficult for user to read stats from ethtool\nand map to their definition.\n\nNew debugfs file is already added to read RPM stats along with their\ndefinition. This patch adds proper checks such that RPM stats will not\nbe part of ethtool.\n\nSigned-off-by: Hariprasad Kelam \u003chkelam@marvell.com\u003e\nSigned-off-by: Sunil Goutham \u003csgoutham@marvell.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "40ee363c844fcb6ae0f1f5cfea68aed7e268c2f4", "tree": "285ddadae367d3ab807a08605e895976f7dadfee", "parents": [ "7c3a0a018e672a9723a79b128227272562300055" ], "author": { "name": "Paolo Abeni", "email": "pabeni@redhat.com", "time": "Wed Sep 15 10:19:07 2021 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 16 14:29:58 2021 +0100" }, "message": "igc: fix tunnel offloading\n\nChecking tunnel offloading, it turns out that offloading doesn\u0027t work\nas expected. The following script allows to reproduce the issue.\nCall it as `testscript DEVICE LOCALIP REMOTEIP NETMASK\u0027\n\n\u003d\u003d\u003d SNIP \u003d\u003d\u003d\nif [ $# -ne 4 ]\nthen\n echo \"Usage $0 DEVICE LOCALIP REMOTEIP NETMASK\"\n exit 1\nfi\nDEVICE\u003d\"$1\"\nLOCAL_ADDRESS\u003d\"$2\"\nREMOTE_ADDRESS\u003d\"$3\"\nNWMASK\u003d\"$4\"\necho \"Driver: $(ethtool -i ${DEVICE} | awk \u0027/^driver:/{print $2}\u0027) \"\nethtool -k \"${DEVICE}\" | grep tx-udp\necho\necho \"Set up NIC and tunnel...\"\nip addr add \"${LOCAL_ADDRESS}/${NWMASK}\" dev \"${DEVICE}\"\nip link set \"${DEVICE}\" up\nsleep 2\nip link add vxlan1 type vxlan id 42 \\\n\t\t remote \"${REMOTE_ADDRESS}\" \\\n\t\t local \"${LOCAL_ADDRESS}\" \\\n\t\t dstport 0 \\\n\t\t dev \"${DEVICE}\"\nip addr add fc00::1/64 dev vxlan1\nip link set vxlan1 up\nsleep 2\nrm -f vxlan.pcap\necho \"Running tcpdump and iperf3...\"\n( nohup tcpdump -i any -w vxlan.pcap \u003e/dev/null 2\u003e\u00261 ) \u0026\nsleep 2\niperf3 -c fc00::2 \u003e/dev/null\npkill tcpdump\necho\necho -n \"Max. Paket Size: \"\ntcpdump -r vxlan.pcap -nnle 2\u003e/dev/null \\\n| grep \"${LOCAL_ADDRESS}.*\u003e ${REMOTE_ADDRESS}.*OTV\" \\\n| awk \u0027{print $8}\u0027 | awk -F \u0027:\u0027 \u0027{print $1}\u0027 \\\n| sort -n | tail -1\necho\nip link del vxlan1\nip addr del ${LOCAL_ADDRESS}/${NWMASK} dev \"${DEVICE}\"\n\u003d\u003d\u003d SNAP \u003d\u003d\u003d\n\nThe expected outcome is\n\n Max. Paket Size: 64904\n\nThis is what you see on igb, the code igc has been taken from.\nHowever, on igc the output is\n\n Max. Paket Size: 1516\n\nso the GSO aggregate packets are segmented by the kernel before calling\nigc_xmit_frame. Inside the subsequent call to igc_tso, the check for\nskb_is_gso(skb) fails and the function returns prematurely.\n\nIt turns out that this occurs because the feature flags aren\u0027t set\nentirely correctly in igc_probe. In contrast to the original code\nfrom igb_probe, igc_probe neglects to set the flags required to allow\ntunnel offloading.\n\nSetting the same flags as igb fixes the issue on igc.\n\nFixes: 34428dff3679 (\"igc: Add GSO partial support\")\nSigned-off-by: Paolo Abeni \u003cpabeni@redhat.com\u003e\nTested-by: Corinna Vinschen \u003cvinschen@redhat.com\u003e\nAcked-by: Sasha Neftin \u003csasha.neftin@intel.com\u003e\nTested-by: Nechama Kraus \u003cnechamax.kraus@linux.intel.com\u003e\nSigned-off-by: Tony Nguyen \u003canthony.l.nguyen@intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "7c3a0a018e672a9723a79b128227272562300055", "tree": "95b4fc0f899cc2e4c6e568241e5fa32f76fd2185", "parents": [ "84fb7dfc7463afcba61281f36535576a7f7b0626" ], "author": { "name": "Eli Cohen", "email": "elic@nvidia.com", "time": "Wed Sep 15 07:47:27 2021 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 16 14:09:30 2021 +0100" }, "message": "net/{mlx5|nfp|bnxt}: Remove unnecessary RTNL lock assert\n\nRemove the assert from the callback priv lookup function since it does\nnot require RTNL lock and is already protected by flow_indr_block_lock.\n\nThis will avoid warnings from being emitted to dmesg if the driver\nregisters its callback after an ingress qdisc was created for a\nnetdevice.\n\nThe warnings started after the following patch was merged:\ncommit 74fc4f828769 (\"net: Fix offloading indirect devices dependency on qdisc order creation\")\n\nSigned-off-by: Eli Cohen \u003celic@nvidia.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "84fb7dfc7463afcba61281f36535576a7f7b0626", "tree": "5d04ba98161fe5bcf7d1a177b56a1b19bb7e6165", "parents": [ "98dc68f8b0c2248bdc3688c90d2499247b9432e4" ], "author": { "name": "Adam Borowski", "email": "kilobyte@angband.pl", "time": "Sun Sep 12 23:23:21 2021 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 16 14:08:04 2021 +0100" }, "message": "net: wan: wanxl: define CROSS_COMPILE_M68K\n\nIt was used but never set. The hardcoded value from before the dawn of\ntime was non-standard; the usual name for cross-tools is $TRIPLET-$TOOL\n\nSigned-off-by: Adam Borowski \u003ckilobyte@angband.pl\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "98dc68f8b0c2248bdc3688c90d2499247b9432e4", "tree": "cd39925cc6a153df985d8f6caadd14c11f474021", "parents": [ "a57d8c217aadac75530b8e7ffb3a3e1b7bfd0330" ], "author": { "name": "Xiang wangx", "email": "wangxiang@cdjrlc.com", "time": "Thu Sep 16 20:24:42 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 16 13:55:51 2021 +0100" }, "message": "selftests: nci: replace unsigned int with int\n\nShould not use comparison of unsigned expressions \u003c 0.\n\nSigned-off-by: Xiang wangx \u003cwangxiang@cdjrlc.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "52583c8d8b12f232bd1128de9f5895bcdad7d7a3", "tree": "011c1350e58207a1bf5675e226ab3af89b36134d", "parents": [ "4fd3ff3b29aebd271db5b74275adfa5d1505883d" ], "author": { "name": "Cai Huoqing", "email": "caihuoqing@baidu.com", "time": "Wed Sep 15 22:58:42 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 16 13:35:35 2021 +0100" }, "message": "net: thunderx: Make use of the helper function dev_err_probe()\n\nWhen possible use dev_err_probe help to properly deal with the\nPROBE_DEFER error, the benefit is that DEFER issue will be logged\nin the devices_deferred debugfs file.\nAnd using dev_err_probe() can reduce code size, and simplify the code.\n\nSigned-off-by: Cai Huoqing \u003ccaihuoqing@baidu.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4fd3ff3b29aebd271db5b74275adfa5d1505883d", "tree": "df753029ffb86520dc972f03b052cc781768210d", "parents": [ "015a22f46b25d12f28efcce28039019ec504cb13" ], "author": { "name": "Cai Huoqing", "email": "caihuoqing@baidu.com", "time": "Wed Sep 15 22:58:34 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 16 13:35:35 2021 +0100" }, "message": "net: hinic: Make use of the helper function dev_err_probe()\n\nWhen possible use dev_err_probe help to properly deal with the\nPROBE_DEFER error, the benefit is that DEFER issue will be logged\nin the devices_deferred debugfs file.\nAnd using dev_err_probe() can reduce code size, and simplify the code.\n\nSigned-off-by: Cai Huoqing \u003ccaihuoqing@baidu.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "015a22f46b25d12f28efcce28039019ec504cb13", "tree": "540cb03dbeddbc914b0c426a6974528f8f5ea6b7", "parents": [ "a72691ee19ca9763fca4a841843e83e88690aa97" ], "author": { "name": "Cai Huoqing", "email": "caihuoqing@baidu.com", "time": "Wed Sep 15 22:58:27 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 16 13:35:35 2021 +0100" }, "message": "net: ethoc: Make use of the helper function dev_err_probe()\n\nWhen possible use dev_err_probe help to properly deal with the\nPROBE_DEFER error, the benefit is that DEFER issue will be logged\nin the devices_deferred debugfs file.\nAnd using dev_err_probe() can reduce code size, and simplify the code.\n\nSigned-off-by: Cai Huoqing \u003ccaihuoqing@baidu.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "a72691ee19ca9763fca4a841843e83e88690aa97", "tree": "17e6c7b4737e4664f3a2d82c016bd706ed76e458", "parents": [ "9eda994d4b57e177a1e05416458e95ae7cac6c40" ], "author": { "name": "Cai Huoqing", "email": "caihuoqing@baidu.com", "time": "Wed Sep 15 22:58:19 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 16 13:35:35 2021 +0100" }, "message": "net: enetc: Make use of the helper function dev_err_probe()\n\nWhen possible use dev_err_probe help to properly deal with the\nPROBE_DEFER error, the benefit is that DEFER issue will be logged\nin the devices_deferred debugfs file.\nAnd using dev_err_probe() can reduce code size, and simplify the code.\n\nSigned-off-by: Cai Huoqing \u003ccaihuoqing@baidu.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "9eda994d4b57e177a1e05416458e95ae7cac6c40", "tree": "fbe47c1239fcb7b6d84f4fdb8dda86356039bb1b", "parents": [ "b0ab7096dd9bf4d6d5c9d889ab216eb78f7e0b04" ], "author": { "name": "Cai Huoqing", "email": "caihuoqing@baidu.com", "time": "Wed Sep 15 22:58:11 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 16 13:35:35 2021 +0100" }, "message": "net: chelsio: cxgb4vf: Make use of the helper function dev_err_probe()\n\nWhen possible use dev_err_probe help to properly deal with the\nPROBE_DEFER error, the benefit is that DEFER issue will be logged\nin the devices_deferred debugfs file.\nAnd using dev_err_probe() can reduce code size, and simplify the code.\n\nSigned-off-by: Cai Huoqing \u003ccaihuoqing@baidu.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "b0ab7096dd9bf4d6d5c9d889ab216eb78f7e0b04", "tree": "2f5573bff7d63c09359a619ba18e9f43516e5d73", "parents": [ "d502933c30c6417edfc3ca2babd50343ad24b896" ], "author": { "name": "Cai Huoqing", "email": "caihuoqing@baidu.com", "time": "Wed Sep 15 22:57:56 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 16 13:35:35 2021 +0100" }, "message": "net: atl1e: Make use of the helper function dev_err_probe()\n\nWhen possible use dev_err_probe help to properly deal with the\nPROBE_DEFER error, the benefit is that DEFER issue will be logged\nin the devices_deferred debugfs file.\nAnd using dev_err_probe() can reduce code size, and simplify the code.\n\nSigned-off-by: Cai Huoqing \u003ccaihuoqing@baidu.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "d502933c30c6417edfc3ca2babd50343ad24b896", "tree": "5983ef37003fde24e3777bcedd3a868b00f09bef", "parents": [ "95b5fc03c189e4ea5c63785274cc0b77fcc3a818" ], "author": { "name": "Cai Huoqing", "email": "caihuoqing@baidu.com", "time": "Wed Sep 15 22:57:48 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 16 13:35:35 2021 +0100" }, "message": "net: atl1c: Make use of the helper function dev_err_probe()\n\nWhen possible use dev_err_probe help to properly deal with the\nPROBE_DEFER error, the benefit is that DEFER issue will be logged\nin the devices_deferred debugfs file.\nAnd using dev_err_probe() can reduce code size, and simplify the code.\n\nSigned-off-by: Cai Huoqing \u003ccaihuoqing@baidu.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "95b5fc03c189e4ea5c63785274cc0b77fcc3a818", "tree": "3c81a0128d17b5ccc5aaf6fb6b8c34f830cd3bdb", "parents": [ "5706383b30cfd511022b42d583d2196b5ec8f8a3" ], "author": { "name": "Cai Huoqing", "email": "caihuoqing@baidu.com", "time": "Wed Sep 15 22:57:41 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 16 13:35:35 2021 +0100" }, "message": "net: arc_emac: Make use of the helper function dev_err_probe()\n\nWhen possible use dev_err_probe help to properly deal with the\nPROBE_DEFER error, the benefit is that DEFER issue will be logged\nin the devices_deferred debugfs file.\nAnd using dev_err_probe() can reduce code size, and simplify the code.\n\nSigned-off-by: Cai Huoqing \u003ccaihuoqing@baidu.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "90cc7bed1ed19f869ae7221a6b41887fe762a6a3", "tree": "8d9cb1aaa00c3c34b6365b92c9fb655798fe1ba7", "parents": [ "ff1ffd71d5f0612cf194f5705c671d6b64bf5f91" ], "author": { "name": "Helge Deller", "email": "deller@gmx.de", "time": "Thu Sep 16 08:35:42 2021 +0200" }, "committer": { "name": "Helge Deller", "email": "deller@gmx.de", "time": "Thu Sep 16 08:35:42 2021 +0200" }, "message": "parisc: Use absolute_pointer() to define PAGE0\n\nUse absolute_pointer() wrapper for PAGE0 to avoid this compiler warning:\n\n arch/parisc/kernel/setup.c: In function \u0027start_parisc\u0027:\n error: \u0027__builtin_memcmp_eq\u0027 specified bound 8 exceeds source size 0\n\nSigned-off-by: Helge Deller \u003cdeller@gmx.de\u003e\nCo-Developed-by: Guenter Roeck \u003clinux@roeck-us.net\u003e\nSuggested-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "ff1ffd71d5f0612cf194f5705c671d6b64bf5f91", "tree": "f89226ca43252465d5908d5ceb57cb87c3327b5c", "parents": [ "453fa43cdb8e0f4231ab84755fd2fc562823541b", "dfb5c1e12c28e35e4d4e5bc8022b0e9d585b89a7" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 15 17:18:56 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 15 17:18:56 2021 -0700" }, "message": "Merge tag \u0027hyperv-fixes-signed-20210915\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/hyperv/linux\n\nPull hyperv fixes from Wei Liu:\n\n - Fix kernel crash caused by uio driver (Vitaly Kuznetsov)\n\n - Remove on-stack cpumask from HV APIC code (Wei Liu)\n\n* tag \u0027hyperv-fixes-signed-20210915\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/hyperv/linux:\n x86/hyperv: remove on-stack cpumask from hv_send_ipi_mask_allbutself\n asm-generic/hyperv: provide cpumask_to_vpset_noself\n Drivers: hv: vmbus: Fix kernel crash upon unbinding a device from uio_hv_generic driver\n" }, { "commit": "453fa43cdb8e0f4231ab84755fd2fc562823541b", "tree": "9ac900fb5461e1d931bdad99df13fa0df9cb34c5", "parents": [ "b7213ffa0e585feb1aee3e7173e965e66ee0abaa", "13be2efc390acd2a46a69a359f6efc00ca434599" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 15 17:06:01 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 15 17:06:01 2021 -0700" }, "message": "Merge tag \u0027rtc-5.15-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/abelloni/linux\n\nPull RTC fix from Alexandre Belloni:\n \"Fix a locking issue in the cmos rtc driver\"\n\n* tag \u0027rtc-5.15-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/abelloni/linux:\n rtc: cmos: Disable irq around direct invocation of cmos_interrupt()\n" }, { "commit": "a57d8c217aadac75530b8e7ffb3a3e1b7bfd0330", "tree": "939447e3f1734c7111123dfcb619ddb3fd389454", "parents": [ "301de697d869be6564aebeb5ab811c84c0a7abed" ], "author": { "name": "Vladimir Oltean", "email": "vladimir.oltean@nxp.com", "time": "Tue Sep 14 16:47:26 2021 +0300" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Sep 15 15:09:46 2021 -0700" }, "message": "net: dsa: flush switchdev workqueue before tearing down CPU/DSA ports\n\nSometimes when unbinding the mv88e6xxx driver on Turris MOX, these error\nmessages appear:\n\nmv88e6085 d0032004.mdio-mii:12: port 1 failed to delete be:79:b4:9e:9e:96 vid 1 from fdb: -2\nmv88e6085 d0032004.mdio-mii:12: port 1 failed to delete be:79:b4:9e:9e:96 vid 0 from fdb: -2\nmv88e6085 d0032004.mdio-mii:12: port 1 failed to delete d8:58:d7:00:ca:6d vid 100 from fdb: -2\nmv88e6085 d0032004.mdio-mii:12: port 1 failed to delete d8:58:d7:00:ca:6d vid 1 from fdb: -2\nmv88e6085 d0032004.mdio-mii:12: port 1 failed to delete d8:58:d7:00:ca:6d vid 0 from fdb: -2\n\n(and similarly for other ports)\n\nWhat happens is that DSA has a policy \"even if there are bugs, let\u0027s at\nleast not leak memory\" and dsa_port_teardown() clears the dp-\u003efdbs and\ndp-\u003emdbs lists, which are supposed to be empty.\n\nBut deleting that cleanup code, the warnings go away.\n\n\u003d\u003e the FDB and MDB lists (used for refcounting on shared ports, aka CPU\nand DSA ports) will eventually be empty, but are not empty by the time\nwe tear down those ports. Aka we are deleting them too soon.\n\nThe addresses that DSA complains about are host-trapped addresses: the\nlocal addresses of the ports, and the MAC address of the bridge device.\n\nThe problem is that offloading those entries happens from a deferred\nwork item scheduled by the SWITCHDEV_FDB_DEL_TO_DEVICE handler, and this\nraces with the teardown of the CPU and DSA ports where the refcounting\nis kept.\n\nIn fact, not only it races, but fundamentally speaking, if we iterate\nthrough the port list linearly, we might end up tearing down the shared\nports even before we delete a DSA user port which has a bridge upper.\n\nSo as it turns out, we need to first tear down the user ports (and the\nunused ones, for no better place of doing that), then the shared ports\n(the CPU and DSA ports). In between, we need to ensure that all work\nitems scheduled by our switchdev handlers (which only run for user\nports, hence the reason why we tear them down first) have finished.\n\nFixes: 161ca59d39e9 (\"net: dsa: reference count the MDB entries at the cross-chip notifier level\")\nSigned-off-by: Vladimir Oltean \u003cvladimir.oltean@nxp.com\u003e\nReviewed-by: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nLink: https://lore.kernel.org/r/20210914134726.2305133-1-vladimir.oltean@nxp.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "301de697d869be6564aebeb5ab811c84c0a7abed", "tree": "76f496de565d714f9fff0de43eecbe02a0d0afff", "parents": [ "6a52e73368038f47f6618623d75061dc263b26ae" ], "author": { "name": "Vladimir Oltean", "email": "vladimir.oltean@nxp.com", "time": "Tue Sep 14 17:05:15 2021 +0300" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Sep 15 15:06:46 2021 -0700" }, "message": "Revert \"net: phy: Uniform PHY driver access\"\n\nThis reverts commit 3ac8eed62596387214869319379c1fcba264d8c6, which did\nmore than it said on the box, and not only it replaced to_phy_driver\nwith phydev-\u003edrv, but it also removed the \"!drv\" check, without actually\nexplaining why that is fine.\n\nThat patch in fact breaks suspend/resume on any system which has PHY\ndevices with no drivers bound.\n\nThe stack trace is:\n\nUnable to handle kernel NULL pointer dereference at virtual address 00000000000000e8\npc : mdio_bus_phy_suspend+0xd8/0xec\nlr : dpm_run_callback+0x38/0x90\nCall trace:\n mdio_bus_phy_suspend+0xd8/0xec\n dpm_run_callback+0x38/0x90\n __device_suspend+0x108/0x3cc\n dpm_suspend+0x140/0x210\n dpm_suspend_start+0x7c/0xa0\n suspend_devices_and_enter+0x13c/0x540\n pm_suspend+0x2a4/0x330\n\nExamples why that assumption is not fine:\n\n- There is an MDIO bus with a PHY device that doesn\u0027t have a specific\n PHY driver loaded, because mdiobus_register() automatically creates a\n PHY device for it but there is no specific PHY driver in the system.\n Normally under those circumstances, the generic PHY driver will be\n bound lazily to it (at phy_attach_direct time). But some Ethernet\n drivers attach to their PHY at .ndo_open time. Until then it, the\n to-be-driven-by-genphy PHY device will not have a driver. The blamed\n patch amounts to saying \"you need to open all net devices before the\n system can suspend, to avoid the NULL pointer dereference\".\n\n- There is any raw MDIO device which has \u0027plausible\u0027 values in the PHY\n ID registers 2 and 3, which is located on an MDIO bus whose driver\n does not set bus-\u003ephy_mask \u003d ~0 (which prevents auto-scanning of PHY\n devices). An example could be a MAC\u0027s internal MDIO bus with PCS\n devices on it, for serial links such as SGMII. PHY devices will get\n created for those PCSes too, due to that MDIO bus auto-scanning, and\n although those PHY devices are not used, they do not bother anybody\n either. PCS devices are usually managed in Linux as raw MDIO devices.\n Nonetheless, they do not have a PHY driver, nor does anybody attempt\n to connect to them (because they are not a PHY), and therefore this\n patch breaks that.\n\nThe goal itself of the patch is questionable, so I am going for a\nstraight revert. to_phy_driver does not seem to have a need to be\nreplaced by phydev-\u003edrv, in fact that might even trigger code paths\nwhich were not given too deep of a thought.\n\nFor instance:\n\nphy_probe populates phydev-\u003edrv at the beginning, but does not clean it\nup on any error (including EPROBE_DEFER). So if the phydev driver\nrequests probe deferral, phydev-\u003edrv will remain populated despite there\nbeing no driver bound.\n\nIf a system suspend starts in between the initial probe deferral request\nand the subsequent probe retry, we will be calling the phydev-\u003edrv-\u003esuspend\nmethod, but _before_ any phydev-\u003edrv-\u003eprobe call has succeeded.\n\nThat is to say, if the phydev-\u003edrv is allocating any driver-private data\nstructure in -\u003eprobe, it pretty much expects that data structure to be\navailable in -\u003esuspend. But it may not. That is a pretty insane\nenvironment to present to PHY drivers.\n\nIn the code structure before the blamed patch, mdio_bus_phy_may_suspend\nwould just say \"no, don\u0027t suspend\" to any PHY device which does not have\na driver pointer _in_the_device_structure_ (not the phydev-\u003edrv). That\nwould essentially ensure that -\u003esuspend will never get called for a\ndevice that has not yet successfully completed probe. This is the code\nstructure the patch is returning to, via the revert.\n\nFixes: 3ac8eed62596 (\"net: phy: Uniform PHY driver access\")\nSigned-off-by: Vladimir Oltean \u003cvladimir.oltean@nxp.com\u003e\nAcked-by: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nLink: https://lore.kernel.org/r/20210914140515.2311548-1-vladimir.oltean@nxp.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "6a52e73368038f47f6618623d75061dc263b26ae", "tree": "ac6a10ada0817a9c9f979350a31efcbca7a84f1f", "parents": [ "7366c23ff492ad260776a3ee1aaabba9fc773a8b" ], "author": { "name": "Vladimir Oltean", "email": "vladimir.oltean@nxp.com", "time": "Tue Sep 14 16:43:31 2021 +0300" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Sep 15 15:03:36 2021 -0700" }, "message": "net: dsa: destroy the phylink instance on any error in dsa_slave_phy_setup\n\nDSA supports connecting to a phy-handle, and has a fallback to a non-OF\nbased method of connecting to an internal PHY on the switch\u0027s own MDIO\nbus, if no phy-handle and no fixed-link nodes were present.\n\nThe -ENODEV error code from the first attempt (phylink_of_phy_connect)\nis what triggers the second attempt (phylink_connect_phy).\n\nHowever, when the first attempt returns a different error code than\n-ENODEV, this results in an unbalance of calls to phylink_create and\nphylink_destroy by the time we exit the function. The phylink instance\nhas leaked.\n\nThere are many other error codes that can be returned by\nphylink_of_phy_connect. For example, phylink_validate returns -EINVAL.\nSo this is a practical issue too.\n\nFixes: aab9c4067d23 (\"net: dsa: Plug in PHYLINK support\")\nSigned-off-by: Vladimir Oltean \u003cvladimir.oltean@nxp.com\u003e\nReviewed-by: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nReviewed-by: Russell King (Oracle) \u003crmk+kernel@armlinux.org.uk\u003e\nLink: https://lore.kernel.org/r/20210914134331.2303380-1-vladimir.oltean@nxp.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "b7213ffa0e585feb1aee3e7173e965e66ee0abaa", "tree": "a9a4e9e96aea41389182aa7eba385521a106aa9a", "parents": [ "fc7c028dcdbfe981bca75d2a7b95f363eb691ef3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 15 13:56:37 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 15 13:56:37 2021 -0700" }, "message": "qnx4: avoid stringop-overread errors\n\nThe qnx4 directory entries are 64-byte blocks that have different\ncontents depending on the a status byte that is in the last byte of the\nblock.\n\nIn particular, a directory entry can be either a \"link info\" entry with\na 48-byte name and pointers to the real inode information, or an \"inode\nentry\" with a smaller 16-byte name and the full inode information.\n\nBut the code was written to always just treat the directory name as if\nit was part of that \"inode entry\", and just extend the name to the\nlonger case if the status byte said it was a link entry.\n\nThat work just fine and gives the right results, but now that gcc is\ntracking data structure accesses much more, the code can trigger a\ncompiler error about using up to 48 bytes (the long name) in a structure\nthat only has that shorter name in it:\n\n fs/qnx4/dir.c: In function ‘qnx4_readdir’:\n fs/qnx4/dir.c:51:32: error: ‘strnlen’ specified bound 48 exceeds source size 16 [-Werror\u003dstringop-overread]\n 51 | size \u003d strnlen(de-\u003edi_fname, size);\n | ^~~~~~~~~~~~~~~~~~~~~~~~~~~\n In file included from fs/qnx4/qnx4.h:3,\n from fs/qnx4/dir.c:16:\n include/uapi/linux/qnx4_fs.h:45:25: note: source object declared here\n 45 | char di_fname[QNX4_SHORT_NAME_MAX];\n | ^~~~~~~~\n\nwhich is because the source code doesn\u0027t really make this whole \"one of\ntwo different types\" explicit.\n\nFix this by introducing a very explicit union of the two types, and\nbasically explaining to the compiler what is really going on.\n\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "fc7c028dcdbfe981bca75d2a7b95f363eb691ef3", "tree": "7066ffe8658976fd44124247d8eee9fc4d14bde8", "parents": [ "d6efd3f18763ac84098fa318742cd2a3bfdf4d72" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Sep 06 16:06:04 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 15 13:42:33 2021 -0700" }, "message": "sparc: avoid stringop-overread errors\n\nThe sparc mdesc code does pointer games with \u0027struct mdesc_hdr\u0027, but\ndidn\u0027t describe to the compiler how that header is then followed by the\ndata that the header describes.\n\nAs a result, gcc is now unhappy since it does stricter pointer range\ntracking, and doesn\u0027t understand about how these things work. This\nresults in various errors like:\n\n arch/sparc/kernel/mdesc.c: In function ‘mdesc_node_by_name’:\n arch/sparc/kernel/mdesc.c:647:22: error: ‘strcmp’ reading 1 or more bytes from a region of size 0 [-Werror\u003dstringop-overread]\n 647 | if (!strcmp(names + ep[ret].name_offset, name))\n | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n\nwhich are easily avoided by just describing \u0027struct mdesc_hdr\u0027 better,\nand making the node_block() helper function look into that unsized\ndata[] that follows the header.\n\nThis makes the sparc64 build happy again at least for my cross-compiler\nversion (gcc version 11.2.1).\n\nLink: https://lore.kernel.org/lkml/CAHk-\u003dwi4NW3NC0xWykkw\u003d6LnjQD6D_rtRtxY9g8gQAJXtQMi8A@mail.gmail.com/\nCc: Guenter Roeck \u003clinux@roeck-us.net\u003e\nCc: David S. Miller \u003cdavem@davemloft.net\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "336562752acc1a723f9a24b5b8129ae22e0478c6", "tree": "c9c87cd366e52a7632c629603538b3239c48e3d6", "parents": [ "69cd823956ba8ce266a901170b1060db8073bddd" ], "author": { "name": "Matteo Croce", "email": "mcroce@microsoft.com", "time": "Wed Sep 15 01:54:00 2021 +0200" }, "committer": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Wed Sep 15 22:39:55 2021 +0200" }, "message": "bpf: Update bpf_get_smp_processor_id() documentation\n\nBPF programs run with migration disabled regardless of preemption, as\nthey are protected by migrate_disable(). Update the uapi documentation\naccordingly.\n\nSigned-off-by: Matteo Croce \u003cmcroce@microsoft.com\u003e\nSigned-off-by: Daniel Borkmann \u003cdaniel@iogearbox.net\u003e\nAcked-by: Yonghong Song \u003cyhs@fb.com\u003e\nLink: https://lore.kernel.org/bpf/20210914235400.59427-1-mcroce@linux.microsoft.com\n" }, { "commit": "69cd823956ba8ce266a901170b1060db8073bddd", "tree": "9a46ab82d2648d0c51a7efec2985b0d710bc1d8f", "parents": [ "2220ecf55c1b7aa36e99b00a6b964f4e5333f9bf" ], "author": { "name": "Grant Seltzer", "email": "grantseltzer@gmail.com", "time": "Tue Sep 14 22:19:52 2021 -0400" }, "committer": { "name": "Andrii Nakryiko", "email": "andrii@kernel.org", "time": "Wed Sep 15 13:16:02 2021 -0700" }, "message": "libbpf: Add sphinx code documentation comments\n\nThis adds comments above five functions in btf.h which document\ntheir uses. These comments are of a format that doxygen and sphinx\ncan pick up and render. These are rendered by libbpf.readthedocs.org\n\nSigned-off-by: Grant Seltzer \u003cgrantseltzer@gmail.com\u003e\nSigned-off-by: Andrii Nakryiko \u003candrii@kernel.org\u003e\nLink: https://lore.kernel.org/bpf/20210915021951.117186-1-grantseltzer@gmail.com\n" }, { "commit": "d6efd3f18763ac84098fa318742cd2a3bfdf4d72", "tree": "3369fa7a752f9793166f93191588b4fb266b20fa", "parents": [ "80be5998ad6339e3e804a772723390cb50b96428", "ebdc20d7bc74e8b6a242ff1f30e9017ffca9092c" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 15 12:11:48 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 15 12:11:48 2021 -0700" }, "message": "Merge branch \u0027absolute-pointer\u0027 (patches from Guenter)\n\nMerge absolute_pointer macro series from Guenter Roeck:\n \"Kernel test builds currently fail for several architectures with error\n messages such as the following.\n\n drivers/net/ethernet/i825xx/82596.c: In function \u0027i82596_probe\u0027:\n arch/m68k/include/asm/string.h:72:25: error:\n \u0027__builtin_memcpy\u0027 reading 6 bytes from a region of size 0\n [-Werror\u003dstringop-overread]\n\n Such warnings may be reported by gcc 11.x for string and memory\n operations on fixed addresses if gcc\u0027s builtin functions are used for\n those operations.\n\n This series introduces absolute_pointer() to fix the problem.\n absolute_pointer() disassociates a pointer from its originating symbol\n type and context, and thus prevents gcc from making assumptions about\n pointers passed to memory operations\"\n\n* emailed patches from Guenter Roeck \u003clinux@roeck-us.net\u003e:\n alpha: Use absolute_pointer to define COMMAND_LINE\n alpha: Move setup.h out of uapi\n net: i825xx: Use absolute_pointer for memcpy from fixed memory location\n compiler.h: Introduce absolute_pointer macro\n" }, { "commit": "ebdc20d7bc74e8b6a242ff1f30e9017ffca9092c", "tree": "9b342cb4470066c4d5e23ca9d5c9c605ab3c52f4", "parents": [ "3cb8b1537f8a89a681d2548ded5526280846f6db" ], "author": { "name": "Guenter Roeck", "email": "linux@roeck-us.net", "time": "Tue Sep 14 20:52:27 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 15 12:04:28 2021 -0700" }, "message": "alpha: Use absolute_pointer to define COMMAND_LINE\n\nalpha:allmodconfig fails to build with the following error\nwhen using gcc 11.x.\n\n arch/alpha/kernel/setup.c: In function \u0027setup_arch\u0027:\n arch/alpha/kernel/setup.c:493:13: error:\n\t\u0027strcmp\u0027 reading 1 or more bytes from a region of size 0\n\nAvoid the problem by declaring COMMAND_LINE as absolute_pointer().\n\nSigned-off-by: Guenter Roeck \u003clinux@roeck-us.net\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "3cb8b1537f8a89a681d2548ded5526280846f6db", "tree": "55ae5bfcf585eb0e7405713f4bbcd26c4fcdb69b", "parents": [ "dff2d13114f0beec448da9b3716204eb34b0cf41" ], "author": { "name": "Guenter Roeck", "email": "linux@roeck-us.net", "time": "Tue Sep 14 20:52:26 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 15 12:04:28 2021 -0700" }, "message": "alpha: Move setup.h out of uapi\n\nMost of the contents of setup.h have no value for userspace\napplications. The file was probably moved to uapi accidentally.\n\nKeep the file in uapi to define the alpha-specific COMMAND_LINE_SIZE.\nMove all other defines to arch/alpha/include/asm/setup.h.\n\nSuggested-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nSigned-off-by: Guenter Roeck \u003clinux@roeck-us.net\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" } ], "next": "dff2d13114f0beec448da9b3716204eb34b0cf41" }