)]}' { "log": [ { "commit": "70575e77839f4c5337ce2653b39b86bb365a870e", "tree": "c89c019fdb0360d4861a67580066a03c7d434b45", "parents": [ "7bc6e90d7aa4170039abe80b9f4e8c8e4eb35091", "a43ae8057cc154fd26a3a23c0e8643bef104d995" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 09:41:34 2022 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 09:41:34 2022 -0700" }, "message": "Merge tag \u0027for_linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost\n\nPull virtio fixes from Michael Tsirkin:\n \"Some last minute fixes.\n\n The virtio-blk one is the most important one since it was actually\n seen in the field, but the rest of them are small and clearly safe,\n everything here has been in next for a while\"\n\n* tag \u0027for_linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost:\n vdpa/mlx5: Fix MQ to support non power of two num queues\n vduse: prevent uninitialized memory accesses\n virtio-blk: Fix WARN_ON_ONCE in virtio_queue_rq()\n virtio_test: fixup for vq reset\n virtio-crypto: fix memory-leak\n vdpa/ifcvf: fix the calculation of queuepair\n" }, { "commit": "7bc6e90d7aa4170039abe80b9f4e8c8e4eb35091", "tree": "14bc5b5b614437063a144a439ab529e1887b003b", "parents": [ "a0debc4c7a7e1a4adf0866b105e9e4f29002c2ca", "6c84501a3c38adaf1c3486fb80b972f728ad8fd1" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 09:33:33 2022 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 09:33:33 2022 -0700" }, "message": "Merge tag \u0027block-6.0-2022-09-29\u0027 of git://git.kernel.dk/linux\n\nPull block fixes from Jens Axboe:\n \"A single NVMe pull request via Christoph with a few fixes that should\n go into the 6.0 release:\n\n - Fix IOC_PR_CLEAR and IOC_PR_RELEASE ioctls for nvme devices\n (Michael Kelley)\n\n - Disable Write Zeroes on Phison E3C/E4C (Tina Hsu)\"\n\n* tag \u0027block-6.0-2022-09-29\u0027 of git://git.kernel.dk/linux:\n nvme-pci: disable Write Zeroes on Phison E3C/E4C\n nvme: Fix IOC_PR_CLEAR and IOC_PR_RELEASE ioctls for nvme devices\n" }, { "commit": "a0debc4c7a7e1a4adf0866b105e9e4f29002c2ca", "tree": "1e7ee8804585088491736384a6a3107e2e2576cc", "parents": [ "d649d2c49baa4336bf65a02fc8e9d42c8ef2db6b", "d59bd748db0a97a5d6a33b284b6c58b7f6f4f768" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 09:28:39 2022 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 09:28:39 2022 -0700" }, "message": "Merge tag \u0027io_uring-6.0-2022-09-29\u0027 of git://git.kernel.dk/linux\n\nPull io_uring fixes from Jens Axboe:\n \"Two fixes that should go into 6.0:\n\n - Tweak the single issuer logic to register the task at creation,\n rather than at first submit. SINGLE_ISSUER was added for 6.0, and\n after some discussion on this, we decided to make it a bit stricter\n while it\u0027s still possible to do so (Dylan).\n\n - Stefan from Samba had some doubts on the level triggered poll that\n was added for this release. Rather than attempt to mess around with\n it now, just do the quick one-liner to disable it for release and\n we have time to discuss and change it for 6.1 instead (me)\"\n\n* tag \u0027io_uring-6.0-2022-09-29\u0027 of git://git.kernel.dk/linux:\n io_uring/poll: disable level triggered poll\n io_uring: register single issuer task at creation\n" }, { "commit": "d649d2c49baa4336bf65a02fc8e9d42c8ef2db6b", "tree": "9b9a18fef898edf7ed80089f77d886fc31a8ccdf", "parents": [ "89e10b860cbf8015dd77aaa90839fb0ef65e0619", "40158dbf7eb2b13d8851fe0b875b4c3170ea15db" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 08:54:14 2022 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 08:54:14 2022 -0700" }, "message": "Merge tag \u0027pstore-v6.0-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux\n\nPull pstore revert from Kees Cook:\n \"A misbehavior with some compression backends in pstore was just\n discovered due to the recent crypto acomp migration.\n\n Since we\u0027re so close to release, it seems better to just simply revert\n it, and we can figure out what\u0027s going on without leaving it broken\n for a release.\n\n - Revert crypto acomp migration (Guilherme G. Piccoli)\"\n\n* tag \u0027pstore-v6.0-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux:\n Revert \"pstore: migrate to crypto acomp interface\"\n" }, { "commit": "89e10b860cbf8015dd77aaa90839fb0ef65e0619", "tree": "f91e2478387f6f852161a0c5e595d9cfe6bd470c", "parents": [ "5a77386984b513ebfb2700e70dac44509fc81aa9", "4335417da2b8d6d9b2d4411b5f9e248e5bb2d380" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 08:30:10 2022 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 08:30:10 2022 -0700" }, "message": "Merge tag \u0027gpio-fixes-for-v6.0\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux\n\nPull gpio fixes from Bartosz Golaszewski:\n \"One more fix for the upcoming release:\n\n - fix the check for pwm support on non-A8K platforms in gpio-mvebu\"\n\n* tag \u0027gpio-fixes-for-v6.0\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux:\n gpio: mvebu: Fix check for pwm support on non-A8K platforms\n" }, { "commit": "40158dbf7eb2b13d8851fe0b875b4c3170ea15db", "tree": "d9b0e163ac2fa7179650c5538ec7c1962040518e", "parents": [ "568035b01cfb107af8d2e4bd2fb9aea22cf5b868" ], "author": { "name": "Guilherme G. Piccoli", "email": "gpiccoli@igalia.com", "time": "Thu Sep 29 18:55:15 2022 -0300" }, "committer": { "name": "Kees Cook", "email": "keescook@chromium.org", "time": "Fri Sep 30 08:16:06 2022 -0700" }, "message": "Revert \"pstore: migrate to crypto acomp interface\"\n\nThis reverts commit e4f0a7ec586b7644107839f5394fb685cf1aadcc.\n\nWhen using this new interface, both efi_pstore and ramoops\nbackends are unable to properly decompress dmesg if using\nzstd, lz4 and lzo algorithms (and maybe more). It does succeed\nwith deflate though.\n\nThe message observed in the kernel log is:\n\n[2.328828] pstore: crypto_acomp_decompress failed, ret \u003d -22!\n\nThe pstore infrastructure is able to collect the dmesg with\nboth backends tested, but since decompression fails it\u0027s\nunreadable. With this revert everything is back to normal.\n\nFixes: e4f0a7ec586b (\"pstore: migrate to crypto acomp interface\")\nCc: Ard Biesheuvel \u003cardb@kernel.org\u003e\nSigned-off-by: Guilherme G. Piccoli \u003cgpiccoli@igalia.com\u003e\nSigned-off-by: Kees Cook \u003ckeescook@chromium.org\u003e\nLink: https://lore.kernel.org/r/20220929215515.276486-1-gpiccoli@igalia.com\n" }, { "commit": "5a77386984b513ebfb2700e70dac44509fc81aa9", "tree": "e3b3bcda3b258300bd57d49e7cec1f0d6e4569e8", "parents": [ "987a926c1d8a40e4256953b04771fbdb63bc7938", "6643b3836f3908c4f77883b2fae72451e85cf3ca" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 08:12:04 2022 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Sep 30 08:12:04 2022 -0700" }, "message": "Merge tag \u0027drm-fixes-2022-09-30-1\u0027 of git://anongit.freedesktop.org/drm/drm\n\nPull drm fixes from Dave Airlie:\n \"Last set of fixes for 6.0 hopefully - minor bridge fixes, i915 fixes,\n and a bunch of amdgpu fixes for new IP blocks, along with a couple of\n regression fixes. Should be all set for merge window next week.\n\n amdgpu:\n - GC 11.x fixes\n - SMU 13.x fixes\n - DCN 3.1.4 fixes\n - DCN 3.2.x fixes\n - GC 9.x fix\n - Fence fix\n - SR-IOV supend/resume fix\n - PSR regression fix\n\n i915:\n - Restrict forced preemption to the active context\n - Restrict perf_limit_reasons to the supported platforms - gen11+\n\n bridge:\n - analogix: Revert earlier suspend fix\n - lt8912b: Fix corrupt display output\"\n\n* tag \u0027drm-fixes-2022-09-30-1\u0027 of git://anongit.freedesktop.org/drm/drm: (26 commits)\n drm/amd/display: Prevent OTG shutdown during PSR SU\n drm/i915/gt: Perf_limit_reasons are only available for Gen11+\n drm/amdgpu: Add amdgpu suspend-resume code path under SRIOV\n drm/amdgpu: Remove fence_process in count_emitted\n drm/amdgpu: Correct the position in patch_cond_exec\n drm/amd/display: fill in clock values when DPM is not enabled\n drm/amd/display: Avoid unnecessary pixel rate divider programming\n drm/amd/display: Remove assert for odm transition case\n drm/amd/display: Fix typo in get_pixel_rate_div\n drm/amd/display: Fix audio on display after unplugging another\n drm/amd/display: Add explicit FIFO disable for DP blank\n drm/amd/display: Wrap OTG disable workaround with FIFO control\n drm/amd/display: Do DIO FIFO enable after DP video stream enable\n drm/amd/display: Update DCN32 to use new SR latencies\n drm/amd/display: Avoid avoid unnecessary pixel rate divider programming\n drm/amdkfd: fix dropped interrupt in kfd_int_process_v11\n drm/amdgpu: pass queue size and is_aql_queue to MES\n drm/amdkfd: fix MQD init for GFX11 in init_mqd\n drm/amd/pm: use adverse selection for dpm features unsupported by driver\n drm/amd/pm: enable gfxoff feature for SMU 13.0.0\n ...\n" }, { "commit": "6643b3836f3908c4f77883b2fae72451e85cf3ca", "tree": "7792da8562cbee892aaf56e9ddf1b3517a17b3dc", "parents": [ "91462afa42e6b8c5a148a44482c893f1800c513d", "7738be973fc4e2ba22154fafd3a5d7b9666f9abf" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Sep 30 09:28:00 2022 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Sep 30 09:28:58 2022 +1000" }, "message": "Merge tag \u0027drm-intel-fixes-2022-09-29\u0027 of git://anongit.freedesktop.org/drm/drm-intel into drm-fixes\n\n- Restrict forced preemption to the active context (Chris)\n- Restrict perf_limit_reasons to the supported platforms - gen11+ (Ashutosh)\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n\nFrom: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/YzXAkH1a32pYJD33@intel.com\n" }, { "commit": "91462afa42e6b8c5a148a44482c893f1800c513d", "tree": "e45a625d2b7dc4c196ddb9250ea536df92781fe3", "parents": [ "153a97b4e3de713c05b6bd42d734a2f9eeaf7968", "83ca5fb40e758e0a0257bf4e3a1148dd52c6d0f2" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Sep 30 09:03:03 2022 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Sep 30 09:03:03 2022 +1000" }, "message": "Merge tag \u0027amd-drm-fixes-6.0-2022-09-29\u0027 of https://gitlab.freedesktop.org/agd5f/linux into drm-fixes\n\namd-drm-fixes-6.0-2022-09-29:\n\namdgpu:\n- GC 11.x fixes\n- SMU 13.x fixes\n- DCN 3.1.4 fixes\n- DCN 3.2.x fixes\n- GC 9.x fix\n- Fence fix\n- SR-IOV supend/resume fix\n- PSR regression fix\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\nFrom: Alex Deucher \u003calexander.deucher@amd.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20220929144003.8363-1-alexander.deucher@amd.com\n" }, { "commit": "153a97b4e3de713c05b6bd42d734a2f9eeaf7968", "tree": "3131fa1777b3b03d4ccf59c68f86f0b95fbd773d", "parents": [ "f76349cf41451c5c42a99f18a9163377e4b364ff", "cc62d98bd56d45de4531844ca23913a15136c05b" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Sep 30 08:50:29 2022 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Sep 30 08:50:34 2022 +1000" }, "message": "Merge tag \u0027drm-misc-fixes-2022-09-29\u0027 of git://anongit.freedesktop.org/drm/drm-misc into drm-fixes\n\nShort summary of fixes pull:\n\n * bridge/analogix: Revert earlier suspend fix\n * bridge/lt8912b: Fix corrupt display output\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n\nFrom: Thomas Zimmermann \u003ctzimmermann@suse.de\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/YzWvHhaqHhYirn4L@linux-uq9g\n" }, { "commit": "987a926c1d8a40e4256953b04771fbdb63bc7938", "tree": "10001b1d338c7bb060fa72e845f61bb99c842c72", "parents": [ "511cce163b75bc3933fa3de769a82bb7e8663f2b", "06bbaa6dc53cb72040db952053432541acb9adc7" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 29 14:37:45 2022 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 29 14:37:45 2022 -0700" }, "message": "Merge tag \u0027pull-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs\n\nPull coredump fix from Al Viro:\n \"Fix for breakage in dump_user_range()\"\n\n* tag \u0027pull-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:\n [coredump] don\u0027t use __kernel_write() on kmap_local_page()\n" }, { "commit": "511cce163b75bc3933fa3de769a82bb7e8663f2b", "tree": "a8ed2893dc998233395689e72aad141974e9d916", "parents": [ "da9eede6b2423381d592b832a7c5ba0082f2e905", "3b04cba7add093d0d8267cf70a333ca8fe8233ca" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 29 08:32:53 2022 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 29 08:32:53 2022 -0700" }, "message": "Merge tag \u0027net-6.0-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net\n\nPull networking fixes from Paolo Abeni:\n \"Including fixes from wifi and can.\n\n Current release - regressions:\n\n - phy: don\u0027t WARN for PHY_UP state in mdio_bus_phy_resume()\n\n - wifi: fix locking in mac80211 mlme\n\n - eth:\n - revert \"net: mvpp2: debugfs: fix memory leak when using debugfs_lookup()\"\n - mlxbf_gige: fix an IS_ERR() vs NULL bug in mlxbf_gige_mdio_probe\n\n Previous releases - regressions:\n\n - wifi: fix regression with non-QoS drivers\n\n Previous releases - always broken:\n\n - mptcp: fix unreleased socket in accept queue\n\n - wifi:\n - don\u0027t start TX with fq-\u003elock to fix deadlock\n - fix memory corruption in minstrel_ht_update_rates()\n\n - eth:\n - macb: fix ZynqMP SGMII non-wakeup source resume failure\n - mt7531: only do PLL once after the reset\n - usbnet: fix memory leak in usbnet_disconnect()\n\n Misc:\n\n - usb: qmi_wwan: add new usb-id for Dell branded EM7455\"\n\n* tag \u0027net-6.0-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (30 commits)\n mptcp: fix unreleased socket in accept queue\n mptcp: factor out __mptcp_close() without socket lock\n net: ethernet: mtk_eth_soc: fix mask of RX_DMA_GET_SPORT{,_V2}\n net: mscc: ocelot: fix tagged VLAN refusal while under a VLAN-unaware bridge\n can: c_can: don\u0027t cache TX messages for C_CAN cores\n ice: xsk: drop power of 2 ring size restriction for AF_XDP\n ice: xsk: change batched Tx descriptor cleaning\n net: usb: qmi_wwan: Add new usb-id for Dell branded EM7455\n selftests: Fix the if conditions of in test_extra_filter()\n net: phy: Don\u0027t WARN for PHY_UP state in mdio_bus_phy_resume()\n net: stmmac: power up/down serdes in stmmac_open/release\n wifi: mac80211: mlme: Fix double unlock on assoc success handling\n wifi: mac80211: mlme: Fix missing unlock on beacon RX\n wifi: mac80211: fix memory corruption in minstrel_ht_update_rates()\n wifi: mac80211: fix regression with non-QoS drivers\n wifi: mac80211: ensure vif queues are operational after start\n wifi: mac80211: don\u0027t start TX with fq-\u003elock to fix deadlock\n wifi: cfg80211: fix MCS divisor value\n net: hippi: Add missing pci_disable_device() in rr_init_one()\n net/mlxbf_gige: Fix an IS_ERR() vs NULL bug in mlxbf_gige_mdio_probe\n ...\n" }, { "commit": "da9eede6b2423381d592b832a7c5ba0082f2e905", "tree": "950687e26fba47140fc4c070af5e8233d1cad8a6", "parents": [ "71f187570592e74d32db4f52e7f50a64c11ee621", "e62563db857f81d75c5726a35bc0180bed6d1540" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 29 08:22:53 2022 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 29 08:22:53 2022 -0700" }, "message": "Merge tag \u0027input-for-v6.0-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input\n\nPull input fixes from Dmitry Torokhov:\n\n - small fixes for iqs62x-keys and melfas_mip4 drivers\n\n - corrected register address in snvs_pwrkey driver\n\n - Synaptic driver will stop trying to use intertouch (native) mode on\n some Lenovo AMD devices\n\n* tag \u0027input-for-v6.0-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:\n Input: snvs_pwrkey - fix SNVS_HPVIDR1 register address\n Input: synaptics - disable Intertouch for Lenovo T14 and P14s AMD G1\n Input: iqs62x-keys - drop unused device node references\n Input: melfas_mip4 - fix return value check in mip4_probe()\n" }, { "commit": "6c84501a3c38adaf1c3486fb80b972f728ad8fd1", "tree": "c86021bca20a8597bebd4b9bbadd29d6a621ab16", "parents": [ "4c66a326b5ab784cddd72de07ac5b6210e9e1b06", "d14c273132aec81a1a8107c9ab4865b89e7910a7" ], "author": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Thu Sep 29 09:04:02 2022 -0600" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Thu Sep 29 09:04:02 2022 -0600" }, "message": "Merge tag \u0027nvme-6.0-2022-09-29\u0027 of git://git.infradead.org/nvme into block-6.0\n\nPull NVMe fixes from Christoph:\n\n\"nvme fixes for Linux 6.1\n\n - fix IOC_PR_CLEAR and IOC_PR_RELEASE ioctls for nvme devices\n (Michael Kelley)\n - disable Write Zeroes on Phison E3C/E4C (Tina Hsu)\"\n\n* tag \u0027nvme-6.0-2022-09-29\u0027 of git://git.infradead.org/nvme:\n nvme-pci: disable Write Zeroes on Phison E3C/E4C\n nvme: Fix IOC_PR_CLEAR and IOC_PR_RELEASE ioctls for nvme devices\n" }, { "commit": "83ca5fb40e758e0a0257bf4e3a1148dd52c6d0f2", "tree": "190fc36c3852081ecf6342c5a24b714eb4a1c253", "parents": [ "3b7329cf5a767c1be38352d43066012e220ad43c" ], "author": { "name": "Leo Li", "email": "sunpeng.li@amd.com", "time": "Mon Sep 26 17:53:55 2022 -0400" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Thu Sep 29 10:07:42 2022 -0400" }, "message": "drm/amd/display: Prevent OTG shutdown during PSR SU\n\n[Why]\n\nEnabling Z10 optimizations allows DMUB to disable the OTG during PSR\nlink-off. This theoretically saves power by putting more of the display\nhardware to sleep. However, we observe that with PSR SU, it causes\nvisual artifacts, higher power usage, and potential system hang.\n\nThis is partly due to an odd behavior with the VStartup interrupt used\nto signal DRM vblank events. If the OTG is toggled on/off during a PSR\nlink on/off cycle, the vstartup interrupt fires twice in quick\nsuccession. This generates incorrectly timed vblank events.\nAdditionally, it can cause cursor updates to generate visual artifacts.\n\nNote that this is not observed with PSR1 since PSR is fully disabled\nwhen there are vblank event requestors. Cursor updates are also\nartifact-free, likely because there are no selectively-updated (SU)\nframes that can generate artifacts.\n\n[How]\n\nA potential solution is to disable z10 idle optimizations only when fast\nupdates (flips \u0026 cursor updates) are committed. A mechanism to do so\nwould require some thoughtful design. Let\u0027s just disable idle\noptimizations for PSR2 for now.\n\nFixes: 7cc191ee7621 (\"drm/amd/display: Implement MPO PSR SU\")\nReported-by: August Wikerfors \u003cgit@augustwikerfors.se\u003e\nLink: https://lore.kernel.org/r/c1f8886a-5624-8f49-31b1-e42b6d20dcf5@augustwikerfors.se/\nTested-by: August Wikerfors \u003cgit@augustwikerfors.se\u003e\nReviewed-by: Harry Wentland \u003charry.wentland@amd.com\u003e\nSigned-off-by: Leo Li \u003csunpeng.li@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "71f187570592e74d32db4f52e7f50a64c11ee621", "tree": "d7aad29d1cf6a0ddca68466bd262b5a45000daf3", "parents": [ "81bcd4b522cf1dc312385b79b144153aff0cf0c6", "141f3d6256e58103ece1c3dd2835e871f1dde240" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 29 05:40:59 2022 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 29 05:40:59 2022 -0700" }, "message": "Merge tag \u0027ata-6.0-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dlemoal/libata\n\nPull ATA fixes from Damien Le Moal:\n \"Three late patches to fix problems discovered recently:\n\n - Add a horkage to disable link power management by default for the\n Pioneer BDR-207M and BDR-205 DVD drives (from Niklas)\n\n - Two patches to fix setting the maximum queue depth of libsas owned\n ATA devices (from me)\"\n\n* tag \u0027ata-6.0-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dlemoal/libata:\n ata: libata-sata: Fix device queue depth control\n ata: libata-scsi: Fix initialization of device queue depth\n libata: add ATA_HORKAGE_NOLPM for Pioneer BDR-207M and BDR-205\n" }, { "commit": "81bcd4b522cf1dc312385b79b144153aff0cf0c6", "tree": "e3643352a4efce3e1c8c9a04ccf69fd49225701c", "parents": [ "c3e0e1e23c70455916ff3472072437b3605c6cfe", "4f196cb64b693c64f8f981432b1ff326b7ea1c28" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 29 05:35:32 2022 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 29 05:35:32 2022 -0700" }, "message": "Merge tag \u0027loongarch-fixes-6.0-3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/chenhuacai/linux-loongson\n\nPull LoongArch fixes from Huacai Chen:\n \"Some trivial fixes and cleanup\"\n\n* tag \u0027loongarch-fixes-6.0-3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/chenhuacai/linux-loongson:\n LoongArch: Clean up loongson3_smp_ops declaration\n LoongArch: Fix and cleanup csr_era handling in do_ri()\n LoongArch: Align the address of kernel_entry to 4KB\n" }, { "commit": "4f196cb64b693c64f8f981432b1ff326b7ea1c28", "tree": "87f8da8af35803a133ce4d27604df4cd457d3eb5", "parents": [ "06e76acec70d1381b9a96e6014d5587d84201f82" ], "author": { "name": "Yanteng Si", "email": "siyanteng@loongson.cn", "time": "Tue Sep 27 12:34:56 2022 +0800" }, "committer": { "name": "Huacai Chen", "email": "chenhuacai@loongson.cn", "time": "Thu Sep 29 10:15:00 2022 +0800" }, "message": "LoongArch: Clean up loongson3_smp_ops declaration\n\nSince loongson3_smp_ops is not used in LoongArch anymore, let\u0027s remove\nit for cleanup.\n\nFixes: f2ac457a6138 (\"LoongArch: Add CPU definition headers\")\nSigned-off-by: Yanteng Si \u003csiyanteng@loongson.cn\u003e\nSigned-off-by: Huacai Chen \u003cchenhuacai@loongson.cn\u003e\n" }, { "commit": "06e76acec70d1381b9a96e6014d5587d84201f82", "tree": "b08e7b19c316b38fcf878a78f8324c966f4a7e1f", "parents": [ "2938431e9338bf4e8d434513d3a5832d64c0fa8b" ], "author": { "name": "Huacai Chen", "email": "chenhuacai@loongson.cn", "time": "Mon Sep 26 22:33:39 2022 +0800" }, "committer": { "name": "Huacai Chen", "email": "chenhuacai@loongson.cn", "time": "Thu Sep 29 10:15:00 2022 +0800" }, "message": "LoongArch: Fix and cleanup csr_era handling in do_ri()\n\nWe don\u0027t emulate reserved instructions and just send a signal to the\ncurrent process now. So we don\u0027t need to call compute_return_era() to\nadd 4 (point to the next instruction) to csr_era in pt_regs. RA/ERA\u0027s\nbackup/restore is cleaned up as well.\n\nSigned-off-by: Jun Yi \u003cyijun@loongson.cn\u003e\nSigned-off-by: Huacai Chen \u003cchenhuacai@loongson.cn\u003e\n" }, { "commit": "2938431e9338bf4e8d434513d3a5832d64c0fa8b", "tree": "9fb271169040fe5aee6eeda917117edee258ceb2", "parents": [ "f76349cf41451c5c42a99f18a9163377e4b364ff" ], "author": { "name": "Huacai Chen", "email": "chenhuacai@loongson.cn", "time": "Mon Sep 26 22:25:29 2022 +0800" }, "committer": { "name": "Huacai Chen", "email": "chenhuacai@loongson.cn", "time": "Thu Sep 29 10:15:00 2022 +0800" }, "message": "LoongArch: Align the address of kernel_entry to 4KB\n\nAlign the address of kernel_entry to 4KB, to avoid early tlb miss\nexception in case the entry code crosses page boundary.\n\nSigned-off-by: Huacai Chen \u003cchenhuacai@loongson.cn\u003e\n" }, { "commit": "3b04cba7add093d0d8267cf70a333ca8fe8233ca", "tree": "65446d1ed1c352092f390758f324cfdf4fc56f29", "parents": [ "3e1308a7c88ba149baa08ddffedee30c6b3224a0", "30e51b923e436b631e8d5b77fa5e318c6b066dc7" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Sep 28 19:05:39 2022 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Sep 28 19:05:40 2022 -0700" }, "message": "Merge branch \u0027mptcp-properly-clean-up-unaccepted-subflows\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: Properly clean up unaccepted subflows\n\nPatch 1 factors out part of the mptcp_close() function for use by a caller\nthat already owns the socket lock. This is a prerequisite for patch 2.\n\nPatch 2 is the fix that fully cleans up the unaccepted subflow sockets.\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/20220927193158.195729-1-mathew.j.martineau@linux.intel.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "30e51b923e436b631e8d5b77fa5e318c6b066dc7", "tree": "65446d1ed1c352092f390758f324cfdf4fc56f29", "parents": [ "26d3e21ce1aab6cb19069c510fac8e7474445b18" ], "author": { "name": "Menglong Dong", "email": "imagedong@tencent.com", "time": "Tue Sep 27 12:31:58 2022 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Sep 28 19:05:21 2022 -0700" }, "message": "mptcp: fix unreleased socket in accept queue\n\nThe mptcp socket and its subflow sockets in accept queue can\u0027t be\nreleased after the process exit.\n\nWhile the release of a mptcp socket in listening state, the\ncorresponding tcp socket will be released too. Meanwhile, the tcp\nsocket in the unaccept queue will be released too. However, only init\nsubflow is in the unaccept queue, and the joined subflow is not in the\nunaccept queue, which makes the joined subflow won\u0027t be released, and\ntherefore the corresponding unaccepted mptcp socket will not be released\nto.\n\nThis can be reproduced easily with following steps:\n\n1. create 2 namespace and veth:\n $ ip netns add mptcp-client\n $ ip netns add mptcp-server\n $ sysctl -w net.ipv4.conf.all.rp_filter\u003d0\n $ ip netns exec mptcp-client sysctl -w net.mptcp.enabled\u003d1\n $ ip netns exec mptcp-server sysctl -w net.mptcp.enabled\u003d1\n $ ip link add red-client netns mptcp-client type veth peer red-server \\\n netns mptcp-server\n $ ip -n mptcp-server address add 10.0.0.1/24 dev red-server\n $ ip -n mptcp-server address add 192.168.0.1/24 dev red-server\n $ ip -n mptcp-client address add 10.0.0.2/24 dev red-client\n $ ip -n mptcp-client address add 192.168.0.2/24 dev red-client\n $ ip -n mptcp-server link set red-server up\n $ ip -n mptcp-client link set red-client up\n\n2. configure the endpoint and limit for client and server:\n $ ip -n mptcp-server mptcp endpoint flush\n $ ip -n mptcp-server mptcp limits set subflow 2 add_addr_accepted 2\n $ ip -n mptcp-client mptcp endpoint flush\n $ ip -n mptcp-client mptcp limits set subflow 2 add_addr_accepted 2\n $ ip -n mptcp-client mptcp endpoint add 192.168.0.2 dev red-client id \\\n 1 subflow\n\n3. listen and accept on a port, such as 9999. The nc command we used\n here is modified, which makes it use mptcp protocol by default.\n $ ip netns exec mptcp-server nc -l -k -p 9999\n\n4. open another *two* terminal and use each of them to connect to the\n server with the following command:\n $ ip netns exec mptcp-client nc 10.0.0.1 9999\n Input something after connect to trigger the connection of the second\n subflow. So that there are two established mptcp connections, with the\n second one still unaccepted.\n\n5. exit all the nc command, and check the tcp socket in server namespace.\n And you will find that there is one tcp socket in CLOSE_WAIT state\n and can\u0027t release forever.\n\nFix this by closing all of the unaccepted mptcp socket in\nmptcp_subflow_queue_clean() with __mptcp_close().\n\nNow, we can ensure that all unaccepted mptcp sockets will be cleaned by\n__mptcp_close() before they are released, so mptcp_sock_destruct(), which\nis used to clean the unaccepted mptcp socket, is not needed anymore.\n\nThe selftests for mptcp is ran for this commit, and no new failures.\n\nFixes: f296234c98a8 (\"mptcp: Add handling of incoming MP_JOIN requests\")\nFixes: 6aeed9045071 (\"mptcp: fix race on unaccepted mptcp sockets\")\nCc: stable@vger.kernel.org\nReviewed-by: Jiang Biao \u003cbenbjiang@tencent.com\u003e\nReviewed-by: Mengen Sun \u003cmengensun@tencent.com\u003e\nAcked-by: Paolo Abeni \u003cpabeni@redhat.com\u003e\nSigned-off-by: Menglong Dong \u003cimagedong@tencent.com\u003e\nSigned-off-by: Mat Martineau \u003cmathew.j.martineau@linux.intel.com\u003e\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "26d3e21ce1aab6cb19069c510fac8e7474445b18", "tree": "6c17560f92ea8d98d1bd6722ded76537c43678ab", "parents": [ "3e1308a7c88ba149baa08ddffedee30c6b3224a0" ], "author": { "name": "Menglong Dong", "email": "imagedong@tencent.com", "time": "Tue Sep 27 12:31:57 2022 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Sep 28 19:05:21 2022 -0700" }, "message": "mptcp: factor out __mptcp_close() without socket lock\n\nFactor out __mptcp_close() from mptcp_close(). The caller of\n__mptcp_close() should hold the socket lock, and cancel mptcp work when\n__mptcp_close() returns true.\n\nThis function will be used in the next commit.\n\nFixes: f296234c98a8 (\"mptcp: Add handling of incoming MP_JOIN requests\")\nFixes: 6aeed9045071 (\"mptcp: fix race on unaccepted mptcp sockets\")\nCc: stable@vger.kernel.org\nReviewed-by: Jiang Biao \u003cbenbjiang@tencent.com\u003e\nReviewed-by: Mengen Sun \u003cmengensun@tencent.com\u003e\nAcked-by: Paolo Abeni \u003cpabeni@redhat.com\u003e\nSigned-off-by: Menglong Dong \u003cimagedong@tencent.com\u003e\nSigned-off-by: Mat Martineau \u003cmathew.j.martineau@linux.intel.com\u003e\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "3e1308a7c88ba149baa08ddffedee30c6b3224a0", "tree": "984d98bd044368c3f3769aaec0a638ceb413c973", "parents": [ "c9da02bfb1112461e048d3b736afb1873f6f4ccf", "b3056ae2b57858b02b376b3fed6077040caf14b4" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Sep 28 19:04:43 2022 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Sep 28 19:04:43 2022 -0700" }, "message": "Merge branch \u0027100GbE\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue\n\nTony Nguyen says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nice: xsk: ZC changes\n\nMaciej Fijalkowski says:\n\nThis set consists of two fixes to issues that were either pointed out on\nindirectly (John was reviewing AF_XDP selftests that were testing ice\u0027s\nZC support) mailing list or were directly reported by customers.\n\nFirst patch allows user space to see done descriptor in CQ even after a\nsingle frame being transmitted and second patch removes the need for\nhaving HW rings sized to power of 2 number of descriptors when used\nagainst AF_XDP.\n\nI also forgot to mention that due to the current Tx cleaning algorithm,\n4k HW ring was broken and these two patches bring it back to life, so we\nkill two birds with one stone.\n\n* \u0027100GbE\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue:\n ice: xsk: drop power of 2 ring size restriction for AF_XDP\n ice: xsk: change batched Tx descriptor cleaning\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/20220927164112.4011983-1-anthony.l.nguyen@intel.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "c9da02bfb1112461e048d3b736afb1873f6f4ccf", "tree": "ce65954992ce016280ed7a2f50fff8f681d021ed", "parents": [ "276d37eb449133bc22872b8f0a6f878e120deeff" ], "author": { "name": "Daniel Golle", "email": "daniel@makrotopia.org", "time": "Tue Sep 27 16:30:02 2022 +0100" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Sep 28 19:03:57 2022 -0700" }, "message": "net: ethernet: mtk_eth_soc: fix mask of RX_DMA_GET_SPORT{,_V2}\n\nThe bitmasks applied in RX_DMA_GET_SPORT and RX_DMA_GET_SPORT_V2 macros\nwere swapped. Fix that.\n\nReported-by: Chen Minqiang \u003cptpt52@gmail.com\u003e\nFixes: 160d3a9b192985 (\"net: ethernet: mtk_eth_soc: introduce MTK_NETSYS_V2 support\")\nAcked-by: Lorenzo Bianconi \u003clorenzo@kernel.org\u003e\nSigned-off-by: Daniel Golle \u003cdaniel@makrotopia.org\u003e\nLink: https://lore.kernel.org/r/YzMW+mg9UsaCdKRQ@makrotopia.org\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "276d37eb449133bc22872b8f0a6f878e120deeff", "tree": "ceba374e6313c7e04828a9326f0b6fef4cec6c8a", "parents": [ "af2faee54d591904050d69a55b6e429565bbe389" ], "author": { "name": "Vladimir Oltean", "email": "vladimir.oltean@nxp.com", "time": "Tue Sep 27 15:20:42 2022 +0300" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Sep 28 19:03:28 2022 -0700" }, "message": "net: mscc: ocelot: fix tagged VLAN refusal while under a VLAN-unaware bridge\n\nCurrently the following set of commands fails:\n\n$ ip link add br0 type bridge # vlan_filtering 0\n$ ip link set swp0 master br0\n$ bridge vlan\nport vlan-id\nswp0 1 PVID Egress Untagged\n$ bridge vlan add dev swp0 vid 10\nError: mscc_ocelot_switch_lib: Port with more than one egress-untagged VLAN cannot have egress-tagged VLANs.\n\nDumping ocelot-\u003evlans, one can see that the 2 egress-untagged VLANs on swp0 are\nvid 1 (the bridge PVID) and vid 4094, a PVID used privately by the driver for\nVLAN-unaware bridging. So this is why bridge vid 10 is refused, despite\n\u0027bridge vlan\u0027 showing a single egress untagged VLAN.\n\nAs mentioned in the comment added, having this private VLAN does not impose\nrestrictions to the hardware configuration, yet it is a bookkeeping problem.\n\nThere are 2 possible solutions.\n\nOne is to make the functions that operate on VLAN-unaware pvids:\n- ocelot_add_vlan_unaware_pvid()\n- ocelot_del_vlan_unaware_pvid()\n- ocelot_port_setup_dsa_8021q_cpu()\n- ocelot_port_teardown_dsa_8021q_cpu()\ncall something different than ocelot_vlan_member_(add|del)(), the latter being\nthe real problem, because it allocates a struct ocelot_bridge_vlan *vlan which\nit adds to ocelot-\u003evlans. We don\u0027t really *need* the private VLANs in\nocelot-\u003evlans, it\u0027s just that we have the extra convenience of having the\nvlan-\u003eportmask cached in software (whereas without these structures, we\u0027d have\nto create a raw ocelot_vlant_rmw_mask() procedure which reads back the current\nport mask from hardware).\n\nThe other solution is to filter out the private VLANs from\nocelot_port_num_untagged_vlans(), since they aren\u0027t what callers care about.\nWe only need to do this to the mentioned function and not to\nocelot_port_num_tagged_vlans(), because private VLANs are never egress-tagged.\n\nNothing else seems to be broken in either solution, but the first one requires\nmore rework which will conflict with the net-next change 36a0bf443585 (\"net:\nmscc: ocelot: set up tag_8021q CPU ports independent of user port affinity\"),\nand I\u0027d like to avoid that. So go with the other one.\n\nFixes: 54c319846086 (\"net: mscc: ocelot: enforce FDB isolation when VLAN-unaware\")\nSigned-off-by: Vladimir Oltean \u003cvladimir.oltean@nxp.com\u003e\nLink: https://lore.kernel.org/r/20220927122042.1100231-1-vladimir.oltean@nxp.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "d59bd748db0a97a5d6a33b284b6c58b7f6f4f768", "tree": "90598965f64b482935193f09ce422d922dc8fd2b", "parents": [ "7cae596bc31f900bb72492ff40c7f5addf72fa19" ], "author": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Wed Sep 28 09:49:28 2022 -0600" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Wed Sep 28 19:27:11 2022 -0600" }, "message": "io_uring/poll: disable level triggered poll\n\nStefan reports that there are issues with the level triggered\nnotification. Since we\u0027re late in the cycle, and it was introduced for\nthe 6.0 release, just disable it at prep time and we can bring this\nback when Samba is happy with it.\n\nReported-by: Stefan Metzmacher \u003cmetze@samba.org\u003e\nReviewed-by: Stefan Metzmacher \u003cmetze@samba.org\u003e\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "7738be973fc4e2ba22154fafd3a5d7b9666f9abf", "tree": "d5149c11dfc508e58f495d0f447bf37e2b87c213", "parents": [ "6ef7d362123ecb5bf6d163bb9c7fd6ba2d8c968c" ], "author": { "name": "Ashutosh Dixit", "email": "ashutosh.dixit@intel.com", "time": "Wed Sep 28 12:02:12 2022 -0700" }, "committer": { "name": "Rodrigo Vivi", "email": "rodrigo.vivi@intel.com", "time": "Wed Sep 28 16:39:26 2022 -0400" }, "message": "drm/i915/gt: Perf_limit_reasons are only available for Gen11+\n\nRegister GT0_PERF_LIMIT_REASONS (0x1381a8) is available only for\nGen11+. Therefore ensure perf_limit_reasons sysfs files are created only\nfor Gen11+. Otherwise on Gen \u003c 5 accessing these files results in the\nfollowing oops:\n\n\u003c1\u003e [88.829420] BUG: unable to handle page fault for address: ffffc90000bb81a8\n\u003c1\u003e [88.829438] #PF: supervisor read access in kernel mode\n\u003c1\u003e [88.829447] #PF: error_code(0x0000) - not-present page\n\nThis patch is a backport of the drm-tip commit 0d2d201095e9\n(\"drm/i915: Perf_limit_reasons are only available for Gen11+\") to\ndrm-intel-fixes. The backport is not identical to the original, it only\nincludes the sysfs portions of if. The debugfs portion is not available\nin drm-intel-fixes so has not been backported.\n\nBspec: 20008\nBug: https://gitlab.freedesktop.org/drm/intel/-/issues/6863\nFixes: fa68bff7cf27 (\"drm/i915/gt: Add sysfs throttle frequency interfaces\")\nSigned-off-by: Ashutosh Dixit \u003cashutosh.dixit@intel.com\u003e\nReviewed-by: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\nSigned-off-by: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20220919162401.2077713-1-ashutosh.dixit@intel.com\n(backported from commit 0d2d201095e9f141d6a9fb44320afce761f8b5c2)\n" }, { "commit": "c3e0e1e23c70455916ff3472072437b3605c6cfe", "tree": "1bffbcf0e3958451b4bba3750e8685328573848b", "parents": [ "e817c070fd9f813774defd8ef5e0349108e924e0", "c0cca6a66458a0daa627774de7ca2b678a6bb3d8" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 28 11:54:57 2022 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 28 11:54:57 2022 -0700" }, "message": "Merge tag \u0027irq_urgent_for_v6.0\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull more irqchip fixes from Borislav Petkov:\n \"More irqchip fixes for 6.0 from Marc Zyngier. Stuff got left hanging\n due to the whole Plumbers and vacations commotion.\n\n - A couple of configuration fixes for the recently merged Loongarch\n drivers\n\n - A fix to avoid dynamic allocation of a cpumask which was causing\n issues with PREEMPT_RT and the GICv3 ITS\n\n - A tightening of an error check in the stm32 exti driver\"\n\n* tag \u0027irq_urgent_for_v6.0\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n irqchip/loongson-pch-lpc: Add dependence on LoongArch\n irqchip: Select downstream irqchip drivers for LoongArch CPU\n irqchip/gic-v3-its: Remove cpumask_var_t allocation\n irqchip/stm32-exti: Remove check on always false condition\n" }, { "commit": "e817c070fd9f813774defd8ef5e0349108e924e0", "tree": "653f2a8b4692f103fcc59d41792a2783d2f3e5dd", "parents": [ "b1cab78ba392162a5ef11173d02db6e182e04edd", "e7afa79a3b35a27a046a2139f8b20bd6b98155c2" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 28 11:36:31 2022 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 28 11:36:31 2022 -0700" }, "message": "Merge tag \u0027mmc-v6.0-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc\n\nPull MMC fixes from Ulf Hansson:\n \"A couple of MMC fixes. This time there is also a fix for the ARM SCMI\n firmware driver, which has been acked by Sudeep Holla, the maintainer.\n\n MMC core:\n - Terminate infinite loop in SD-UHS voltage switch\n\n MMC host:\n - hsq: Fix kernel crash in the recovery path\n - moxart: Fix bus width configurations\n - sdhci: Fix kernel panic for cqe irq\n\n ARM_SCMI:\n - Fixup clock management by reverting \u0027firmware: arm_scmi: Add clock\n management to the SCMI power domain\u0027\"\n\n* tag \u0027mmc-v6.0-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc:\n mmc: hsq: Fix data stomping during mmc recovery\n Revert \"firmware: arm_scmi: Add clock management to the SCMI power domain\"\n mmc: core: Terminate infinite loop in SD-UHS voltage switch\n mmc: moxart: fix 4-bit bus width and remove 8-bit bus width\n mmc: sdhci: Fix host-\u003ecmd is null\n" }, { "commit": "06bbaa6dc53cb72040db952053432541acb9adc7", "tree": "113842569565963748044a4b00498e31051a68c1", "parents": [ "bfbfb6182ad1d7d184b16f25165faad879147f79" ], "author": { "name": "Al Viro", "email": "viro@zeniv.linux.org.uk", "time": "Mon Sep 26 11:59:14 2022 -0400" }, "committer": { "name": "Al Viro", "email": "viro@zeniv.linux.org.uk", "time": "Wed Sep 28 14:28:40 2022 -0400" }, "message": "[coredump] don\u0027t use __kernel_write() on kmap_local_page()\n\npassing kmap_local_page() result to __kernel_write() is unsafe -\nrandom -\u003ewrite_iter() might (and 9p one does) get unhappy when\npassed ITER_KVEC with pointer that came from kmap_local_page().\n\nFix by providing a variant of __kernel_write() that takes an iov_iter\nfrom caller (__kernel_write() becomes a trivial wrapper) and adding\ndump_emit_page() that parallels dump_emit(), except that instead of\n__kernel_write() it uses __kernel_write_iter() with ITER_BVEC source.\n\nFixes: 3159ed57792b \"fs/coredump: use kmap_local_page()\"\nSigned-off-by: Al Viro \u003cviro@zeniv.linux.org.uk\u003e\n" }, { "commit": "af2faee54d591904050d69a55b6e429565bbe389", "tree": "2d0b8ff8cbd043fc2db81c6a64576bdb6b72e153", "parents": [ "44d70bb561dac9363f45787aa93dfca36877ee01", "81d192c2ce74157e717e1fc4b68791f82f7499d4" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Sep 28 10:48:38 2022 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Sep 28 10:48:38 2022 -0700" }, "message": "Merge tag \u0027linux-can-fixes-for-6.0-20220928\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can\n\nMarc Kleine-Budde 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: can 2022-09-28\n\nThe patch is by me and targets the c_can driver. It disables an\noptimization in the TX path of C_CAN cores which causes problems.\n\n* tag \u0027linux-can-fixes-for-6.0-20220928\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can:\n can: c_can: don\u0027t cache TX messages for C_CAN cores\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/20220928090629.1124190-1-mkl@pengutronix.de\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "b1cab78ba392162a5ef11173d02db6e182e04edd", "tree": "8d8eb72f10dead061c6c19e7f230990fe8c3bbff", "parents": [ "49c13ed0316d55d73f1c81c66a7e2abd743f9ae6" ], "author": { "name": "Shakeel Butt", "email": "shakeelb@google.com", "time": "Mon Sep 26 19:59:31 2022 +0000" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Sep 28 09:20:37 2022 -0700" }, "message": "Revert \"net: set proper memcg for net_init hooks allocations\"\n\nThis reverts commit 1d0403d20f6c281cb3d14c5f1db5317caeec48e9.\n\nAnatoly Pugachev reported that the commit 1d0403d20f6c (\"net: set proper\nmemcg for net_init hooks allocations\") is somehow causing the sparc64\nVMs failed to boot and the VMs boot fine with that patch reverted. So,\nrevert the patch for now and later we can debug the issue.\n\nLink: https://lore.kernel.org/all/20220918092849.GA10314@u164.east.ru/\nReported-by: Anatoly Pugachev \u003cmatorola@gmail.com\u003e\nSigned-off-by: Shakeel Butt \u003cshakeelb@google.com\u003e\nCc: Vasily Averin \u003cvvs@openvz.org\u003e\nCc: Jakub Kicinski \u003ckuba@kernel.org\u003e\nCc: Michal Koutný \u003cmkoutny@suse.com\u003e\nCc: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nCc: cgroups@vger.kernel.org\nCc: sparclinux@vger.kernel.org\nCc: linux-mm@kvack.org\nCc: linux-kernel@vger.kernel.org\nTested-by: Anatoly Pugachev \u003cmatorola@gmail.com\u003e\nAcked-by: Johannes Weiner \u003channes@cmpxchg.org\u003e\nFixes: 1d0403d20f6c (\"net: set proper memcg for net_init hooks allocations\")\nReviewed-by: Muchun Song \u003csongmuchun@bytedance.com\u003e\nAcked-by: Roman Gushchin \u003croman.gushchin@linux.dev\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "141f3d6256e58103ece1c3dd2835e871f1dde240", "tree": "14c975e3ab5818be33bf4f59dd2ff3be7407c52e", "parents": [ "6a8438de524346f2ac73b0b493980c336ebce688" ], "author": { "name": "Damien Le Moal", "email": "damien.lemoal@opensource.wdc.com", "time": "Sat Sep 24 15:18:26 2022 +0900" }, "committer": { "name": "Damien Le Moal", "email": "damien.lemoal@opensource.wdc.com", "time": "Wed Sep 28 20:47:31 2022 +0900" }, "message": "ata: libata-sata: Fix device queue depth control\n\nThe function __ata_change_queue_depth() uses the helper\nata_scsi_find_dev() to get the ata_device structure of a scsi device and\nset that device maximum queue depth. However, when the ata device is\nmanaged by libsas, ata_scsi_find_dev() returns NULL, turning\n__ata_change_queue_depth() into a nop, which prevents the user from\nsetting the maximum queue depth of ATA devices used with libsas based\nHBAs.\n\nFix this by renaming __ata_change_queue_depth() to\nata_change_queue_depth() and adding a pointer to the ata_device\nstructure of the target device as argument. This pointer is provided by\nata_scsi_change_queue_depth() using ata_scsi_find_dev() in the case of\na libata managed device and by sas_change_queue_depth() using\nsas_to_ata_dev() in the case of a libsas managed ata device.\n\nSigned-off-by: Damien Le Moal \u003cdamien.lemoal@opensource.wdc.com\u003e\nTested-by: John Garry \u003cjohn.garry@huawei.com\u003e\n" }, { "commit": "6a8438de524346f2ac73b0b493980c336ebce688", "tree": "449bc4f791f4cfca63a2d6a960359876acd85a77", "parents": [ "ea08aec7e77bfd6599489ec430f9f859ab84575a" ], "author": { "name": "Damien Le Moal", "email": "damien.lemoal@opensource.wdc.com", "time": "Sat Sep 24 14:44:11 2022 +0900" }, "committer": { "name": "Damien Le Moal", "email": "damien.lemoal@opensource.wdc.com", "time": "Wed Sep 28 20:47:26 2022 +0900" }, "message": "ata: libata-scsi: Fix initialization of device queue depth\n\nFor SATA devices supporting NCQ, drivers using libsas first initialize a\nscsi device queue depth based on the controller and device capabilities,\nleading to the scsi device queue_depth field being 32 (ATA maximum queue\ndepth) for most setup. However, if libata was loaded using the\nforce\u003d[ID]]noncq argument, the default queue depth should be set to 1 to\nreflect the fact that queuable commands will never be used. This is\nconsistent with manually setting a device queue depth to 1 through sysfs\nas that disables NCQ use for the device.\n\nFix ata_scsi_dev_config() to honor the noncq parameter by sertting the\ndevice queue depth to 1 for devices that do not have the ATA_DFLAG_NCQ\nflag set.\n\nSigned-off-by: Damien Le Moal \u003cdamien.lemoal@opensource.wdc.com\u003e\nTested-by: John Garry \u003cjohn.garry@huawei.com\u003e\n" }, { "commit": "81d192c2ce74157e717e1fc4b68791f82f7499d4", "tree": "2d0b8ff8cbd043fc2db81c6a64576bdb6b72e153", "parents": [ "44d70bb561dac9363f45787aa93dfca36877ee01" ], "author": { "name": "Marc Kleine-Budde", "email": "mkl@pengutronix.de", "time": "Fri Sep 23 13:42:23 2022 +0200" }, "committer": { "name": "Marc Kleine-Budde", "email": "mkl@pengutronix.de", "time": "Wed Sep 28 10:34:04 2022 +0200" }, "message": "can: c_can: don\u0027t cache TX messages for C_CAN cores\n\nAs Jacob noticed, the optimization introduced in 387da6bc7a82 (\"can:\nc_can: cache frames to operate as a true FIFO\") doesn\u0027t properly work\non C_CAN, but on D_CAN IP cores. The exact reasons are still unknown.\n\nFor now disable caching if CAN frames in the TX path for C_CAN cores.\n\nFixes: 387da6bc7a82 (\"can: c_can: cache frames to operate as a true FIFO\")\nLink: https://lore.kernel.org/all/20220928083354.1062321-1-mkl@pengutronix.de\nLink: https://lore.kernel.org/all/15a8084b-9617-2da1-6704-d7e39d60643b@gmail.com\nReported-by: Jacob Kroon \u003cjacob.kroon@gmail.com\u003e\nTested-by: Jacob Kroon \u003cjacob.kroon@gmail.com\u003e\nCc: stable@vger.kernel.org # v5.15\nSigned-off-by: Marc Kleine-Budde \u003cmkl@pengutronix.de\u003e\n" }, { "commit": "44d70bb561dac9363f45787aa93dfca36877ee01", "tree": "0439f04fa6da835ceb892338bd0d6a65a364589e", "parents": [ "797666cd5af041ffb66642fff62f7389f08566a2", "6546646a7fb0d7fe1caef947889497c16aaecc8c" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Sep 27 16:52:45 2022 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Sep 27 16:52:45 2022 -0700" }, "message": "Merge tag \u0027wireless-2022-09-27\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless\n\nJohannes Berg says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nA few late-comer fixes:\n * locking in mac80211 MLME\n * non-QoS driver crash/regression\n * minstrel memory corruption\n * TX deadlock\n * TX queues not always enabled\n * HE/EHT bitrate calculation\n\n* tag \u0027wireless-2022-09-27\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless:\n wifi: mac80211: mlme: Fix double unlock on assoc success handling\n wifi: mac80211: mlme: Fix missing unlock on beacon RX\n wifi: mac80211: fix memory corruption in minstrel_ht_update_rates()\n wifi: mac80211: fix regression with non-QoS drivers\n wifi: mac80211: ensure vif queues are operational after start\n wifi: mac80211: don\u0027t start TX with fq-\u003elock to fix deadlock\n wifi: cfg80211: fix MCS divisor value\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/20220927135923.45312-1-johannes@sipsolutions.net\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "49c13ed0316d55d73f1c81c66a7e2abd743f9ae6", "tree": "f0e37ed7639664f20ec78561bf204dabe1214f1d", "parents": [ "46452d3786a82bd732ba73fb308ae5cbe4e1e591", "7bea67a994300c7dff827d1eb2344c180d1ffd93" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Sep 27 16:49:42 2022 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Sep 27 16:49:42 2022 -0700" }, "message": "Merge tag \u0027soc-fixes-6.0-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc\n\nPull ARM SoC fixes from Arnd Bergmann:\n \"This should be the last set of bugfixes in the SoC tree:\n\n - Two fixes for Arm integrator, dealing with a regression caused by\n invalid DT properties combined with a change in dma address\n translation, and missing device_type annotations on the PCI bus\n\n - Fixes for drivers/reset/, addressing bugs in i.MX8MP, Sparx5 and\n NPCM8XX platforms\n\n - Bjorn Andersson\u0027s email address changes in the MAINTAINERS file\n\n - Multiple minor fixes to Qualcomm dts files, and a change to the\n remoteproc firmware filename that did not match the actual path in\n the linux-firmware package\n\n - Minor code fixes for the Allwinner/sunxi SRAM driver, and the\n broadcom STB Bus Interface Unit driver\n\n - A build fix for the sunplus sp7021 platform\n\n - Two dts fixes for TI OMAP family SoCs, addressing an extraneous\n usb4 device node and an incorrect DMA handle\"\n\n* tag \u0027soc-fixes-6.0-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc:\n ARM: dts: integrator: Fix DMA ranges\n ARM: dts: integrator: Tag PCI host with device_type\n ARM: sunplus: fix serial console kconfig and build problems\n reset: npcm: fix iprst2 and iprst4 setting\n arm64: dts: qcom: sm8350: fix UFS PHY serdes size\n soc: bcm: brcmstb: biuctrl: Avoid double of_node_put()\n arm64: dts: qcom: sc8280xp-x13s: Update firmware location\n soc: sunxi: sram: Fix debugfs info for A64 SRAM C\n soc: sunxi: sram: Fix probe function ordering issues\n soc: sunxi: sram: Prevent the driver from being unbound\n soc: sunxi: sram: Actually claim SRAM regions\n ARM: dts: am5748: keep usb4_tm disabled\n reset: microchip-sparx5: issue a reset on startup\n reset: imx7: Fix the iMX8MP PCIe PHY PERST support\n MAINTAINERS: Update Bjorn\u0027s email address\n arm64: dts: qcom: sc7280: move USB wakeup-source property\n arm64: dts: qcom: thinkpad-x13s: Fix firmware location\n arm64: dts: qcom: sm8150: Fix fastrpc iommu values\n ARM: dts: am33xx: Fix MMCHS0 dma properties\n" }, { "commit": "a43ae8057cc154fd26a3a23c0e8643bef104d995", "tree": "322fb3cf0275d89bdb24516dc537bf806acbe090", "parents": [ "46f8a29272e51b6df7393d58fc5cb8967397ef2b" ], "author": { "name": "Eli Cohen", "email": "elic@nvidia.com", "time": "Mon Sep 12 15:50:19 2022 +0300" }, "committer": { "name": "Michael S. Tsirkin", "email": "mst@redhat.com", "time": "Tue Sep 27 18:32:45 2022 -0400" }, "message": "vdpa/mlx5: Fix MQ to support non power of two num queues\n\nRQT objects require that a power of two value be configured for both\nrqt_max_size and rqt_actual size.\n\nFor create_rqt, make sure to round up to the power of two the value of\ngiven by the user who created the vdpa device and given by\nndev-\u003erqt_size. The actual size is also rounded up to the power of two\nusing the current number of VQs given by ndev-\u003ecur_num_vqs.\n\nSame goes with modify_rqt where we need to make sure act size is power\nof two based on the new number of QPs.\n\nWithout this patch, attempt to create a device with non power of two QPs\nwould result in error from firmware.\n\nFixes: 52893733f2c5 (\"vdpa/mlx5: Add multiqueue support\")\nSigned-off-by: Eli Cohen \u003celic@nvidia.com\u003e\nMessage-Id: \u003c20220912125019.833708-1-elic@nvidia.com\u003e\nSigned-off-by: Michael S. Tsirkin \u003cmst@redhat.com\u003e\n" }, { "commit": "46f8a29272e51b6df7393d58fc5cb8967397ef2b", "tree": "c0b4ecf355de210b969535f63d017c7dec0fe64e", "parents": [ "37fafe6b61e4f15d977982635bb785f4e605f7cd" ], "author": { "name": "Maxime Coquelin", "email": "maxime.coquelin@redhat.com", "time": "Wed Aug 31 17:49:23 2022 +0200" }, "committer": { "name": "Michael S. Tsirkin", "email": "mst@redhat.com", "time": "Tue Sep 27 18:32:45 2022 -0400" }, "message": "vduse: prevent uninitialized memory accesses\n\nIf the VDUSE application provides a smaller config space\nthan the driver expects, the driver may use uninitialized\nmemory from the stack.\n\nThis patch prevents it by initializing the buffer passed by\nthe driver to store the config value.\n\nThis fix addresses CVE-2022-2308.\n\nCc: stable@vger.kernel.org # v5.15+\nFixes: c8a6153b6c59 (\"vduse: Introduce VDUSE - vDPA Device in Userspace\")\nReviewed-by: Xie Yongji \u003cxieyongji@bytedance.com\u003e\nAcked-by: Jason Wang \u003cjasowang@redhat.com\u003e\nSigned-off-by: Maxime Coquelin \u003cmaxime.coquelin@redhat.com\u003e\nMessage-Id: \u003c20220831154923.97809-1-maxime.coquelin@redhat.com\u003e\nSigned-off-by: Michael S. Tsirkin \u003cmst@redhat.com\u003e\nReviewed-by: Chaitanya Kulkarni \u003ckch@nvidia.com\u003e\n" }, { "commit": "37fafe6b61e4f15d977982635bb785f4e605f7cd", "tree": "a0306305742e74b4f9428dc09d8221a01ae81b84", "parents": [ "dbe449d8f8f2ef91af729f0adcc9bef785873168" ], "author": { "name": "Suwan Kim", "email": "suwan.kim027@gmail.com", "time": "Wed Aug 31 00:01:53 2022 +0900" }, "committer": { "name": "Michael S. Tsirkin", "email": "mst@redhat.com", "time": "Tue Sep 27 18:30:49 2022 -0400" }, "message": "virtio-blk: Fix WARN_ON_ONCE in virtio_queue_rq()\n\nIf a request fails at virtio_queue_rqs(), it is inserted to requeue_list\nand passed to virtio_queue_rq(). Then blk_mq_start_request() can be called\nagain at virtio_queue_rq() and trigger WARN_ON_ONCE like below trace because\nrequest state was already set to MQ_RQ_IN_FLIGHT in virtio_queue_rqs()\ndespite the failure.\n\n[ 1.890468] ------------[ cut here ]------------\n[ 1.890776] WARNING: CPU: 2 PID: 122 at block/blk-mq.c:1143\nblk_mq_start_request+0x8a/0xe0\n[ 1.891045] Modules linked in:\n[ 1.891250] CPU: 2 PID: 122 Comm: journal-offline Not tainted 5.19.0+ #44\n[ 1.891504] Hardware name: ChromiumOS crosvm, BIOS 0\n[ 1.891739] RIP: 0010:blk_mq_start_request+0x8a/0xe0\n[ 1.891961] Code: 12 80 74 22 48 8b 4b 10 8b 89 64 01 00 00 8b 53\n20 83 fa ff 75 08 ba 00 00 00 80 0b 53 24 c1 e1 10 09 d1 89 48 34 5b\n41 5e c3 \u003c0f\u003e 0b eb b8 65 8b 05 2b 39 b6 7e 89 c0 48 0f a3 05 39 77 5b\n01 0f\n[ 1.892443] RSP: 0018:ffffc900002777b0 EFLAGS: 00010202\n[ 1.892673] RAX: 0000000000000000 RBX: ffff888004bc0000 RCX: 0000000000000000\n[ 1.892952] RDX: 0000000000000000 RSI: ffff888003d7c200 RDI: ffff888004bc0000\n[ 1.893228] RBP: 0000000000000000 R08: 0000000000000001 R09: ffff888004bc0100\n[ 1.893506] R10: ffffffffffffffff R11: ffffffff8185ca10 R12: ffff888004bc0000\n[ 1.893797] R13: ffffc90000277900 R14: ffff888004ab2340 R15: ffff888003d86e00\n[ 1.894060] FS: 00007ffa143a4640(0000) GS:ffff88807dd00000(0000)\nknlGS:0000000000000000\n[ 1.894412] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033\n[ 1.894682] CR2: 00005648577d9088 CR3: 00000000053da004 CR4: 0000000000170ee0\n[ 1.894953] Call Trace:\n[ 1.895139] \u003cTASK\u003e\n[ 1.895303] virtblk_prep_rq+0x1e5/0x280\n[ 1.895509] virtio_queue_rq+0x5c/0x310\n[ 1.895710] ? virtqueue_add_sgs+0x95/0xb0\n[ 1.895905] ? _raw_spin_unlock_irqrestore+0x16/0x30\n[ 1.896133] ? virtio_queue_rqs+0x340/0x390\n[ 1.896453] ? sbitmap_get+0xfa/0x220\n[ 1.896678] __blk_mq_issue_directly+0x41/0x180\n[ 1.896906] blk_mq_plug_issue_direct+0xd8/0x2c0\n[ 1.897115] blk_mq_flush_plug_list+0x115/0x180\n[ 1.897342] blk_add_rq_to_plug+0x51/0x130\n[ 1.897543] blk_mq_submit_bio+0x3a1/0x570\n[ 1.897750] submit_bio_noacct_nocheck+0x418/0x520\n[ 1.897985] ? submit_bio_noacct+0x1e/0x260\n[ 1.897989] ext4_bio_write_page+0x222/0x420\n[ 1.898000] mpage_process_page_bufs+0x178/0x1c0\n[ 1.899451] mpage_prepare_extent_to_map+0x2d2/0x440\n[ 1.899603] ext4_writepages+0x495/0x1020\n[ 1.899733] do_writepages+0xcb/0x220\n[ 1.899871] ? __seccomp_filter+0x171/0x7e0\n[ 1.900006] file_write_and_wait_range+0xcd/0xf0\n[ 1.900167] ext4_sync_file+0x72/0x320\n[ 1.900308] __x64_sys_fsync+0x66/0xa0\n[ 1.900449] do_syscall_64+0x31/0x50\n[ 1.900595] entry_SYSCALL_64_after_hwframe+0x63/0xcd\n[ 1.900747] RIP: 0033:0x7ffa16ec96ea\n[ 1.900883] Code: b8 4a 00 00 00 0f 05 48 3d 00 f0 ff ff 77 41 c3\n48 83 ec 18 89 7c 24 0c e8 e3 02 f8 ff 8b 7c 24 0c 89 c2 b8 4a 00 00\n00 0f 05 \u003c48\u003e 3d 00 f0 ff ff 77 36 89 d7 89 44 24 0c e8 43 03 f8 ff 8b\n44 24\n[ 1.901302] RSP: 002b:00007ffa143a3ac0 EFLAGS: 00000293 ORIG_RAX:\n000000000000004a\n[ 1.901499] RAX: ffffffffffffffda RBX: 0000560277ec6fe0 RCX: 00007ffa16ec96ea\n[ 1.901696] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000016\n[ 1.901884] RBP: 0000560277ec5910 R08: 0000000000000000 R09: 00007ffa143a4640\n[ 1.902082] R10: 00007ffa16e4d39e R11: 0000000000000293 R12: 00005602773f59e0\n[ 1.902459] R13: 0000000000000000 R14: 00007fffbfc007ff R15: 00007ffa13ba4000\n[ 1.902763] \u003c/TASK\u003e\n[ 1.902877] ---[ end trace 0000000000000000 ]---\n\nTo avoid calling blk_mq_start_request() twice, This patch moves the\nexecution of blk_mq_start_request() to the end of virtblk_prep_rq().\nAnd instead of requeuing failed request to plug list in the error path of\nvirtblk_add_req_batch(), it uses blk_mq_requeue_request() to change failed\nrequest state to MQ_RQ_IDLE. Then virtblk can safely handle the request\non the next trial.\n\nFixes: 0e9911fa768f (\"virtio-blk: support mq_ops-\u003equeue_rqs()\")\nReported-by: Alexandre Courbot \u003cacourbot@chromium.org\u003e\nTested-by: Alexandre Courbot \u003cacourbot@chromium.org\u003e\nSigned-off-by: Suwan Kim \u003csuwan.kim027@gmail.com\u003e\nMessage-Id: \u003c20220830150153.12627-1-suwan.kim027@gmail.com\u003e\nSigned-off-by: Michael S. Tsirkin \u003cmst@redhat.com\u003e\nAcked-by: Stefan Hajnoczi \u003cstefanha@redhat.com\u003e\nReviewed-by: Pankaj Raghav \u003cp.raghav@samsung.com\u003e\n" }, { "commit": "dbe449d8f8f2ef91af729f0adcc9bef785873168", "tree": "59777b2ded3c2e1f5573f1990f9087462cda8ec1", "parents": [ "1bedcf22c081a6e9943f09937b2da8d3ef52d20d" ], "author": { "name": "Xuan Zhuo", "email": "xuanzhuo@linux.alibaba.com", "time": "Tue Aug 30 19:05:49 2022 +0800" }, "committer": { "name": "Michael S. Tsirkin", "email": "mst@redhat.com", "time": "Tue Sep 27 18:30:49 2022 -0400" }, "message": "virtio_test: fixup for vq reset\n\nFix virtio test compilation failure caused by vq reset.\n\n../../drivers/virtio/virtio_ring.c: In function ‘vring_create_virtqueue_packed’:\n../../drivers/virtio/virtio_ring.c:1999:8: error: ‘struct virtqueue’ has no member named ‘reset’\n 1999 | vq-\u003evq.reset \u003d false;\n | ^\n../../drivers/virtio/virtio_ring.c: In function ‘__vring_new_virtqueue’:\n../../drivers/virtio/virtio_ring.c:2493:8: error: ‘struct virtqueue’ has no member named ‘reset’\n 2493 | vq-\u003evq.reset \u003d false;\n | ^\n../../drivers/virtio/virtio_ring.c: In function ‘virtqueue_resize’:\n../../drivers/virtio/virtio_ring.c:2587:18: error: ‘struct virtqueue’ has no member named ‘num_max’\n 2587 | if (num \u003e vq-\u003evq.num_max)\n | ^\n../../drivers/virtio/virtio_ring.c:2596:11: error: ‘struct virtio_device’ has no member named ‘config’\n 2596 | if (!vdev-\u003econfig-\u003edisable_vq_and_reset)\n | ^~\n../../drivers/virtio/virtio_ring.c:2599:11: error: ‘struct virtio_device’ has no member named ‘config’\n 2599 | if (!vdev-\u003econfig-\u003eenable_vq_after_reset)\n | ^~\n../../drivers/virtio/virtio_ring.c:2602:12: error: ‘struct virtio_device’ has no member named ‘config’\n 2602 | err \u003d vdev-\u003econfig-\u003edisable_vq_and_reset(_vq);\n | ^~\n../../drivers/virtio/virtio_ring.c:2614:10: error: ‘struct virtio_device’ has no member named ‘config’\n 2614 | if (vdev-\u003econfig-\u003eenable_vq_after_reset(_vq))\n | ^~\nmake: *** [\u003cbuiltin\u003e: virtio_ring.o] Error 1\n\nSigned-off-by: Xuan Zhuo \u003cxuanzhuo@linux.alibaba.com\u003e\nMessage-Id: \u003c20220830110549.103168-1-xuanzhuo@linux.alibaba.com\u003e\nSigned-off-by: Michael S. Tsirkin \u003cmst@redhat.com\u003e\n" }, { "commit": "1bedcf22c081a6e9943f09937b2da8d3ef52d20d", "tree": "04118a25e3d0dd0779f38965185392aad910b8ca", "parents": [ "db5db1a00d0816207be3a0166fcb4f523eaf3b52" ], "author": { "name": "lei he", "email": "helei.sig11@bytedance.com", "time": "Mon Sep 19 15:51:58 2022 +0800" }, "committer": { "name": "Michael S. Tsirkin", "email": "mst@redhat.com", "time": "Tue Sep 27 18:30:48 2022 -0400" }, "message": "virtio-crypto: fix memory-leak\n\nFix memory-leak for virtio-crypto akcipher request, this problem is\nintroduced by 59ca6c93387d3(virtio-crypto: implement RSA algorithm).\nThe leak can be reproduced and tested with the following script\ninside virtual machine:\n\n#!/bin/bash\n\nLOOP_TIMES\u003d10000\n\n# required module: pkcs8_key_parser, virtio_crypto\nmodprobe pkcs8_key_parser # if CONFIG_PKCS8_PRIVATE_KEY_PARSER\u003dm\nmodprobe virtio_crypto # if CONFIG_CRYPTO_DEV_VIRTIO\u003dm\nrm -rf /tmp/data\ndd if\u003d/dev/random of\u003d/tmp/data count\u003d1 bs\u003d230\n\n# generate private key and self-signed cert\nopenssl req -nodes -x509 -newkey rsa:2048 -keyout key.pem \\\n\t\t-outform der -out cert.der \\\n\t\t-subj \"/C\u003dCN/ST\u003dGD/L\u003dSZ/O\u003dvihoo/OU\u003ddev/CN\u003dalways.com/emailAddress\u003dyy@always.com\"\n# convert private key from pem to der\nopenssl pkcs8 -in key.pem -topk8 -nocrypt -outform DER -out key.der\n\n# add key\nPRIV_KEY_ID\u003d`cat key.der | keyctl padd asymmetric test_priv_key @s`\necho \"priv key id \u003d \"$PRIV_KEY_ID\nPUB_KEY_ID\u003d`cat cert.der | keyctl padd asymmetric test_pub_key @s`\necho \"pub key id \u003d \"$PUB_KEY_ID\n\n# query key\nkeyctl pkey_query $PRIV_KEY_ID 0\nkeyctl pkey_query $PUB_KEY_ID 0\n\n# here we only run pkey_encrypt becasuse it is the fastest interface\nfunction bench_pub() {\n\tkeyctl pkey_encrypt $PUB_KEY_ID 0 /tmp/data enc\u003dpkcs1 \u003e/tmp/enc.pub\n}\n\n# do bench_pub in loop to obtain the memory leak\nfor (( i \u003d 0; i \u003c ${LOOP_TIMES}; ++i )); do\n\tbench_pub\ndone\n\nSigned-off-by: lei he \u003chelei.sig11@bytedance.com\u003e\nAcked-by: Michael S. Tsirkin \u003cmst@redhat.com\u003e\nReviewed-by: Gonglei \u003carei.gonglei@huawei.com\u003e\nMessage-Id: \u003c20220919075158.3625-1-helei.sig11@bytedance.com\u003e\nSigned-off-by: Michael S. Tsirkin \u003cmst@redhat.com\u003e\n" }, { "commit": "db5db1a00d0816207be3a0166fcb4f523eaf3b52", "tree": "46cf3685026075732ecae76a573f70514cf75217", "parents": [ "f76349cf41451c5c42a99f18a9163377e4b364ff" ], "author": { "name": "Angus Chen", "email": "angus.chen@jaguarmicro.com", "time": "Fri Sep 23 17:10:13 2022 +0800" }, "committer": { "name": "Michael S. Tsirkin", "email": "mst@redhat.com", "time": "Tue Sep 27 18:30:48 2022 -0400" }, "message": "vdpa/ifcvf: fix the calculation of queuepair\n\nThe q_pair_id to address a queue pair in the lm bar should be\ncalculated by queue_id / 2 rather than queue_id / nr_vring.\n\nFixes: 2ddae773c93b (\"vDPA/ifcvf: detect and use the onboard number of queues directly\")\nSigned-off-by: Angus Chen \u003cangus.chen@jaguarmicro.com\u003e\nReviewed-by: Jason Wang \u003cjasowang@redhat.com\u003e\nReviewed-by: Michael S. Tsirkin \u003cmst@redhat.com\u003e\nAcked-by: Zhu Lingshan \u003clingshan.zhu@intel.com\u003e\nMessage-Id: \u003c20220923091013.191-1-angus.chen@jaguarmicro.com\u003e\nSigned-off-by: Michael S. Tsirkin \u003cmst@redhat.com\u003e\n" }, { "commit": "3b7329cf5a767c1be38352d43066012e220ad43c", "tree": "7bf5f687a801980a777419990e14a8037874a840", "parents": [ "11e38360cc15e2d5995449d19030b13a67d73cb3" ], "author": { "name": "Bokun Zhang", "email": "Bokun.Zhang@amd.com", "time": "Wed Sep 28 00:30:04 2022 +0800" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 18:03:36 2022 -0400" }, "message": "drm/amdgpu: Add amdgpu suspend-resume code path under SRIOV\n\n- Under SRIOV, we need to send REQ_GPU_FINI to the hypervisor\n during the suspend time. Furthermore, we cannot request a\n mode 1 reset under SRIOV as VF. Therefore, we will skip it\n as it is called in suspend_noirq() function.\n\n- In the resume code path, we need to send REQ_GPU_INIT to the\n hypervisor and also resume PSP IP block under SRIOV.\n\nSigned-off-by: Bokun Zhang \u003cBokun.Zhang@amd.com\u003e\nReviewed-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\nCc: stable@vger.kernel.org\n" }, { "commit": "11e38360cc15e2d5995449d19030b13a67d73cb3", "tree": "7e2fba3b4f5bf2296484a5c852f774f0638d7d4b", "parents": [ "b3e45b18e5c40f1f7e5f6383953343f96f963b13" ], "author": { "name": "Jiadong.Zhu", "email": "Jiadong.Zhu@amd.com", "time": "Fri Sep 23 21:17:53 2022 +0800" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 18:03:09 2022 -0400" }, "message": "drm/amdgpu: Remove fence_process in count_emitted\n\nThe function amdgpu_fence_count_emitted used in work_hander should not call\namdgpu_fence_process which must be used in irq handler.\n\nReviewed-by: Christian König \u003cchristian.koenig@amd.com\u003e\nSigned-off-by: Jiadong.Zhu \u003cJiadong.Zhu@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "b3e45b18e5c40f1f7e5f6383953343f96f963b13", "tree": "f8cdd0fa3818f93432ee73db759a9424d5f037f2", "parents": [ "2d3907c152611a0d65efe54b93972320dcce1565" ], "author": { "name": "Jiadong.Zhu", "email": "Jiadong.Zhu@amd.com", "time": "Thu Sep 15 15:19:20 2022 +0800" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 18:02:58 2022 -0400" }, "message": "drm/amdgpu: Correct the position in patch_cond_exec\n\nThe current position calulated in gfx_v9_0_ring_emit_patch_cond_exec\nunderflows when the wptr is divisible by ring-\u003ebuf_mask + 1.\n\nReviewed-by: Christian König \u003cchristian.koenig@amd.com\u003e\nSigned-off-by: Jiadong.Zhu \u003cJiadong.Zhu@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "2d3907c152611a0d65efe54b93972320dcce1565", "tree": "c4b10f6cf37817c8d52e5a06f15d81202216d41a", "parents": [ "a28acf7091babf4248b042326946cee7736f796a" ], "author": { "name": "Samson Tam", "email": "samson.tam@amd.com", "time": "Fri Sep 09 17:16:32 2022 -0400" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 18:00:01 2022 -0400" }, "message": "drm/amd/display: fill in clock values when DPM is not enabled\n\n[Why]\nFor individual feature testing, PMFW may not report all clock\nvalues back. Driver will default them to 0 but this will\ncause the BB table to be skipped and default to one state\nwith max clocks.\n\n[How]\nAdd helper function to scan through initial clock values and\npopulate them with default clock limits so that BB table\ncan be built.\nAdd dpm_enabled flag to check when DPM is not enabled and\nto trigger helper function.\n\nReviewed-by: Jun Lei \u003cJun.Lei@amd.com\u003e\nAcked-by: Jasdeep Dhillon \u003cjdhillon@amd.com\u003e\nSigned-off-by: Samson Tam \u003csamson.tam@amd.com\u003e\nTested-by: Daniel Wheeler \u003cdaniel.wheeler@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "a28acf7091babf4248b042326946cee7736f796a", "tree": "c00e02c8a6d6b373432a8aa84c6edf9145afb5b3", "parents": [ "8dd2e87dbd49ba0b6afc1dd22a5930d797cf423f" ], "author": { "name": "Taimur Hassan", "email": "Syed.Hassan@amd.com", "time": "Tue Sep 13 18:35:20 2022 -0400" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 17:59:07 2022 -0400" }, "message": "drm/amd/display: Avoid unnecessary pixel rate divider programming\n\n[Why]\nProgramming pixel rate divider when FIFO is enabled can cause FIFO error.\n\n[How]\nSkip divider programming when divider values are the same to prevent FIFO\nerror.\n\nReviewed-by: Alvin Lee \u003calvin.lee2@amd.com\u003e\nAcked-by: Jasdeep Dhillon \u003cjdhillon@amd.com\u003e\nSigned-off-by: Taimur Hassan \u003cSyed.Hassan@amd.com\u003e\nTested-by: Daniel Wheeler \u003cdaniel.wheeler@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "8dd2e87dbd49ba0b6afc1dd22a5930d797cf423f", "tree": "ecdb81046ed0fba62ffdbaad30f52312778b4d96", "parents": [ "391ce355e72bd5713afc85ec5acc955294499dcf" ], "author": { "name": "Eric Bernstein", "email": "eric.bernstein@amd.com", "time": "Mon Mar 21 10:42:34 2022 -0400" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 17:58:48 2022 -0400" }, "message": "drm/amd/display: Remove assert for odm transition case\n\nRemove assert that will hit during odm transition case, since this is a\nvalid case.\n\nSigned-off-by: Eric Bernstein \u003ceric.bernstein@amd.com\u003e\nReviewed-by: Rodrigo Siqueira \u003cRodrigo.Siqueira@amd.com\u003e\nTested-by: Daniel Wheeler \u003cdaniel.wheeler@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "391ce355e72bd5713afc85ec5acc955294499dcf", "tree": "6989182a39fe06e83e56b9ea01a0cda3369cdc84", "parents": [ "1ec5c0e1289c70c8e7dd93a988d945ad5bc3f4b1" ], "author": { "name": "Taimur Hassan", "email": "Syed.Hassan@amd.com", "time": "Wed Sep 14 16:25:31 2022 -0400" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 17:58:07 2022 -0400" }, "message": "drm/amd/display: Fix typo in get_pixel_rate_div\n\n[Why \u0026 How]\nSome FIFO errors still occur due to reading wrong pixel rate divider.\nFix typo to prevent FIFO error.\n\nReviewed-by: Nicholas Kazlauskas \u003cNicholas.Kazlauskas@amd.com\u003e\nAcked-by: Jasdeep Dhillon \u003cjdhillon@amd.com\u003e\nSigned-off-by: Taimur Hassan \u003cSyed.Hassan@amd.com\u003e\nTested-by: Daniel Wheeler \u003cdaniel.wheeler@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "1ec5c0e1289c70c8e7dd93a988d945ad5bc3f4b1", "tree": "b0fa504b8f3dbe1b182e30f69a05538c296831fa", "parents": [ "85096a73f4dc223b5848e41ce40afc5ee913079e" ], "author": { "name": "Aric Cyr", "email": "aric.cyr@amd.com", "time": "Wed Sep 14 12:54:52 2022 -0400" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 17:57:46 2022 -0400" }, "message": "drm/amd/display: Fix audio on display after unplugging another\n\nRevert \"dc: skip audio setup when audio stream is enabled\"\n\nThis reverts commit 65fbfb02c2734cacffec5e3f492e1b4f1dabcf98\n\n[why]\nWe have minimal pipe split transition method to avoid pipe\nallocation outage.However, this method will invoke audio setup\nwhich cause audio output stuck once pipe reallocate.\n\n[how]\nskip audio setup for pipelines which audio stream has been enabled\n\nReviewed-by: Martin Leung \u003cMartin.Leung@amd.com\u003e\nAcked-by: Jasdeep Dhillon \u003cjdhillon@amd.com\u003e\nSigned-off-by: Aric Cyr \u003caric.cyr@amd.com\u003e\nTested-by: Daniel Wheeler \u003cdaniel.wheeler@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "85096a73f4dc223b5848e41ce40afc5ee913079e", "tree": "0b3254b53c073973efdf299583cecec5e72685a1", "parents": [ "ec669ef2e2cf277f3e73d842bb3bda1c3ea2ea73" ], "author": { "name": "Nicholas Kazlauskas", "email": "nicholas.kazlauskas@amd.com", "time": "Thu Sep 08 11:11:47 2022 -0400" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 17:56:51 2022 -0400" }, "message": "drm/amd/display: Add explicit FIFO disable for DP blank\n\n[Why]\nWe rely on DMCUB to do this when disabling the link but it should\nactually come before we disable the DP VID stream.\n\nIf we don\u0027t then the FIFO can end up with underflow that persists\nthe next time it\u0027s enabled.\n\n[How]\nAdd a DCN314 specific blank sequence that will disable the DIG FIFO\nfirst.\n\nReviewed-by: Syed Hassan \u003cSyed.Hassan@amd.com\u003e\nAcked-by: Jasdeep Dhillon \u003cjdhillon@amd.com\u003e\nSigned-off-by: Nicholas Kazlauskas \u003cnicholas.kazlauskas@amd.com\u003e\nTested-by: Daniel Wheeler \u003cdaniel.wheeler@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "ec669ef2e2cf277f3e73d842bb3bda1c3ea2ea73", "tree": "471a59939447f571fe7f0597924426ad3db4a63a", "parents": [ "dd37fba90e72a48998d9f0eb5497088ea2b02ad1" ], "author": { "name": "Nicholas Kazlauskas", "email": "nicholas.kazlauskas@amd.com", "time": "Thu Sep 08 14:17:01 2022 -0400" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 17:56:29 2022 -0400" }, "message": "drm/amd/display: Wrap OTG disable workaround with FIFO control\n\n[Why]\nThe DIO FIFO will underflow if we turn off the OTG before we turn\noff the FIFO.\n\nSince this happens as part of the OTG workaround and we don\u0027t reset\nthe FIFO afterwards we see the error persist.\n\n[How]\nAdd disable FIFO before the disable CRTC and enable FIFO after enabling\nthe CRTC.\n\nReviewed-by: Syed Hassan \u003cSyed.Hassan@amd.com\u003e\nAcked-by: Jasdeep Dhillon \u003cjdhillon@amd.com\u003e\nSigned-off-by: Nicholas Kazlauskas \u003cnicholas.kazlauskas@amd.com\u003e\nTested-by: Daniel Wheeler \u003cdaniel.wheeler@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "dd37fba90e72a48998d9f0eb5497088ea2b02ad1", "tree": "140d113cca74ec639a45c4f5f5888a4935ad7508", "parents": [ "b6d1c39bc637fe4879d610e6687b1d4224c4b2fb" ], "author": { "name": "Nicholas Kazlauskas", "email": "nicholas.kazlauskas@amd.com", "time": "Wed Sep 07 10:11:34 2022 -0400" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 17:55:44 2022 -0400" }, "message": "drm/amd/display: Do DIO FIFO enable after DP video stream enable\n\n[Why]\nAvoids a race condition where DIO FIFO can underflow due to no incoming\ndata available.\n\n[How]\nShift the FIFO enable below stream enable.\n\nMake sure fullness level is written before the DIO reset takes place\nand that we\u0027re not doing it twice.\n\nReviewed-by: Syed Hassan \u003cSyed.Hassan@amd.com\u003e\nAcked-by: Jasdeep Dhillon \u003cjdhillon@amd.com\u003e\nSigned-off-by: Nicholas Kazlauskas \u003cnicholas.kazlauskas@amd.com\u003e\nTested-by: Daniel Wheeler \u003cdaniel.wheeler@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "b6d1c39bc637fe4879d610e6687b1d4224c4b2fb", "tree": "f81d4669f2f654feeedb4e242679a4e85b60aa3c", "parents": [ "783d446749c09eb34178cb8600a9a93d3b99e35d" ], "author": { "name": "Alvin Lee", "email": "Alvin.Lee2@amd.com", "time": "Tue Sep 13 11:06:31 2022 -0400" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 17:55:11 2022 -0400" }, "message": "drm/amd/display: Update DCN32 to use new SR latencies\n\n[Description]\nUpdate to new SR latencies for DCN32\n\nReviewed-by: Nevenko Stupar \u003cNevenko.Stupar@amd.com\u003e\nReviewed-by: Jun Lei \u003cJun.Lei@amd.com\u003e\nAcked-by: Jasdeep Dhillon \u003cjdhillon@amd.com\u003e\nSigned-off-by: Alvin Lee \u003cAlvin.Lee2@amd.com\u003e\nTested-by: Daniel Wheeler \u003cdaniel.wheeler@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "783d446749c09eb34178cb8600a9a93d3b99e35d", "tree": "c5807abd4df70c443fa2cbbf95aadcd455ab513b", "parents": [ "664883ddde67971d59764f2dda855183ecf8bc46" ], "author": { "name": "Taimur Hassan", "email": "Syed.Hassan@amd.com", "time": "Tue Sep 13 16:04:06 2022 -0400" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 17:54:56 2022 -0400" }, "message": "drm/amd/display: Avoid avoid unnecessary pixel rate divider programming\n\n[Why]\nProgramming pixel rate divider when FIFO is enabled can cause FIFO error.\n\n[How]\nSkip divider programming when divider values are the same to prevent FIFO\nerror.\n\nReviewed-by: Nicholas Kazlauskas \u003cNicholas.Kazlauskas@amd.com\u003e\nAcked-by: Jasdeep Dhillon \u003cjdhillon@amd.com\u003e\nSigned-off-by: Taimur Hassan \u003cSyed.Hassan@amd.com\u003e\nTested-by: Daniel Wheeler \u003cdaniel.wheeler@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "664883ddde67971d59764f2dda855183ecf8bc46", "tree": "ac76dc75d58a3afc9248c58802dd0becdd9c9b40", "parents": [ "91ef6cfd30bf75763f41e7a4bc119883fe190606" ], "author": { "name": "Graham Sider", "email": "Graham.Sider@amd.com", "time": "Fri Sep 23 10:07:15 2022 -0400" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 17:54:25 2022 -0400" }, "message": "drm/amdkfd: fix dropped interrupt in kfd_int_process_v11\n\nShader wave interrupts were getting dropped in event_interrupt_wq_v11\nif the PRIV bit was set to 1. This would often lead to a hang. Until\ndebugger logic is upstreamed, expand comment to stop early return.\n\nSigned-off-by: Graham Sider \u003cGraham.Sider@amd.com\u003e\nReviewed-by: Harish Kasiviswanathan \u003cHarish.Kasiviswanathan@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "91ef6cfd30bf75763f41e7a4bc119883fe190606", "tree": "0f0f48eef3c091edfddf31d90832dc0e505796dd", "parents": [ "a9b4700279f21c5c1a3425630dea86d338ab13b2" ], "author": { "name": "Graham Sider", "email": "Graham.Sider@amd.com", "time": "Mon Sep 19 13:57:14 2022 -0400" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 17:54:12 2022 -0400" }, "message": "drm/amdgpu: pass queue size and is_aql_queue to MES\n\nUpdate mes_v11_api_def.h add_queue API with is_aql_queue parameter. Also\nre-use gds_size for the queue size (unused for KFD). MES requires the\nqueue size in order to compute the actual wptr offset within the queue\nRB since it increases monotonically for AQL queues.\n\nv2: Make is_aql_queue assign clearer\n\nSigned-off-by: Graham Sider \u003cGraham.Sider@amd.com\u003e\nReviewed-by: Felix Kuehling \u003cFelix.Kuehling@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "a9b4700279f21c5c1a3425630dea86d338ab13b2", "tree": "545e349e5a3f2c40041bcd4df67b42f233aca9d5", "parents": [ "b9caeadf5cb89566412b94b14a1583419aeba45e" ], "author": { "name": "Graham Sider", "email": "Graham.Sider@amd.com", "time": "Tue Sep 20 17:16:13 2022 -0400" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 17:54:01 2022 -0400" }, "message": "drm/amdkfd: fix MQD init for GFX11 in init_mqd\n\nSet remaining compute_static_thread_mgmt_se* accordingly.\n\nSigned-off-by: Graham Sider \u003cGraham.Sider@amd.com\u003e\nAcked-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\nReviewed-by: Felix Kuehling \u003cFelix.Kuehling@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "b9caeadf5cb89566412b94b14a1583419aeba45e", "tree": "b55448aee8e310f703c37d84a1b31ee8eed17791", "parents": [ "3c480531a4b1652561fd16824c16be8b8a3ba8f1" ], "author": { "name": "Evan Quan", "email": "evan.quan@amd.com", "time": "Thu Sep 01 11:45:02 2022 +0800" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 17:47:07 2022 -0400" }, "message": "drm/amd/pm: use adverse selection for dpm features unsupported by driver\n\nIt\u0027s vbios and pmfw instead of driver who decide whether some dpm features\nis supported or not. Driver just de-selects those features which are not\npermitted on user\u0027s request. Thus, we use adverse selects model.\n\nSigned-off-by: Evan Quan \u003cevan.quan@amd.com\u003e\nReviewed-by: Lijo Lazar \u003clijo.lazar@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "3c480531a4b1652561fd16824c16be8b8a3ba8f1", "tree": "f514945158e78618327d21b6c64a638e579f6003", "parents": [ "7516777434570833469252c16e5fd7d8f73c45be" ], "author": { "name": "Evan Quan", "email": "evan.quan@amd.com", "time": "Wed Aug 03 16:46:47 2022 +0800" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 17:47:00 2022 -0400" }, "message": "drm/amd/pm: enable gfxoff feature for SMU 13.0.0\n\nThe feature is ready with latest 78.58.0 PMFW.\n\nSigned-off-by: Evan Quan \u003cevan.quan@amd.com\u003e\nReviewed-by: Lijo Lazar \u003clijo.lazar@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "7516777434570833469252c16e5fd7d8f73c45be", "tree": "508c0b3ee632c3ed33d0a4702e6f4ef60a572b44", "parents": [ "f76349cf41451c5c42a99f18a9163377e4b364ff" ], "author": { "name": "Evan Quan", "email": "evan.quan@amd.com", "time": "Fri Aug 26 19:30:20 2022 +0800" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Sep 27 17:46:52 2022 -0400" }, "message": "drm/amdgpu: avoid gfx register accessing during gfxoff\n\nMake sure gfxoff is disabled before gfx register accessing.\n\nSigned-off-by: Evan Quan \u003cevan.quan@amd.com\u003e\nReviewed-by: Lijo Lazar \u003clijo.lazar@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "b3056ae2b57858b02b376b3fed6077040caf14b4", "tree": "12bed1da7950ccc5cda023358170b6747097d668", "parents": [ "29322791bc8b4f42fc65734840826e3ddc30921e" ], "author": { "name": "Maciej Fijalkowski", "email": "maciej.fijalkowski@intel.com", "time": "Thu Sep 01 12:40:40 2022 +0200" }, "committer": { "name": "Tony Nguyen", "email": "anthony.l.nguyen@intel.com", "time": "Tue Sep 27 09:01:01 2022 -0700" }, "message": "ice: xsk: drop power of 2 ring size restriction for AF_XDP\n\nWe had multiple customers in the past months that reported commit\n296f13ff3854 (\"ice: xsk: Force rings to be sized to power of 2\")\nmakes them unable to use ring size of 8160 in conjunction with AF_XDP.\nRemove this restriction.\n\nFixes: 296f13ff3854 (\"ice: xsk: Force rings to be sized to power of 2\")\nCC: Alasdair McWilliam \u003calasdair.mcwilliam@outlook.com\u003e\nSigned-off-by: Maciej Fijalkowski \u003cmaciej.fijalkowski@intel.com\u003e\nTested-by: George Kuruvinakunnel \u003cgeorge.kuruvinakunnel@intel.com\u003e\nSigned-off-by: Tony Nguyen \u003canthony.l.nguyen@intel.com\u003e\n" }, { "commit": "29322791bc8b4f42fc65734840826e3ddc30921e", "tree": "3a3b4416a7b01217dfa82d79dc5fade72024de3a", "parents": [ "797666cd5af041ffb66642fff62f7389f08566a2" ], "author": { "name": "Maciej Fijalkowski", "email": "maciej.fijalkowski@intel.com", "time": "Thu Sep 01 12:40:39 2022 +0200" }, "committer": { "name": "Tony Nguyen", "email": "anthony.l.nguyen@intel.com", "time": "Tue Sep 27 08:11:02 2022 -0700" }, "message": "ice: xsk: change batched Tx descriptor cleaning\n\nAF_XDP Tx descriptor cleaning in ice driver currently works in a \"lazy\"\nway - descriptors are not cleaned immediately after send. We rather hold\non with cleaning until we see that free space in ring drops below\nparticular threshold. This was supposed to reduce the amount of\nunnecessary work related to cleaning and instead of keeping the ring\nempty, ring was rather saturated.\n\nIn AF_XDP realm cleaning Tx descriptors implies producing them to CQ.\nThis is a way of letting know user space that particular descriptor has\nbeen sent, as John points out in [0].\n\nWe tried to implement serial descriptor cleaning which would be used in\nconjunction with batched cleaning but it made code base more convoluted\nand probably harder to maintain in future. Therefore we step away from\nbatched cleaning in a current form in favor of an approach where we set\nRS bit on every last descriptor from a batch and clean always at the\nbeginning of ice_xmit_zc().\n\nThis means that we give up a bit of Tx performance, but this doesn\u0027t\nhurt l2fwd scenario which is way more meaningful than txonly as this can\nbe treaten as AF_XDP based packet generator. l2fwd is not hurt due to\nthe fact that Tx side is much faster than Rx and Rx is the one that has\nto catch Tx up.\n\nFWIW Tx descriptors are still produced in a batched way.\n\n[0]: https://lore.kernel.org/bpf/62b0a20232920_3573208ab@john.notmuch/\n\nFixes: 126cdfe1007a (\"ice: xsk: Improve AF_XDP ZC Tx and use batching API\")\nSigned-off-by: Maciej Fijalkowski \u003cmaciej.fijalkowski@intel.com\u003e\nTested-by: George Kuruvinakunnel \u003cgeorge.kuruvinakunnel@intel.com\u003e\nSigned-off-by: Tony Nguyen \u003canthony.l.nguyen@intel.com\u003e\n" }, { "commit": "797666cd5af041ffb66642fff62f7389f08566a2", "tree": "26f2d3522b329a79404ce095c2d5153913182efb", "parents": [ "bc7a319844891746135dc1f34ab9df78d636a3ac" ], "author": { "name": "Frank Wunderlich", "email": "frank-w@public-files.de", "time": "Mon Sep 26 17:07:40 2022 +0200" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Sep 27 07:32:11 2022 -0700" }, "message": "net: usb: qmi_wwan: Add new usb-id for Dell branded EM7455\n\nAdd support for Dell 5811e (EM7455) with USB-id 0x413c:0x81c2.\n\nSigned-off-by: Frank Wunderlich \u003cfrank-w@public-files.de\u003e\nCc: stable@vger.kernel.org\nAcked-by: Bjørn Mork \u003cbjorn@mork.no\u003e\nLink: https://lore.kernel.org/r/20220926150740.6684-3-linux@fw-web.de\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "e62563db857f81d75c5726a35bc0180bed6d1540", "tree": "b28c8636e790e1e3f2a0cfc998796d9c76844d16", "parents": [ "2fd003ee8ade6b7f2777becc6f9917d7c313784f" ], "author": { "name": "Sebastian Krzyszkowiak", "email": "sebastian.krzyszkowiak@puri.sm", "time": "Tue Sep 27 07:15:45 2022 -0700" }, "committer": { "name": "Dmitry Torokhov", "email": "dmitry.torokhov@gmail.com", "time": "Tue Sep 27 07:18:30 2022 -0700" }, "message": "Input: snvs_pwrkey - fix SNVS_HPVIDR1 register address\n\nBoth i.MX6 and i.MX8 reference manuals list 0xBF8 as SNVS_HPVIDR1\n(chapters 57.9 and 6.4.5 respectively).\n\nWithout this, trying to read the revision number results in 0 on\nall revisions, causing the i.MX6 quirk to apply on all platforms,\nwhich in turn causes the driver to synthesise power button release\nevents instead of passing the real one as they happen even on\nplatforms like i.MX8 where that\u0027s not wanted.\n\nFixes: 1a26c920717a (\"Input: snvs_pwrkey - send key events for i.MX6 S, DL and Q\")\nTested-by: Martin Kepplinger \u003cmartin.kepplinger@puri.sm\u003e\nSigned-off-by: Sebastian Krzyszkowiak \u003csebastian.krzyszkowiak@puri.sm\u003e\nReviewed-by: Mattijs Korpershoek \u003cmkorpershoek@baylibre.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\nLink: https://lore.kernel.org/r/4599101.ElGaqSPkdT@pliszka\nSigned-off-by: Dmitry Torokhov \u003cdmitry.torokhov@gmail.com\u003e\n" }, { "commit": "46452d3786a82bd732ba73fb308ae5cbe4e1e591", "tree": "7e30d93366aa69f12886675bb5cfb6911359bf76", "parents": [ "a1375562c0a87f0fa2eaf3e8ce15824696d4170a", "c35fbea48659ec99a4f532c9ee9e8692405afdd0" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Sep 27 07:14:58 2022 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Sep 27 07:14:58 2022 -0700" }, "message": "Merge tag \u0027sound-6.0-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound\n\nPull sound fixes from Takashi Iwai:\n \"A few device-specific fixes, mostly for ASoC. All look small / trivial\n enough\"\n\n* tag \u0027sound-6.0-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:\n ALSA: hda: intel-dsp-config: add missing RaptorLake PCI IDs\n ASoC: tas2770: Reinit regcache on reset\n ASoC: nau8824: Fix semaphore is released unexpectedly\n ASoC: Intel: sof_sdw: add support for Dell SKU 0AFF\n ASoC: imx-card: Fix refcount issue with of_node_put\n ASoC: rt5640: Fix the issue of the abnormal JD2 status\n" }, { "commit": "c0cca6a66458a0daa627774de7ca2b678a6bb3d8", "tree": "a3fc9914daf2ca2fdedc7e4810302affdeb5680b", "parents": [ "57646d6769f13f9484ffc6869c493e25a6568073", "e7ccba7728cff0e0f1299951571f209fcadcb7b1" ], "author": { "name": "Borislav Petkov", "email": "bp@suse.de", "time": "Tue Sep 27 15:29:33 2022 +0200" }, "committer": { "name": "Borislav Petkov", "email": "bp@suse.de", "time": "Tue Sep 27 15:29:33 2022 +0200" }, "message": "Merge tag \u0027irqchip-fixes-6.0-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/maz/arm-platforms into irq/urgent\n\nPull more irqchip fixes for 6.0 from Marc Zyngier:\n\n - A couple of configuration fixes for the recently merged Loongarch drivers\n\n - A fix to avoid dynamic allocation of a cpumask which was causing issues\n with PREEMPT_RT and the GICv3 ITS\n\n - A tightening of an error check in the stm32 exti driver\n\nLink: https://lore.kernel.org/r/20220916085158.2592518-1-maz@kernel.org\n" }, { "commit": "e7afa79a3b35a27a046a2139f8b20bd6b98155c2", "tree": "e1c1ece24aeb1f015f395c691d695171da9b9b24", "parents": [ "3c6656337852e9f1a4079d172f3fddfbf00868f9" ], "author": { "name": "Wenchao Chen", "email": "wenchao.chen@unisoc.com", "time": "Fri Sep 16 17:05:06 2022 +0800" }, "committer": { "name": "Ulf Hansson", "email": "ulf.hansson@linaro.org", "time": "Tue Sep 27 12:38:29 2022 +0200" }, "message": "mmc: hsq: Fix data stomping during mmc recovery\n\nThe block device uses multiple queues to access emmc. There will be up to 3\nrequests in the hsq of the host. The current code will check whether there\nis a request doing recovery before entering the queue, but it will not check\nwhether there is a request when the lock is issued. The request is in recovery\nmode. If there is a request in recovery, then a read and write request is\ninitiated at this time, and the conflict between the request and the recovery\nrequest will cause the data to be trampled.\n\nSigned-off-by: Wenchao Chen \u003cwenchao.chen@unisoc.com\u003e\nFixes: 511ce378e16f (\"mmc: Add MMC host software queue support\")\nCc: stable@vger.kernel.org\nLink: https://lore.kernel.org/r/20220916090506.10662-1-wenchao.chen666@gmail.com\nSigned-off-by: Ulf Hansson \u003culf.hansson@linaro.org\u003e\n" }, { "commit": "bc7a319844891746135dc1f34ab9df78d636a3ac", "tree": "83d238648eb98980f1104e9b8f3f81c78e74a708", "parents": [ "ea64cdfad124922c931633e39287c5a31a9b14a1" ], "author": { "name": "Wang Yufen", "email": "wangyufen@huawei.com", "time": "Fri Sep 23 15:02:37 2022 +0800" }, "committer": { "name": "Paolo Abeni", "email": "pabeni@redhat.com", "time": "Tue Sep 27 11:00:02 2022 +0200" }, "message": "selftests: Fix the if conditions of in test_extra_filter()\n\nThe socket 2 bind the addr in use, bind should fail with EADDRINUSE. So\nif bind success or errno !\u003d EADDRINUSE, testcase should be failed.\n\nFixes: 3ca8e4029969 (\"soreuseport: BPF selection functional test\")\nSigned-off-by: Wang Yufen \u003cwangyufen@huawei.com\u003e\nLink: https://lore.kernel.org/r/1663916557-10730-1-git-send-email-wangyufen@huawei.com\nSigned-off-by: Paolo Abeni \u003cpabeni@redhat.com\u003e\n" }, { "commit": "ea64cdfad124922c931633e39287c5a31a9b14a1", "tree": "b2a567d79f950aaaf9d1900090afb6ccc485e80e", "parents": [ "49725ffc15fc4e9fae68c55b691fd25168cbe5c1" ], "author": { "name": "Lukas Wunner", "email": "lukas@wunner.de", "time": "Fri Sep 23 06:09:52 2022 +0200" }, "committer": { "name": "Paolo Abeni", "email": "pabeni@redhat.com", "time": "Tue Sep 27 10:48:38 2022 +0200" }, "message": "net: phy: Don\u0027t WARN for PHY_UP state in mdio_bus_phy_resume()\n\nCommit 744d23c71af3 (\"net: phy: Warn about incorrect mdio_bus_phy_resume()\nstate\") introduced a WARN() on resume from system sleep if a PHY is not\nin PHY_HALTED state.\n\nCommit 6dbe852c379f (\"net: phy: Don\u0027t WARN for PHY_READY state in\nmdio_bus_phy_resume()\") added an exemption for PHY_READY state from\nthe WARN().\n\nIt turns out PHY_UP state needs to be exempted as well because the\nfollowing may happen on suspend:\n\n mdio_bus_phy_suspend()\n phy_stop_machine()\n phydev-\u003estate \u003d PHY_UP # if (phydev-\u003estate \u003e\u003d PHY_UP)\n\nFixes: 744d23c71af3 (\"net: phy: Warn about incorrect mdio_bus_phy_resume() state\")\nReported-by: Marek Szyprowski \u003cm.szyprowski@samsung.com\u003e\nTested-by: Marek Szyprowski \u003cm.szyprowski@samsung.com\u003e\nLink: https://lore.kernel.org/netdev/2b1a1588-505e-dff3-301d-bfc1fb14d685@samsung.com/\nSigned-off-by: Lukas Wunner \u003clukas@wunner.de\u003e\nAcked-by: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nCc: Xiaolei Wang \u003cxiaolei.wang@windriver.com\u003e\nLink: https://lore.kernel.org/r/8128fdb51eeebc9efbf3776a4097363a1317aaf1.1663905575.git.lukas@wunner.de\nSigned-off-by: Paolo Abeni \u003cpabeni@redhat.com\u003e\n" }, { "commit": "49725ffc15fc4e9fae68c55b691fd25168cbe5c1", "tree": "fc7fab1f93f6ca62e540490ff4ca8d07b48fbbcf", "parents": [ "0dc383796f9c59d7297190aa864b986eaf2f73e6" ], "author": { "name": "Junxiao Chang", "email": "junxiao.chang@intel.com", "time": "Fri Sep 23 13:04:48 2022 +0800" }, "committer": { "name": "Paolo Abeni", "email": "pabeni@redhat.com", "time": "Tue Sep 27 10:38:11 2022 +0200" }, "message": "net: stmmac: power up/down serdes in stmmac_open/release\n\nThis commit fixes DMA engine reset timeout issue in suspend/resume\nwith ADLink I-Pi SMARC Plus board which dmesg shows:\n...\n[ 54.678271] PM: suspend exit\n[ 54.754066] intel-eth-pci 0000:00:1d.2 enp0s29f2: PHY [stmmac-3:01] driver [Maxlinear Ethernet GPY215B] (irq\u003dPOLL)\n[ 54.755808] intel-eth-pci 0000:00:1d.2 enp0s29f2: Register MEM_TYPE_PAGE_POOL RxQ-0\n...\n[ 54.780482] intel-eth-pci 0000:00:1d.2 enp0s29f2: Register MEM_TYPE_PAGE_POOL RxQ-7\n[ 55.784098] intel-eth-pci 0000:00:1d.2: Failed to reset the dma\n[ 55.784111] intel-eth-pci 0000:00:1d.2 enp0s29f2: stmmac_hw_setup: DMA engine initialization failed\n[ 55.784115] intel-eth-pci 0000:00:1d.2 enp0s29f2: stmmac_open: Hw setup failed\n...\n\nThe issue is related with serdes which impacts clock. There is\nserdes in ADLink I-Pi SMARC board ethernet controller. Please refer to\ncommit b9663b7ca6ff78 (\"net: stmmac: Enable SERDES power up/down sequence\")\nfor detial. When issue is reproduced, DMA engine clock is not ready\nbecause serdes is not powered up.\n\nTo reproduce DMA engine reset timeout issue with hardware which has\nserdes in GBE controller, install Ubuntu. In Ubuntu GUI, click\n\"Power Off/Log Out\" -\u003e \"Suspend\" menu, it disables network interface,\nthen goes to sleep mode. When it wakes up, it enables network\ninterface again. Stmmac driver is called in this way:\n\n1. stmmac_release: Stop network interface. In this function, it\n disables DMA engine and network interface;\n2. stmmac_suspend: It is called in kernel suspend flow. But because\n network interface has been disabled(netif_running(ndev) is\n false), it does nothing and returns directly;\n3. System goes into S3 or S0ix state. Some time later, system is\n waken up by keyboard or mouse;\n4. stmmac_resume: It does nothing because network interface has\n been disabled;\n5. stmmac_open: It is called to enable network interace again. DMA\n engine is initialized in this API, but serdes is not power on so\n there will be DMA engine reset timeout issue.\n\nSimilarly, serdes powerdown should be added in stmmac_release.\nNetwork interface might be disabled by cmd \"ifconfig eth0 down\",\nDMA engine, phy and mac have been disabled in ndo_stop callback,\nserdes should be powered down as well. It doesn\u0027t make sense that\nserdes is on while other components have been turned off.\n\nIf ethernet interface is in enabled state(netif_running(ndev) is true)\nbefore suspend/resume, the issue couldn\u0027t be reproduced because serdes\ncould be powered up in stmmac_resume.\n\nBecause serdes_powerup is added in stmmac_open, it doesn\u0027t need to be\ncalled in probe function.\n\nFixes: b9663b7ca6ff78 (\"net: stmmac: Enable SERDES power up/down sequence\")\nSigned-off-by: Junxiao Chang \u003cjunxiao.chang@intel.com\u003e\nReviewed-by: Voon Weifeng \u003cweifeng.voon@intel.com\u003e\nTested-by: Jimmy JS Chen \u003cjimmyjs.chen@adlinktech.com\u003e\nTested-by: Looi, Hong Aun \u003chong.aun.looi@intel.com\u003e\nLink: https://lore.kernel.org/r/20220923050448.1220250-1-junxiao.chang@intel.com\nSigned-off-by: Paolo Abeni \u003cpabeni@redhat.com\u003e\n" }, { "commit": "6546646a7fb0d7fe1caef947889497c16aaecc8c", "tree": "35ea08ffdbee0f2e83dda1225864da757006c9ee", "parents": [ "883b8dc1a8766464d5bde4d97e1d7c795d990d31" ], "author": { "name": "Rafael Mendonca", "email": "rafaelmendsr@gmail.com", "time": "Sun Sep 25 11:34:19 2022 -0300" }, "committer": { "name": "Johannes Berg", "email": "johannes.berg@intel.com", "time": "Tue Sep 27 10:34:45 2022 +0200" }, "message": "wifi: mac80211: mlme: Fix double unlock on assoc success handling\n\nCommit 6911458dc428 (\"wifi: mac80211: mlme: refactor assoc success\nhandling\") moved the per-link setup out of ieee80211_assoc_success() into a\nnew function ieee80211_assoc_config_link() but missed to remove the unlock\nof \u0027sta_mtx\u0027 in case of HE capability/operation missing on HE AP, which\nleads to a double unlock:\n\nieee80211_assoc_success() {\n ...\n ieee80211_assoc_config_link() {\n ...\n if (!(link-\u003eu.mgd.conn_flags \u0026 IEEE80211_CONN_DISABLE_HE) \u0026\u0026\n (!elems-\u003ehe_cap || !elems-\u003ehe_operation)) {\n mutex_unlock(\u0026sdata-\u003elocal-\u003esta_mtx);\n ...\n }\n ...\n }\n ...\n mutex_unlock(\u0026sdata-\u003elocal-\u003esta_mtx);\n ...\n}\n\nFixes: 6911458dc428 (\"wifi: mac80211: mlme: refactor assoc success handling\")\nSigned-off-by: Rafael Mendonca \u003crafaelmendsr@gmail.com\u003e\nLink: https://lore.kernel.org/r/20220925143420.784975-1-rafaelmendsr@gmail.com\nSigned-off-by: Johannes Berg \u003cjohannes.berg@intel.com\u003e\n" }, { "commit": "883b8dc1a8766464d5bde4d97e1d7c795d990d31", "tree": "38e85cd1180ede7584b82ff36531ce5419731d24", "parents": [ "be92292b90bfdc31f332c962882b6d3ea0285fdf" ], "author": { "name": "Rafael Mendonca", "email": "rafaelmendsr@gmail.com", "time": "Sat Sep 24 15:40:41 2022 -0300" }, "committer": { "name": "Johannes Berg", "email": "johannes.berg@intel.com", "time": "Tue Sep 27 10:33:51 2022 +0200" }, "message": "wifi: mac80211: mlme: Fix missing unlock on beacon RX\n\nCommit 98b0b467466c (\"wifi: mac80211: mlme: use correct link_sta\")\nswitched to link station instead of deflink and added some checks to do\nthat, which are done with the \u0027sta_mtx\u0027 mutex held. However, the error\npath of these checks does not unlock \u0027sta_mtx\u0027 before returning.\n\nFixes: 98b0b467466c (\"wifi: mac80211: mlme: use correct link_sta\")\nSigned-off-by: Rafael Mendonca \u003crafaelmendsr@gmail.com\u003e\nLink: https://lore.kernel.org/r/20220924184042.778676-1-rafaelmendsr@gmail.com\nSigned-off-by: Johannes Berg \u003cjohannes.berg@intel.com\u003e\n" }, { "commit": "be92292b90bfdc31f332c962882b6d3ea0285fdf", "tree": "7c409356625af2b08b77943cc499edf86cf5b4d3", "parents": [ "d873697ef2b7e1b6fdd8e9d449d9354bd5d29a4a" ], "author": { "name": "Paweł Lenkow", "email": "pawel.lenkow@camlingroup.com", "time": "Mon Sep 19 17:01:35 2022 +0200" }, "committer": { "name": "Johannes Berg", "email": "johannes.berg@intel.com", "time": "Tue Sep 27 10:33:10 2022 +0200" }, "message": "wifi: mac80211: fix memory corruption in minstrel_ht_update_rates()\n\nDuring our testing of WFM200 module over SDIO on i.MX6Q-based platform,\nwe discovered a memory corruption on the system, tracing back to the wfx\ndriver. Using kfence, it was possible to trace it back to the root\ncause, which is hw-\u003emax_rates set to 8 in wfx_init_common,\nwhile the maximum defined by IEEE80211_TX_TABLE_SIZE is 4.\n\nThis causes array out-of-bounds writes during updates of the rate table,\nas seen below:\n\nBUG: KFENCE: memory corruption in kfree_rcu_work+0x320/0x36c\n\nCorrupted memory at 0xe0a4ffe0 [ 0x03 0x03 0x03 0x03 0x01 0x00 0x00\n0x02 0x02 0x02 0x09 0x00 0x21 0xbb 0xbb 0xbb ] (in kfence-#81):\nkfree_rcu_work+0x320/0x36c\nprocess_one_work+0x3ec/0x920\nworker_thread+0x60/0x7a4\nkthread+0x174/0x1b4\nret_from_fork+0x14/0x2c\n0x0\n\nkfence-#81: 0xe0a4ffc0-0xe0a4ffdf, size\u003d32, cache\u003dkmalloc-64\n\nallocated by task 297 on cpu 0 at 631.039555s:\nminstrel_ht_update_rates+0x38/0x2b0 [mac80211]\nrate_control_tx_status+0xb4/0x148 [mac80211]\nieee80211_tx_status_ext+0x364/0x1030 [mac80211]\nieee80211_tx_status+0xe0/0x118 [mac80211]\nieee80211_tasklet_handler+0xb0/0xe0 [mac80211]\ntasklet_action_common.constprop.0+0x11c/0x148\n__do_softirq+0x1a4/0x61c\nirq_exit+0xcc/0x104\ncall_with_stack+0x18/0x20\n__irq_svc+0x80/0xb0\nwq_worker_sleeping+0x10/0x100\nwq_worker_sleeping+0x10/0x100\nschedule+0x50/0xe0\nschedule_timeout+0x2e0/0x474\nwait_for_completion+0xdc/0x1ec\nmmc_wait_for_req_done+0xc4/0xf8\nmmc_io_rw_extended+0x3b4/0x4ec\nsdio_io_rw_ext_helper+0x290/0x384\nsdio_memcpy_toio+0x30/0x38\nwfx_sdio_copy_to_io+0x88/0x108 [wfx]\nwfx_data_write+0x88/0x1f0 [wfx]\nbh_work+0x1c8/0xcc0 [wfx]\nprocess_one_work+0x3ec/0x920\nworker_thread+0x60/0x7a4\nkthread+0x174/0x1b4\nret_from_fork+0x14/0x2c 0x0\n\nAfter discussion on the wireless mailing list it was clarified\nthat the issue has been introduced by:\ncommit ee0e16ab756a (\"mac80211: minstrel_ht: fill all requested rates\")\nand fix shall be in minstrel_ht_update_rates in rc80211_minstrel_ht.c.\n\nFixes: ee0e16ab756a (\"mac80211: minstrel_ht: fill all requested rates\")\nLink: https://lore.kernel.org/all/12e5adcd-8aed-f0f7-70cc-4fb7b656b829@camlingroup.com/\nLink: https://lore.kernel.org/linux-wireless/20220915131445.30600-1-lech.perczak@camlingroup.com/\nCc: Jérôme Pouiller \u003cjerome.pouiller@silabs.com\u003e\nCc: Johannes Berg \u003cjohannes@sipsolutions.net\u003e\nCc: Peter Seiderer \u003cps.report@gmx.net\u003e\nCc: Kalle Valo \u003ckvalo@kernel.org\u003e\nCc: Krzysztof Drobiński \u003ckrzysztof.drobinski@camlingroup.com\u003e,\nSigned-off-by: Paweł Lenkow \u003cpawel.lenkow@camlingroup.com\u003e\nSigned-off-by: Lech Perczak \u003clech.perczak@camlingroup.com\u003e\nReviewed-by: Peter Seiderer \u003cps.report@gmx.net\u003e\nReviewed-by: Jérôme Pouiller \u003cjerome.pouiller@silabs.com\u003e\nAcked-by: Felix Fietkau \u003cnbd@nbd.name\u003e\nSigned-off-by: Johannes Berg \u003cjohannes.berg@intel.com\u003e\n" }, { "commit": "d873697ef2b7e1b6fdd8e9d449d9354bd5d29a4a", "tree": "795f61f199eb76a79d0e0c32d96bb53354fd58a7", "parents": [ "527008e5e87600a389feb8a57042c928ecca195d" ], "author": { "name": "Hans de Goede", "email": "hdegoede@redhat.com", "time": "Sun Sep 18 21:20:52 2022 +0200" }, "committer": { "name": "Johannes Berg", "email": "johannes.berg@intel.com", "time": "Tue Sep 27 10:32:36 2022 +0200" }, "message": "wifi: mac80211: fix regression with non-QoS drivers\n\nCommit 10cb8e617560 (\"mac80211: enable QoS support for nl80211 ctrl port\")\nchanged ieee80211_tx_control_port() to aways call\n__ieee80211_select_queue() without checking local-\u003ehw.queues.\n\n__ieee80211_select_queue() returns a queue-id between 0 and 3, which means\nthat now ieee80211_tx_control_port() may end up setting the queue-mapping\nfor a skb to a value higher then local-\u003ehw.queues if local-\u003ehw.queues\nis less then 4.\n\nSpecifically this is a problem for ralink rt2500-pci cards where\nlocal-\u003ehw.queues is 2. There this causes rt2x00queue_get_tx_queue() to\nreturn NULL and the following error to be logged: \"ieee80211 phy0:\nrt2x00mac_tx: Error - Attempt to send packet over invalid queue 2\",\nafter which association with the AP fails.\n\nOther callers of __ieee80211_select_queue() skip calling it when\nlocal-\u003ehw.queues \u003c IEEE80211_NUM_ACS, add the same check to\nieee80211_tx_control_port(). This fixes ralink rt2500-pci and\nsimilar cards when less then 4 tx-queues no longer working.\n\nFixes: 10cb8e617560 (\"mac80211: enable QoS support for nl80211 ctrl port\")\nCc: Markus Theil \u003cmarkus.theil@tu-ilmenau.de\u003e\nSuggested-by: Stanislaw Gruszka \u003cstf_xl@wp.pl\u003e\nSigned-off-by: Hans de Goede \u003chdegoede@redhat.com\u003e\nLink: https://lore.kernel.org/r/20220918192052.443529-1-hdegoede@redhat.com\nSigned-off-by: Johannes Berg \u003cjohannes.berg@intel.com\u003e\n" }, { "commit": "527008e5e87600a389feb8a57042c928ecca195d", "tree": "37607409fa25fd21d784a19c896d4d0d6f1f7adb", "parents": [ "b7ce33df1ce2b6631234233c5367bcfe9c67fbe9" ], "author": { "name": "Alexander Wetzel", "email": "alexander@wetzel-home.de", "time": "Thu Sep 15 15:09:46 2022 +0200" }, "committer": { "name": "Johannes Berg", "email": "johannes.berg@intel.com", "time": "Tue Sep 27 10:31:52 2022 +0200" }, "message": "wifi: mac80211: ensure vif queues are operational after start\n\nMake sure local-\u003equeue_stop_reasons and vif.txqs_stopped stay in sync.\n\nWhen a new vif is created the queues may end up in an inconsistent state\nand be inoperable:\nCommunication not using iTXQ will work, allowing to e.g. complete the\nassociation. But the 4-way handshake will time out. The sta will not\nsend out any skbs queued in iTXQs.\n\nAll normal attempts to start the queues will fail when reaching this\nstate.\nlocal-\u003equeue_stop_reasons will have marked all queues as operational but\nvif.txqs_stopped will still be set, creating an inconsistent internal\nstate.\n\nIn reality this seems to be race between the mac80211 function\nieee80211_do_open() setting SDATA_STATE_RUNNING and the wake_txqs_tasklet:\nDepending on the driver and the timing the queues may end up to be\noperational or not.\n\nCc: stable@vger.kernel.org\nFixes: f856373e2f31 (\"wifi: mac80211: do not wake queues on a vif that is being stopped\")\nSigned-off-by: Alexander Wetzel \u003calexander@wetzel-home.de\u003e\nAcked-by: Felix Fietkau \u003cnbd@nbd.name\u003e\nLink: https://lore.kernel.org/r/20220915130946.302803-1-alexander@wetzel-home.de\nSigned-off-by: Johannes Berg \u003cjohannes.berg@intel.com\u003e\n" }, { "commit": "b7ce33df1ce2b6631234233c5367bcfe9c67fbe9", "tree": "3aaea8c0553c5f810cb2fec3a953fd0a2d9b3121", "parents": [ "64e966d1e84b29c9fa916cfeaabbf4013703942e" ], "author": { "name": "Alexander Wetzel", "email": "alexander@wetzel-home.de", "time": "Thu Sep 15 14:41:20 2022 +0200" }, "committer": { "name": "Johannes Berg", "email": "johannes.berg@intel.com", "time": "Tue Sep 27 10:29:04 2022 +0200" }, "message": "wifi: mac80211: don\u0027t start TX with fq-\u003elock to fix deadlock\n\nieee80211_txq_purge() calls fq_tin_reset() and\nieee80211_purge_tx_queue(); Both are then calling\nieee80211_free_txskb(). Which can decide to TX the skb again.\n\nThere are at least two ways to get a deadlock:\n\n1) When we have a TDLS teardown packet queued in either tin or frags\n ieee80211_tdls_td_tx_handle() will call ieee80211_subif_start_xmit()\n while we still hold fq-\u003elock. ieee80211_txq_enqueue() will thus\n deadlock.\n\n2) A variant of the above happens if aggregation is up and running:\n In that case ieee80211_iface_work() will deadlock with the original\n task: The original tasks already holds fq-\u003elock and tries to get\n sta-\u003elock after kicking off ieee80211_iface_work(). But the worker\n can get sta-\u003elock prior to the original task and will then spin for\n fq-\u003elock.\n\nAvoid these deadlocks by not sending out any skbs when called via\nieee80211_free_txskb().\n\nSigned-off-by: Alexander Wetzel \u003calexander@wetzel-home.de\u003e\nLink: https://lore.kernel.org/r/20220915124120.301918-1-alexander@wetzel-home.de\nSigned-off-by: Johannes Berg \u003cjohannes.berg@intel.com\u003e\n" }, { "commit": "64e966d1e84b29c9fa916cfeaabbf4013703942e", "tree": "acdd546d969a8f83b640e1f7ae9915fc3b99990e", "parents": [ "781b80f452fcc1cfc16ee41f12556626a9ced049" ], "author": { "name": "Tamizh Chelvam Raja", "email": "quic_tamizhr@quicinc.com", "time": "Thu Sep 08 23:40:34 2022 +0530" }, "committer": { "name": "Johannes Berg", "email": "johannes.berg@intel.com", "time": "Tue Sep 27 10:26:55 2022 +0200" }, "message": "wifi: cfg80211: fix MCS divisor value\n\nThe Bitrate for HE/EHT MCS6 is calculated wrongly due to the\nincorrect MCS divisor value for mcs6. Fix it with the proper\nvalue.\n\nprevious mcs_divisor value \u003d (11769/6144) \u003d 1.915527\n\nfixed mcs_divisor value \u003d (11377/6144) \u003d 1.851725\n\nFixes: 9c97c88d2f4b (\"cfg80211: Add support to calculate and report 4096-QAM HE rates\")\nSigned-off-by: Tamizh Chelvam Raja \u003cquic_tamizhr@quicinc.com\u003e\nLink: https://lore.kernel.org/r/20220908181034.9936-1-quic_tamizhr@quicinc.com\nSigned-off-by: Johannes Berg \u003cjohannes.berg@intel.com\u003e\n" }, { "commit": "d14c273132aec81a1a8107c9ab4865b89e7910a7", "tree": "c86021bca20a8597bebd4b9bbadd29d6a621ab16", "parents": [ "c292a337d0e45a292c301e3cd51c35aa0ae91e95" ], "author": { "name": "Tina Hsu", "email": "tina_hsu@phison.corp-partner.google.com", "time": "Thu Sep 22 14:16:30 2022 +0800" }, "committer": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Tue Sep 27 09:20:30 2022 +0200" }, "message": "nvme-pci: disable Write Zeroes on Phison E3C/E4C\n\nE3C/E4C SSDs do support the Write Zeroes command in theory, but have very\nbad performance when using it. As the firmware has been frozen for these\nproducts we can not expect firmware improvements for it, so disable\nWrite Zeroes.\n\nSigned-off-by: Tina Hsu \u003ctina_hsu@phison.corp-partner.google.com\u003e\n[hch: update the commit message]\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\n" }, { "commit": "c292a337d0e45a292c301e3cd51c35aa0ae91e95", "tree": "299f0901115dc602e4e6043bc9d4cd0ad85d33bf", "parents": [ "4c66a326b5ab784cddd72de07ac5b6210e9e1b06" ], "author": { "name": "Michael Kelley", "email": "mikelley@microsoft.com", "time": "Thu Sep 22 21:49:09 2022 -0700" }, "committer": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Tue Sep 27 09:20:29 2022 +0200" }, "message": "nvme: Fix IOC_PR_CLEAR and IOC_PR_RELEASE ioctls for nvme devices\n\nThe IOC_PR_CLEAR and IOC_PR_RELEASE ioctls are\nnon-functional on NVMe devices because the nvme_pr_clear()\nand nvme_pr_release() functions set the IEKEY field incorrectly.\nThe IEKEY field should be set only when the key is zero (i.e,\nnot specified). The current code does it backwards.\n\nFurthermore, the NVMe spec describes the persistent\nreservation \"clear\" function as an option on the reservation\nrelease command. The current implementation of nvme_pr_clear()\nerroneously uses the reservation register command.\n\nFix these errors. Note that NVMe version 1.3 and later specify\nthat setting the IEKEY field will return an error of Invalid\nField in Command. The fix will set IEKEY when the key is zero,\nwhich is appropriate as these ioctls consider a zero key to\nbe \"unspecified\", and the intention of the spec change is\nto require a valid key.\n\nTested on a version 1.4 PCI NVMe device in an Azure VM.\n\nFixes: 1673f1f08c88 (\"nvme: move block_device_operations and ns/ctrl freeing to common code\")\nFixes: 1d277a637a71 (\"NVMe: Add persistent reservation ops\")\nSigned-off-by: Michael Kelley \u003cmikelley@microsoft.com\u003e\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\n" }, { "commit": "ea08aec7e77bfd6599489ec430f9f859ab84575a", "tree": "8f9a3c77ce210c9324d2baa3eb90e7469233f1a4", "parents": [ "a357f7b4583ebf81d19c95aef57497ae81c5f63c" ], "author": { "name": "Niklas Cassel", "email": "niklas.cassel@wdc.com", "time": "Mon Sep 26 18:38:09 2022 +0000" }, "committer": { "name": "Damien Le Moal", "email": "damien.lemoal@opensource.wdc.com", "time": "Tue Sep 27 08:20:37 2022 +0900" }, "message": "libata: add ATA_HORKAGE_NOLPM for Pioneer BDR-207M and BDR-205\n\nCommit 1527f69204fe (\"ata: ahci: Add Green Sardine vendor ID as\nboard_ahci_mobile\") added an explicit entry for AMD Green Sardine\nAHCI controller using the board_ahci_mobile configuration (this\nconfiguration has later been renamed to board_ahci_low_power).\n\nThe board_ahci_low_power configuration enables support for low power\nmodes.\n\nThis explicit entry takes precedence over the generic AHCI controller\nentry, which does not enable support for low power modes.\n\nTherefore, when commit 1527f69204fe (\"ata: ahci: Add Green Sardine\nvendor ID as board_ahci_mobile\") was backported to stable kernels,\nit make some Pioneer optical drives, which was working perfectly fine\nbefore the commit was backported, stop working.\n\nThe real problem is that the Pioneer optical drives do not handle low\npower modes correctly. If these optical drives would have been tested\non another AHCI controller using the board_ahci_low_power configuration,\nthis issue would have been detected earlier.\n\nUnfortunately, the board_ahci_low_power configuration is only used in\nless than 15% of the total AHCI controller entries, so many devices\nhave never been tested with an AHCI controller with low power modes.\n\nFixes: 1527f69204fe (\"ata: ahci: Add Green Sardine vendor ID as board_ahci_mobile\")\nCc: stable@vger.kernel.org\nReported-by: Jaap Berkhout \u003cj.j.berkhout@staalenberk.nl\u003e\nSigned-off-by: Niklas Cassel \u003cniklas.cassel@wdc.com\u003e\nReviewed-by: Mario Limonciello \u003cmario.limonciello@amd.com\u003e\nSigned-off-by: Damien Le Moal \u003cdamien.lemoal@opensource.wdc.com\u003e\n" }, { "commit": "a1375562c0a87f0fa2eaf3e8ce15824696d4170a", "tree": "c56e40167a7531d26d19433566533b7510a6f9ba", "parents": [ "3800a713b6070d4f03fb43613a7b7d536a99b2b7", "e400ad8b7e6a1b9102123c6240289a811501f7d9" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Sep 26 14:53:38 2022 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Sep 26 14:53:38 2022 -0700" }, "message": "Merge tag \u0027x86_urgent_for_v6.0-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull x86 fixes from Dave Hansen:\n\n - A performance fix for recent large AMD systems that avoids an ancient\n cpu idle hardware workaround\n\n - A new Intel model number. Folks like these upstream as soon as\n possible so that each developer doing feature development doesn\u0027t\n need to carry their own #define\n\n - SGX fixes for a userspace crash and a rare kernel warning\n\n* tag \u0027x86_urgent_for_v6.0-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n ACPI: processor idle: Practically limit \"Dummy wait\" workaround to old Intel systems\n x86/sgx: Handle VA page allocation failure for EAUG on PF.\n x86/sgx: Do not fail on incomplete sanitization on premature stop of ksgxd\n x86/cpu: Add CPU model numbers for Meteor Lake\n" }, { "commit": "7bea67a994300c7dff827d1eb2344c180d1ffd93", "tree": "99dd15bd12b36b229a4652ba0d74aee7cd8d3410", "parents": [ "2653853e2a2a21a9b87d976044863cf776035b8e" ], "author": { "name": "Linus Walleij", "email": "linus.walleij@linaro.org", "time": "Mon Sep 26 09:33:11 2022 +0200" }, "committer": { "name": "Arnd Bergmann", "email": "arnd@arndb.de", "time": "Mon Sep 26 23:05:58 2022 +0200" }, "message": "ARM: dts: integrator: Fix DMA ranges\n\nA recent change affecting the behaviour of phys_to_dma() to\nactually require the device tree ranges to work unmasked a\nbug in the Integrator DMA ranges.\n\nThe PL110 uses the CMA allocator to obtain coherent allocations\nfrom a dedicated 1MB video memory, leading to the following\ncall chain:\n\ndrm_gem_cma_create()\n dma_alloc_attrs()\n dma_alloc_from_dev_coherent()\n __dma_alloc_from_coherent()\n dma_get_device_base()\n phys_to_dma()\n translate_phys_to_dma()\n\nphys_to_dma() by way of translate_phys_to_dma() will nowadays not\nprovide 1:1 mappings unless the ranges are properly defined in\nthe device tree and reflected into the dev-\u003edma_range_map.\n\nThere is a bug in the device trees because the DMA ranges are\nincorrectly specified, and the patch uncovers this bug.\n\nSolution:\n\n- Fix the LB (logic bus) ranges to be 1-to-1 like they should\n have always been.\n- Provide a 1:1 dma-ranges attribute to the PL110.\n- Mark the PL110 display controller as DMA coherent.\n\nThis makes the DMA ranges work right and makes the PL110\nframebuffer work again.\n\nFixes: af6f23b88e95 (\"ARM/dma-mapping: use the generic versions of dma_to_phys/phys_to_dma by default\")\nSigned-off-by: Linus Walleij \u003clinus.walleij@linaro.org\u003e\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Arnd Bergmann \u003carnd@arndb.de\u003e\nCc: stable@vger.kernel.org\nLink: https://lore.kernel.org/r/20220926073311.1610568-1-linus.walleij@linaro.org\u0027\nSigned-off-by: Arnd Bergmann \u003carnd@arndb.de\u003e\n" }, { "commit": "3800a713b6070d4f03fb43613a7b7d536a99b2b7", "tree": "28250666422f7316cb2434a985137efa90023a3e", "parents": [ "3a710532cfe586710738670fb7e1ee44fc3c9b5c", "59298997df89e19aad426d4ae0a7e5037074da5a" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Sep 26 13:23:15 2022 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Sep 26 13:23:15 2022 -0700" }, "message": "Merge tag \u0027mm-hotfixes-stable-2022-09-26\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm\n\nPull last (?) hotfixes from Andrew Morton:\n \"26 hotfixes.\n\n 8 are for issues which were introduced during this -rc cycle, 18 are\n for earlier issues, and are cc:stable\"\n\n* tag \u0027mm-hotfixes-stable-2022-09-26\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm: (26 commits)\n x86/uaccess: avoid check_object_size() in copy_from_user_nmi()\n mm/page_isolation: fix isolate_single_pageblock() isolation behavior\n mm,hwpoison: check mm when killing accessing process\n mm/hugetlb: correct demote page offset logic\n mm: prevent page_frag_alloc() from corrupting the memory\n mm: bring back update_mmu_cache() to finish_fault()\n frontswap: don\u0027t call -\u003einit if no ops are registered\n mm/huge_memory: use pfn_to_online_page() in split_huge_pages_all()\n mm: fix madivse_pageout mishandling on non-LRU page\n powerpc/64s/radix: don\u0027t need to broadcast IPI for radix pmd collapse flush\n mm: gup: fix the fast GUP race against THP collapse\n mm: fix dereferencing possible ERR_PTR\n vmscan: check folio_test_private(), not folio_get_private()\n mm: fix VM_BUG_ON in __delete_from_swap_cache()\n tools: fix compilation after gfp_types.h split\n mm/damon/dbgfs: fix memory leak when using debugfs_lookup()\n mm/migrate_device.c: copy pte dirty bit to page\n mm/migrate_device.c: add missing flush_cache_page()\n mm/migrate_device.c: flush TLB while holding PTL\n x86/mm: disable instrumentations of mm/pgprot.c\n ...\n" }, { "commit": "0dc383796f9c59d7297190aa864b986eaf2f73e6", "tree": "f7a45fc3c6f5de68ab04494080a734e5cf379e6c", "parents": [ "4774db8dfc6a2e6649920ebb2fc8e2f062c2080d" ], "author": { "name": "ruanjinjie", "email": "ruanjinjie@huawei.com", "time": "Fri Sep 23 17:43:20 2022 +0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Mon Sep 26 13:22:39 2022 -0700" }, "message": "net: hippi: Add missing pci_disable_device() in rr_init_one()\n\nAdd missing pci_disable_device() if rr_init_one() fails\n\nSigned-off-by: ruanjinjie \u003cruanjinjie@huawei.com\u003e\nLink: https://lore.kernel.org/r/20220923094320.3109154-1-ruanjinjie@huawei.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "4774db8dfc6a2e6649920ebb2fc8e2f062c2080d", "tree": "008f299193ea46f890bfc5ff2f55d3b53f54fd87", "parents": [ "c635ebe8d911a93bd849a9419b01a58783de76f1" ], "author": { "name": "Peng Wu", "email": "wupeng58@huawei.com", "time": "Fri Sep 23 02:36:40 2022 +0000" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Mon Sep 26 13:20:23 2022 -0700" }, "message": "net/mlxbf_gige: Fix an IS_ERR() vs NULL bug in mlxbf_gige_mdio_probe\n\nThe devm_ioremap() function returns NULL on error, it doesn\u0027t return\nerror pointers.\n\nFixes: 3a1a274e933f (\"mlxbf_gige: compute MDIO period based on i1clk\")\nSigned-off-by: Peng Wu \u003cwupeng58@huawei.com\u003e\nLink: https://lore.kernel.org/r/20220923023640.116057-1-wupeng58@huawei.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "c635ebe8d911a93bd849a9419b01a58783de76f1", "tree": "4df3f44d53e6e48622658155e07fc2391c50fc16", "parents": [ "6e23ec0ba92d426c77a73a9ccab16346e5e0ef49" ], "author": { "name": "Rafael Mendonca", "email": "rafaelmendsr@gmail.com", "time": "Thu Sep 22 14:51:08 2022 -0300" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Mon Sep 26 13:17:53 2022 -0700" }, "message": "cxgb4: fix missing unlock on ETHOFLD desc collect fail path\n\nThe label passed to the QDESC_GET for the ETHOFLD TXQ, RXQ, and FLQ, is the\n\u0027out\u0027 one, which skips the \u0027out_unlock\u0027 label, and thus doesn\u0027t unlock the\n\u0027uld_mutex\u0027 before returning. Additionally, since commit 5148e5950c67\n(\"cxgb4: add EOTID tracking and software context dump\"), the access to\nthese ETHOFLD hardware queues should be protected by the \u0027mqprio_mutex\u0027\ninstead.\n\nFixes: 2d0cb84dd973 (\"cxgb4: add ETHOFLD hardware queue support\")\nFixes: 5148e5950c67 (\"cxgb4: add EOTID tracking and software context dump\")\nSigned-off-by: Rafael Mendonca \u003crafaelmendsr@gmail.com\u003e\nReviewed-by: Rahul Lakkireddy \u003crahul.lakkireddy@chelsio.com\u003e\nLink: https://lore.kernel.org/r/20220922175109.764898-1-rafaelmendsr@gmail.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "3a710532cfe586710738670fb7e1ee44fc3c9b5c", "tree": "51c69189d3c164f85fccecc237de2b3caa528097", "parents": [ "f76349cf41451c5c42a99f18a9163377e4b364ff", "a078dff870136090b5779ca2831870a6c5539d36" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Sep 26 13:10:11 2022 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Sep 26 13:10:11 2022 -0700" }, "message": "Merge tag \u0027ext4_for_linus_fixes2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4\n\nPull missed ext4 fix from Ted Ts\u0027o:\n \"Fix an potential unitialzied variable bug; this was a fixup that I had\n forgotten to apply before the last pull request for ext4. My bad\"\n\n* tag \u0027ext4_for_linus_fixes2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:\n ext4: fixup possible uninitialized variable access in ext4_mb_choose_next_group_cr1()\n" }, { "commit": "6e23ec0ba92d426c77a73a9ccab16346e5e0ef49", "tree": "6d79f7e1c215272c2ca3f7b6bbad5f9c5e9f848e", "parents": [ "a43206156263fbaf1f2b7f96257441f331e91bb7" ], "author": { "name": "Hangyu Hua", "email": "hbh25y@gmail.com", "time": "Fri Sep 23 10:00:46 2022 +0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Mon Sep 26 12:40:39 2022 -0700" }, "message": "net: sched: act_ct: fix possible refcount leak in tcf_ct_init()\n\nnf_ct_put need to be called to put the refcount got by tcf_ct_fill_params\nto avoid possible refcount leak when tcf_ct_flow_table_get fails.\n\nFixes: c34b961a2492 (\"net/sched: act_ct: Create nf flow table per zone\")\nSigned-off-by: Hangyu Hua \u003chbh25y@gmail.com\u003e\nLink: https://lore.kernel.org/r/20220923020046.8021-1-hbh25y@gmail.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "59298997df89e19aad426d4ae0a7e5037074da5a", "tree": "bd519e75ebd1570b528f9c7e3e2741c7d11055ea", "parents": [ "80e2b584f3abfc31c3fe5573007f0d1d10810fde" ], "author": { "name": "Kees Cook", "email": "keescook@chromium.org", "time": "Mon Sep 19 13:16:48 2022 -0700" }, "committer": { "name": "Andrew Morton", "email": "akpm@linux-foundation.org", "time": "Mon Sep 26 12:14:35 2022 -0700" }, "message": "x86/uaccess: avoid check_object_size() in copy_from_user_nmi()\n\nThe check_object_size() helper under CONFIG_HARDENED_USERCOPY is designed\nto skip any checks where the length is known at compile time as a\nreasonable heuristic to avoid \"likely known-good\" cases. However, it can\nonly do this when the copy_*_user() helpers are, themselves, inline too.\n\nUsing find_vmap_area() requires taking a spinlock. The\ncheck_object_size() helper can call find_vmap_area() when the destination\nis in vmap memory. If show_regs() is called in interrupt context, it will\nattempt a call to copy_from_user_nmi(), which may call check_object_size()\nand then find_vmap_area(). If something in normal context happens to be\nin the middle of calling find_vmap_area() (with the spinlock held), the\ninterrupt handler will hang forever.\n\nThe copy_from_user_nmi() call is actually being called with a fixed-size\nlength, so check_object_size() should never have been called in the first\nplace. Given the narrow constraints, just replace the\n__copy_from_user_inatomic() call with an open-coded version that calls\nonly into the sanitizers and not check_object_size(), followed by a call\nto raw_copy_from_user().\n\n[akpm@linux-foundation.org: no instrument_copy_from_user() in my tree...]\nLink: https://lkml.kernel.org/r/20220919201648.2250764-1-keescook@chromium.org\nLink: https://lore.kernel.org/all/CAOUHufaPshtKrTWOz7T7QFYUNVGFm0JBjvM700Nhf9qEL9b3EQ@mail.gmail.com\nFixes: 0aef499f3172 (\"mm/usercopy: Detect vmalloc overruns\")\nSigned-off-by: Kees Cook \u003ckeescook@chromium.org\u003e\nReported-by: Yu Zhao \u003cyuzhao@google.com\u003e\nReported-by: Florian Lehner \u003cdev@der-flo.net\u003e\nSuggested-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nAcked-by: Peter Zijlstra (Intel) \u003cpeterz@infradead.org\u003e\nTested-by: Florian Lehner \u003cdev@der-flo.net\u003e\nCc: Matthew Wilcox \u003cwilly@infradead.org\u003e\nCc: Josh Poimboeuf \u003cjpoimboe@kernel.org\u003e\nCc: Dave Hansen \u003cdave.hansen@linux.intel.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\n" }, { "commit": "80e2b584f3abfc31c3fe5573007f0d1d10810fde", "tree": "583b75d0f740677022c1221491672e99dff4f999", "parents": [ "77677cdbc2aa4b5d5d839562793d3d126201d18d" ], "author": { "name": "Zi Yan", "email": "ziy@nvidia.com", "time": "Tue Sep 13 22:39:13 2022 -0400" }, "committer": { "name": "Andrew Morton", "email": "akpm@linux-foundation.org", "time": "Mon Sep 26 12:14:34 2022 -0700" }, "message": "mm/page_isolation: fix isolate_single_pageblock() isolation behavior\n\nset_migratetype_isolate() does not allow isolating MIGRATE_CMA pageblocks\nunless it is used for CMA allocation. isolate_single_pageblock() did not\nhave the same behavior when it is used together with\nset_migratetype_isolate() in start_isolate_page_range(). This allows\nalloc_contig_range() with migratetype other than MIGRATE_CMA, like\nMIGRATE_MOVABLE (used by alloc_contig_pages()), to isolate first and last\npageblock but fail the rest. The failure leads to changing migratetype of\nthe first and last pageblock to MIGRATE_MOVABLE from MIGRATE_CMA,\ncorrupting the CMA region. This can happen during gigantic page\nallocations.\n\nLike Doug said here:\nhttps://lore.kernel.org/linux-mm/a3363a52-883b-dcd1-b77f-f2bb378d6f2d@gmail.com/T/#u,\nfor gigantic page allocations, the user would notice no difference,\nsince the allocation on CMA region will fail as well as it did before. \nBut it might hurt the performance of device drivers that use CMA, since\nCMA region size decreases.\n\nFix it by passing migratetype into isolate_single_pageblock(), so that\nset_migratetype_isolate() used by isolate_single_pageblock() will prevent\nthe isolation happening.\n\nLink: https://lkml.kernel.org/r/20220914023913.1855924-1-zi.yan@sent.com\nFixes: b2c9e2fbba32 (\"mm: make alloc_contig_range work at pageblock granularity\")\nSigned-off-by: Zi Yan \u003cziy@nvidia.com\u003e\nReported-by: Doug Berger \u003copendmb@gmail.com\u003e\nCc: David Hildenbrand \u003cdavid@redhat.com\u003e\nCc: Doug Berger \u003copendmb@gmail.com\u003e\nCc: Mike Kravetz \u003cmike.kravetz@oracle.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\n" }, { "commit": "77677cdbc2aa4b5d5d839562793d3d126201d18d", "tree": "aa8a70e7ea30366a2e207000f3923e01515e7cbd", "parents": [ "317314527d173e1f139ceaf8cb87cb1746abf240" ], "author": { "name": "Shuai Xue", "email": "xueshuai@linux.alibaba.com", "time": "Wed Sep 14 14:49:35 2022 +0800" }, "committer": { "name": "Andrew Morton", "email": "akpm@linux-foundation.org", "time": "Mon Sep 26 12:14:34 2022 -0700" }, "message": "mm,hwpoison: check mm when killing accessing process\n\nThe GHES code calls memory_failure_queue() from IRQ context to queue work\ninto workqueue and schedule it on the current CPU. Then the work is\nprocessed in memory_failure_work_func() by kworker and calls\nmemory_failure().\n\nWhen a page is already poisoned, commit a3f5d80ea401 (\"mm,hwpoison: send\nSIGBUS with error virutal address\") make memory_failure() call\nkill_accessing_process() that:\n\n - holds mmap locking of current-\u003emm\n - does pagetable walk to find the error virtual address\n - and sends SIGBUS to the current process with error info.\n\nHowever, the mm of kworker is not valid, resulting in a null-pointer\ndereference. So check mm when killing the accessing process.\n\n[akpm@linux-foundation.org: remove unrelated whitespace alteration]\nLink: https://lkml.kernel.org/r/20220914064935.7851-1-xueshuai@linux.alibaba.com\nFixes: a3f5d80ea401 (\"mm,hwpoison: send SIGBUS with error virutal address\")\nSigned-off-by: Shuai Xue \u003cxueshuai@linux.alibaba.com\u003e\nReviewed-by: Miaohe Lin \u003clinmiaohe@huawei.com\u003e\nAcked-by: Naoya Horiguchi \u003cnaoya.horiguchi@nec.com\u003e\nCc: Huang Ying \u003cying.huang@intel.com\u003e\nCc: Baolin Wang \u003cbaolin.wang@linux.alibaba.com\u003e\nCc: Bixuan Cui \u003ccuibixuan@linux.alibaba.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\n" }, { "commit": "317314527d173e1f139ceaf8cb87cb1746abf240", "tree": "f244845d33ab1e283634f49a5ef224ee85aa2700", "parents": [ "dac22531bbd4af2426c4e29e05594415ccfa365d" ], "author": { "name": "Doug Berger", "email": "opendmb@gmail.com", "time": "Wed Sep 14 12:09:17 2022 -0700" }, "committer": { "name": "Andrew Morton", "email": "akpm@linux-foundation.org", "time": "Mon Sep 26 12:14:34 2022 -0700" }, "message": "mm/hugetlb: correct demote page offset logic\n\nWith gigantic pages it may not be true that struct page structures are\ncontiguous across the entire gigantic page. The nth_page macro is used\nhere in place of direct pointer arithmetic to correct for this.\n\nMike said:\n\n: This error could cause addressing exceptions. However, this is only\n: possible in configurations where CONFIG_SPARSEMEM \u0026\u0026\n: !CONFIG_SPARSEMEM_VMEMMAP. Such a configuration option is rare and\n: unknown to be the default anywhere.\n\nLink: https://lkml.kernel.org/r/20220914190917.3517663-1-opendmb@gmail.com\nFixes: 8531fc6f52f5 (\"hugetlb: add hugetlb demote page support\")\nSigned-off-by: Doug Berger \u003copendmb@gmail.com\u003e\nReviewed-by: Mike Kravetz \u003cmike.kravetz@oracle.com\u003e\nReviewed-by: Oscar Salvador \u003cosalvador@suse.de\u003e\nReviewed-by: Anshuman Khandual \u003canshuman.khandual@arm.com\u003e\nCc: Muchun Song \u003csongmuchun@bytedance.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\n" }, { "commit": "dac22531bbd4af2426c4e29e05594415ccfa365d", "tree": "a72a209c803887d244c57b5c72d20a4d33b3419d", "parents": [ "70427f6e9ecfc8c5f977b21dd9f846b3bda02500" ], "author": { "name": "Maurizio Lombardi", "email": "mlombard@redhat.com", "time": "Fri Jul 15 14:50:13 2022 +0200" }, "committer": { "name": "Andrew Morton", "email": "akpm@linux-foundation.org", "time": "Mon Sep 26 12:14:34 2022 -0700" }, "message": "mm: prevent page_frag_alloc() from corrupting the memory\n\nA number of drivers call page_frag_alloc() with a fragment\u0027s size \u003e\nPAGE_SIZE.\n\nIn low memory conditions, __page_frag_cache_refill() may fail the order\n3 cache allocation and fall back to order 0; In this case, the cache\nwill be smaller than the fragment, causing memory corruptions.\n\nPrevent this from happening by checking if the newly allocated cache is\nlarge enough for the fragment; if not, the allocation will fail and\npage_frag_alloc() will return NULL.\n\nLink: https://lkml.kernel.org/r/20220715125013.247085-1-mlombard@redhat.com\nFixes: b63ae8ca096d (\"mm/net: Rename and move page fragment handling from net/ to mm/\")\nSigned-off-by: Maurizio Lombardi \u003cmlombard@redhat.com\u003e\nReviewed-by: Alexander Duyck \u003calexanderduyck@fb.com\u003e\nCc: Chen Lin \u003cchen45464546@163.com\u003e\nCc: Jakub Kicinski \u003ckuba@kernel.org\u003e\nCc: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\n" }, { "commit": "70427f6e9ecfc8c5f977b21dd9f846b3bda02500", "tree": "b3efec847dc58187babf995c52b1ab083f4f027f", "parents": [ "37dcc673d065d9823576cd9f2484a72531e1cba6" ], "author": { "name": "Sergei Antonov", "email": "saproj@gmail.com", "time": "Thu Sep 08 23:48:09 2022 +0300" }, "committer": { "name": "Andrew Morton", "email": "akpm@linux-foundation.org", "time": "Mon Sep 26 12:14:34 2022 -0700" }, "message": "mm: bring back update_mmu_cache() to finish_fault()\n\nRunning this test program on ARMv4 a few times (sometimes just once)\nreproduces the bug.\n\nint main()\n{\n unsigned i;\n char paragon[SIZE];\n void* ptr;\n\n memset(paragon, 0xAA, SIZE);\n ptr \u003d mmap(NULL, SIZE, PROT_READ | PROT_WRITE,\n MAP_ANON | MAP_SHARED, -1, 0);\n if (ptr \u003d\u003d MAP_FAILED) return 1;\n printf(\"ptr \u003d %p\\n\", ptr);\n for (i\u003d0;i\u003c10000;i++){\n memset(ptr, 0xAA, SIZE);\n if (memcmp(ptr, paragon, SIZE)) {\n printf(\"Unexpected bytes on iteration %u!!!\\n\", i);\n break;\n }\n }\n munmap(ptr, SIZE);\n}\n\nIn the \"ptr\" buffer there appear runs of zero bytes which are aligned\nby 16 and their lengths are multiple of 16.\n\nLinux v5.11 does not have the bug, \"git bisect\" finds the first bad commit:\nf9ce0be71d1f (\"mm: Cleanup faultaround and finish_fault() codepaths\")\n\nBefore the commit update_mmu_cache() was called during a call to\nfilemap_map_pages() as well as finish_fault(). After the commit\nfinish_fault() lacks it.\n\nBring back update_mmu_cache() to finish_fault() to fix the bug.\nAlso call update_mmu_tlb() only when returning VM_FAULT_NOPAGE to more\nclosely reproduce the code of alloc_set_pte() function that existed before\nthe commit.\n\nOn many platforms update_mmu_cache() is nop:\n x86, see arch/x86/include/asm/pgtable\n ARMv6+, see arch/arm/include/asm/tlbflush.h\nSo, it seems, few users ran into this bug.\n\nLink: https://lkml.kernel.org/r/20220908204809.2012451-1-saproj@gmail.com\nFixes: f9ce0be71d1f (\"mm: Cleanup faultaround and finish_fault() codepaths\")\nSigned-off-by: Sergei Antonov \u003csaproj@gmail.com\u003e\nAcked-by: Kirill A. Shutemov \u003ckirill.shutemov@linux.intel.com\u003e\nCc: Will Deacon \u003cwill@kernel.org\u003e\nCc: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\n" }, { "commit": "37dcc673d065d9823576cd9f2484a72531e1cba6", "tree": "66ddedfb64543ffe7ecae65534be3340e71846a2", "parents": [ "2b7aa91ba0e86b8643f5d3c83874c80599c731d7" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Fri Sep 09 15:08:29 2022 +0200" }, "committer": { "name": "Andrew Morton", "email": "akpm@linux-foundation.org", "time": "Mon Sep 26 12:14:34 2022 -0700" }, "message": "frontswap: don\u0027t call -\u003einit if no ops are registered\n\nIf no frontswap module (i.e. zswap) was registered, frontswap_ops will be\nNULL. In such situation, swapon crashes with the following stack trace:\n\n Unable to handle kernel access to user memory outside uaccess routines at virtual address 0000000000000000\n Mem abort info:\n ESR \u003d 0x0000000096000004\n EC \u003d 0x25: DABT (current EL), IL \u003d 32 bits\n SET \u003d 0, FnV \u003d 0\n EA \u003d 0, S1PTW \u003d 0\n FSC \u003d 0x04: level 0 translation fault\n Data abort info:\n ISV \u003d 0, ISS \u003d 0x00000004\n CM \u003d 0, WnR \u003d 0\n user pgtable: 4k pages, 48-bit VAs, pgdp\u003d00000020a4fab000\n [0000000000000000] pgd\u003d0000000000000000, p4d\u003d0000000000000000\n Internal error: Oops: 96000004 [#1] SMP\n Modules linked in: zram fsl_dpaa2_eth pcs_lynx phylink ahci_qoriq crct10dif_ce ghash_ce sbsa_gwdt fsl_mc_dpio nvme lm90 nvme_core at803x xhci_plat_hcd rtc_fsl_ftm_alarm xgmac_mdio ahci_platform i2c_imx ip6_tables ip_tables fuse\n Unloaded tainted modules: cppc_cpufreq():1\n CPU: 10 PID: 761 Comm: swapon Not tainted 6.0.0-rc2-00454-g22100432cf14 #1\n Hardware name: SolidRun Ltd. SolidRun CEX7 Platform, BIOS EDK II Jun 21 2022\n pstate: 00400005 (nzcv daif +PAN -UAO -TCO -DIT -SSBS BTYPE\u003d--)\n pc : frontswap_init+0x38/0x60\n lr : __do_sys_swapon+0x8a8/0x9f4\n sp : ffff80000969bcf0\n x29: ffff80000969bcf0 x28: ffff37bee0d8fc00 x27: ffff80000a7f5000\n x26: fffffcdefb971e80 x25: ffffaba797453b90 x24: 0000000000000064\n x23: ffff37c1f209d1a8 x22: ffff37bee880e000 x21: ffffaba797748560\n x20: ffff37bee0d8fce4 x19: ffffaba797748488 x18: 0000000000000014\n x17: 0000000030ec029a x16: ffffaba795a479b0 x15: 0000000000000000\n x14: 0000000000000000 x13: 0000000000000030 x12: 0000000000000001\n x11: ffff37c63c0aba18 x10: 0000000000000000 x9 : ffffaba7956b8c88\n x8 : ffff80000969bcd0 x7 : 0000000000000000 x6 : 0000000000000000\n x5 : 0000000000000001 x4 : 0000000000000000 x3 : ffffaba79730f000\n x2 : ffff37bee0d8fc00 x1 : 0000000000000000 x0 : 0000000000000000\n Call trace:\n frontswap_init+0x38/0x60\n __do_sys_swapon+0x8a8/0x9f4\n __arm64_sys_swapon+0x28/0x3c\n invoke_syscall+0x78/0x100\n el0_svc_common.constprop.0+0xd4/0xf4\n do_el0_svc+0x38/0x4c\n el0_svc+0x34/0x10c\n el0t_64_sync_handler+0x11c/0x150\n el0t_64_sync+0x190/0x194\n Code: d000e283 910003fd f9006c41 f946d461 (f9400021)\n ---[ end trace 0000000000000000 ]---\n\nLink: https://lkml.kernel.org/r/20220909130829.3262926-1-hch@lst.de\nFixes: 1da0d94a3ec8 (\"frontswap: remove support for multiple ops\")\nReported-by: Nathan Chancellor \u003cnathan@kernel.org\u003e\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Liu Shixin \u003cliushixin2@huawei.com\u003e\nCc: Konrad Rzeszutek Wilk \u003ckonrad.wilk@oracle.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\n" }, { "commit": "2b7aa91ba0e86b8643f5d3c83874c80599c731d7", "tree": "19349cd5a50e1f060e328646a9d7214c8b9cd906", "parents": [ "58d426a7ba92870d489686dfdb9d06b66815a2ab" ], "author": { "name": "Naoya Horiguchi", "email": "naoya.horiguchi@nec.com", "time": "Thu Sep 08 13:11:50 2022 +0900" }, "committer": { "name": "Andrew Morton", "email": "akpm@linux-foundation.org", "time": "Mon Sep 26 12:14:33 2022 -0700" }, "message": "mm/huge_memory: use pfn_to_online_page() in split_huge_pages_all()\n\nNULL pointer dereference is triggered when calling thp split via debugfs\non the system with offlined memory blocks. With debug option enabled, the\nfollowing kernel messages are printed out:\n\n page:00000000467f4890 refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x121c000\n flags: 0x17fffc00000000(node\u003d0|zone\u003d2|lastcpupid\u003d0x1ffff)\n raw: 0017fffc00000000 0000000000000000 dead000000000122 0000000000000000\n raw: 0000000000000000 0000000000000000 00000001ffffffff 0000000000000000\n page dumped because: unmovable page\n page:000000007d7ab72e is uninitialized and poisoned\n page dumped because: VM_BUG_ON_PAGE(PagePoisoned(p))\n ------------[ cut here ]------------\n kernel BUG at include/linux/mm.h:1248!\n invalid opcode: 0000 [#1] PREEMPT SMP PTI\n CPU: 16 PID: 20964 Comm: bash Tainted: G I 6.0.0-rc3-foll-numa+ #41\n ...\n RIP: 0010:split_huge_pages_write+0xcf4/0xe30\n\nThis shows that page_to_nid() in page_zone() is unexpectedly called for an\nofflined memmap.\n\nUse pfn_to_online_page() to get struct page in PFN walker.\n\nLink: https://lkml.kernel.org/r/20220908041150.3430269-1-naoya.horiguchi@linux.dev\nFixes: f1dd2cd13c4b (\"mm, memory_hotplug: do not associate hotadded memory to zones until online\") [visible after d0dc12e86b319]\nSigned-off-by: Naoya Horiguchi \u003cnaoya.horiguchi@nec.com\u003e\nCo-developed-by: David Hildenbrand \u003cdavid@redhat.com\u003e\nSigned-off-by: David Hildenbrand \u003cdavid@redhat.com\u003e\nReviewed-by: Yang Shi \u003cshy828301@gmail.com\u003e\nAcked-by: Michal Hocko \u003cmhocko@suse.com\u003e\nReviewed-by: Miaohe Lin \u003clinmiaohe@huawei.com\u003e\nReviewed-by: Oscar Salvador \u003cosalvador@suse.de\u003e\nAcked-by: Kirill A. Shutemov \u003ckirill.shutemov@linux.intel.com\u003e\nCc: Matthew Wilcox \u003cwilly@infradead.org\u003e\nCc: Muchun Song \u003csongmuchun@bytedance.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\t[5.10+]\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\n" }, { "commit": "58d426a7ba92870d489686dfdb9d06b66815a2ab", "tree": "d0be23f5a797e2f42856dcde0ff12062aea9d5f9", "parents": [ "bedf03416913d88c796288f9dca109a53608c745" ], "author": { "name": "Minchan Kim", "email": "minchan@kernel.org", "time": "Thu Sep 08 08:12:04 2022 -0700" }, "committer": { "name": "Andrew Morton", "email": "akpm@linux-foundation.org", "time": "Mon Sep 26 12:14:33 2022 -0700" }, "message": "mm: fix madivse_pageout mishandling on non-LRU page\n\nMADV_PAGEOUT tries to isolate non-LRU pages and gets a warning from\nisolate_lru_page below.\n\nFix it by checking PageLRU in advance.\n\n------------[ cut here ]------------\ntrying to isolate tail page\nWARNING: CPU: 0 PID: 6175 at mm/folio-compat.c:158 isolate_lru_page+0x130/0x140\nModules linked in:\nCPU: 0 PID: 6175 Comm: syz-executor.0 Not tainted 5.18.12 #1\nHardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.13.0-1ubuntu1.1 04/01/2014\nRIP: 0010:isolate_lru_page+0x130/0x140\n\nLink: https://lore.kernel.org/linux-mm/485f8c33.2471b.182d5726afb.Coremail.hantianshuo@iie.ac.cn/\nLink: https://lkml.kernel.org/r/20220908151204.762596-1-minchan@kernel.org\nFixes: 1a4e58cce84e (\"mm: introduce MADV_PAGEOUT\")\nSigned-off-by: Minchan Kim \u003cminchan@kernel.org\u003e\nReported-by: 韩天ç`• \u003chantianshuo@iie.ac.cn\u003e\nSuggested-by: Yang Shi \u003cshy828301@gmail.com\u003e\nAcked-by: Yang Shi \u003cshy828301@gmail.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\n" } ], "next": "bedf03416913d88c796288f9dca109a53608c745" }