)]}' { "log": [ { "commit": "4d02da974ea85a62074efedf354e82778f910d82", "tree": "e5338041c62c5e382e1dcf9c4a173bfd8450bb6c", "parents": [ "3be28e93cd88fbcbe97cabcbe92b1ccc9f830450", "e6ea60bac1ee28bb46232f8c2ecd3a3fbb9011e0" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 19 13:33:16 2020 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 19 13:33:16 2020 -0800" }, "message": "Merge tag \u0027net-5.10-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net\n\nPull networking fixes from Jakub Kicinski:\n \"Networking fixes for 5.10-rc5, including fixes from the WiFi\n (mac80211), can and bpf (including the strncpy_from_user fix).\n\n Current release - regressions:\n\n - mac80211: fix memory leak of filtered powersave frames\n\n - mac80211: free sta in sta_info_insert_finish() on errors to avoid\n sleeping in atomic context\n\n - netlabel: fix an uninitialized variable warning added in -rc4\n\n Previous release - regressions:\n\n - vsock: forward all packets to the host when no H2G is registered,\n un-breaking AWS Nitro Enclaves\n\n - net: Exempt multicast addresses from five-second neighbor lifetime\n requirement, decreasing the chances neighbor tables fill up\n\n - net/tls: fix corrupted data in recvmsg\n\n - qed: fix ILT configuration of SRC block\n\n - can: m_can: process interrupt only when not runtime suspended\n\n Previous release - always broken:\n\n - page_frag: Recover from memory pressure by not recycling pages\n allocating from the reserves\n\n - strncpy_from_user: Mask out bytes after NUL terminator\n\n - ip_tunnels: Set tunnel option flag only when tunnel metadata is\n present, always setting it confuses Open vSwitch\n\n - bpf, sockmap:\n - Fix partial copy_page_to_iter so progress can still be made\n - Fix socket memory accounting and obeying SO_RCVBUF\n\n - net: Have netpoll bring-up DSA management interface\n\n - net: bridge: add missing counters to ndo_get_stats64 callback\n\n - tcp: brr: only postpone PROBE_RTT if RTT is \u003c current min_rtt\n\n - enetc: Workaround MDIO register access HW bug\n\n - net/ncsi: move netlink family registration to a subsystem init,\n instead of tying it to driver probe\n\n - net: ftgmac100: unregister NC-SI when removing driver to avoid\n crash\n\n - lan743x:\n - prevent interrupt storm on open\n - fix freeing skbs in the wrong context\n\n - net/mlx5e: Fix socket refcount leak on kTLS RX resync\n\n - net: dsa: mv88e6xxx: Avoid VLAN database corruption on 6097\n\n - fix 21 unset return codes and other mistakes on error paths, mostly\n detected by the Hulk Robot\"\n\n* tag \u0027net-5.10-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (115 commits)\n fail_function: Remove a redundant mutex unlock\n selftest/bpf: Test bpf_probe_read_user_str() strips trailing bytes after NUL\n lib/strncpy_from_user.c: Mask out bytes after NUL terminator.\n net/smc: fix direct access to ib_gid_addr-\u003endev in smc_ib_determine_gid()\n net/smc: fix matching of existing link groups\n ipv6: Remove dependency of ipv6_frag_thdr_truncated on ipv6 module\n libbpf: Fix VERSIONED_SYM_COUNT number parsing\n net/mlx4_core: Fix init_hca fields offset\n atm: nicstar: Unmap DMA on send error\n page_frag: Recover from memory pressure\n net: dsa: mv88e6xxx: Wait for EEPROM done after HW reset\n mlxsw: core: Use variable timeout for EMAD retries\n mlxsw: Fix firmware flashing\n net: Have netpoll bring-up DSA management interface\n atl1e: fix error return code in atl1e_probe()\n atl1c: fix error return code in atl1c_probe()\n ah6: fix error return code in ah6_input()\n net: usb: qmi_wwan: Set DTR quirk for MR400\n can: m_can: process interrupt only when not runtime suspended\n can: flexcan: flexcan_chip_start(): fix erroneous flexcan_transceiver_enable() during bus-off recovery\n ...\n" }, { "commit": "3be28e93cd88fbcbe97cabcbe92b1ccc9f830450", "tree": "6a0483c8d204515da84d11ea8cd7825804e97adc", "parents": [ "dda3f4252e6c8b833a2ef164afd3da9808d0f07c", "ee415d73dcc24caef7f6bbf292dcc365613d2188" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 19 13:01:53 2020 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 19 13:01:53 2020 -0800" }, "message": "Merge tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma\n\nPull rdma fixes from Jason Gunthorpe:\n \"The last two weeks have been quiet here, just the usual smattering of\n long standing bug fixes.\n\n A collection of error case bug fixes:\n\n - Improper nesting of spinlock types in cm\n\n - Missing error codes and kfree()\n\n - Ensure dma_virt_ops users have the right kconfig symbols to work\n properly\n\n - Compilation failure of tools/testing\"\n\n* tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:\n tools/testing/scatterlist: Fix test to compile and run\n IB/hfi1: Fix error return code in hfi1_init_dd()\n RMDA/sw: Don\u0027t allow drivers using dma_virt_ops on highmem configs\n RDMA/pvrdma: Fix missing kfree() in pvrdma_register_device()\n RDMA/cm: Make the local_id_table xarray non-irq\n" }, { "commit": "e6ea60bac1ee28bb46232f8c2ecd3a3fbb9011e0", "tree": "f119a3b701852f184b6131405470de0f3969c783", "parents": [ "90b49784814c10c0f028646f95cc2c2848696712", "2801a5da5b25b7af9dd2addd19b2315c02d17b64" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Nov 19 12:26:09 2020 -0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Nov 19 12:26:10 2020 -0800" }, "message": "Merge https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf\n\nAlexei Starovoitov says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n1) libbpf should not attempt to load unused subprogs, from Andrii.\n\n2) Make strncpy_from_user() mask out bytes after NUL terminator, from Daniel.\n\n3) Relax return code check for subprograms in the BPF verifier, from Dmitrii.\n\n4) Fix several sockmap issues, from John.\n\n* https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:\n fail_function: Remove a redundant mutex unlock\n selftest/bpf: Test bpf_probe_read_user_str() strips trailing bytes after NUL\n lib/strncpy_from_user.c: Mask out bytes after NUL terminator.\n libbpf: Fix VERSIONED_SYM_COUNT number parsing\n bpf, sockmap: Avoid failures from skb_to_sgvec when skb has frag_list\n bpf, sockmap: Handle memory acct if skb_verdict prog redirects to self\n bpf, sockmap: Avoid returning unneeded EAGAIN when redirecting to self\n bpf, sockmap: Use truesize with sk_rmem_schedule()\n bpf, sockmap: Ensure SO_RCVBUF memory is observed on ingress redirect\n bpf, sockmap: Fix partial copy_page_to_iter so progress can still be made\n selftests/bpf: Fix error return code in run_getsockopt_test()\n bpf: Relax return code check for subprograms\n tools, bpftool: Add missing close before bpftool net attach exit\n MAINTAINERS/bpf: Update Andrii\u0027s entry.\n selftests/bpf: Fix unused attribute usage in subprogs_unused test\n bpf: Fix unsigned \u0027datasec_id\u0027 compared with zero in check_pseudo_btf_id\n bpf: Fix passing zero to PTR_ERR() in bpf_btf_printf_prepare\n libbpf: Don\u0027t attempt to load unused subprog as an entry-point BPF program\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/20201119200721.288-1-alexei.starovoitov@gmail.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "2801a5da5b25b7af9dd2addd19b2315c02d17b64", "tree": "dbabe546198f7e2d4386a36345730541e747ab58", "parents": [ "14d6d86c210aea1a83c19a8f6391ecabcbefed94" ], "author": { "name": "Luo Meng", "email": "luomeng12@huawei.com", "time": "Wed Nov 18 22:49:31 2020 +0900" }, "committer": { "name": "Alexei Starovoitov", "email": "ast@kernel.org", "time": "Thu Nov 19 11:58:16 2020 -0800" }, "message": "fail_function: Remove a redundant mutex unlock\n\nFix a mutex_unlock() issue where before copy_from_user() is\nnot called mutex_locked.\n\nFixes: 4b1a29a7f542 (\"error-injection: Support fault injection framework\")\nReported-by: Hulk Robot \u003chulkci@huawei.com\u003e\nSigned-off-by: Luo Meng \u003cluomeng12@huawei.com\u003e\nSigned-off-by: Masami Hiramatsu \u003cmhiramat@kernel.org\u003e\nSigned-off-by: Alexei Starovoitov \u003cast@kernel.org\u003e\nAcked-by: Masami Hiramatsu \u003cmhiramat@kernel.org\u003e\nLink: https://lore.kernel.org/bpf/160570737118.263807.8358435412898356284.stgit@devnote2\n" }, { "commit": "14d6d86c210aea1a83c19a8f6391ecabcbefed94", "tree": "5eadde2770ff55b99a77d39e56e9b3dc322ce628", "parents": [ "1fd6cee127e2ddff36d648573d7566aafb0d0b77", "c8a36aedf3e24768e94d87fdcdd37684bd241c44" ], "author": { "name": "Alexei Starovoitov", "email": "ast@kernel.org", "time": "Thu Nov 19 10:45:55 2020 -0800" }, "committer": { "name": "Alexei Starovoitov", "email": "ast@kernel.org", "time": "Thu Nov 19 11:58:15 2020 -0800" }, "message": "Merge branch \u0027Fix bpf_probe_read_user_str() overcopying\u0027\n\nDaniel Xu says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\n6ae08ae3dea2 (\"bpf: Add probe_read_{user, kernel} and probe_read_{user,\nkernel}_str helpers\") introduced a subtle bug where\nbpf_probe_read_user_str() would potentially copy a few extra bytes after\nthe NUL terminator.\n\nThis issue is particularly nefarious when strings are used as map keys,\nas seemingly identical strings can occupy multiple entries in a map.\n\nThis patchset fixes the issue and introduces a selftest to prevent\nfuture regressions.\n\nv6 -\u003e v7:\n* Add comments\n\nv5 -\u003e v6:\n* zero-pad up to sizeof(unsigned long) after NUL\n\nv4 -\u003e v5:\n* don\u0027t read potentially uninitialized memory\n\nv3 -\u003e v4:\n* directly pass userspace pointer to prog\n* test more strings of different length\n\nv2 -\u003e v3:\n* set pid filter before attaching prog in selftest\n* use long instead of int as bpf_probe_read_user_str() retval\n* style changes\n\nv1 -\u003e v2:\n* add Fixes: tag\n* add selftest\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: Alexei Starovoitov \u003cast@kernel.org\u003e\n" }, { "commit": "c8a36aedf3e24768e94d87fdcdd37684bd241c44", "tree": "5eadde2770ff55b99a77d39e56e9b3dc322ce628", "parents": [ "6fa6d28051e9fcaa1570e69648ea13a353a5d218" ], "author": { "name": "Daniel Xu", "email": "dxu@dxuuu.xyz", "time": "Tue Nov 17 12:05:46 2020 -0800" }, "committer": { "name": "Alexei Starovoitov", "email": "ast@kernel.org", "time": "Thu Nov 19 11:58:15 2020 -0800" }, "message": "selftest/bpf: Test bpf_probe_read_user_str() strips trailing bytes after NUL\n\nPreviously, bpf_probe_read_user_str() could potentially overcopy the\ntrailing bytes after the NUL due to how do_strncpy_from_user() does the\ncopy in long-sized strides. The issue has been fixed in the previous\ncommit.\n\nThis commit adds a selftest that ensures we don\u0027t regress\nbpf_probe_read_user_str() again.\n\nSigned-off-by: Daniel Xu \u003cdxu@dxuuu.xyz\u003e\nSigned-off-by: Alexei Starovoitov \u003cast@kernel.org\u003e\nAcked-by: Song Liu \u003csongliubraving@fb.com\u003e\nAcked-by: Andrii Nakryiko \u003candrii@kernel.org\u003e\nLink: https://lore.kernel.org/bpf/4d977508fab4ec5b7b574b85bdf8b398868b6ee9.1605642949.git.dxu@dxuuu.xyz\n" }, { "commit": "6fa6d28051e9fcaa1570e69648ea13a353a5d218", "tree": "1fd6809c9d91663c9bf31062dcf6d58ef49d374b", "parents": [ "1fd6cee127e2ddff36d648573d7566aafb0d0b77" ], "author": { "name": "Daniel Xu", "email": "dxu@dxuuu.xyz", "time": "Tue Nov 17 12:05:45 2020 -0800" }, "committer": { "name": "Alexei Starovoitov", "email": "ast@kernel.org", "time": "Thu Nov 19 11:56:16 2020 -0800" }, "message": "lib/strncpy_from_user.c: Mask out bytes after NUL terminator.\n\ndo_strncpy_from_user() may copy some extra bytes after the NUL\nterminator into the destination buffer. This usually does not matter for\nnormal string operations. However, when BPF programs key BPF maps with\nstrings, this matters a lot.\n\nA BPF program may read strings from user memory by calling the\nbpf_probe_read_user_str() helper which eventually calls\ndo_strncpy_from_user(). The program can then key a map with the\ndestination buffer. BPF map keys are fixed-width and string-agnostic,\nmeaning that map keys are treated as a set of bytes.\n\nThe issue is when do_strncpy_from_user() overcopies bytes after the NUL\nterminator, it can result in seemingly identical strings occupying\nmultiple slots in a BPF map. This behavior is subtle and totally\nunexpected by the user.\n\nThis commit masks out the bytes following the NUL while preserving\nlong-sized stride in the fast path.\n\nFixes: 6ae08ae3dea2 (\"bpf: Add probe_read_{user, kernel} and probe_read_{user, kernel}_str helpers\")\nSigned-off-by: Daniel Xu \u003cdxu@dxuuu.xyz\u003e\nSigned-off-by: Alexei Starovoitov \u003cast@kernel.org\u003e\nLink: https://lore.kernel.org/bpf/21efc982b3e9f2f7b0379eed642294caaa0c27a7.1605642949.git.dxu@dxuuu.xyz\n" }, { "commit": "dda3f4252e6c8b833a2ef164afd3da9808d0f07c", "tree": "9a294782b7d3d4fd3c48f16bf39d57b34be44ad1", "parents": [ "3494d58865ad4a47611dbb427b214cc5227fa5eb", "da631f7fd623b6c180c8d93a93040d1e0d61291f" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 19 11:32:31 2020 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 19 11:32:31 2020 -0800" }, "message": "Merge tag \u0027powerpc-cve-2020-4788\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux\n\nPull powerpc fixes from Michael Ellerman:\n \"Fixes for CVE-2020-4788.\n\n From Daniel\u0027s cover letter:\n\n IBM Power9 processors can speculatively operate on data in the L1\n cache before it has been completely validated, via a way-prediction\n mechanism. It is not possible for an attacker to determine the\n contents of impermissible memory using this method, since these\n systems implement a combination of hardware and software security\n measures to prevent scenarios where protected data could be leaked.\n\n However these measures don\u0027t address the scenario where an attacker\n induces the operating system to speculatively execute instructions\n using data that the attacker controls. This can be used for example to\n speculatively bypass \"kernel user access prevention\" techniques, as\n discovered by Anthony Steinhauser of Google\u0027s Safeside Project. This\n is not an attack by itself, but there is a possibility it could be\n used in conjunction with side-channels or other weaknesses in the\n privileged code to construct an attack.\n\n This issue can be mitigated by flushing the L1 cache between privilege\n boundaries of concern.\n\n This patch series flushes the L1 cache on kernel entry (patch 2) and\n after the kernel performs any user accesses (patch 3). It also adds a\n self-test and performs some related cleanups\"\n\n* tag \u0027powerpc-cve-2020-4788\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:\n powerpc/64s: rename pnv|pseries_setup_rfi_flush to _setup_security_mitigations\n selftests/powerpc: refactor entry and rfi_flush tests\n selftests/powerpc: entry flush test\n powerpc: Only include kup-radix.h for 64-bit Book3S\n powerpc/64s: flush L1D after user accesses\n powerpc/64s: flush L1D on kernel entry\n selftests/powerpc: rfi_flush: disable entry flush if present\n" }, { "commit": "3494d58865ad4a47611dbb427b214cc5227fa5eb", "tree": "a3eefe871506736788cf5a68c62887e845ca362d", "parents": [ "131ad0b6f5294dd4eb1c8e5e5c611a85cbae0c4e", "dc293f2106903ab9c24e9cea18c276e32c394c33" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 19 11:22:33 2020 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 19 11:22:33 2020 -0800" }, "message": "Merge tag \u0027xtensa-20201119\u0027 of git://github.com/jcmvbkbc/linux-xtensa\n\nPull xtensa fixes from Max Filippov:\n\n - fix placement of cache alias remapping area\n\n - disable preemption around cache alias management calls\n\n - add missing __user annotation to strncpy_from_user argument\n\n* tag \u0027xtensa-20201119\u0027 of git://github.com/jcmvbkbc/linux-xtensa:\n xtensa: uaccess: Add missing __user to strncpy_from_user() prototype\n xtensa: disable preemption around cache alias management calls\n xtensa: fix TLBTEMP area placement\n" }, { "commit": "131ad0b6f5294dd4eb1c8e5e5c611a85cbae0c4e", "tree": "27c6182e48404ce93d1b59061cef5b994b33419b", "parents": [ "4ca35b4f45092634df21dada47746571a34cc09c", "de15e20f50b126b3a5a753dd259df775e6a0ea5c" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 19 11:15:15 2020 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 19 11:15:15 2020 -0800" }, "message": "Merge tag \u0027acpi-5.10-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm\n\nPull ACPI fixes from Rafael Wysocki:\n \"These fix recent regression in the APEI code and initialization issue\n in the ACPI fan driver.\n\n Specifics:\n\n - Make the APEI code avoid attempts to obtain logical addresses for\n registers located in the I/O address space to fix initialization\n issues (Aili Yao)\n\n - Fix sysfs attribute initialization in the ACPI fan driver (Guenter\n Roeck)\"\n\n* tag \u0027acpi-5.10-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:\n ACPI, APEI, Fix error return value in apei_map_generic_address()\n ACPI: fan: Initialize performance state sysfs attribute\n" }, { "commit": "4ca35b4f45092634df21dada47746571a34cc09c", "tree": "054bae0e499efc9344b4374adf9f678caced7adc", "parents": [ "fee3c824edf24fd4c497bb3b887d1447150ec4ec", "3a8ac4d39651c71d6d2f3376b13d0bad1484249a" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 19 11:11:50 2020 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 19 11:11:50 2020 -0800" }, "message": "Merge tag \u0027pm-5.10-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm\n\nPull power management fixes from Rafael Wysocki:\n \"These fix two issues in ARM cpufreq drivers and one cpuidle driver\n issue.\n\n Specifics:\n\n - Add missing RCU_NONIDLE() annotations to the Tegra cpuidle driver\n (Dmitry Osipenko)\n\n - Fix boot frequency computation in the tegra186 cpufreq driver (Jon\n Hunter)\n\n - Make the SCMI cpufreq driver register a dummy clock provider to\n avoid OPP addition failures (Sudeep Holla)\"\n\n* tag \u0027pm-5.10-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:\n cpufreq: scmi: Fix OPP addition failure with a dummy clock provider\n cpufreq: tegra186: Fix get frequency callback\n cpuidle: tegra: Annotate tegra_pm_set_cpu_in_lp2() with RCU_NONIDLE\n" }, { "commit": "fee3c824edf24fd4c497bb3b887d1447150ec4ec", "tree": "cc29ba7b56153ab21f5fe8607665a9feb2e7ff7c", "parents": [ "d748287a28de8955e1756a454e81ab1a73d45ecf", "04a9cd51d3f3308a98cbc6adc07acb12fbade011" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 19 11:05:28 2020 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 19 11:05:28 2020 -0800" }, "message": "Merge tag \u0027spi-fix-v5.10-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi\n\nPull spi fixes from Mark Brown:\n \"This is a relatively large set of fixes, the bulk of it being a series\n from Lukas Wunner which fixes confusion with the lifetime of driver\n data allocated along with the SPI controller structure that\u0027s been\n created as part of the conversion to devm APIs.\n\n The simplest fix, explained in detail in Lukas\u0027 commit message, is to\n move to a devm_ function for allocation of the controller and hence\n driver data in order to push the free of that after anything tries to\n reference the driver data in the remove path. This results in a\n relatively large diff due to the addition of a new function but isn\u0027t\n particularly complex.\n\n There\u0027s also a fix from Sven van Asbroeck which fixes yet more fallout\n from the conflicts between the various different places one can\n configure the polarity of GPIOs in modern systems.\n\n Otherwise everything is fairly small and driver specific\"\n\n* tag \u0027spi-fix-v5.10-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi:\n spi: npcm-fiu: Don\u0027t leak SPI master in probe error path\n spi: dw: Set transfer handler before unmasking the IRQs\n spi: cadence-quadspi: Fix error return code in cqspi_probe\n spi: bcm2835aux: Restore err assignment in bcm2835aux_spi_probe\n spi: lpspi: Fix use-after-free on unbind\n spi: bcm-qspi: Fix use-after-free on unbind\n spi: bcm2835aux: Fix use-after-free on unbind\n spi: bcm2835: Fix use-after-free on unbind\n spi: Introduce device-managed SPI controller allocation\n spi: fsi: Fix transfer returning without finalizing message\n spi: fix client driver breakages when using GPIO descriptors\n" }, { "commit": "90b49784814c10c0f028646f95cc2c2848696712", "tree": "0e58091a226c305fefdc2b7b832ccb885f41571f", "parents": [ "2d8f6481c17db9fa5238b277cdbc392084060b09", "41a0be3f8f6be893860b991eb10c47fc3ee09d7f" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Nov 19 10:59:22 2020 -0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Nov 19 10:59:22 2020 -0800" }, "message": "Merge branch \u0027net-smc-fixes-2020-11-18\u0027\n\nKarsten Graul says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nnet/smc: fixes 2020-11-18\n\nPatch 1 fixes the matching of link groups because with SMC-Dv2 the vlanid\nshould no longer be part of this matching. Patch 2 removes a sparse message.\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/20201118214038.24039-1-kgraul@linux.ibm.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "41a0be3f8f6be893860b991eb10c47fc3ee09d7f", "tree": "0e58091a226c305fefdc2b7b832ccb885f41571f", "parents": [ "0530bd6e6a3d5a0292a1a5f33ea980ae7e8b56ca" ], "author": { "name": "Karsten Graul", "email": "kgraul@linux.ibm.com", "time": "Wed Nov 18 22:40:38 2020 +0100" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Nov 19 10:59:19 2020 -0800" }, "message": "net/smc: fix direct access to ib_gid_addr-\u003endev in smc_ib_determine_gid()\n\nSparse complaints 3 times about:\nnet/smc/smc_ib.c:203:52: warning: incorrect type in argument 1 (different address spaces)\nnet/smc/smc_ib.c:203:52: expected struct net_device const *dev\nnet/smc/smc_ib.c:203:52: got struct net_device [noderef] __rcu *const ndev\n\nFix that by using the existing and validated ndev variable instead of\naccessing attr-\u003endev directly.\n\nFixes: 5102eca9039b (\"net/smc: Use rdma_read_gid_l2_fields to L2 fields\")\nSigned-off-by: Karsten Graul \u003ckgraul@linux.ibm.com\u003e\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "0530bd6e6a3d5a0292a1a5f33ea980ae7e8b56ca", "tree": "262be117490f7112e6058e20fcb2af7bdc41e855", "parents": [ "2d8f6481c17db9fa5238b277cdbc392084060b09" ], "author": { "name": "Karsten Graul", "email": "kgraul@linux.ibm.com", "time": "Wed Nov 18 22:40:37 2020 +0100" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Nov 19 10:59:19 2020 -0800" }, "message": "net/smc: fix matching of existing link groups\n\nWith the multi-subnet support of SMC-Dv2 the match for existing link\ngroups should not include the vlanid of the network device.\nSet ini-\u003esmcd_version accordingly before the call to smc_conn_create()\nand use this value in smc_conn_create() to skip the vlanid check.\n\nFixes: 5c21c4ccafe8 (\"net/smc: determine accepted ISM devices\")\nSigned-off-by: Karsten Graul \u003ckgraul@linux.ibm.com\u003e\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "d748287a28de8955e1756a454e81ab1a73d45ecf", "tree": "2e49d50315fd330e048429ac036ce23abce92f3e", "parents": [ "841d6e9ec9e345054dd4a28ef5feceb73e9702be", "2ba546ebe0ce2af47833d8912ced9b4a579f13cb" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 19 10:55:54 2020 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 19 10:55:54 2020 -0800" }, "message": "Merge tag \u0027regulator-fix-v5.10-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator\n\nPull regulator fixes from Mark Brown:\n \"Mostly core fixes here, one set from Michał Mirosław which cleans up\n some issues introduced as part of the coupled regulators work, one\n memory leak during probe and two due to regulators which have an input\n supply name and regulator name which are identical, which is very\n unusual.\n\n There\u0027s also a fix for our handling of the similarly unusual case\n where we can\u0027t determine if a regulator is enabled during boot\"\n\n* tag \u0027regulator-fix-v5.10-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator:\n regulator: ti-abb: Fix array out of bound read access on the first transition\n regulator: workaround self-referent regulators\n regulator: avoid resolve_supply() infinite recursion\n regulator: fix memory leak with repeated set_machine_constraints()\n regulator: pfuze100: limit pfuze-support-disable-sw to pfuze{100,200}\n regulator: core: don\u0027t disable regulator if is_enabled return error.\n" }, { "commit": "2d8f6481c17db9fa5238b277cdbc392084060b09", "tree": "8d0cd40c597a93ec03b13fc00add9ec4656284c1", "parents": [ "f0b0a2d8a1fd4e862f8f5bf8bf449d116264ce5c" ], "author": { "name": "Georg Kohmann", "email": "geokohma@cisco.com", "time": "Thu Nov 19 10:58:33 2020 +0100" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Thu Nov 19 10:49:50 2020 -0800" }, "message": "ipv6: Remove dependency of ipv6_frag_thdr_truncated on ipv6 module\n\nIPV6\u003dm\nNF_DEFRAG_IPV6\u003dy\n\nld: net/ipv6/netfilter/nf_conntrack_reasm.o: in function\n`nf_ct_frag6_gather\u0027:\nnet/ipv6/netfilter/nf_conntrack_reasm.c:462: undefined reference to\n`ipv6_frag_thdr_truncated\u0027\n\nNetfilter is depending on ipv6 symbol ipv6_frag_thdr_truncated. This\ndependency is forcing IPV6\u003dy.\n\nRemove this dependency by moving ipv6_frag_thdr_truncated out of ipv6. This\nis the same solution as used with a similar issues: Referring to\ncommit 70b095c843266 (\"ipv6: remove dependency of nf_defrag_ipv6 on ipv6\nmodule\")\n\nFixes: 9d9e937b1c8b (\"ipv6/netfilter: Discard first fragment not including all headers\")\nReported-by: Randy Dunlap \u003crdunlap@infradead.org\u003e\nReported-by: kernel test robot \u003clkp@intel.com\u003e\nSigned-off-by: Georg Kohmann \u003cgeokohma@cisco.com\u003e\nAcked-by: Pablo Neira Ayuso \u003cpablo@netfilter.org\u003e\nAcked-by: Randy Dunlap \u003crdunlap@infradead.org\u003e # build-tested\nLink: https://lore.kernel.org/r/20201119095833.8409-1-geokohma@cisco.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "841d6e9ec9e345054dd4a28ef5feceb73e9702be", "tree": "98032041f914869a01211a4619c152d6d26f967a", "parents": [ "c2e7554e1b85935d962127efa3c2a76483b0b3b6", "b98467fe96d2415836d154ecfe1cd389bf4147b5" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 19 10:49:35 2020 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 19 10:49:35 2020 -0800" }, "message": "Merge tag \u0027thermal-v5.10-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/thermal/linux\n\nPull thermal fix from Daniel Lezcano:\n \"Disable the CPU PM notifier for OMAP4430 for suspend in order to\n prevent wrong temperature leading to a critical shutdown (Peter\n Ujfalusi)\"\n\n* tag \u0027thermal-v5.10-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/thermal/linux:\n thermal: ti-soc-thermal: Disable the CPU PM notifier for OMAP4430\n" }, { "commit": "1fd6cee127e2ddff36d648573d7566aafb0d0b77", "tree": "ccc7cce8627207c7c69c0d1945b1c2f6e8ea961c", "parents": [ "4363023d2668e621b0743db351a9555d6e6ea57e" ], "author": { "name": "Jiri Olsa", "email": "jolsa@kernel.org", "time": "Wed Nov 18 22:13:50 2020 +0100" }, "committer": { "name": "Alexei Starovoitov", "email": "ast@kernel.org", "time": "Thu Nov 19 08:45:12 2020 -0800" }, "message": "libbpf: Fix VERSIONED_SYM_COUNT number parsing\n\nWe remove \"other info\" from \"readelf -s --wide\" output when\nparsing GLOBAL_SYM_COUNT variable, which was added in [1].\nBut we don\u0027t do that for VERSIONED_SYM_COUNT and it\u0027s failing\nthe check_abi target on powerpc Fedora 33.\n\nThe extra \"other info\" wasn\u0027t problem for VERSIONED_SYM_COUNT\nparsing until commit [2] added awk in the pipe, which assumes\nthat the last column is symbol, but it can be \"other info\".\n\nAdding \"other info\" removal for VERSIONED_SYM_COUNT the same\nway as we did for GLOBAL_SYM_COUNT parsing.\n\n[1] aa915931ac3e (\"libbpf: Fix readelf output parsing for Fedora\")\n[2] 746f534a4809 (\"tools/libbpf: Avoid counting local symbols in ABI check\")\n\nFixes: 746f534a4809 (\"tools/libbpf: Avoid counting local symbols in ABI check\")\nSigned-off-by: Jiri Olsa \u003cjolsa@kernel.org\u003e\nSigned-off-by: Alexei Starovoitov \u003cast@kernel.org\u003e\nAcked-by: Andrii Nakryiko \u003candrii@kernel.org\u003e\nLink: https://lore.kernel.org/bpf/20201118211350.1493421-1-jolsa@kernel.org\n" }, { "commit": "de15e20f50b126b3a5a753dd259df775e6a0ea5c", "tree": "731680d5386b8dba213792dba8260d6f8bc162cd", "parents": [ "4fba15fbb8106e8db17f486d653484e64969eb87", "7dc7a8b04f3da8aa3c3be514e155e2fa094e976f" ], "author": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Thu Nov 19 14:03:52 2020 +0100" }, "committer": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Thu Nov 19 14:03:52 2020 +0100" }, "message": "Merge branch \u0027acpi-fan\u0027\n\n* acpi-fan:\n ACPI: fan: Initialize performance state sysfs attribute\n" }, { "commit": "3a8ac4d39651c71d6d2f3376b13d0bad1484249a", "tree": "d7ae66b344b0fe49e3a83187a4ed46fdcdd63bb4", "parents": [ "14c620cf2ee81349527491110a47a157ac6d511c", "c39de538a06e76d89b7e598a71e16688009cd56c" ], "author": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Thu Nov 19 14:02:18 2020 +0100" }, "committer": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Thu Nov 19 14:02:18 2020 +0100" }, "message": "Merge branch \u0027pm-cpuidle\u0027\n\n* pm-cpuidle:\n cpuidle: tegra: Annotate tegra_pm_set_cpu_in_lp2() with RCU_NONIDLE\n" }, { "commit": "da631f7fd623b6c180c8d93a93040d1e0d61291f", "tree": "180b2b205e3ba6b0a91eec963761bfb77b0984ab", "parents": [ "0d239f3b03efc78fb5b290aff6c747fecd3b98cb" ], "author": { "name": "Daniel Axtens", "email": "dja@axtens.net", "time": "Tue Nov 17 16:59:16 2020 +1100" }, "committer": { "name": "Michael Ellerman", "email": "mpe@ellerman.id.au", "time": "Thu Nov 19 23:47:25 2020 +1100" }, "message": "powerpc/64s: rename pnv|pseries_setup_rfi_flush to _setup_security_mitigations\n\npseries|pnv_setup_rfi_flush already does the count cache flush setup, and\nwe just added entry and uaccess flushes. So the name is not very accurate\nany more. In both platforms we then also immediately setup the STF flush.\n\nRename them to _setup_security_mitigations and fold the STF flush in.\n\nSigned-off-by: Daniel Axtens \u003cdja@axtens.net\u003e\nSigned-off-by: Michael Ellerman \u003cmpe@ellerman.id.au\u003e\n" }, { "commit": "0d239f3b03efc78fb5b290aff6c747fecd3b98cb", "tree": "394f601c1777fb63ba77e10e92ad1d4d155c5c4d", "parents": [ "89a83a0c69c81a25ce91002b90ca27ed86132a0a" ], "author": { "name": "Daniel Axtens", "email": "dja@axtens.net", "time": "Tue Nov 17 16:59:15 2020 +1100" }, "committer": { "name": "Michael Ellerman", "email": "mpe@ellerman.id.au", "time": "Thu Nov 19 23:47:23 2020 +1100" }, "message": "selftests/powerpc: refactor entry and rfi_flush tests\n\nFor simplicity in backporting, the original entry_flush test contained\na lot of duplicated code from the rfi_flush test. De-duplicate that code.\n\nSigned-off-by: Daniel Axtens \u003cdja@axtens.net\u003e\nSigned-off-by: Michael Ellerman \u003cmpe@ellerman.id.au\u003e\n" }, { "commit": "89a83a0c69c81a25ce91002b90ca27ed86132a0a", "tree": "59b04b6e9d8b44d75faa31db6d54f14910a65e2d", "parents": [ "178d52c6e89c38d0553b0ac8b99927b11eb995b0" ], "author": { "name": "Daniel Axtens", "email": "dja@axtens.net", "time": "Tue Nov 17 16:59:14 2020 +1100" }, "committer": { "name": "Michael Ellerman", "email": "mpe@ellerman.id.au", "time": "Thu Nov 19 23:47:20 2020 +1100" }, "message": "selftests/powerpc: entry flush test\n\nAdd a test modelled on the RFI flush test which counts the number\nof L1D misses doing a simple syscall with the entry flush on and off.\n\nFor simplicity of backporting, this test duplicates a lot of code from\nrfi_flush. We clean that up in the next patch.\n\nSigned-off-by: Daniel Axtens \u003cdja@axtens.net\u003e\nSigned-off-by: Michael Ellerman \u003cmpe@ellerman.id.au\u003e\n" }, { "commit": "178d52c6e89c38d0553b0ac8b99927b11eb995b0", "tree": "5be07736829467698d8a7f4ed691c8d9b0487db1", "parents": [ "9a32a7e78bd0cd9a9b6332cbdc345ee5ffd0c5de" ], "author": { "name": "Michael Ellerman", "email": "mpe@ellerman.id.au", "time": "Thu Nov 19 23:43:53 2020 +1100" }, "committer": { "name": "Michael Ellerman", "email": "mpe@ellerman.id.au", "time": "Thu Nov 19 23:47:20 2020 +1100" }, "message": "powerpc: Only include kup-radix.h for 64-bit Book3S\n\nIn kup.h we currently include kup-radix.h for all 64-bit builds, which\nincludes Book3S and Book3E. The latter doesn\u0027t make sense, Book3E\nnever uses the Radix MMU.\n\nThis has worked up until now, but almost by accident, and the recent\nuaccess flush changes introduced a build breakage on Book3E because of\nthe bad structure of the code.\n\nSo disentangle things so that we only use kup-radix.h for Book3S. This\nrequires some more stubs in kup.h and fixing an include in\nsyscall_64.c.\n\nSigned-off-by: Michael Ellerman \u003cmpe@ellerman.id.au\u003e\n" }, { "commit": "9a32a7e78bd0cd9a9b6332cbdc345ee5ffd0c5de", "tree": "e9f7d6a2b72ddc59cb30cc87a36e635c2e0ab243", "parents": [ "f79643787e0a0762d2409b7b8334e83f22d85695" ], "author": { "name": "Nicholas Piggin", "email": "npiggin@gmail.com", "time": "Tue Nov 17 16:59:13 2020 +1100" }, "committer": { "name": "Michael Ellerman", "email": "mpe@ellerman.id.au", "time": "Thu Nov 19 23:47:18 2020 +1100" }, "message": "powerpc/64s: flush L1D after user accesses\n\nIBM Power9 processors can speculatively operate on data in the L1 cache\nbefore it has been completely validated, via a way-prediction mechanism. It\nis not possible for an attacker to determine the contents of impermissible\nmemory using this method, since these systems implement a combination of\nhardware and software security measures to prevent scenarios where\nprotected data could be leaked.\n\nHowever these measures don\u0027t address the scenario where an attacker induces\nthe operating system to speculatively execute instructions using data that\nthe attacker controls. This can be used for example to speculatively bypass\n\"kernel user access prevention\" techniques, as discovered by Anthony\nSteinhauser of Google\u0027s Safeside Project. This is not an attack by itself,\nbut there is a possibility it could be used in conjunction with\nside-channels or other weaknesses in the privileged code to construct an\nattack.\n\nThis issue can be mitigated by flushing the L1 cache between privilege\nboundaries of concern. This patch flushes the L1 cache after user accesses.\n\nThis is part of the fix for CVE-2020-4788.\n\nSigned-off-by: Nicholas Piggin \u003cnpiggin@gmail.com\u003e\nSigned-off-by: Daniel Axtens \u003cdja@axtens.net\u003e\nSigned-off-by: Michael Ellerman \u003cmpe@ellerman.id.au\u003e\n" }, { "commit": "f79643787e0a0762d2409b7b8334e83f22d85695", "tree": "ed004d08dce4f9fa4a8f42024d63cf376adb5a9f", "parents": [ "fcb48454c23c5679d1a2e252f127642e91b05cbe" ], "author": { "name": "Nicholas Piggin", "email": "npiggin@gmail.com", "time": "Tue Nov 17 16:59:12 2020 +1100" }, "committer": { "name": "Michael Ellerman", "email": "mpe@ellerman.id.au", "time": "Thu Nov 19 23:47:15 2020 +1100" }, "message": "powerpc/64s: flush L1D on kernel entry\n\nIBM Power9 processors can speculatively operate on data in the L1 cache\nbefore it has been completely validated, via a way-prediction mechanism. It\nis not possible for an attacker to determine the contents of impermissible\nmemory using this method, since these systems implement a combination of\nhardware and software security measures to prevent scenarios where\nprotected data could be leaked.\n\nHowever these measures don\u0027t address the scenario where an attacker induces\nthe operating system to speculatively execute instructions using data that\nthe attacker controls. This can be used for example to speculatively bypass\n\"kernel user access prevention\" techniques, as discovered by Anthony\nSteinhauser of Google\u0027s Safeside Project. This is not an attack by itself,\nbut there is a possibility it could be used in conjunction with\nside-channels or other weaknesses in the privileged code to construct an\nattack.\n\nThis issue can be mitigated by flushing the L1 cache between privilege\nboundaries of concern. This patch flushes the L1 cache on kernel entry.\n\nThis is part of the fix for CVE-2020-4788.\n\nSigned-off-by: Nicholas Piggin \u003cnpiggin@gmail.com\u003e\nSigned-off-by: Daniel Axtens \u003cdja@axtens.net\u003e\nSigned-off-by: Michael Ellerman \u003cmpe@ellerman.id.au\u003e\n" }, { "commit": "fcb48454c23c5679d1a2e252f127642e91b05cbe", "tree": "dce94d41fdb52ee4373e3017840f673957ed8c73", "parents": [ "3cea11cd5e3b00d91caf0b4730194039b45c5891" ], "author": { "name": "Russell Currey", "email": "ruscur@russell.cc", "time": "Tue Nov 17 16:59:11 2020 +1100" }, "committer": { "name": "Michael Ellerman", "email": "mpe@ellerman.id.au", "time": "Thu Nov 19 23:47:12 2020 +1100" }, "message": "selftests/powerpc: rfi_flush: disable entry flush if present\n\nWe are about to add an entry flush. The rfi (exit) flush test measures\nthe number of L1D flushes over a syscall with the RFI flush enabled and\ndisabled. But if the entry flush is also enabled, the effect of enabling\nand disabling the RFI flush is masked.\n\nIf there is a debugfs entry for the entry flush, disable it during the RFI\nflush and restore it later.\n\nReported-by: Spoorthy S \u003cspoorts2@in.ibm.com\u003e\nSigned-off-by: Russell Currey \u003cruscur@russell.cc\u003e\nSigned-off-by: Daniel Axtens \u003cdja@axtens.net\u003e\nSigned-off-by: Michael Ellerman \u003cmpe@ellerman.id.au\u003e\n" }, { "commit": "f0b0a2d8a1fd4e862f8f5bf8bf449d116264ce5c", "tree": "4d6f08c92640b194edcdc568f923849a4b62e569", "parents": [ "6d9c8d15af0ef20a66a0b432cac0d08319920602", "a1f634463aaf2c94dfa13001dbdea011303124cc" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Nov 18 17:49:24 2020 -0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Nov 18 17:49:24 2020 -0800" }, "message": "Merge tag \u0027linux-can-fixes-for-5.10-20201118\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 2020-11-18\n\nJimmy Assarsson provides two patches for the kvaser_pciefd and kvaser_usb\ndrivers, where the can_bittiming_const are fixed.\n\nThe next patch is by me and fixes an erroneous flexcan_transceiver_enable()\nduring bus-off recovery in the flexcan driver.\n\nJarkko Nikula\u0027s patch for the m_can driver fixes the IRQ handler to only\nprocess the interrupts if the device is not suspended.\n\n* tag \u0027linux-can-fixes-for-5.10-20201118\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can:\n can: m_can: process interrupt only when not runtime suspended\n can: flexcan: flexcan_chip_start(): fix erroneous flexcan_transceiver_enable() during bus-off recovery\n can: kvaser_usb: kvaser_usb_hydra: Fix KCAN bittiming limits\n can: kvaser_pciefd: Fix KCAN bittiming limits\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/20201118160414.2731659-1-mkl@pengutronix.de\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "6d9c8d15af0ef20a66a0b432cac0d08319920602", "tree": "bfb80ce2ddafe292ca580c3cf5d1d3e49f316134", "parents": [ "f93e8497a9bc6d4b7a0efcb9f54e36f806544a38" ], "author": { "name": "Aya Levin", "email": "ayal@nvidia.com", "time": "Wed Nov 18 10:19:22 2020 +0200" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Nov 18 17:46:20 2020 -0800" }, "message": "net/mlx4_core: Fix init_hca fields offset\n\nSlave function read the following capabilities from the wrong offset:\n1. log_mc_entry_sz\n2. fs_log_entry_sz\n3. log_mc_hash_sz\n\nFix that by adjusting these capabilities offset to match firmware\nlayout.\n\nDue to the wrong offset read, the following issues might occur:\n1+2. Negative value reported at max_mcast_qp_attach.\n3. Driver to init FW with multicast hash size of zero.\n\nFixes: a40ded604365 (\"net/mlx4_core: Add masking for a few queries on HCA caps\")\nSigned-off-by: Aya Levin \u003cayal@nvidia.com\u003e\nReviewed-by: Moshe Shemesh \u003cmoshe@nvidia.com\u003e\nReviewed-by: Eran Ben Elisha \u003ceranbe@nvidia.com\u003e\nSigned-off-by: Tariq Toukan \u003ctariqt@nvidia.com\u003e\nLink: https://lore.kernel.org/r/20201118081922.553-1-tariqt@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "f93e8497a9bc6d4b7a0efcb9f54e36f806544a38", "tree": "924744d385e42da2322c748d63951e25da72f789", "parents": [ "6dceaa9f56e22d0f9b4c4ad2ed9e04e315ce7fe5", "68ec32daf7d50a9f7425f8607a7402c13aa0c587" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Nov 18 17:17:32 2020 -0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Nov 18 17:17:33 2020 -0800" }, "message": "Merge tag \u0027mlx5-fixes-2020-11-17\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux\n\nSaeed Mahameed says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nmlx5 fixes 2020-11-17\n\nThis series introduces some fixes to mlx5 driver.\n\n* tag \u0027mlx5-fixes-2020-11-17\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux:\n net/mlx5: fix error return code in mlx5e_tc_nic_init()\n net/mlx5: E-Switch, Fail mlx5_esw_modify_vport_rate if qos disabled\n net/mlx5: Disable QoS when min_rates on all VFs are zero\n net/mlx5: Clear bw_share upon VF disable\n net/mlx5: Add handling of port type in rule deletion\n net/mlx5e: Fix check if netdev is bond slave\n net/mlx5e: Fix IPsec packet drop by mlx5e_tc_update_skb\n net/mlx5e: Set IPsec WAs only in IP\u0027s non checksum partial case.\n net/mlx5e: Fix refcount leak on kTLS RX resync\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/20201117195702.386113-1-saeedm@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "6dceaa9f56e22d0f9b4c4ad2ed9e04e315ce7fe5", "tree": "64bcf02a80ddf3c915c55c92af4cd3b8f374c052", "parents": [ "d8c19014bba8f565d8a2f1f46b4e38d1d97bf1a7" ], "author": { "name": "Sebastian Andrzej Siewior", "email": "bigeasy@linutronix.de", "time": "Mon Nov 16 17:21:14 2020 +0100" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Nov 18 16:42:07 2020 -0800" }, "message": "atm: nicstar: Unmap DMA on send error\n\nThe `skb\u0027 is mapped for DMA in ns_send() but does not unmap DMA in case\npush_scqe() fails to submit the `skb\u0027. The memory of the `skb\u0027 is\nreleased so only the DMA mapping is leaking.\n\nUnmap the DMA mapping in case push_scqe() failed.\n\nFixes: 864a3ff635fa7 (\"atm: [nicstar] remove virt_to_bus() and support 64-bit platforms\")\nCc: Chas Williams \u003c3chas3@gmail.com\u003e\nSigned-off-by: Sebastian Andrzej Siewior \u003cbigeasy@linutronix.de\u003e\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "d8c19014bba8f565d8a2f1f46b4e38d1d97bf1a7", "tree": "178b41464b312cbea27acd19bf381880fb90755e", "parents": [ "a3dcb3e7e70c72a68a79b30fc3a3adad5612731c" ], "author": { "name": "Dongli Zhang", "email": "dongli.zhang@oracle.com", "time": "Sun Nov 15 12:10:29 2020 -0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Nov 18 15:21:56 2020 -0800" }, "message": "page_frag: Recover from memory pressure\n\nThe ethernet driver may allocate skb (and skb-\u003edata) via napi_alloc_skb().\nThis ends up to page_frag_alloc() to allocate skb-\u003edata from\npage_frag_cache-\u003eva.\n\nDuring the memory pressure, page_frag_cache-\u003eva may be allocated as\npfmemalloc page. As a result, the skb-\u003epfmemalloc is always true as\nskb-\u003edata is from page_frag_cache-\u003eva. The skb will be dropped if the\nsock (receiver) does not have SOCK_MEMALLOC. This is expected behaviour\nunder memory pressure.\n\nHowever, once kernel is not under memory pressure any longer (suppose large\namount of memory pages are just reclaimed), the page_frag_alloc() may still\nre-use the prior pfmemalloc page_frag_cache-\u003eva to allocate skb-\u003edata. As a\nresult, the skb-\u003epfmemalloc is always true unless page_frag_cache-\u003eva is\nre-allocated, even if the kernel is not under memory pressure any longer.\n\nHere is how kernel runs into issue.\n\n1. The kernel is under memory pressure and allocation of\nPAGE_FRAG_CACHE_MAX_ORDER in __page_frag_cache_refill() will fail. Instead,\nthe pfmemalloc page is allocated for page_frag_cache-\u003eva.\n\n2: All skb-\u003edata from page_frag_cache-\u003eva (pfmemalloc) will have\nskb-\u003epfmemalloc\u003dtrue. The skb will always be dropped by sock without\nSOCK_MEMALLOC. This is an expected behaviour.\n\n3. Suppose a large amount of pages are reclaimed and kernel is not under\nmemory pressure any longer. We expect skb-\u003epfmemalloc drop will not happen.\n\n4. Unfortunately, page_frag_alloc() does not proactively re-allocate\npage_frag_alloc-\u003eva and will always re-use the prior pfmemalloc page. The\nskb-\u003epfmemalloc is always true even kernel is not under memory pressure any\nlonger.\n\nFix this by freeing and re-allocating the page instead of recycling it.\n\nReferences: https://lore.kernel.org/lkml/20201103193239.1807-1-dongli.zhang@oracle.com/\nReferences: https://lore.kernel.org/linux-mm/20201105042140.5253-1-willy@infradead.org/\nSuggested-by: Matthew Wilcox (Oracle) \u003cwilly@infradead.org\u003e\nCc: Aruna Ramakrishna \u003caruna.ramakrishna@oracle.com\u003e\nCc: Bert Barbe \u003cbert.barbe@oracle.com\u003e\nCc: Rama Nichanamatlu \u003crama.nichanamatlu@oracle.com\u003e\nCc: Venkat Venkatsubra \u003cvenkat.x.venkatsubra@oracle.com\u003e\nCc: Manjunath Patil \u003cmanjunath.b.patil@oracle.com\u003e\nCc: Joe Jin \u003cjoe.jin@oracle.com\u003e\nCc: SRINIVAS \u003csrinivas.eeda@oracle.com\u003e\nFixes: 79930f5892e1 (\"net: do not deplete pfmemalloc reserve\")\nSigned-off-by: Dongli Zhang \u003cdongli.zhang@oracle.com\u003e\nAcked-by: Vlastimil Babka \u003cvbabka@suse.cz\u003e\nReviewed-by: Eric Dumazet \u003cedumazet@google.com\u003e\nLink: https://lore.kernel.org/r/20201115201029.11903-1-dongli.zhang@oracle.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "c2e7554e1b85935d962127efa3c2a76483b0b3b6", "tree": "6fae06eb24edbeb0f249ee0f883ac2aa79319394", "parents": [ "ce228d45942492580968d698b0216d3356d75226", "20b329129009caf1c646152abe09b697227e1c37" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Nov 18 12:12:18 2020 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Nov 18 12:12:18 2020 -0800" }, "message": "Merge tag \u0027gfs2-v5.10-rc4-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2\n\nPull gfs2 fix from Andreas Gruenbacher:\n \"Fix gfs2 freeze/thaw\"\n\n* tag \u0027gfs2-v5.10-rc4-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2:\n gfs2: Fix regression in freeze_go_sync\n" }, { "commit": "ce228d45942492580968d698b0216d3356d75226", "tree": "f86887e7b129385f294ff3cd5a806aa636049e51", "parents": [ "f86fee1845eec29bed735e7120e6993ad8c4e295", "c3213d260a23e263ef85ba21ac68c9e7578020b5" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Nov 18 12:06:34 2020 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Nov 18 12:06:34 2020 -0800" }, "message": "Merge tag \u0027nfsd-5.10-2\u0027 of git://linux-nfs.org/~bfields/linux\n\nPull nfsd fix from Bruce Fields:\n \"Just one quick fix for a tracing oops\"\n\n* tag \u0027nfsd-5.10-2\u0027 of git://linux-nfs.org/~bfields/linux:\n SUNRPC: Fix oops in the rpc_xdr_buf event class\n" }, { "commit": "f86fee1845eec29bed735e7120e6993ad8c4e295", "tree": "deea8ce84e2b7f8184f2f324176a14da4a1541e8", "parents": [ "0fa8ee0d9ab95c9350b8b84574824d9a384a9f7d", "3084db0e0d5076cd48408274ab0911cd3ccdae88" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Nov 18 11:57:55 2020 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Nov 18 11:57:55 2020 -0800" }, "message": "Merge tag \u0027linux-kselftest-kunit-fixes-5.10-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest\n\nPull Kunit fixes from Shuah Khan:\n \"Several fixes to Kunit documentation and tools, and to not pollute\n the source directory.\n\n Also remove the incorrect kunit .gitattributes file\"\n\n* tag \u0027linux-kselftest-kunit-fixes-5.10-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest:\n kunit: fix display of failed expectations for strings\n kunit: tool: fix extra trailing \\n in raw + parsed test output\n kunit: tool: print out stderr from make (like build warnings)\n KUnit: Docs: usage: wording fixes\n KUnit: Docs: style: fix some Kconfig example issues\n KUnit: Docs: fix a wording typo\n kunit: Do not pollute source directory with generated files (test.log)\n kunit: Do not pollute source directory with generated files (.kunitconfig)\n kunit: tool: fix pre-existing python type annotation errors\n kunit: Fix kunit.py parse subcommand (use null build_dir)\n kunit: tool: unmark test_data as binary blobs\n" }, { "commit": "a3dcb3e7e70c72a68a79b30fc3a3adad5612731c", "tree": "140bb8b5445fdb32c135012e182d2891f7bb0ec9", "parents": [ "fc2635bff24c6b75291ead011c16da7a053fb4f5" ], "author": { "name": "Andrew Lunn", "email": "andrew@lunn.ch", "time": "Mon Nov 16 08:43:01 2020 -0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Nov 18 11:24:44 2020 -0800" }, "message": "net: dsa: mv88e6xxx: Wait for EEPROM done after HW reset\n\nWhen the switch is hardware reset, it reads the contents of the\nEEPROM. This can contain instructions for programming values into\nregisters and to perform waits between such programming. Reading the\nEEPROM can take longer than the 100ms mv88e6xxx_hardware_reset() waits\nafter deasserting the reset GPIO. So poll the EEPROM done bit to\nensure it is complete.\n\nSigned-off-by: Andrew Lunn \u003candrew@lunn.ch\u003e\nSigned-off-by: Ruslan Sushko \u003crus@sushko.dev\u003e\nLink: https://lore.kernel.org/r/20201116164301.977661-1-rus@sushko.dev\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "fc2635bff24c6b75291ead011c16da7a053fb4f5", "tree": "1bbd0f3e23b03326453d93ba3d00f7dd5e8442f7", "parents": [ "1532b9778478577152201adbafa7738b1e844868", "1f492eab67bced119a0ac7db75ef2047e29a30c6" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Nov 18 11:07:03 2020 -0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Nov 18 11:07:04 2020 -0800" }, "message": "Merge branch \u0027mlxsw-couple-of-fixes\u0027\n\nIdo Schimmel says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nmlxsw: Couple of fixes\n\nPatch #1 fixes firmware flashing when CONFIG_MLXSW_CORE\u003dy and\nCONFIG_MLXFW\u003dm.\n\nPatch #2 prevents EMAD transactions from needlessly failing when the\nsystem is under heavy load by using exponential backoff.\n\nPlease consider patch #2 for stable.\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/20201117173352.288491-1-idosch@idosch.org\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "1f492eab67bced119a0ac7db75ef2047e29a30c6", "tree": "1bbd0f3e23b03326453d93ba3d00f7dd5e8442f7", "parents": [ "fb738b99ef229bd3d25f1b3e5503925dba9b1a7c" ], "author": { "name": "Ido Schimmel", "email": "idosch@nvidia.com", "time": "Tue Nov 17 19:33:52 2020 +0200" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Nov 18 11:07:00 2020 -0800" }, "message": "mlxsw: core: Use variable timeout for EMAD retries\n\nThe driver sends Ethernet Management Datagram (EMAD) packets to the\ndevice for configuration purposes and waits for up to 200ms for a reply.\nA request is retried up to 5 times.\n\nWhen the system is under heavy load, replies are not always processed in\ntime and EMAD transactions fail.\n\nMake the process more robust to such delays by using exponential\nbackoff. First wait for up to 200ms, then retransmit and wait for up to\n400ms and so on.\n\nFixes: caf7297e7ab5 (\"mlxsw: core: Introduce support for asynchronous EMAD register access\")\nReported-by: Denis Yulevich \u003cdenisyu@nvidia.com\u003e\nTested-by: Denis Yulevich \u003cdenisyu@nvidia.com\u003e\nSigned-off-by: Ido Schimmel \u003cidosch@nvidia.com\u003e\nReviewed-by: Jiri Pirko \u003cjiri@nvidia.com\u003e\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "fb738b99ef229bd3d25f1b3e5503925dba9b1a7c", "tree": "aa1e20e2e2d9d50193e55180101c89a7669e49ea", "parents": [ "1532b9778478577152201adbafa7738b1e844868" ], "author": { "name": "Ido Schimmel", "email": "idosch@nvidia.com", "time": "Tue Nov 17 19:33:51 2020 +0200" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Nov 18 11:07:00 2020 -0800" }, "message": "mlxsw: Fix firmware flashing\n\nThe commit cited below moved firmware flashing functionality from\nmlxsw_spectrum to mlxsw_core, but did not adjust the Kconfig\ndependencies. This makes it possible to have mlxsw_core as built-in and\nmlxfw as a module. The mlxfw code is therefore not reachable from\nmlxsw_core and firmware flashing fails:\n\n# devlink dev flash pci/0000:01:00.0 file mellanox/mlxsw_spectrum-13.2008.1310.mfa2\ndevlink answers: Operation not supported\n\nFix by having mlxsw_core select mlxfw.\n\nFixes: b79cb787ac70 (\"mlxsw: Move fw flashing code into core.c\")\nSigned-off-by: Ido Schimmel \u003cidosch@nvidia.com\u003e\nReported-by: Vadim Pasternak \u003cvadimp@nvidia.com\u003e\nTested-by: Vadim Pasternak \u003cvadimp@nvidia.com\u003e\nReviewed-by: Jiri Pirko \u003cjiri@nvidia.com\u003e\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "1532b9778478577152201adbafa7738b1e844868", "tree": "6c168fb5db64a50ddf77f2c156cd3db670481ed1", "parents": [ "3a36060bf294e7b7e33c5dddcc4f5d2c1c834e56" ], "author": { "name": "Florian Fainelli", "email": "f.fainelli@gmail.com", "time": "Mon Nov 16 19:52:34 2020 -0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Nov 18 11:04:11 2020 -0800" }, "message": "net: Have netpoll bring-up DSA management interface\n\nDSA network devices rely on having their DSA management interface up and\nrunning otherwise their ndo_open() will return -ENETDOWN. Without doing\nthis it would not be possible to use DSA devices as netconsole when\nconfigured on the command line. These devices also do not utilize the\nupper/lower linking so the check about the netpoll device having upper\nis not going to be a problem.\n\nThe solution adopted here is identical to the one done for\nnet/ipv4/ipconfig.c with 728c02089a0e (\"net: ipv4: handle DSA enabled\nmaster network devices\"), with the network namespace scope being\nrestricted to that of the process configuring netpoll.\n\nFixes: 04ff53f96a93 (\"net: dsa: Add netconsole support\")\nTested-by: Vladimir Oltean \u003colteanv@gmail.com\u003e\nSigned-off-by: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nLink: https://lore.kernel.org/r/20201117035236.22658-1-f.fainelli@gmail.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "3a36060bf294e7b7e33c5dddcc4f5d2c1c834e56", "tree": "3329a9555153bed1c2e3bb5d08db0c4367c08a96", "parents": [ "537a14726582c4e7bfe4dff9cb7fca19dc912cf6" ], "author": { "name": "Zhang Changzhong", "email": "zhangchangzhong@huawei.com", "time": "Tue Nov 17 10:57:55 2020 +0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Nov 18 11:02:15 2020 -0800" }, "message": "atl1e: fix error return code in atl1e_probe()\n\nFix to return a negative error code from the error handling\ncase instead of 0, as done elsewhere in this function.\n\nFixes: a6a5325239c2 (\"atl1e: Atheros L1E Gigabit Ethernet driver\")\nReported-by: Hulk Robot \u003chulkci@huawei.com\u003e\nSigned-off-by: Zhang Changzhong \u003czhangchangzhong@huawei.com\u003e\nLink: https://lore.kernel.org/r/1605581875-36281-1-git-send-email-zhangchangzhong@huawei.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "537a14726582c4e7bfe4dff9cb7fca19dc912cf6", "tree": "56521ed18ad8e14518abadbfaf19d5337e03b2b9", "parents": [ "a5ebcbdf34b65fcc07f38eaf2d60563b42619a59" ], "author": { "name": "Zhang Changzhong", "email": "zhangchangzhong@huawei.com", "time": "Tue Nov 17 10:55:21 2020 +0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Nov 18 11:00:17 2020 -0800" }, "message": "atl1c: fix error return code in atl1c_probe()\n\nFix to return a negative error code from the error handling\ncase instead of 0, as done elsewhere in this function.\n\nFixes: 43250ddd75a3 (\"atl1c: Atheros L1C Gigabit Ethernet driver\")\nReported-by: Hulk Robot \u003chulkci@huawei.com\u003e\nSigned-off-by: Zhang Changzhong \u003czhangchangzhong@huawei.com\u003e\nLink: https://lore.kernel.org/r/1605581721-36028-1-git-send-email-zhangchangzhong@huawei.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "a5ebcbdf34b65fcc07f38eaf2d60563b42619a59", "tree": "9f744b49e91c52559864c87506e9bf9fa61a42da", "parents": [ "df8d85d8c69d6837817e54dcb73c84a8b5a13877" ], "author": { "name": "Zhang Changzhong", "email": "zhangchangzhong@huawei.com", "time": "Tue Nov 17 10:45:05 2020 +0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Nov 18 10:53:16 2020 -0800" }, "message": "ah6: fix error return code in ah6_input()\n\nFix to return a negative error code from the error handling\ncase instead of 0, as done elsewhere in this function.\n\nFixes: 1da177e4c3f4 (\"Linux-2.6.12-rc2\")\nReported-by: Hulk Robot \u003chulkci@huawei.com\u003e\nSigned-off-by: Zhang Changzhong \u003czhangchangzhong@huawei.com\u003e\nLink: https://lore.kernel.org/r/1605581105-35295-1-git-send-email-zhangchangzhong@huawei.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "df8d85d8c69d6837817e54dcb73c84a8b5a13877", "tree": "db3df5aa7cdd57067711d4a451a4337105f54b1f", "parents": [ "c09c8a27b9baa417864b9adc3228b10ae5eeec93" ], "author": { "name": "Filip Moc", "email": "dev@moc6.cz", "time": "Tue Nov 17 18:36:31 2020 +0100" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Nov 18 10:49:15 2020 -0800" }, "message": "net: usb: qmi_wwan: Set DTR quirk for MR400\n\nLTE module MR400 embedded in TL-MR6400 v4 requires DTR to be set.\n\nSigned-off-by: Filip Moc \u003cdev@moc6.cz\u003e\nAcked-by: Bjørn Mork \u003cbjorn@mork.no\u003e\nLink: https://lore.kernel.org/r/20201117173631.GA550981@moc6.cz\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "2ba546ebe0ce2af47833d8912ced9b4a579f13cb", "tree": "70cf6426654a8d633a5512d671f241d39bd6f27e", "parents": [ "f5c042b23f7429e5c2ac987b01a31c69059a978b" ], "author": { "name": "Nishanth Menon", "email": "nm@ti.com", "time": "Wed Nov 18 08:50:09 2020 -0600" }, "committer": { "name": "Mark Brown", "email": "broonie@kernel.org", "time": "Wed Nov 18 17:59:24 2020 +0000" }, "message": "regulator: ti-abb: Fix array out of bound read access on the first transition\n\nAt the start of driver initialization, we do not know what bias\nsetting the bootloader has configured the system for and we only know\nfor certain the very first time we do a transition.\n\nHowever, since the initial value of the comparison index is -EINVAL,\nthis negative value results in an array out of bound access on the\nvery first transition.\n\nSince we don\u0027t know what the setting is, we just set the bias\nconfiguration as there is nothing to compare against. This prevents\nthe array out of bound access.\n\nNOTE: Even though we could use a more relaxed check of \"\u003c 0\" the only\nvalid values(ignoring cosmic ray induced bitflips) are -EINVAL, 0+.\n\nFixes: 40b1936efebd (\"regulator: Introduce TI Adaptive Body Bias(ABB) on-chip LDO driver\")\nLink: https://lore.kernel.org/linux-mm/CA+G9fYuk4imvhyCN7D7T6PMDH6oNp6HDCRiTUKMQ6QXXjBa4ag@mail.gmail.com/\nReported-by: Naresh Kamboju \u003cnaresh.kamboju@linaro.org\u003e\nReviewed-by: Arnd Bergmann \u003carnd@arndb.de\u003e\nSigned-off-by: Nishanth Menon \u003cnm@ti.com\u003e\nLink: https://lore.kernel.org/r/20201118145009.10492-1-nm@ti.com\nSigned-off-by: Mark Brown \u003cbroonie@kernel.org\u003e\n" }, { "commit": "a1f634463aaf2c94dfa13001dbdea011303124cc", "tree": "3a790dcb197057001adea5a423db03a3e128c7f4", "parents": [ "cd9f13c59461351d7a5fd07924264fb49b287359" ], "author": { "name": "Jarkko Nikula", "email": "jarkko.nikula@linux.intel.com", "time": "Tue Sep 15 16:47:15 2020 +0300" }, "committer": { "name": "Marc Kleine-Budde", "email": "mkl@pengutronix.de", "time": "Wed Nov 18 16:37:32 2020 +0100" }, "message": "can: m_can: process interrupt only when not runtime suspended\n\nAvoid processing bogus interrupt statuses when the HW is runtime suspended and\nthe M_CAN_IR register read may get all bits 1\u0027s. Handler can be called if the\ninterrupt request is shared with other peripherals or at the end of free_irq().\n\nTherefore check the runtime suspended status before processing.\n\nFixes: cdf8259d6573 (\"can: m_can: Add PM Support\")\nSigned-off-by: Jarkko Nikula \u003cjarkko.nikula@linux.intel.com\u003e\nLink: https://lore.kernel.org/r/20200915134715.696303-1-jarkko.nikula@linux.intel.com\nSigned-off-by: Marc Kleine-Budde \u003cmkl@pengutronix.de\u003e\n" }, { "commit": "20b329129009caf1c646152abe09b697227e1c37", "tree": "585cd61c8decf037015a12773334c0b847471894", "parents": [ "09162bc32c880a791c6c0668ce0745cf7958f576" ], "author": { "name": "Bob Peterson", "email": "rpeterso@redhat.com", "time": "Wed Nov 18 08:54:31 2020 -0500" }, "committer": { "name": "Andreas Gruenbacher", "email": "agruenba@redhat.com", "time": "Wed Nov 18 16:28:11 2020 +0100" }, "message": "gfs2: Fix regression in freeze_go_sync\n\nPatch 541656d3a513 (\"gfs2: freeze should work on read-only mounts\") changed\nthe check for glock state in function freeze_go_sync() from \"gl-\u003egl_state\n\u003d\u003d LM_ST_SHARED\" to \"gl-\u003egl_req \u003d\u003d LM_ST_EXCLUSIVE\". That\u0027s wrong and it\nregressed gfs2\u0027s freeze/thaw mechanism because it caused only the freezing\nnode (which requests the glock in EX) to queue freeze work.\n\nAll nodes go through this go_sync code path during the freeze to drop their\nSHared hold on the freeze glock, allowing the freezing node to acquire it\nin EXclusive mode. But all the nodes must freeze access to the file system\nlocally, so they ALL must queue freeze work. The freeze_work calls\nfreeze_func, which makes a request to reacquire the freeze glock in SH,\neffectively blocking until the thaw from the EX holder. Once thawed, the\nfreezing node drops its EX hold on the freeze glock, then the (blocked)\nfreeze_func reacquires the freeze glock in SH again (on all nodes, including\nthe freezer) so all nodes go back to a thawed state.\n\nThis patch changes the check back to gl_state \u003d\u003d LM_ST_SHARED like it was\nprior to 541656d3a513.\n\nFixes: 541656d3a513 (\"gfs2: freeze should work on read-only mounts\")\nCc: stable@vger.kernel.org # v5.8+\nSigned-off-by: Bob Peterson \u003crpeterso@redhat.com\u003e\nSigned-off-by: Andreas Gruenbacher \u003cagruenba@redhat.com\u003e\n" }, { "commit": "cd9f13c59461351d7a5fd07924264fb49b287359", "tree": "23383f4ee21765eeaf6ea1f1d9a3494d13025684", "parents": [ "d003868d7f8579838ed58b6429af91844039b6f8" ], "author": { "name": "Marc Kleine-Budde", "email": "mkl@pengutronix.de", "time": "Wed Nov 18 16:01:48 2020 +0100" }, "committer": { "name": "Marc Kleine-Budde", "email": "mkl@pengutronix.de", "time": "Wed Nov 18 16:02:34 2020 +0100" }, "message": "can: flexcan: flexcan_chip_start(): fix erroneous flexcan_transceiver_enable() during bus-off recovery\n\nIf the CAN controller goes into bus off, the do_set_mode() callback with\nCAN_MODE_START can be used to recover the controller, which then calls\nflexcan_chip_start(). If configured, this is done automatically by the\nframework or manually by the user.\n\nIn flexcan_chip_start() there is an explicit call to\nflexcan_transceiver_enable(), which does a regulator_enable() on the\ntransceiver regulator. This results in a net usage counter increase, as there\nis no corresponding flexcan_transceiver_disable() in the bus off code path.\nThis further leads to the transceiver stuck enabled, even if the CAN interface\nis shut down.\n\nTo fix this problem the\nflexcan_transceiver_enable()/flexcan_transceiver_disable() are moved out of\nflexcan_chip_start()/flexcan_chip_stop() into flexcan_open()/flexcan_close().\n\nFixes: e955cead0311 (\"CAN: Add Flexcan CAN controller driver\")\nLink: https://lore.kernel.org/r/20201118150148.2664024-1-mkl@pengutronix.de\nSigned-off-by: Marc Kleine-Budde \u003cmkl@pengutronix.de\u003e\n" }, { "commit": "d003868d7f8579838ed58b6429af91844039b6f8", "tree": "0b6649cb5d0fb344d1f78b4ef2baa0fb4251dd8f", "parents": [ "470e14c00c63752466ac44de392f584dfdddd82e" ], "author": { "name": "Jimmy Assarsson", "email": "extja@kvaser.com", "time": "Sun Nov 15 17:30:23 2020 +0100" }, "committer": { "name": "Marc Kleine-Budde", "email": "mkl@pengutronix.de", "time": "Wed Nov 18 09:06:35 2020 +0100" }, "message": "can: kvaser_usb: kvaser_usb_hydra: Fix KCAN bittiming limits\n\nUse correct bittiming limits for the KCAN CAN controller.\n\nFixes: aec5fb2268b7 (\"can: kvaser_usb: Add support for Kvaser USB hydra family\")\nSigned-off-by: Jimmy Assarsson \u003cextja@kvaser.com\u003e\nLink: https://lore.kernel.org/r/20201115163027.16851-2-jimmyassarsson@gmail.com\nSigned-off-by: Marc Kleine-Budde \u003cmkl@pengutronix.de\u003e\n" }, { "commit": "470e14c00c63752466ac44de392f584dfdddd82e", "tree": "6266214d7ccb581c5755314cc5b629ba3468c93f", "parents": [ "c09c8a27b9baa417864b9adc3228b10ae5eeec93" ], "author": { "name": "Jimmy Assarsson", "email": "extja@kvaser.com", "time": "Sun Nov 15 17:30:22 2020 +0100" }, "committer": { "name": "Marc Kleine-Budde", "email": "mkl@pengutronix.de", "time": "Wed Nov 18 09:06:35 2020 +0100" }, "message": "can: kvaser_pciefd: Fix KCAN bittiming limits\n\nUse correct bittiming limits for the KCAN CAN controller.\n\nFixes: 26ad340e582d (\"can: kvaser_pciefd: Add driver for Kvaser PCIEcan devices\")\nSigned-off-by: Jimmy Assarsson \u003cextja@kvaser.com\u003e\nLink: https://lore.kernel.org/r/20201115163027.16851-1-jimmyassarsson@gmail.com\nSigned-off-by: Marc Kleine-Budde \u003cmkl@pengutronix.de\u003e\n" }, { "commit": "c09c8a27b9baa417864b9adc3228b10ae5eeec93", "tree": "e21b3702dc1bb7eb66a8b250a63f781eeabce8c7", "parents": [ "93be52612431e71ee8cb980ef11468997857e4c4" ], "author": { "name": "Florian Klink", "email": "flokli@flokli.de", "time": "Sun Nov 15 23:45:09 2020 +0100" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Nov 17 17:02:03 2020 -0800" }, "message": "ipv4: use IS_ENABLED instead of ifdef\n\nChecking for ifdef CONFIG_x fails if CONFIG_x\u003dm.\n\nUse IS_ENABLED instead, which is true for both built-ins and modules.\n\nOtherwise, a\n\u003e ip -4 route add 1.2.3.4/32 via inet6 fe80::2 dev eth1\nfails with the message \"Error: IPv6 support not enabled in kernel.\" if\nCONFIG_IPV6 is `m`.\n\nIn the spirit of b8127113d01e53adba15b41aefd37b90ed83d631.\n\nFixes: d15662682db2 (\"ipv4: Allow ipv6 gateway with ipv4 routes\")\nCc: Kim Phillips \u003ckim.phillips@arm.com\u003e\nSigned-off-by: Florian Klink \u003cflokli@flokli.de\u003e\nReviewed-by: David Ahern \u003cdsahern@kernel.org\u003e\nLink: https://lore.kernel.org/r/20201115224509.2020651-1-flokli@flokli.de\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "93be52612431e71ee8cb980ef11468997857e4c4", "tree": "132a01192ade703e87a4166b7581335d57398ea8", "parents": [ "e33de7c5317e2827b2ba6fd120a505e9eb727b05" ], "author": { "name": "Dmitry Bogdanov", "email": "dbogdanov@marvell.com", "time": "Mon Nov 16 16:29:44 2020 +0300" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Nov 17 16:35:32 2020 -0800" }, "message": "qed: fix ILT configuration of SRC block\n\nThe code refactoring of ILT configuration was not complete, the old\nunused variables were used for the SRC block. That could lead to the memory\ncorruption by HW when rx filters are configured.\nThis patch completes that refactoring.\n\nFixes: 8a52bbab39c9 (qed: Debug feature: ilt and mdump)\nSigned-off-by: Igor Russkikh \u003cirusskikh@marvell.com\u003e\nSigned-off-by: Ariel Elior \u003caelior@marvell.com\u003e\nSigned-off-by: Dmitry Bogdanov \u003cdbogdanov@marvell.com\u003e\nLink: https://lore.kernel.org/r/20201116132944.2055-1-dbogdanov@marvell.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "e33de7c5317e2827b2ba6fd120a505e9eb727b05", "tree": "007a281c9797d9377c3187441a777d079d11a70b", "parents": [ "138559b9f99d3b6b1d5e75c78facc067a23871c6" ], "author": { "name": "Wang Hai", "email": "wanghai38@huawei.com", "time": "Mon Nov 16 16:20:18 2020 +0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Nov 17 16:08:36 2020 -0800" }, "message": "inet_diag: Fix error path to cancel the meseage in inet_req_diag_fill()\n\nnlmsg_cancel() needs to be called in the error path of\ninet_req_diag_fill to cancel the message.\n\nFixes: d545caca827b (\"net: inet: diag: expose the socket mark to privileged processes.\")\nReported-by: Hulk Robot \u003chulkci@huawei.com\u003e\nSigned-off-by: Wang Hai \u003cwanghai38@huawei.com\u003e\nLink: https://lore.kernel.org/r/20201116082018.16496-1-wanghai38@huawei.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "ee415d73dcc24caef7f6bbf292dcc365613d2188", "tree": "5b4e911c9db9ab7dc194733ad31f82988aa57c78", "parents": [ "dabbd6abcdbeb1358a53ec28a244429320eb0e3a" ], "author": { "name": "Maor Gottlieb", "email": "maorg@nvidia.com", "time": "Sun Nov 15 14:06:23 2020 +0200" }, "committer": { "name": "Jason Gunthorpe", "email": "jgg@nvidia.com", "time": "Tue Nov 17 20:02:20 2020 -0400" }, "message": "tools/testing/scatterlist: Fix test to compile and run\n\nAdd missing define of ALIGN_DOWN to make the test build and run. In\naddition, __sg_alloc_table_from_pages now support unaligned maximum\nsegment, so adapt the test result accordingly.\n\nFixes: 07da1223ec93 (\"lib/scatterlist: Add support in dynamic allocation of SG table from pages\")\nLink: https://lore.kernel.org/r/20201115120623.139113-1-leon@kernel.org\nSigned-off-by: Maor Gottlieb \u003cmaorg@nvidia.com\u003e\nSigned-off-by: Leon Romanovsky \u003cleonro@nvidia.com\u003e\nSigned-off-by: Jason Gunthorpe \u003cjgg@nvidia.com\u003e\n" }, { "commit": "4363023d2668e621b0743db351a9555d6e6ea57e", "tree": "af1536961c439659efc1310db414eaece5c8a849", "parents": [ "2443ca66676d50a4eb3305c236bccd84a9828ce2" ], "author": { "name": "John Fastabend", "email": "john.fastabend@gmail.com", "time": "Mon Nov 16 14:29:28 2020 -0800" }, "committer": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Wed Nov 18 00:14:04 2020 +0100" }, "message": "bpf, sockmap: Avoid failures from skb_to_sgvec when skb has frag_list\n\nWhen skb has a frag_list its possible for skb_to_sgvec() to fail. This\nhappens when the scatterlist has fewer elements to store pages than would\nbe needed for the initial skb plus any of its frags.\n\nThis case appears rare, but is possible when running an RX parser/verdict\nprograms exposed to the internet. Currently, when this happens we throw\nan error, break the pipe, and kfree the msg. This effectively breaks the\napplication or forces it to do a retry.\n\nLets catch this case and handle it by doing an skb_linearize() on any\nskb we receive with frags. At this point skb_to_sgvec should not fail\nbecause the failing conditions would require frags to be in place.\n\nFixes: 604326b41a6fb (\"bpf, sockmap: convert to generic sk_msg interface\")\nSigned-off-by: John Fastabend \u003cjohn.fastabend@gmail.com\u003e\nSigned-off-by: Daniel Borkmann \u003cdaniel@iogearbox.net\u003e\nReviewed-by: Jakub Sitnicki \u003cjakub@cloudflare.com\u003e\nLink: https://lore.kernel.org/bpf/160556576837.73229.14800682790808797635.stgit@john-XPS-13-9370\n" }, { "commit": "2443ca66676d50a4eb3305c236bccd84a9828ce2", "tree": "20b3d1690d7bbb17bf291282c8f657eabc46a2c5", "parents": [ "6fa9201a898983da731fca068bb4b5c941537588" ], "author": { "name": "John Fastabend", "email": "john.fastabend@gmail.com", "time": "Mon Nov 16 14:29:08 2020 -0800" }, "committer": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Wed Nov 18 00:13:27 2020 +0100" }, "message": "bpf, sockmap: Handle memory acct if skb_verdict prog redirects to self\n\nIf the skb_verdict_prog redirects an skb knowingly to itself, fix your\nBPF program this is not optimal and an abuse of the API please use\nSK_PASS. That said there may be cases, such as socket load balancing,\nwhere picking the socket is hashed based or otherwise picks the same\nsocket it was received on in some rare cases. If this happens we don\u0027t\nwant to confuse userspace giving them an EAGAIN error if we can avoid\nit.\n\nTo avoid double accounting in these cases. At the moment even if the\nskb has already been charged against the sockets rcvbuf and forward\nalloc we check it again and do set_owner_r() causing it to be orphaned\nand recharged. For one this is useless work, but more importantly we\ncan have a case where the skb could be put on the ingress queue, but\nbecause we are under memory pressure we return EAGAIN. The trouble\nhere is the skb has already been accounted for so any rcvbuf checks\ninclude the memory associated with the packet already. This rolls\nup and can result in unnecessary EAGAIN errors in userspace read()\ncalls.\n\nFix by doing an unlikely check and skipping checks if skb-\u003esk \u003d\u003d sk.\n\nFixes: 51199405f9672 (\"bpf: skb_verdict, support SK_PASS on RX BPF path\")\nSigned-off-by: John Fastabend \u003cjohn.fastabend@gmail.com\u003e\nSigned-off-by: Daniel Borkmann \u003cdaniel@iogearbox.net\u003e\nReviewed-by: Jakub Sitnicki \u003cjakub@cloudflare.com\u003e\nLink: https://lore.kernel.org/bpf/160556574804.73229.11328201020039674147.stgit@john-XPS-13-9370\n" }, { "commit": "6fa9201a898983da731fca068bb4b5c941537588", "tree": "f3fb30f4e3eec00a5affaaf6794166b1494b2e4d", "parents": [ "70796fb751f1d34cc650e640572a174faf009cd4" ], "author": { "name": "John Fastabend", "email": "john.fastabend@gmail.com", "time": "Mon Nov 16 14:28:46 2020 -0800" }, "committer": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Wed Nov 18 00:12:41 2020 +0100" }, "message": "bpf, sockmap: Avoid returning unneeded EAGAIN when redirecting to self\n\nIf a socket redirects to itself and it is under memory pressure it is\npossible to get a socket stuck so that recv() returns EAGAIN and the\nsocket can not advance for some time. This happens because when\nredirecting a skb to the same socket we received the skb on we first\ncheck if it is OK to enqueue the skb on the receiving socket by checking\nmemory limits. But, if the skb is itself the object holding the memory\nneeded to enqueue the skb we will keep retrying from kernel side\nand always fail with EAGAIN. Then userspace will get a recv() EAGAIN\nerror if there are no skbs in the psock ingress queue. This will continue\nuntil either some skbs get kfree\u0027d causing the memory pressure to\nreduce far enough that we can enqueue the pending packet or the\nsocket is destroyed. In some cases its possible to get a socket\nstuck for a noticeable amount of time if the socket is only receiving\nskbs from sk_skb verdict programs. To reproduce I make the socket\nmemory limits ridiculously low so sockets are always under memory\npressure. More often though if under memory pressure it looks like\na spurious EAGAIN error on user space side causing userspace to retry\nand typically enough has moved on the memory side that it works.\n\nTo fix skip memory checks and skb_orphan if receiving on the same\nsock as already assigned.\n\nFor SK_PASS cases this is easy, its always the same socket so we\ncan just omit the orphan/set_owner pair.\n\nFor backlog cases we need to check skb-\u003esk and decide if the orphan\nand set_owner pair are needed.\n\nFixes: 51199405f9672 (\"bpf: skb_verdict, support SK_PASS on RX BPF path\")\nSigned-off-by: John Fastabend \u003cjohn.fastabend@gmail.com\u003e\nSigned-off-by: Daniel Borkmann \u003cdaniel@iogearbox.net\u003e\nReviewed-by: Jakub Sitnicki \u003cjakub@cloudflare.com\u003e\nLink: https://lore.kernel.org/bpf/160556572660.73229.12566203819812939627.stgit@john-XPS-13-9370\n" }, { "commit": "70796fb751f1d34cc650e640572a174faf009cd4", "tree": "8d41d820e002f1bb34fc94adecb835a01ef0883d", "parents": [ "36cd0e696a832a00247fca522034703566ac8885" ], "author": { "name": "John Fastabend", "email": "john.fastabend@gmail.com", "time": "Mon Nov 16 14:28:26 2020 -0800" }, "committer": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Wed Nov 18 00:12:37 2020 +0100" }, "message": "bpf, sockmap: Use truesize with sk_rmem_schedule()\n\nWe use skb-\u003esize with sk_rmem_scheduled() which is not correct. Instead\nuse truesize to align with socket and tcp stack usage of sk_rmem_schedule.\n\nSuggested-by: Daniel Borkman \u003cdaniel@iogearbox.net\u003e\nSigned-off-by: John Fastabend \u003cjohn.fastabend@gmail.com\u003e\nSigned-off-by: Daniel Borkmann \u003cdaniel@iogearbox.net\u003e\nReviewed-by: Jakub Sitnicki \u003cjakub@cloudflare.com\u003e\nLink: https://lore.kernel.org/bpf/160556570616.73229.17003722112077507863.stgit@john-XPS-13-9370\n" }, { "commit": "36cd0e696a832a00247fca522034703566ac8885", "tree": "28849770d1faadc2117ca1fd60799173cf46231d", "parents": [ "c9c89dcd872ea33327673fcb97398993a1f22736" ], "author": { "name": "John Fastabend", "email": "john.fastabend@gmail.com", "time": "Mon Nov 16 14:28:06 2020 -0800" }, "committer": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Wed Nov 18 00:12:34 2020 +0100" }, "message": "bpf, sockmap: Ensure SO_RCVBUF memory is observed on ingress redirect\n\nFix sockmap sk_skb programs so that they observe sk_rcvbuf limits. This\nallows users to tune SO_RCVBUF and sockmap will honor them.\n\nWe can refactor the if(charge) case out in later patches. But, keep this\nfix to the point.\n\nFixes: 51199405f9672 (\"bpf: skb_verdict, support SK_PASS on RX BPF path\")\nSuggested-by: Jakub Sitnicki \u003cjakub@cloudflare.com\u003e\nSigned-off-by: John Fastabend \u003cjohn.fastabend@gmail.com\u003e\nSigned-off-by: Daniel Borkmann \u003cdaniel@iogearbox.net\u003e\nReviewed-by: Jakub Sitnicki \u003cjakub@cloudflare.com\u003e\nLink: https://lore.kernel.org/bpf/160556568657.73229.8404601585878439060.stgit@john-XPS-13-9370\n" }, { "commit": "c9c89dcd872ea33327673fcb97398993a1f22736", "tree": "69be08077f4173e230befb8ae57df301e6e5c35e", "parents": [ "2acc3c1bc8e98bc66b1badec42e9ea205b4fcdaa" ], "author": { "name": "John Fastabend", "email": "john.fastabend@gmail.com", "time": "Mon Nov 16 14:27:46 2020 -0800" }, "committer": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Wed Nov 18 00:12:10 2020 +0100" }, "message": "bpf, sockmap: Fix partial copy_page_to_iter so progress can still be made\n\nIf copy_page_to_iter() fails or even partially completes, but with fewer\nbytes copied than expected we currently reset sg.start and return EFAULT.\nThis proves problematic if we already copied data into the user buffer\nbefore we return an error. Because we leave the copied data in the user\nbuffer and fail to unwind the scatterlist so kernel side believes data\nhas been copied and user side believes data has _not_ been received.\n\nExpected behavior should be to return number of bytes copied and then\non the next read we need to return the error assuming its still there. This\ncan happen if we have a copy length spanning multiple scatterlist elements\nand one or more complete before the error is hit.\n\nThe error is rare enough though that my normal testing with server side\nprograms, such as nginx, httpd, envoy, etc., I have never seen this. The\nonly reliable way to reproduce that I\u0027ve found is to stream movies over\nmy browser for a day or so and wait for it to hang. Not very scientific,\nbut with a few extra WARN_ON()s in the code the bug was obvious.\n\nWhen we review the errors from copy_page_to_iter() it seems we are hitting\na page fault from copy_page_to_iter_iovec() where the code checks\nfault_in_pages_writeable(buf, copy) where buf is the user buffer. It\nalso seems typical server applications don\u0027t hit this case.\n\nThe other way to try and reproduce this is run the sockmap selftest tool\ntest_sockmap with data verification enabled, but it doesn\u0027t reproduce the\nfault. Perhaps we can trigger this case artificially somehow from the\ntest tools. I haven\u0027t sorted out a way to do that yet though.\n\nFixes: 604326b41a6fb (\"bpf, sockmap: convert to generic sk_msg interface\")\nSigned-off-by: John Fastabend \u003cjohn.fastabend@gmail.com\u003e\nSigned-off-by: Daniel Borkmann \u003cdaniel@iogearbox.net\u003e\nReviewed-by: Jakub Sitnicki \u003cjakub@cloudflare.com\u003e\nLink: https://lore.kernel.org/bpf/160556566659.73229.15694973114605301063.stgit@john-XPS-13-9370\n" }, { "commit": "138559b9f99d3b6b1d5e75c78facc067a23871c6", "tree": "53999ef38799ce1c226ba835fdcbc83c8a56f2fe", "parents": [ "a5bbcbf29089a1252c201b1a7fd38151de355db9" ], "author": { "name": "Tariq Toukan", "email": "tariqt@nvidia.com", "time": "Sun Nov 15 15:14:48 2020 +0200" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Nov 17 14:41:20 2020 -0800" }, "message": "net/tls: Fix wrong record sn in async mode of device resync\n\nIn async_resync mode, we log the TCP seq of records until the async request\nis completed. Later, in case one of the logged seqs matches the resync\nrequest, we return it, together with its record serial number. Before this\nfix, we mistakenly returned the serial number of the current record\ninstead.\n\nFixes: ed9b7646b06a (\"net/tls: Add asynchronous resync\")\nSigned-off-by: Tariq Toukan \u003ctariqt@nvidia.com\u003e\nReviewed-by: Boris Pismenny \u003cborisp@nvidia.com\u003e\nLink: https://lore.kernel.org/r/20201115131448.2702-1-tariqt@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "a5bbcbf29089a1252c201b1a7fd38151de355db9", "tree": "701f2df0e13232e7cb174988cb6d283bf205157b", "parents": [ "fd5736bf9f235d26c83cac8a16c70bbdafa55abe" ], "author": { "name": "Taehee Yoo", "email": "ap420073@gmail.com", "time": "Sun Nov 15 10:30:41 2020 +0000" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Nov 17 14:08:59 2020 -0800" }, "message": "netdevsim: set .owner to THIS_MODULE\n\nIf THIS_MODULE is not set, the module would be removed while debugfs is\nbeing used.\nIt eventually makes kernel panic.\n\nFixes: 82c93a87bf8b (\"netdevsim: implement couple of testing devlink health reporters\")\nFixes: 424be63ad831 (\"netdevsim: add UDP tunnel port offload support\")\nFixes: 4418f862d675 (\"netdevsim: implement support for devlink region and snapshots\")\nFixes: d3cbb907ae57 (\"netdevsim: add ACL trap reporting cookie as a metadata\")\nSigned-off-by: Taehee Yoo \u003cap420073@gmail.com\u003e\nLink: https://lore.kernel.org/r/20201115103041.30701-1-ap420073@gmail.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "fd5736bf9f235d26c83cac8a16c70bbdafa55abe", "tree": "e356f0422f596f701c2f166a05a0a27a6e0d1958", "parents": [ "1b9e2a8c99a5c021041bfb2d512dc3ed92a94ffd" ], "author": { "name": "Alex Marginean", "email": "alexandru.marginean@nxp.com", "time": "Thu Nov 12 20:26:08 2020 +0200" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Nov 17 12:12:12 2020 -0800" }, "message": "enetc: Workaround for MDIO register access issue\n\nDue to a hardware issue, an access to MDIO registers\nthat is concurrent with other ENETC register accesses\nmay lead to the MDIO access being dropped or corrupted.\nThe workaround introduces locking for all register accesses\nto the ENETC register space. To reduce performance impact,\na readers-writers locking scheme has been implemented.\nThe writer in this case is the MDIO access code (irrelevant\nwhether that MDIO access is a register read or write), and\nthe reader is any access code to non-MDIO ENETC registers.\nAlso, the datapath functions acquire the read lock fewer times\nand use _hot accessors. All the rest of the code uses the _wa\naccessors which lock every register access.\nThe commit introducing MDIO support is -\ncommit ebfcb23d62ab (\"enetc: Add ENETC PF level external MDIO support\")\nbut due to subsequent refactoring this patch is applicable on\ntop of a later commit.\n\nFixes: 6517798dd343 (\"enetc: Make MDIO accessors more generic and export to include/linux/fsl\")\nSigned-off-by: Alex Marginean \u003calexandru.marginean@nxp.com\u003e\nSigned-off-by: Vladimir Oltean \u003cvladimir.oltean@nxp.com\u003e\nSigned-off-by: Claudiu Manoil \u003cclaudiu.manoil@nxp.com\u003e\nLink: https://lore.kernel.org/r/20201112182608.26177-1-claudiu.manoil@nxp.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "0fa8ee0d9ab95c9350b8b84574824d9a384a9f7d", "tree": "e93b1f28886347874c3ec40d763735bb9af07fef", "parents": [ "111e91a6df505e532a3809ead372787a01e23e0c", "ae3d6083acf60116d4f409677452399547ed2009" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Nov 17 12:00:40 2020 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Nov 17 12:00:40 2020 -0800" }, "message": "Merge branch \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input\n\nPull input fixes from Dmitry Torokhov:\n \"A fix for use-after-free in the Sun keyboard driver, a fix to firmware\n updates on newer ICs in the Elan touchpad diver, and a couple misc\n driver fixes\"\n\n* \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:\n Input: elan_i2c - fix firmware update on newer ICs\n Input: resistive-adc-touch - fix kconfig dependency on IIO_BUFFER\n Input: sunkbd - avoid use-after-free in teardown paths\n Input: i8042 - allow insmod to succeed on devices without an i8042 controller\n Input: adxl34x - clean up a data type in adxl34x_probe()\n" }, { "commit": "68ec32daf7d50a9f7425f8607a7402c13aa0c587", "tree": "df19fc020aa20245f169f06b604a410a2c5aa912", "parents": [ "5b8631c7b21ca8bc039f0bc030048973b039e0d2" ], "author": { "name": "Wang Hai", "email": "wanghai38@huawei.com", "time": "Sat Nov 14 19:52:23 2020 +0800" }, "committer": { "name": "Saeed Mahameed", "email": "saeedm@nvidia.com", "time": "Tue Nov 17 11:50:54 2020 -0800" }, "message": "net/mlx5: fix error return code in mlx5e_tc_nic_init()\n\nFix to return a negative error code from the error handling\ncase instead of 0, as done elsewhere in this function.\n\nFixes: aedd133d17bc (\"net/mlx5e: Support CT offload for tc nic flows\")\nReported-by: Hulk Robot \u003chulkci@huawei.com\u003e\nSigned-off-by: Wang Hai \u003cwanghai38@huawei.com\u003e\nSigned-off-by: Saeed Mahameed \u003csaeedm@nvidia.com\u003e\n" }, { "commit": "5b8631c7b21ca8bc039f0bc030048973b039e0d2", "tree": "0abf2462259bd9fbb7c73897bc595a9ea691df7e", "parents": [ "470b74758260e4abc2508cf1614573c00a00465c" ], "author": { "name": "Eli Cohen", "email": "elic@nvidia.com", "time": "Mon Nov 09 11:35:52 2020 +0200" }, "committer": { "name": "Saeed Mahameed", "email": "saeedm@nvidia.com", "time": "Tue Nov 17 11:50:53 2020 -0800" }, "message": "net/mlx5: E-Switch, Fail mlx5_esw_modify_vport_rate if qos disabled\n\nAvoid calling mlx5_esw_modify_vport_rate() if qos is not enabled and\navoid unnecessary syndrome messages from firmware.\n\nFixes: fcb64c0f5640 (\"net/mlx5: E-Switch, add ingress rate support\")\nSigned-off-by: Eli Cohen \u003celic@nvidia.com\u003e\nReviewed-by: Roi Dayan \u003croid@nvidia.com\u003e\nSigned-off-by: Saeed Mahameed \u003csaeedm@nvidia.com\u003e\n" }, { "commit": "470b74758260e4abc2508cf1614573c00a00465c", "tree": "73b3d444f7bfffbccee2c840d34bcffc5614569e", "parents": [ "1ce5fc724a26e0b476e42c5d588bdb80caea003b" ], "author": { "name": "Vladyslav Tarasiuk", "email": "vladyslavt@nvidia.com", "time": "Wed Oct 21 11:05:41 2020 +0300" }, "committer": { "name": "Saeed Mahameed", "email": "saeedm@nvidia.com", "time": "Tue Nov 17 11:50:53 2020 -0800" }, "message": "net/mlx5: Disable QoS when min_rates on all VFs are zero\n\nCurrently when QoS is enabled for VF and any min_rate is configured,\nthe driver sets bw_share value to at least 1 and doesn’t allow to set\nit to 0 to make minimal rate unlimited. It means there is always a\nminimal rate configured for every VF, even if user tries to remove it.\n\nIn order to make QoS disable possible, check whether all vports have\nconfigured min_rate \u003d 0. If this is true, set their bw_share to 0 to\ndisable min_rate limitations.\n\nFixes: c9497c98901c (\"net/mlx5: Add support for setting VF min rate\")\nSigned-off-by: Vladyslav Tarasiuk \u003cvladyslavt@nvidia.com\u003e\nReviewed-by: Moshe Shemesh \u003cmoshe@nvidia.com\u003e\nSigned-off-by: Saeed Mahameed \u003csaeedm@nvidia.com\u003e\n" }, { "commit": "1ce5fc724a26e0b476e42c5d588bdb80caea003b", "tree": "98629bfae3879484c256d5b0e6727cd2de34b3ea", "parents": [ "8cbcc5ef2a281f6bb10099f4572a08cb765ffbf4" ], "author": { "name": "Vladyslav Tarasiuk", "email": "vladyslavt@nvidia.com", "time": "Mon Nov 02 13:45:24 2020 +0200" }, "committer": { "name": "Saeed Mahameed", "email": "saeedm@nvidia.com", "time": "Tue Nov 17 11:50:53 2020 -0800" }, "message": "net/mlx5: Clear bw_share upon VF disable\n\nCurrently, if user disables VFs with some min and max rates configured,\nthey are cleared. But QoS data is not cleared and restored upon next VF\nenable placing limits on minimal rate for given VF, when user expects\nnone.\n\nTo match cleared vport-\u003einfo struct with QoS-related min and max rates\nupon VF disable, clear vport-\u003eqos struct too.\n\nFixes: 556b9d16d3f5 (\"net/mlx5: Clear VF\u0027s configuration on disabling SRIOV\")\nSigned-off-by: Vladyslav Tarasiuk \u003cvladyslavt@nvidia.com\u003e\nReviewed-by: Moshe Shemesh \u003cmoshe@nvidia.com\u003e\nSigned-off-by: Saeed Mahameed \u003csaeedm@nvidia.com\u003e\n" }, { "commit": "8cbcc5ef2a281f6bb10099f4572a08cb765ffbf4", "tree": "b90a016e566292d35e4d5bfbd222b99796183184", "parents": [ "219b3267ca102a35092f5998921a9e6f99074af2" ], "author": { "name": "Michael Guralnik", "email": "michaelgur@nvidia.com", "time": "Mon Nov 02 17:34:44 2020 +0200" }, "committer": { "name": "Saeed Mahameed", "email": "saeedm@nvidia.com", "time": "Tue Nov 17 11:50:52 2020 -0800" }, "message": "net/mlx5: Add handling of port type in rule deletion\n\nHandle destruction of rules with port destination type to enable\nfull destruction of flow.\n\nWithout this handling of TX rules the deletion of these rules fails.\nDmesg of flow destruction failure:\n\n[ 203.714146] mlx5_core 0000:00:0b.0: mlx5_cmd_check:753:(pid 342): SET_FLOW_TABLE_ENTRY(0x936) op_mod(0x0) failed, status bad parameter(0x3), syndrome (0x144b7a)\n[ 210.547387] ------------[ cut here ]------------\n[ 210.548663] refcount_t: decrement hit 0; leaking memory.\n[ 210.550651] WARNING: CPU: 4 PID: 342 at lib/refcount.c:31 refcount_warn_saturate+0x5c/0x110\n[ 210.550654] Modules linked in: mlx5_ib mlx5_core ib_ipoib rdma_ucm rdma_cm iw_cm ib_cm ib_umad ib_uverbs ib_core\n[ 210.550675] CPU: 4 PID: 342 Comm: test Not tainted 5.8.0-rc2+ #116\n[ 210.550678] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.12.1-0-ga5cab58e9a3f-prebuilt.qemu.org 04/01/2014\n[ 210.550680] RIP: 0010:refcount_warn_saturate+0x5c/0x110\n[ 210.550685] Code: c6 d1 1b 01 00 0f 84 ad 00 00 00 5b 5d c3 80 3d b5 d1 1b 01 00 75 f4 48 c7 c7 20 d1 15 82 c6 05 a5 d1 1b 01 01 e8 a7 eb af ff \u003c0f\u003e 0b eb dd 80 3d 99 d1 1b 01 00 75 d4 48 c7 c7 c0 cf 15 82 c6 05\n[ 210.550687] RSP: 0018:ffff8881642e77e8 EFLAGS: 00010282\n[ 210.550691] RAX: 0000000000000000 RBX: 0000000000000004 RCX: 0000000000000000\n[ 210.550694] RDX: 0000000000000027 RSI: 0000000000000004 RDI: ffffed102c85ceef\n[ 210.550696] RBP: ffff888161720428 R08: ffffffff8124c10e R09: ffffed103243beae\n[ 210.550698] R10: ffff8881921df56b R11: ffffed103243bead R12: ffff8881841b4180\n[ 210.550701] R13: ffff888161720428 R14: ffff8881616d0000 R15: ffff888161720380\n[ 210.550704] FS: 00007fc27f025740(0000) GS:ffff888192000000(0000) knlGS:0000000000000000\n[ 210.550706] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033\n[ 210.550708] CR2: 0000557e4b41a6a0 CR3: 0000000002415004 CR4: 0000000000360ea0\n[ 210.550711] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000\n[ 210.550713] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400\n[ 210.550715] Call Trace:\n[ 210.550717] mlx5_del_flow_rules+0x484/0x490 [mlx5_core]\n[ 210.550720] ? mlx5_cmd_set_fte+0xa80/0xa80 [mlx5_core]\n[ 210.550722] mlx5_ib_destroy_flow+0x17f/0x280 [mlx5_ib]\n[ 210.550724] uverbs_free_flow+0x4c/0x90 [ib_uverbs]\n[ 210.550726] destroy_hw_idr_uobject+0x41/0xb0 [ib_uverbs]\n[ 210.550728] uverbs_destroy_uobject+0xaa/0x390 [ib_uverbs]\n[ 210.550731] __uverbs_cleanup_ufile+0x129/0x1b0 [ib_uverbs]\n[ 210.550733] ? uverbs_destroy_uobject+0x390/0x390 [ib_uverbs]\n[ 210.550735] uverbs_destroy_ufile_hw+0x78/0x190 [ib_uverbs]\n[ 210.550737] ib_uverbs_close+0x36/0x140 [ib_uverbs]\n[ 210.550739] __fput+0x181/0x380\n[ 210.550741] task_work_run+0x88/0xd0\n[ 210.550743] do_exit+0x5f6/0x13b0\n[ 210.550745] ? sched_clock_cpu+0x30/0x140\n[ 210.550747] ? is_current_pgrp_orphaned+0x70/0x70\n[ 210.550750] ? lock_downgrade+0x360/0x360\n[ 210.550752] ? mark_held_locks+0x1d/0x90\n[ 210.550754] do_group_exit+0x8a/0x140\n[ 210.550756] get_signal+0x20a/0xf50\n[ 210.550758] do_signal+0x8c/0xbe0\n[ 210.550760] ? hrtimer_nanosleep+0x1d8/0x200\n[ 210.550762] ? nanosleep_copyout+0x50/0x50\n[ 210.550764] ? restore_sigcontext+0x320/0x320\n[ 210.550766] ? __hrtimer_init+0xf0/0xf0\n[ 210.550768] ? timespec64_add_safe+0x150/0x150\n[ 210.550770] ? mark_held_locks+0x1d/0x90\n[ 210.550772] ? lockdep_hardirqs_on_prepare+0x14c/0x240\n[ 210.550774] __prepare_exit_to_usermode+0x119/0x170\n[ 210.550776] do_syscall_64+0x65/0x300\n[ 210.550778] ? trace_hardirqs_off+0x10/0x120\n[ 210.550781] ? mark_held_locks+0x1d/0x90\n[ 210.550783] ? asm_sysvec_apic_timer_interrupt+0xa/0x20\n[ 210.550785] ? lockdep_hardirqs_on+0x112/0x190\n[ 210.550787] entry_SYSCALL_64_after_hwframe+0x44/0xa9\n[ 210.550789] RIP: 0033:0x7fc27f1cd157\n[ 210.550791] Code: Bad RIP value.\n[ 210.550793] RSP: 002b:00007ffd4db27ea8 EFLAGS: 00000246 ORIG_RAX: 0000000000000023\n[ 210.550798] RAX: fffffffffffffdfc RBX: ffffffffffffff80 RCX: 00007fc27f1cd157\n[ 210.550800] RDX: 00007fc27f025740 RSI: 00007ffd4db27eb0 RDI: 00007ffd4db27eb0\n[ 210.550803] RBP: 0000000000000016 R08: 0000000000000000 R09: 000000000000000e\n[ 210.550805] R10: 00007ffd4db27dc7 R11: 0000000000000246 R12: 0000000000400c00\n[ 210.550808] R13: 00007ffd4db285f0 R14: 0000000000000000 R15: 0000000000000000\n[ 210.550809] irq event stamp: 49399\n[ 210.550812] hardirqs last enabled at (49399): [\u003cffffffff81172d36\u003e] console_unlock+0x556/0x6f0\n[ 210.550815] hardirqs last disabled at (49398): [\u003cffffffff81172897\u003e] console_unlock+0xb7/0x6f0\n[ 210.550818] softirqs last enabled at (48706): [\u003cffffffff81e0037b\u003e] __do_softirq+0x37b/0x60c\n[ 210.550820] softirqs last disabled at (48697): [\u003cffffffff81c00e2f\u003e] asm_call_on_stack+0xf/0x20\n[ 210.550822] ---[ end trace ad18c0e6fa846454 ]---\n[ 210.581862] mlx5_core 0000:00:0c.0: mlx5_destroy_flow_table:2132:(pid 342): Flow table 262150 wasn\u0027t destroyed, refcount \u003e 1\n\nFixes: a7ee18bdee83 (\"RDMA/mlx5: Allow creating a matcher for a NIC TX flow table\")\nSigned-off-by: Michael Guralnik \u003cmichaelgur@nvidia.com\u003e\nReviewed-by: Mark Bloch \u003cmbloch@nvidia.com\u003e\nReviewed-by: Maor Gottlieb \u003cmaorg@nvidia.com\u003e\nSigned-off-by: Saeed Mahameed \u003csaeedm@nvidia.com\u003e\n" }, { "commit": "219b3267ca102a35092f5998921a9e6f99074af2", "tree": "5bceb77c038368393a4ac62ec70723f13532d98e", "parents": [ "6248ce991f8eed4f2f0fdec694f5749156105629" ], "author": { "name": "Maor Dickman", "email": "maord@nvidia.com", "time": "Wed Nov 04 14:10:30 2020 +0200" }, "committer": { "name": "Saeed Mahameed", "email": "saeedm@nvidia.com", "time": "Tue Nov 17 11:50:52 2020 -0800" }, "message": "net/mlx5e: Fix check if netdev is bond slave\n\nBond events handler uses bond_slave_get_rtnl to check if net device\nis bond slave. bond_slave_get_rtnl return the rcu rx_handler pointer\nfrom the netdev which exists for bond slaves but also exists for\ndevices that are attached to linux bridge so using it as indication\nfor bond slave is wrong.\n\nFix by using netif_is_lag_port instead.\n\nFixes: 7e51891a237f (\"net/mlx5e: Use netdev events to set/del egress acl forward-to-vport rule\")\nSigned-off-by: Maor Dickman \u003cmaord@nvidia.com\u003e\nReviewed-by: Raed Salem \u003craeds@nvidia.com\u003e\nReviewed-by: Ariel Levkovich \u003clariel@nvidia.com\u003e\nSigned-off-by: Saeed Mahameed \u003csaeedm@nvidia.com\u003e\n" }, { "commit": "6248ce991f8eed4f2f0fdec694f5749156105629", "tree": "95dc9f6a8fb34cf4b4fb2ed2297eebb686793784", "parents": [ "5cfb540ef27b5b763a3b181d142847ef0411728e" ], "author": { "name": "Huy Nguyen", "email": "huyn@mellanox.com", "time": "Tue Nov 03 12:56:18 2020 -0600" }, "committer": { "name": "Saeed Mahameed", "email": "saeedm@nvidia.com", "time": "Tue Nov 17 11:50:52 2020 -0800" }, "message": "net/mlx5e: Fix IPsec packet drop by mlx5e_tc_update_skb\n\nBoth TC and IPsec crypto offload use metadata_regB to store\nprivate information. Since TC does not use bit 31 of regB, IPsec\nwill use bit 31 as the IPsec packet marker. The IPsec\u0027s regB usage\nis changed to:\nBit31: IPsec marker\nBit30-24: IPsec syndrome\nBit23-0: IPsec obj id\n\nFixes: b2ac7541e377 (\"net/mlx5e: IPsec: Add Connect-X IPsec Rx data path offload\")\nSigned-off-by: Huy Nguyen \u003chuyn@mellanox.com\u003e\nReviewed-by: Raed Salem \u003craeds@nvidia.com\u003e\nReviewed-by: Ariel Levkovich \u003clariel@nvidia.com\u003e\nSigned-off-by: Saeed Mahameed \u003csaeedm@nvidia.com\u003e\n" }, { "commit": "5cfb540ef27b5b763a3b181d142847ef0411728e", "tree": "0d19efb580fb0a08887064794aa45fa4ede8b6c6", "parents": [ "ea63609857321c38fd4ad096388b413b66001c6c" ], "author": { "name": "Huy Nguyen", "email": "huyn@mellanox.com", "time": "Thu Oct 22 11:22:56 2020 -0500" }, "committer": { "name": "Saeed Mahameed", "email": "saeedm@nvidia.com", "time": "Tue Nov 17 11:50:52 2020 -0800" }, "message": "net/mlx5e: Set IPsec WAs only in IP\u0027s non checksum partial case.\n\nThe IP\u0027s checksum partial still requires L4 csum flag on Ethernet WQE.\nMake the IPsec WAs only for the IP\u0027s non checksum partial case\n(for example icmd packet)\n\nFixes: 5be019040cb7 (\"net/mlx5e: IPsec: Add Connect-X IPsec Tx data path offload\")\nSigned-off-by: Huy Nguyen \u003chuyn@mellanox.com\u003e\nReviewed-by: Raed Salem \u003craeds@nvidia.com\u003e\nReviewed-by: Alaa Hleihel \u003calaa@nvidia.com\u003e\nSigned-off-by: Saeed Mahameed \u003csaeedm@nvidia.com\u003e\n" }, { "commit": "ea63609857321c38fd4ad096388b413b66001c6c", "tree": "676cc3b1ee6422e364c872dd6dd733d719a6775e", "parents": [ "1b9e2a8c99a5c021041bfb2d512dc3ed92a94ffd" ], "author": { "name": "Maxim Mikityanskiy", "email": "maximmi@mellanox.com", "time": "Thu Oct 08 12:34:10 2020 +0300" }, "committer": { "name": "Saeed Mahameed", "email": "saeedm@nvidia.com", "time": "Tue Nov 17 11:50:51 2020 -0800" }, "message": "net/mlx5e: Fix refcount leak on kTLS RX resync\n\nOn resync, the driver calls inet_lookup_established\n(__inet6_lookup_established) that increases sk_refcnt of the socket. To\ndecrease it, the driver set skb-\u003edestructor to sock_edemux. However, it\ndidn\u0027t work well, because the TCP stack also sets this destructor for\nearly demux, and the refcount gets decreased only once, while increased\ntwo times (in mlx5e and in the TCP stack). It leads to a socket leak, a\nTLS context leak, which in the end leads to calling tls_dev_del twice:\non socket close and on driver unload, which in turn leads to a crash.\n\nThis commit fixes the refcount leak by calling sock_gen_put right away\nafter using the socket, thus fixing all the subsequent issues.\n\nFixes: 0419d8c9d8f8 (\"net/mlx5e: kTLS, Add kTLS RX resync support\")\nSigned-off-by: Maxim Mikityanskiy \u003cmaximmi@mellanox.com\u003e\nReviewed-by: Tariq Toukan \u003ctariqt@nvidia.com\u003e\nSigned-off-by: Saeed Mahameed \u003csaeedm@nvidia.com\u003e\n" }, { "commit": "111e91a6df505e532a3809ead372787a01e23e0c", "tree": "d3ea7beacfa988dd3b25522257d7dfe87dd051bf", "parents": [ "ed129cd75ac1073f32d04d0f2012ede40e86fb77", "78d732e1f326f74f240d416af9484928303d9951" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Nov 17 11:22:03 2020 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Nov 17 11:22:03 2020 -0800" }, "message": "Merge tag \u0027s390-5.10-4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux\n\nPull s390 fixes from Heiko Carstens:\n\n - fix system call exit path; avoid return to user space with any\n TIF/CIF/PIF set\n\n - fix file permission for cpum_sfb_size parameter\n\n - another small defconfig update\n\n* tag \u0027s390-5.10-4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:\n s390/cpum_sf.c: fix file permission for cpum_sfb_size\n s390: update defconfigs\n s390: fix system call exit path\n" }, { "commit": "ed129cd75ac1073f32d04d0f2012ede40e86fb77", "tree": "28056fd9352da597cd2158c261aa45ddeb0ca95f", "parents": [ "be1dd6692adbdb1d70da47da124ac8376bba5ad5", "ac3b57adf87ad9bac7e33ca26bbbb13fae1ed62b" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Nov 17 11:15:08 2020 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Nov 17 11:15:08 2020 -0800" }, "message": "Merge tag \u0027mips_fixes_5.10_1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux\n\nPull MIPS fixes from Thomas Bogendoerfer:\n\n - fix bug preventing booting on several platforms\n\n - fix for build error, when modules need has_transparent_hugepage\n\n - fix for memleak in alchemy clk setup\n\n* tag \u0027mips_fixes_5.10_1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux:\n MIPS: Alchemy: Fix memleak in alchemy_clk_setup_cpu\n MIPS: kernel: Fix for_each_memblock conversion\n MIPS: export has_transparent_hugepage() for modules\n" }, { "commit": "1b9e2a8c99a5c021041bfb2d512dc3ed92a94ffd", "tree": "93791d0c260842e895e449e4406b7f8b49728a70", "parents": [ "3d5179458d22dc0b4fdc724e4bed4231a655112a" ], "author": { "name": "Ryan Sharpelletti", "email": "sharpelletti@google.com", "time": "Mon Nov 16 17:44:13 2020 +0000" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Nov 17 11:03:22 2020 -0800" }, "message": "tcp: only postpone PROBE_RTT if RTT is \u003c current min_rtt estimate\n\nDuring loss recovery, retransmitted packets are forced to use TCP\ntimestamps to calculate the RTT samples, which have a millisecond\ngranularity. BBR is designed using a microsecond granularity. As a\nresult, multiple RTT samples could be truncated to the same RTT value\nduring loss recovery. This is problematic, as BBR will not enter\nPROBE_RTT if the RTT sample is \u003c\u003d the current min_rtt sample, meaning\nthat if there are persistent losses, PROBE_RTT will constantly be\npushed off and potentially never re-entered. This patch makes sure\nthat BBR enters PROBE_RTT by checking if RTT sample is \u003c the current\nmin_rtt sample, rather than \u003c\u003d.\n\nThe Netflix transport/TCP team discovered this bug in the Linux TCP\nBBR code during lab tests.\n\nFixes: 0f8782ea1497 (\"tcp_bbr: add BBR congestion control\")\nSigned-off-by: Ryan Sharpelletti \u003csharpelletti@google.com\u003e\nSigned-off-by: Neal Cardwell \u003cncardwell@google.com\u003e\nSigned-off-by: Soheil Hassas Yeganeh \u003csoheil@google.com\u003e\nSigned-off-by: Yuchung Cheng \u003cycheng@google.com\u003e\nLink: https://lore.kernel.org/r/20201116174412.1433277-1-sharpelletti.kdev@gmail.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "3d5179458d22dc0b4fdc724e4bed4231a655112a", "tree": "f9162b8edd7a2d316242a37ba8aabcc60cd573be", "parents": [ "7b027c249da54f492699c43e26cba486cfd48035" ], "author": { "name": "Joel Stanley", "email": "joel@jms.id.au", "time": "Tue Nov 17 13:14:48 2020 +1030" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Nov 17 10:59:03 2020 -0800" }, "message": "net: ftgmac100: Fix crash when removing driver\n\nWhen removing the driver we would hit BUG_ON(!list_empty(\u0026dev-\u003eptype_specific))\nin net/core/dev.c due to still having the NC-SI packet handler\nregistered.\n\n # echo 1e660000.ethernet \u003e /sys/bus/platform/drivers/ftgmac100/unbind\n ------------[ cut here ]------------\n kernel BUG at net/core/dev.c:10254!\n Internal error: Oops - BUG: 0 [#1] SMP ARM\n CPU: 0 PID: 115 Comm: sh Not tainted 5.10.0-rc3-next-20201111-00007-g02e0365710c4 #46\n Hardware name: Generic DT based system\n PC is at netdev_run_todo+0x314/0x394\n LR is at cpumask_next+0x20/0x24\n pc : [\u003c806f5830\u003e] lr : [\u003c80863cb0\u003e] psr: 80000153\n sp : 855bbd58 ip : 00000001 fp : 855bbdac\n r10: 80c03d00 r9 : 80c06228 r8 : 81158c54\n r7 : 00000000 r6 : 80c05dec r5 : 80c05d18 r4 : 813b9280\n r3 : 813b9054 r2 : 8122c470 r1 : 00000002 r0 : 00000002\n Flags: Nzcv IRQs on FIQs off Mode SVC_32 ISA ARM Segment none\n Control: 00c5387d Table: 85514008 DAC: 00000051\n Process sh (pid: 115, stack limit \u003d 0x7cb5703d)\n ...\n Backtrace:\n [\u003c806f551c\u003e] (netdev_run_todo) from [\u003c80707eec\u003e] (rtnl_unlock+0x18/0x1c)\n r10:00000051 r9:854ed710 r8:81158c54 r7:80c76bb0 r6:81158c10 r5:8115b410\n r4:813b9000\n [\u003c80707ed4\u003e] (rtnl_unlock) from [\u003c806f5db8\u003e] (unregister_netdev+0x2c/0x30)\n [\u003c806f5d8c\u003e] (unregister_netdev) from [\u003c805a8180\u003e] (ftgmac100_remove+0x20/0xa8)\n r5:8115b410 r4:813b9000\n [\u003c805a8160\u003e] (ftgmac100_remove) from [\u003c805355e4\u003e] (platform_drv_remove+0x34/0x4c)\n\nFixes: bd466c3fb5a4 (\"net/faraday: Support NCSI mode\")\nSigned-off-by: Joel Stanley \u003cjoel@jms.id.au\u003e\nLink: https://lore.kernel.org/r/20201117024448.1170761-1-joel@jms.id.au\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "7b027c249da54f492699c43e26cba486cfd48035", "tree": "95afcc72c40dcdca342a9fd332efb72fe21b8e01", "parents": [ "cb47d16ea21045c66eebbf5ed792e74a8537e27a" ], "author": { "name": "Zhang Changzhong", "email": "zhangchangzhong@huawei.com", "time": "Tue Nov 17 11:02:11 2020 +0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Nov 17 10:50:28 2020 -0800" }, "message": "net: b44: fix error return code in b44_init_one()\n\nFix to return a negative error code from the error handling\ncase instead of 0, as done elsewhere in this function.\n\nFixes: 39a6f4bce6b4 (\"b44: replace the ssb_dma API with the generic DMA API\")\nReported-by: Hulk Robot \u003chulkci@huawei.com\u003e\nSigned-off-by: Zhang Changzhong \u003czhangchangzhong@huawei.com\u003e\nReviewed-by: Michael Chan \u003cmichael.chan@broadcom.com\u003e\nLink: https://lore.kernel.org/r/1605582131-36735-1-git-send-email-zhangchangzhong@huawei.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "be1dd6692adbdb1d70da47da124ac8376bba5ad5", "tree": "60c89b11bb1a2a74f37c336bc07d7415d672dd8c", "parents": [ "9dacf44c3837b7f1cf460de904f352714e7cd107", "568beb27959b0515d325ea1c6cf211eed2d66740" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Nov 17 10:47:45 2020 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Nov 17 10:47:45 2020 -0800" }, "message": "Merge tag \u0027perf-tools-fixes-for-v5.10-2020-11-17\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux\n\nPull perf tools fixes from Arnaldo Carvalho de Melo:\n\n - Fix file corruption due to event deletion in \u0027perf inject\u0027.\n\n - Update arch/x86/lib/mem{cpy,set}_64.S copies used in \u0027perf bench mem\n memcpy\u0027, silencing perf build warning.\n\n - Avoid an msan warning in a copied stack in \u0027perf test\u0027.\n\n - Correct tracepoint field name \"flags\" in ARM\u0027s CS-ETM hardware\n tracing \u0027perf test\u0027 entry.\n\n - Update branch sample pattern for cs-etm to cope with excluding guest\n in userspace counting.\n\n - Don\u0027t free \"lock_seq_stat\" if read_count isn\u0027t zero in \u0027perf lock\u0027.\n\n* tag \u0027perf-tools-fixes-for-v5.10-2020-11-17\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux:\n perf test: Avoid an msan warning in a copied stack.\n perf inject: Fix file corruption due to event deletion\n perf test: Update branch sample pattern for cs-etm\n perf test: Fix a typo in cs-etm testing\n tools arch: Update arch/x86/lib/mem{cpy,set}_64.S copies used in \u0027perf bench mem memcpy\u0027\n perf lock: Don\u0027t free \"lock_seq_stat\" if read_count isn\u0027t zero\n perf lock: Correct field name \"flags\"\n" }, { "commit": "cb47d16ea21045c66eebbf5ed792e74a8537e27a", "tree": "68c0c68cd1d95857c24cb9aab9a1a017614595ca", "parents": [ "d5bd32a876c8cb677c9cc88dcc7eb498cd0c9397" ], "author": { "name": "Zhang Changzhong", "email": "zhangchangzhong@huawei.com", "time": "Mon Nov 16 21:07:13 2020 +0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Nov 17 10:40:34 2020 -0800" }, "message": "qed: fix error return code in qed_iwarp_ll2_start()\n\nFix to return a negative error code from the error handling\ncase instead of 0, as done elsewhere in this function.\n\nFixes: 469981b17a4f (\"qed: Add unaligned and packed packet processing\")\nFixes: fcb39f6c10b2 (\"qed: Add mpa buffer descriptors for storing and processing mpa fpdus\")\nFixes: 1e28eaad07ea (\"qed: Add iWARP support for fpdu spanned over more than two tcp packets\")\nReported-by: Hulk Robot \u003chulkci@huawei.com\u003e\nSigned-off-by: Zhang Changzhong \u003czhangchangzhong@huawei.com\u003e\nAcked-by: Michal Kalderon \u003cmichal.kalderon@marvell.com\u003e\nLink: https://lore.kernel.org/r/1605532033-27373-1-git-send-email-zhangchangzhong@huawei.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "9dacf44c3837b7f1cf460de904f352714e7cd107", "tree": "9f60dc452f830d73db20e17d1d73c7dc71712885", "parents": [ "9c87c9f41245baa3fc4716cf39141439cf405b01", "c583bcb8f5edd48c1798798e341f78afb9bf4f6f" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Nov 17 10:31:56 2020 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Nov 17 10:31:56 2020 -0800" }, "message": "Merge branch \u0027urgent-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu\n\nPull RCU fix from Paul McKenney:\n \"A single commit that fixes a bug that was introduced a couple of merge\n windows ago, but which rather more recently converged to an\n agreed-upon fix. The bug is that interrupts can be incorrectly enabled\n while holding an irq-disabled spinlock. This can of course result in\n self-deadlocks.\n\n The bug is a bit difficult to trigger. It requires that a preempted\n task be blocking a preemptible-RCU grace period long enough to trigger\n an RCU CPU stall warning. In addition, an interrupt must occur at just\n the right time, and that interrupt\u0027s handler must acquire that same\n irq-disabled spinlock. Still, a deadlock is a deadlock.\n\n Furthermore, we do now have a fix, and that fix survives kernel test\n robot, -next, and rcutorture testing. It has also been verified by\n Sebastian as fixing the bug. Therefore...\"\n\n* \u0027urgent-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu:\n rcu: Don\u0027t invoke try_invoke_on_locked_down_task() with irqs disabled\n" }, { "commit": "04a9cd51d3f3308a98cbc6adc07acb12fbade011", "tree": "9a8102f729452d30de9d0c068163d8dc6583ded8", "parents": [ "a41b0ad07bfa081584218431cb0cd7e7ecc71210" ], "author": { "name": "Lukas Wunner", "email": "lukas@wunner.de", "time": "Mon Nov 16 09:23:10 2020 +0100" }, "committer": { "name": "Mark Brown", "email": "broonie@kernel.org", "time": "Tue Nov 17 17:08:39 2020 +0000" }, "message": "spi: npcm-fiu: Don\u0027t leak SPI master in probe error path\n\nIf the calls to of_match_device(), of_alias_get_id(),\ndevm_ioremap_resource(), devm_regmap_init_mmio() or devm_clk_get()\nfail on probe of the NPCM FIU SPI driver, the spi_controller struct is\nerroneously not freed.\n\nFix by switching over to the new devm_spi_alloc_master() helper.\n\nFixes: ace55c411b11 (\"spi: npcm-fiu: add NPCM FIU controller driver\")\nSigned-off-by: Lukas Wunner \u003clukas@wunner.de\u003e\nCc: \u003cstable@vger.kernel.org\u003e # v5.4+: 5e844cc37a5c: spi: Introduce device-managed SPI controller allocation\nCc: \u003cstable@vger.kernel.org\u003e # v5.4+\nCc: Tomer Maimon \u003ctmaimon77@gmail.com\u003e\nLink: https://lore.kernel.org/r/a420c23a363a3bc9aa684c6e790c32a8af106d17.1605512876.git.lukas@wunner.de\nSigned-off-by: Mark Brown \u003cbroonie@kernel.org\u003e\n" }, { "commit": "a41b0ad07bfa081584218431cb0cd7e7ecc71210", "tree": "1820a56891a5bdd77b8b6eb8e82e09ae9fe1f6e8", "parents": [ "ac9978fcad3c5abc43cdd225441ce9459c36e16b" ], "author": { "name": "Serge Semin", "email": "Sergey.Semin@baikalelectronics.ru", "time": "Tue Nov 17 12:40:54 2020 +0300" }, "committer": { "name": "Mark Brown", "email": "broonie@kernel.org", "time": "Tue Nov 17 17:08:39 2020 +0000" }, "message": "spi: dw: Set transfer handler before unmasking the IRQs\n\nIt turns out the IRQs most like can be unmasked before the controller is\nenabled with no problematic consequences. The manual doesn\u0027t explicitly\nstate that, but the examples perform the controller initialization\nprocedure in that order. So the commit da8f58909e7e (\"spi: dw: Unmask IRQs\nafter enabling the chip\") hasn\u0027t been that required as I thought. But\nanyway setting the IRQs up after the chip enabling still worth adding\nsince it has simplified the code a bit. The problem is that it has\nintroduced a potential bug. The transfer handler pointer is now\ninitialized after the IRQs are enabled. That may and eventually will cause\nan invalid or uninitialized callback invocation. Fix that just by\nperforming the callback initialization before the IRQ unmask procedure.\n\nFixes: da8f58909e7e (\"spi: dw: Unmask IRQs after enabling the chip\")\nSigned-off-by: Serge Semin \u003cSergey.Semin@baikalelectronics.ru\u003e\nLink: https://lore.kernel.org/r/20201117094054.4696-1-Sergey.Semin@baikalelectronics.ru\nSigned-off-by: Mark Brown \u003cbroonie@kernel.org\u003e\n" }, { "commit": "dc293f2106903ab9c24e9cea18c276e32c394c33", "tree": "dbdffc40bada4b97abc6b15c3ed67bca93d7dd3a", "parents": [ "3a860d165eb5f4d7cf0bf81ef6a5b5c5e1754422" ], "author": { "name": "Laurent Pinchart", "email": "laurent.pinchart@ideasonboard.com", "time": "Tue Sep 01 00:09:37 2020 +0300" }, "committer": { "name": "Max Filippov", "email": "jcmvbkbc@gmail.com", "time": "Tue Nov 17 05:09:28 2020 -0800" }, "message": "xtensa: uaccess: Add missing __user to strncpy_from_user() prototype\n\nWhen adding __user annotations in commit 2adf5352a34a, the\nstrncpy_from_user() function declaration for the\nCONFIG_GENERIC_STRNCPY_FROM_USER case was missed. Fix it.\n\nReported-by: kernel test robot \u003clkp@intel.com\u003e\nSigned-off-by: Laurent Pinchart \u003claurent.pinchart@ideasonboard.com\u003e\nMessage-Id: \u003c20200831210937.17938-1-laurent.pinchart@ideasonboard.com\u003e\nSigned-off-by: Max Filippov \u003cjcmvbkbc@gmail.com\u003e\n" }, { "commit": "14c620cf2ee81349527491110a47a157ac6d511c", "tree": "b6d055c35d328ef3623fd45e54c902f66e650de9", "parents": [ "09162bc32c880a791c6c0668ce0745cf7958f576", "8410e7f3b31e53bfa7a34c282b4313e79ed7ff8d" ], "author": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Tue Nov 17 13:25:06 2020 +0100" }, "committer": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Tue Nov 17 13:25:06 2020 +0100" }, "message": "Merge branch \u0027cpufreq/arm/fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/vireshk/pm\n\nPull cpufreq-arm fixes for 5.10-rc5 from Viresh Kumar:\n\n\"- tegra186: Fix -\u003eget() callback.\n - arm/scmi: Add dummy clock provider to fix failure.\"\n\n* \u0027cpufreq/arm/fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/vireshk/pm:\n cpufreq: scmi: Fix OPP addition failure with a dummy clock provider\n cpufreq: tegra186: Fix get frequency callback\n" }, { "commit": "ac3b57adf87ad9bac7e33ca26bbbb13fae1ed62b", "tree": "bcc067bf053fd37cd99a3806b492c4b4c04f6406", "parents": [ "61a2f1aecf6052f7bcf900829ca2b9d74437ec07" ], "author": { "name": "Zhang Qilong", "email": "zhangqilong3@huawei.com", "time": "Fri Nov 13 21:18:56 2020 +0800" }, "committer": { "name": "Thomas Bogendoerfer", "email": "tsbogend@alpha.franken.de", "time": "Tue Nov 17 12:53:57 2020 +0100" }, "message": "MIPS: Alchemy: Fix memleak in alchemy_clk_setup_cpu\n\nIf the clk_register fails, we should free h before\nfunction returns to prevent memleak.\n\nFixes: 474402291a0ad (\"MIPS: Alchemy: clock framework integration of onchip clocks\")\nReported-by: Hulk Robot \u003chulkci@huawei.com\u003e\nSigned-off-by: Zhang Qilong \u003czhangqilong3@huawei.com\u003e\nSigned-off-by: Thomas Bogendoerfer \u003ctsbogend@alpha.franken.de\u003e\n" }, { "commit": "61a2f1aecf6052f7bcf900829ca2b9d74437ec07", "tree": "4f65019482e1620c76072dd38b68bfcf7f9dfa40", "parents": [ "31b4d8e172f614adc53ddecb4b6b2f6411a49b84" ], "author": { "name": "Thomas Bogendoerfer", "email": "tsbogend@alpha.franken.de", "time": "Mon Nov 16 18:45:15 2020 +0100" }, "committer": { "name": "Thomas Bogendoerfer", "email": "tsbogend@alpha.franken.de", "time": "Tue Nov 17 10:31:31 2020 +0100" }, "message": "MIPS: kernel: Fix for_each_memblock conversion\n\nThe loop over all memblocks works with PFNs and not physical\naddresses, so we need for_each_mem_pfn_range().\n\nFixes: b10d6bca8720 (\"arch, drivers: replace for_each_membock() with for_each_mem_range()\")\nSigned-off-by: Thomas Bogendoerfer \u003ctsbogend@alpha.franken.de\u003e\nReviewed-by: Mike Rapoport \u003crppt@linux.ibm.com\u003e\nReviewed-by: Serge Semin \u003cfancer.lancer@gmail.com\u003e\n" }, { "commit": "8410e7f3b31e53bfa7a34c282b4313e79ed7ff8d", "tree": "6d4bef50b52f5b4227e50c07875f26ef7c740f7d", "parents": [ "e010d1d25e47642fb91023479a4965000cf934a8" ], "author": { "name": "Sudeep Holla", "email": "sudeep.holla@arm.com", "time": "Tue Nov 10 11:10:40 2020 +0000" }, "committer": { "name": "Viresh Kumar", "email": "viresh.kumar@linaro.org", "time": "Tue Nov 17 10:04:28 2020 +0530" }, "message": "cpufreq: scmi: Fix OPP addition failure with a dummy clock provider\n\nCommit dd461cd9183f (\"opp: Allow dev_pm_opp_get_opp_table() to return\n-EPROBE_DEFER\") handles -EPROBE_DEFER for the clock/interconnects within\n_allocate_opp_table() which is called from dev_pm_opp_add and it\nnow propagates the error back to the caller.\n\nSCMI performance domain re-used clock bindings to keep it simple. However\nwith the above mentioned change, if clock property is present in a device\nnode, opps fails to get added with below errors until clk_get succeeds.\n\n cpu0: failed to add opp 450000000Hz\n cpu0: failed to add opps to the device\n ....(errors on cpu1-cpu4)\n cpu5: failed to add opp 450000000Hz\n cpu5: failed to add opps to the device\n\nSo, in order to fix the issue, we need to register dummy clock provider.\nWith the dummy clock provider, clk_get returns NULL(no errors!), then opp\ncore proceeds to add OPPs for the CPUs.\n\nCc: Rafael J. Wysocki \u003crjw@rjwysocki.net\u003e\nFixes: dd461cd9183f (\"opp: Allow dev_pm_opp_get_opp_table() to return -EPROBE_DEFER\")\nSigned-off-by: Sudeep Holla \u003csudeep.holla@arm.com\u003e\nSigned-off-by: Viresh Kumar \u003cviresh.kumar@linaro.org\u003e\n" }, { "commit": "e010d1d25e47642fb91023479a4965000cf934a8", "tree": "64799c3a9d5c5610b4d5d037057d68ba4122cb74", "parents": [ "3650b228f83adda7e5ee532e2b90429c03f7b9ec" ], "author": { "name": "Jon Hunter", "email": "jonathanh@nvidia.com", "time": "Tue Nov 03 11:55:14 2020 +0000" }, "committer": { "name": "Viresh Kumar", "email": "viresh.kumar@linaro.org", "time": "Tue Nov 17 10:04:21 2020 +0530" }, "message": "cpufreq: tegra186: Fix get frequency callback\n\nCommit b89c01c96051 (\"cpufreq: tegra186: Fix initial frequency\")\nimplemented the CPUFREQ \u0027get\u0027 callback to determine the current\noperating frequency for each CPU. This implementation used a simple\nlooked up to determine the current operating frequency. The problem\nwith this is that frequency table for different Tegra186 devices may\nvary and so the default boot frequency for Tegra186 device may or may\nnot be present in the frequency table. If the default boot frequency is\nnot present in the frequency table, this causes the function\ntegra186_cpufreq_get() to return 0 and in turn causes cpufreq_online()\nto fail which prevents CPUFREQ from working.\n\nFix this by always calculating the CPU frequency based upon the current\n\u0027ndiv\u0027 setting for the CPU. Note that the CPU frequency for Tegra186 is\ncalculated by reading the current \u0027ndiv\u0027 setting, multiplying by the\nCPU reference clock and dividing by a constant divisor.\n\nFixes: b89c01c96051 (\"cpufreq: tegra186: Fix initial frequency\")\n\nSigned-off-by: Jon Hunter \u003cjonathanh@nvidia.com\u003e\nSigned-off-by: Viresh Kumar \u003cviresh.kumar@linaro.org\u003e\n" }, { "commit": "d5bd32a876c8cb677c9cc88dcc7eb498cd0c9397", "tree": "251895e2aef16d0a9e6be148b8ff402e391fa25b", "parents": [ "064c9c32b17ca9b36f95eba32ee790dbbebd9a5f", "0ae0a779efb8840a0cdb2d6bd9a5d07663ac3ee2" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Mon Nov 16 17:39:49 2020 -0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Mon Nov 16 17:39:49 2020 -0800" }, "message": "Merge branch \u0027bnxt_en-bug-fixes\u0027\n\nMichael Chan says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nbnxt_en: Bug fixes.\n\nThis first patch fixes a module eeprom A2h addressing issue. The next\n2 patches fix counter related issues. The last one skips an\nunsupported firmware call on the VF to avoid the error log.\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/1605486472-28156-1-git-send-email-michael.chan@broadcom.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "0ae0a779efb8840a0cdb2d6bd9a5d07663ac3ee2", "tree": "251895e2aef16d0a9e6be148b8ff402e391fa25b", "parents": [ "fa97f303fa4cf8469fd3d1ef29da69c0a3f6ddc8" ], "author": { "name": "Vasundhara Volam", "email": "vasundhara-v.volam@broadcom.com", "time": "Sun Nov 15 19:27:52 2020 -0500" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Mon Nov 16 17:39:47 2020 -0800" }, "message": "bnxt_en: Avoid unnecessary NVM_GET_DEV_INFO cmd error log on VFs.\n\nVFs do not have access permissions to issue NVM_GET_DEV_INFO\nfirmware command.\n\nFixes: 4933f6753b50 (\"bnxt_en: Add bnxt_hwrm_nvm_get_dev_info() to query NVM info.\")\nSigned-off-by: Vasundhara Volam \u003cvasundhara-v.volam@broadcom.com\u003e\nSigned-off-by: Michael Chan \u003cmichael.chan@broadcom.com\u003e\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "fa97f303fa4cf8469fd3d1ef29da69c0a3f6ddc8", "tree": "79174f84cf82e8fe05968ff986067ccc189b19a1", "parents": [ "eba93de6d31c1734dee59909020a162de612e41e" ], "author": { "name": "Michael Chan", "email": "michael.chan@broadcom.com", "time": "Sun Nov 15 19:27:51 2020 -0500" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Mon Nov 16 17:39:46 2020 -0800" }, "message": "bnxt_en: Fix counter overflow logic.\n\nbnxt_add_one_ctr() adds a hardware counter to a software counter and\nadjusts for the hardware counter wraparound against the mask. The logic\nassumes that the hardware counter is always smaller than or equal to\nthe mask.\n\nThis assumption is mostly correct. But in some cases if the firmware\nis older and does not provide the accurate mask, the driver can use\na mask that is smaller than the actual hardware mask. This can cause\nsome extra carry bits to be added to the software counter, resulting in\ncounters that far exceed the actual value. Fix it by masking the\nhardware counter with the mask passed into bnxt_add_one_ctr().\n\nFixes: fea6b3335527 (\"bnxt_en: Accumulate all counters.\")\nReviewed-by: Vasundhara Volam \u003cvasundhara-v.volam@broadcom.com\u003e\nReviewed-by: Pavan Chebbi \u003cpavan.chebbi@broadcom.com\u003e\nReviewed-by: Edwin Peer \u003cedwin.peer@broadcom.com\u003e\nSigned-off-by: Michael Chan \u003cmichael.chan@broadcom.com\u003e\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "eba93de6d31c1734dee59909020a162de612e41e", "tree": "f4eac41c3b2190d05cd146d4e93e3de8fd01fbd7", "parents": [ "4260330b32b14330cfe427d568ac5f5b29b5be3d" ], "author": { "name": "Michael Chan", "email": "michael.chan@broadcom.com", "time": "Sun Nov 15 19:27:50 2020 -0500" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Mon Nov 16 17:39:46 2020 -0800" }, "message": "bnxt_en: Free port stats during firmware reset.\n\nFirmware is unable to retain the port counters during any kind of\nfatal or non-fatal resets, so we must clear the port counters to\navoid false detection of port counter overflow.\n\nFixes: fea6b3335527 (\"bnxt_en: Accumulate all counters.\")\nReviewed-by: Edwin Peer \u003cedwin.peer@broadcom.com\u003e\nReviewed-by: Vasundhara Volam \u003cvasundhara-v.volam@broadcom.com\u003e\nSigned-off-by: Michael Chan \u003cmichael.chan@broadcom.com\u003e\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "4260330b32b14330cfe427d568ac5f5b29b5be3d", "tree": "80c868f3dac0d3538177fc2a4e8a42edad660fd3", "parents": [ "064c9c32b17ca9b36f95eba32ee790dbbebd9a5f" ], "author": { "name": "Edwin Peer", "email": "edwin.peer@broadcom.com", "time": "Sun Nov 15 19:27:49 2020 -0500" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Mon Nov 16 17:39:46 2020 -0800" }, "message": "bnxt_en: read EEPROM A2h address using page 0\n\nThe module eeprom address range returned by bnxt_get_module_eeprom()\nshould be 256 bytes of A0h address space, the lower half of the A2h\naddress space, and page 0 for the upper half of the A2h address space.\n\nFix the firmware call by passing page_number 0 for the A2h slave address\nspace.\n\nFixes: 42ee18fe4ca2 (\"bnxt_en: Add Support for ETHTOOL_GMODULEINFO and ETHTOOL_GMODULEEEPRO\")\nSigned-off-by: Edwin Peer \u003cedwin.peer@broadcom.com\u003e\nSigned-off-by: Michael Chan \u003cmichael.chan@broadcom.com\u003e\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "064c9c32b17ca9b36f95eba32ee790dbbebd9a5f", "tree": "618de50234b6fe2e7a487cc9e8e92ee9bd55165b", "parents": [ "3fe16edf6767decd640fa2654308bc64f8d656dc" ], "author": { "name": "Alex Elder", "email": "elder@linaro.org", "time": "Sat Nov 14 12:20:17 2020 -0600" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Mon Nov 16 17:32:22 2020 -0800" }, "message": "net: ipa: lock when freeing transaction\n\nTransactions sit on one of several lists, depending on their state\n(allocated, pending, complete, or polled). A spinlock protects\nagainst concurrent access when transactions are moved between these\nlists.\n\nTransactions are also reference counted. A newly-allocated\ntransaction has an initial count of 1; a transaction is released in\ngsi_trans_free() only if its decremented reference count reaches 0.\nReleasing a transaction includes removing it from the polled (or if\nunused, allocated) list, so the spinlock is acquired when we release\na transaction.\n\nThe reference count is used to allow a caller to synchronously wait\nfor a committed transaction to complete. In this case, the waiter\ntakes an extra reference to the transaction *before* committing it\n(so it won\u0027t be freed), and releases its reference (calls\ngsi_trans_free()) when it is done with it.\n\nSimilarly, gsi_channel_update() takes an extra reference to ensure a\ntransaction isn\u0027t released before the function is done operating on\nit. Until the transaction is moved to the completed list (by this\nfunction) it won\u0027t be freed, so this reference is taken \"safely.\"\n\nBut in the quiesce path, we want to wait for the \"last\" transaction,\nwhich we find in the completed or polled list. Transactions on\nthese lists can be freed at any time, so we (try to) prevent that\nby taking the reference while holding the spinlock.\n\nCurrently gsi_trans_free() decrements a transaction\u0027s reference\ncount unconditionally, acquiring the lock to remove the transaction\nfrom its list *only* when the count reaches 0. This does not\nprotect the quiesce path, which depends on the lock to ensure its\nextra reference prevents release of the transaction.\n\nFix this by only dropping the last reference to a transaction\nin gsi_trans_free() while holding the spinlock.\n\nFixes: 9dd441e4ed575 (\"soc: qcom: ipa: GSI transactions\")\nReported-by: Stephen Boyd \u003cswboyd@chromium.org\u003e\nSigned-off-by: Alex Elder \u003celder@linaro.org\u003e\nLink: https://lore.kernel.org/r/20201114182017.28270-1-elder@linaro.org\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "3fe16edf6767decd640fa2654308bc64f8d656dc", "tree": "3bf4493610baaf5474748c8d38e9305d9c0ed256", "parents": [ "fc70f5bf5e525dde81565f0a30d5e39168062eba" ], "author": { "name": "Vadim Fedorenko", "email": "vfedorenko@novek.ru", "time": "Sun Nov 15 07:16:00 2020 +0300" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Mon Nov 16 17:11:02 2020 -0800" }, "message": "net/tls: fix corrupted data in recvmsg\n\nIf tcp socket has more data than Encrypted Handshake Message then\ntls_sw_recvmsg will try to decrypt next record instead of returning\nfull control message to userspace as mentioned in comment. The next\nmessage - usually Application Data - gets corrupted because it uses\nzero copy for decryption that\u0027s why the data is not stored in skb\nfor next iteration. Revert check to not decrypt next record if\ncurrent is not Application Data.\n\nFixes: 692d7b5d1f91 (\"tls: Fix recvmsg() to be able to peek across multiple records\")\nSigned-off-by: Vadim Fedorenko \u003cvfedorenko@novek.ru\u003e\nLink: https://lore.kernel.org/r/1605413760-21153-1-git-send-email-vfedorenko@novek.ru\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "fc70f5bf5e525dde81565f0a30d5e39168062eba", "tree": "e8e75cd3c40aac3b4aaaa1783656942338890ed0", "parents": [ "9c79a8ab5f124db01eb1d7287454a702f0d4252f" ], "author": { "name": "Subash Abhinov Kasiviswanathan", "email": "subashab@codeaurora.org", "time": "Fri Nov 13 13:12:05 2020 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Mon Nov 16 16:34:49 2020 -0800" }, "message": "net: qualcomm: rmnet: Fix incorrect receive packet handling during cleanup\n\nDuring rmnet unregistration, the real device rx_handler is first cleared\nfollowed by the removal of rx_handler_data after the rcu synchronization.\n\nAny packets in the receive path may observe that the rx_handler is NULL.\nHowever, there is no check when dereferencing this value to use the\nrmnet_port information.\n\nThis fixes following splat by adding the NULL check.\n\nUnable to handle kernel NULL pointer dereference at virtual\naddress 000000000000000d\npc : rmnet_rx_handler+0x124/0x284\nlr : rmnet_rx_handler+0x124/0x284\n rmnet_rx_handler+0x124/0x284\n __netif_receive_skb_core+0x758/0xd74\n __netif_receive_skb+0x50/0x17c\n process_backlog+0x15c/0x1b8\n napi_poll+0x88/0x284\n net_rx_action+0xbc/0x23c\n __do_softirq+0x20c/0x48c\n\nFixes: ceed73a2cf4a (\"drivers: net: ethernet: qualcomm: rmnet: Initial implementation\")\nSigned-off-by: Sean Tranchetti \u003cstranche@codeaurora.org\u003e\nSigned-off-by: Subash Abhinov Kasiviswanathan \u003csubashab@codeaurora.org\u003e\nLink: https://lore.kernel.org/r/1605298325-3705-1-git-send-email-subashab@codeaurora.org\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "9c79a8ab5f124db01eb1d7287454a702f0d4252f", "tree": "bb60e96ba63669a19278e03632817a25a02e8b3d", "parents": [ "8e5debed39017836a850c6c7bfacc93299d19bad" ], "author": { "name": "Lorenzo Bianconi", "email": "lorenzo@kernel.org", "time": "Fri Nov 13 19:16:57 2020 +0100" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Mon Nov 16 16:30:50 2020 -0800" }, "message": "net: mvneta: fix possible memory leak in mvneta_swbm_add_rx_fragment\n\nRecycle the page running page_pool_put_full_page() in\nmvneta_swbm_add_rx_fragment routine when the last descriptor\ncontains just the FCS or if the received packet contains more than\nMAX_SKB_FRAGS fragments\n\nFixes: ca0e014609f0 (\"net: mvneta: move skb build after descriptors processing\")\nSigned-off-by: Lorenzo Bianconi \u003clorenzo@kernel.org\u003e\nLink: https://lore.kernel.org/r/df6a2bad70323ee58d3901491ada31c1ca2a40b9.1605291228.git.lorenzo@kernel.org\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "8e5debed39017836a850c6c7bfacc93299d19bad", "tree": "eddd74caf98f5c54e767df6b53e8909c68d8b0c6", "parents": [ "7a30ecc9237681bb125cbd30eee92bef7e86293d" ], "author": { "name": "Wong Vee Khee", "email": "vee.khee.wong@intel.com", "time": "Sun Nov 15 15:42:10 2020 +0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Mon Nov 16 16:12:36 2020 -0800" }, "message": "net: stmmac: Use rtnl_lock/unlock on netif_set_real_num_rx_queues() call\n\nFix an issue where dump stack is printed on suspend resume flow due to\nnetif_set_real_num_rx_queues() is not called with rtnl_lock held().\n\nFixes: 686cff3d7022 (\"net: stmmac: Fix incorrect location to set real_num_rx|tx_queues\")\nReported-by: Christophe ROULLIER \u003cchristophe.roullier@st.com\u003e\nTested-by: Christophe ROULLIER \u003cchristophe.roullier@st.com\u003e\nCc: Alexandre TORGUE \u003calexandre.torgue@st.com\u003e\nReviewed-by: Ong Boon Leong \u003cboon.leong.ong@intel.com\u003e\nSigned-off-by: Wong Vee Khee \u003cvee.khee.wong@intel.com\u003e\nLink: https://lore.kernel.org/r/20201115074210.23605-1-vee.khee.wong@intel.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" } ], "next": "7a30ecc9237681bb125cbd30eee92bef7e86293d" }