)]}' { "log": [ { "commit": "16fc44d6387e260f4932e9248b985837324705d8", "tree": "21ab55520c6eed6c78801a6f90444793f3abc56b", "parents": [ "1fe5501ba1abf2b7e78295df73675423bd6899a0", "7412dee9f1fd3e224202b633fdfa6eeaebe0307e" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Apr 21 10:22:54 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Apr 21 10:22:54 2021 -0700" }, "message": "Merge tag \u0027mmc-v5.12-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc\n\nPull MMC fix from Ulf Hansson:\n \"Replace WARN_ONCE with dev_warn_once for non-optimal sg-alignment in\n the meson-gx host driver\"\n\n* tag \u0027mmc-v5.12-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc:\n mmc: meson-gx: replace WARN_ONCE with dev_warn_once about scatterlist size alignment in block mode\n" }, { "commit": "1fe5501ba1abf2b7e78295df73675423bd6899a0", "tree": "ade207f81b2d916f131e636b114cf37b27b08f94", "parents": [ "db2e718a47984b9d71ed890eb2ea36ecf150de18", "0e1e71d34901a633825cd5ae78efaf8abd9215c6" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Apr 20 14:38:35 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Apr 20 14:38:35 2021 -0700" }, "message": "Merge tag \u0027trace-v5.12-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace\n\nPull tracing fix from Steven Rostedt:\n \"Fix tp_printk command line and trace events\n\n Masami added a wrapper to be able to unhash trace event pointers as\n they are only read by root anyway, and they can also be extracted by\n the raw trace data buffers. But this wrapper utilized the iterator to\n have a temporary buffer to manipulate the text with.\n\n tp_printk is a kernel command line option that will send the trace\n output of a trace event to the console on boot up (useful when the\n system crashes before finishing the boot). But the code used the same\n wrapper that Masami added, and its iterator did not have a buffer, and\n this caused the system to crash.\n\n Have the wrapper just print the trace event normally if the iterator\n has no temporary buffer\"\n\n* tag \u0027trace-v5.12-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:\n tracing: Fix checking event hash pointer logic when tp_printk is enabled\n" }, { "commit": "db2e718a47984b9d71ed890eb2ea36ecf150de18", "tree": "809bff3a0c6f30c6b64411f10ccf50f076ea1968", "parents": [ "7af08140979a6e7e12b78c93b8625c8d25b084e2" ], "author": { "name": "Serge E. Hallyn", "email": "serge@hallyn.com", "time": "Tue Apr 20 08:43:34 2021 -0500" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Apr 20 14:28:33 2021 -0700" }, "message": "capabilities: require CAP_SETFCAP to map uid 0\n\ncap_setfcap is required to create file capabilities.\n\nSince commit 8db6c34f1dbc (\"Introduce v3 namespaced file capabilities\"),\na process running as uid 0 but without cap_setfcap is able to work\naround this as follows: unshare a new user namespace which maps parent\nuid 0 into the child namespace.\n\nWhile this task will not have new capabilities against the parent\nnamespace, there is a loophole due to the way namespaced file\ncapabilities are represented as xattrs. File capabilities valid in\nuserns 1 are distinguished from file capabilities valid in userns 2 by\nthe kuid which underlies uid 0. Therefore the restricted root process\ncan unshare a new self-mapping namespace, add a namespaced file\ncapability onto a file, then use that file capability in the parent\nnamespace.\n\nTo prevent that, do not allow mapping parent uid 0 if the process which\nopened the uid_map file does not have CAP_SETFCAP, which is the\ncapability for setting file capabilities.\n\nAs a further wrinkle: a task can unshare its user namespace, then open\nits uid_map file itself, and map (only) its own uid. In this case we do\nnot have the credential from before unshare, which was potentially more\nrestricted. So, when creating a user namespace, we record whether the\ncreator had CAP_SETFCAP. Then we can use that during map_write().\n\nWith this patch:\n\n1. Unprivileged user can still unshare -Ur\n\n ubuntu@caps:~$ unshare -Ur\n root@caps:~# logout\n\n2. Root user can still unshare -Ur\n\n ubuntu@caps:~$ sudo bash\n root@caps:/home/ubuntu# unshare -Ur\n root@caps:/home/ubuntu# logout\n\n3. Root user without CAP_SETFCAP cannot unshare -Ur:\n\n root@caps:/home/ubuntu# /sbin/capsh --drop\u003dcap_setfcap --\n root@caps:/home/ubuntu# /sbin/setcap cap_setfcap\u003dp /sbin/setcap\n unable to set CAP_SETFCAP effective capability: Operation not permitted\n root@caps:/home/ubuntu# unshare -Ur\n unshare: write failed /proc/self/uid_map: Operation not permitted\n\nNote: an alternative solution would be to allow uid 0 mappings by\nprocesses without CAP_SETFCAP, but to prevent such a namespace from\nwriting any file capabilities. This approach can be seen at [1].\n\nBackground history: commit 95ebabde382 (\"capabilities: Don\u0027t allow\nwriting ambiguous v3 file capabilities\") tried to fix the issue by\npreventing v3 fscaps to be written to disk when the root uid would map\nto the same uid in nested user namespaces. This led to regressions for\nvarious workloads. For example, see [2]. Ultimately this is a valid\nuse-case we have to support meaning we had to revert this change in\n3b0c2d3eaa83 (\"Revert 95ebabde382c (\"capabilities: Don\u0027t allow writing\nambiguous v3 file capabilities\")\").\n\nLink: https://git.kernel.org/pub/scm/linux/kernel/git/sergeh/linux.git/log/?h\u003d2021-04-15/setfcap-nsfscaps-v4 [1]\nLink: https://github.com/containers/buildah/issues/3071 [2]\nSigned-off-by: Serge Hallyn \u003cserge@hallyn.com\u003e\nReviewed-by: Andrew G. Morgan \u003cmorgan@kernel.org\u003e\nTested-by: Christian Brauner \u003cchristian.brauner@ubuntu.com\u003e\nReviewed-by: Christian Brauner \u003cchristian.brauner@ubuntu.com\u003e\nTested-by: Giuseppe Scrivano \u003cgscrivan@redhat.com\u003e\nCc: Eric Biederman \u003cebiederm@xmission.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "0e1e71d34901a633825cd5ae78efaf8abd9215c6", "tree": "7ac7cfdfc3ba731c4e25735355bdedaf78a596d5", "parents": [ "8db403b9631331ef1d5e302cdf353c48849ca9d5" ], "author": { "name": "Steven Rostedt (VMware)", "email": "rostedt@goodmis.org", "time": "Mon Apr 19 14:23:12 2021 -0400" }, "committer": { "name": "Steven Rostedt (VMware)", "email": "rostedt@goodmis.org", "time": "Tue Apr 20 10:56:58 2021 -0400" }, "message": "tracing: Fix checking event hash pointer logic when tp_printk is enabled\n\nPointers in events that are printed are unhashed if the flags allow it,\nand the logic to do so is called before processing the event output from\nthe raw ring buffer. In most cases, this is done when a user reads one of\nthe trace files.\n\nBut if tp_printk is added on the kernel command line, this logic is done\nfor trace events when they are triggered, and their output goes out via\nprintk. The unhash logic (and even the validation of the output) did not\nsupport the tp_printk output, and would crash.\n\nLink: https://lore.kernel.org/linux-tegra/9835d9f1-8d3a-3440-c53f-516c2606ad07@nvidia.com/\n\nFixes: efbbdaa22bb7 (\"tracing: Show real address for trace event arguments\")\nReported-by: Jon Hunter \u003cjonathanh@nvidia.com\u003e\nTested-by: Jon Hunter \u003cjonathanh@nvidia.com\u003e\nSigned-off-by: Steven Rostedt (VMware) \u003crostedt@goodmis.org\u003e\n" }, { "commit": "7af08140979a6e7e12b78c93b8625c8d25b084e2", "tree": "1c20d2cdb9f746b513146d28830c82a0d01f6785", "parents": [ "bf05bf16c76bb44ab5156223e1e58e26dfe30a88" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Apr 19 15:08:49 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Apr 19 15:08:49 2021 -0700" }, "message": "Revert \"gcov: clang: fix clang-11+ build\"\n\nThis reverts commit 04c53de57cb6435738961dace8b1b71d3ecd3c39.\n\nNathan Chancellor points out that it should not have been merged into\nmainline by itself. It was a fix for \"gcov: use kvmalloc()\", which is\nstill in -mm/-next. Merging it alone has broken the build.\n\nLink: https://github.com/ClangBuiltLinux/continuous-integration2/runs/2384465683?check_suite_focus\u003dtrue\nReported-by: Nathan Chancellor \u003cnathan@kernel.org\u003e\nCc: Johannes Berg \u003cjohannes.berg@intel.com\u003e\nCc: Nick Desaulniers \u003cndesaulniers@google.com\u003e\nCc: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "7412dee9f1fd3e224202b633fdfa6eeaebe0307e", "tree": "cc3ef6188889d93aa19ce62910ff139033481ad5", "parents": [ "a5e13c6df0e41702d2b2c77c8ad41677ebb065b3" ], "author": { "name": "Neil Armstrong", "email": "narmstrong@baylibre.com", "time": "Fri Apr 16 11:43:47 2021 +0200" }, "committer": { "name": "Ulf Hansson", "email": "ulf.hansson@linaro.org", "time": "Mon Apr 19 09:49:27 2021 +0200" }, "message": "mmc: meson-gx: replace WARN_ONCE with dev_warn_once about scatterlist size alignment in block mode\n\nSince commit e085b51c74cc (\"mmc: meson-gx: check for scatterlist size alignment in block mode\"),\nsupport for SDIO SD_IO_RW_EXTENDED transferts are properly filtered but some driver\nlike brcmfmac still gives a block sg buffer size not aligned with SDIO block,\ntriggerring a WARN_ONCE() with scary stacktrace even if the transfer works fine\nbut with possible degraded performances.\n\nSimply replace with dev_warn_once() to inform user this should be fixed to avoid\ndegraded performance.\n\nThis should be ultimately fixed in brcmfmac, but since it\u0027s only a performance issue\nthe warning should be removed.\n\nFixes: e085b51c74cc (\"mmc: meson-gx: check for scatterlist size alignment in block mode\")\nReported-by: Marek Szyprowski \u003cm.szyprowski@samsung.com\u003e\nSigned-off-by: Neil Armstrong \u003cnarmstrong@baylibre.com\u003e\nLink: https://lore.kernel.org/r/20210416094347.2015896-1-narmstrong@baylibre.com\nSigned-off-by: Ulf Hansson \u003culf.hansson@linaro.org\u003e\n" }, { "commit": "bf05bf16c76bb44ab5156223e1e58e26dfe30a88", "tree": "5b5d3b485abdabe3cd36aa81912bfaa16bb3b845", "parents": [ "5ffe04ccd69ae56af304ff7f98fbdaa4478ed47e" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Apr 18 14:45:32 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Apr 18 14:45:32 2021 -0700" }, "message": "Linux 5.12-rc8\n" }, { "commit": "5ffe04ccd69ae56af304ff7f98fbdaa4478ed47e", "tree": "ac9c37e6bc5bda9bc4c1782fd6d989824d014466", "parents": [ "f5ce0466dc96326e07b52b8fc170c91bc234beb3", "b9a9786a13ea158ee3b23614217f2e3158178421" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Apr 18 13:23:26 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Apr 18 13:23:26 2021 -0700" }, "message": "Merge tag \u0027arm-fixes-5.12-3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc\n\nPull ARM SoC fixes from Arnd Bergmann:\n \"Another smaller set of fixes for three of the Arm platforms:\n\n TI OMAP:\n\n Fix swapped mmc device order also for omap3 that got changed with\n the recent PROBE_PREFER_ASYNCHRONOUS changes. While eventually the\n aliases should be board specific, all the mmc device instances are\n all there in the SoC, and we do probe them by default so that PM\n runtime can idle the devices if left enabled from the bootloader.\n\n Qualcomm Snapdragon:\n\n This bypasses the recently introduced interconnect handling in\n the GENI (serial engine) driver when running off ACPI, as this\n causes the GENI probe to fail and the Lenovo Yoga C630 to boot\n without keyboard and touchpad.\n\n Allwinner:\n\n One 32kHz clock fix for the beelink gs1, a CD polarity fix for the\n SoPine, some MAINTAINERS maintainance, and a clk / reset switch to\n our headers\"\n\n* tag \u0027arm-fixes-5.12-3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc:\n arm64: dts: allwinner: h6: beelink-gs1: Remove ext. 32 kHz osc reference\n MAINTAINERS: Match on allwinner keyword\n MAINTAINERS: Add our new mailing-list\n arm64: dts: allwinner: Fix SD card CD GPIO for SOPine systems\n arm64: dts: allwinner: h6: Switch to macros for RSB clock/reset indices\n ARM: OMAP2+: Fix uninitialized sr_inst\n ARM: dts: Fix swapped mmc order for omap3\n ARM: OMAP2+: Fix warning for omap_init_time_of()\n soc: qcom: geni: shield geni_icc_get() for ACPI boot\n" }, { "commit": "f5ce0466dc96326e07b52b8fc170c91bc234beb3", "tree": "57289971b637dd82d5d551e34cda4399cbe2a7b9", "parents": [ "c98ff1d013d2d53911c4b3e8ba14c7cd141cf1ed", "d2f7eca60b29006285d57c7035539e33300e89e5" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Apr 18 11:55:31 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Apr 18 11:55:31 2021 -0700" }, "message": "Merge tag \u0027for-linus\u0027 of git://git.armlinux.org.uk/~rmk/linux-arm\n\nPull ARM fixes from Russell King:\n\n - Halve maximum number of CPUs if DEBUG_KMAP_LOCAL is enabled\n\n - Fix conversion for_each_membock() to for_each_mem_range()\n\n - Fix footbridge PCI mapping\n\n - Avoid uprobes hooking on thumb instructions\n\n* tag \u0027for-linus\u0027 of git://git.armlinux.org.uk/~rmk/linux-arm:\n ARM: 9071/1: uprobes: Don\u0027t hook on thumb instructions\n ARM: footbridge: fix PCI interrupt mapping\n ARM: 9069/1: NOMMU: Fix conversion for_each_membock() to for_each_mem_range()\n ARM: 9063/1: mm: reduce maximum number of CPUs if DEBUG_KMAP_LOCAL is enabled\n" }, { "commit": "d2f7eca60b29006285d57c7035539e33300e89e5", "tree": "0f2f62725529cb8e7131f464df4f94e20366eeea", "parents": [ "30e3b4f256b4e366a61658c294f6a21b8626dda7" ], "author": { "name": "Fredrik Strupe", "email": "fredrik@strupe.net", "time": "Mon Apr 05 21:52:05 2021 +0100" }, "committer": { "name": "Russell King", "email": "rmk+kernel@armlinux.org.uk", "time": "Sun Apr 18 19:14:45 2021 +0100" }, "message": "ARM: 9071/1: uprobes: Don\u0027t hook on thumb instructions\n\nSince uprobes is not supported for thumb, check that the thumb bit is\nnot set when matching the uprobes instruction hooks.\n\nThe Arm UDF instructions used for uprobes triggering\n(UPROBE_SWBP_ARM_INSN and UPROBE_SS_ARM_INSN) coincidentally share the\nsame encoding as a pair of unallocated 32-bit thumb instructions (not\nUDF) when the condition code is 0b1111 (0xf). This in effect makes it\npossible to trigger the uprobes functionality from thumb, and at that\nusing two unallocated instructions which are not permanently undefined.\n\nSigned-off-by: Fredrik Strupe \u003cfredrik@strupe.net\u003e\nCc: stable@vger.kernel.org\nFixes: c7edc9e326d5 (\"ARM: add uprobes support\")\nSigned-off-by: Russell King \u003crmk+kernel@armlinux.org.uk\u003e\n" }, { "commit": "c98ff1d013d2d53911c4b3e8ba14c7cd141cf1ed", "tree": "d81f88120f6972923bc60098cae5323ce6e92f8e", "parents": [ "aba5970c4da0059ed8052188a341a85ac8845d6f", "176ddd89171ddcf661862d90c5d257877f7326d6" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 17 20:25:33 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 17 20:25:33 2021 -0700" }, "message": "Merge tag \u0027scsi-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi\n\nPull SCSI fixes from James Bottomley:\n \"Two fixes: the libsas fix is for a problem that occurs when trying to\n change the cache type of an ATA device and the libiscsi one is a\n regression fix from this merge window\"\n\n* tag \u0027scsi-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:\n scsi: libsas: Reset num_scatter if libata marks qc as NODATA\n scsi: iscsi: Fix iSCSI cls conn state\n" }, { "commit": "aba5970c4da0059ed8052188a341a85ac8845d6f", "tree": "3157a568eb4ae86a544dafe36dec656645f031df", "parents": [ "194cf4825638256e9afe1d360831aa5379b3517a", "796b556cbfdbe0d65b9793e63653ad24f9fa033c" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 17 20:00:36 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 17 20:00:36 2021 -0700" }, "message": "Merge tag \u0027drm-fixes-2021-04-18\u0027 of git://anongit.freedesktop.org/drm/drm\n\nPull vmwgfx fixes from Dave Airlie:\n \"This contains two regression fixes for vmwgfx, one due to a refactor\n which meant locks were being used before initialisation, and the other\n in fixing up some warnings from the core when destroying pinned\n buffers.\n\n vmwgfx:\n\n - fixed unpinning before destruction\n\n - lockdep init reordering\"\n\n* tag \u0027drm-fixes-2021-04-18\u0027 of git://anongit.freedesktop.org/drm/drm:\n drm/vmwgfx: Make sure bo\u0027s are unpinned before putting them back\n drm/vmwgfx: Fix the lockdep breakage\n drm/vmwgfx: Make sure we unpin no longer needed buffers\n" }, { "commit": "796b556cbfdbe0d65b9793e63653ad24f9fa033c", "tree": "cfdc57e0d8343f65a59d31eef1ceb0544beda18e", "parents": [ "4d2e1288372ccc5ac60290bc10cace49c9bfa6d0", "2ef4fb92363c44e8a6f93fd0877b6a7dee6f874d" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sun Apr 18 09:26:54 2021 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sun Apr 18 09:27:00 2021 +1000" }, "message": "Merge tag \u0027vmwgfx-fixes-2021-04-14\u0027 of gitlab.freedesktop.org:zack/vmwgfx into drm-fixes\n\nvmwgfx fixes for regressions in 5.12\n\nHere\u0027s a set of 3 patches fixing ugly regressions\nin the vmwgfx driver. We broke lock initialization\ncode and ended up using spinlocks before initialization\nbreaking lockdep.\nAlso there was a bit of a fallout from drm changes\nwhich made the core validate that unreferenced buffers\nhave been unpinned. vmwgfx pinning code predates a lot\nof the core drm and wasn\u0027t written to account for those\nsemantics. Fortunately changes required to fix it\nare not too intrusive.\nThe changes have been validated by our internal ci.\n\nSigned-off-by: Zack Rusin \u003czackr@vmware.com\u003e\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n\nFrom: Zack Rusin \u003czackr@vmware.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/f7add0a2-162e-3bd2-b1be-344a94f2acbf@vmware.com\n" }, { "commit": "194cf4825638256e9afe1d360831aa5379b3517a", "tree": "5ed04e8f5dfb13bdae489523bd08c5dca10d995c", "parents": [ "0c93ac69407d63a85be0129aa55ffaec27ffebd3", "39930213e7779b9c4257499972b8afb8858f1a2d" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 17 12:01:51 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 17 12:01:51 2021 -0700" }, "message": "Merge branch \u0027i2c/for-current\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux\n\nPull i2c fix from Wolfram Sang:\n \"One more driver bugfix for I2C\"\n\n* \u0027i2c/for-current\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux:\n i2c: mv64xxx: Fix random system lock caused by runtime PM\n" }, { "commit": "0c93ac69407d63a85be0129aa55ffaec27ffebd3", "tree": "77a2a5a721266f517fefba13f75fe2b514b45067", "parents": [ "88a5af943985fb43b4c9472b5abd9c0b9705533d" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 17 09:27:04 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 17 11:39:49 2021 -0700" }, "message": "readdir: make sure to verify directory entry for legacy interfaces too\n\nThis does the directory entry name verification for the legacy\n\"fillonedir\" (and compat) interface that goes all the way back to the\ndark ages before we had a proper dirent, and the readdir() system call\nreturned just a single entry at a time.\n\nNobody should use this interface unless you still have binaries from\n1991, but let\u0027s do it right.\n\nThis came up during discussions about unsafe_copy_to_user() and proper\nchecking of all the inputs to it, as the networking layer is looking to\nuse it in a few new places. So let\u0027s make sure the _old_ users do it\nall right and proper, before we add new ones.\n\nSee also commit 8a23eb804ca4 (\"Make filldir[64]() verify the directory\nentry filename is valid\") which did the proper modern interfaces that\npeople actually use. It had a note:\n\n Note that I didn\u0027t bother adding the checks to any legacy interfaces\n that nobody uses.\n\nwhich this now corrects. Note that we really don\u0027t care about POSIX and\nthe presense of \u0027/\u0027 in a directory entry, but verify_dirent_name() also\nends up doing the proper name length verification which is what the\ninput checking discussion was about.\n\n[ Another option would be to remove the support for this particular very\n old interface: any binaries that use it are likely a.out binaries, and\n they will no longer run anyway since we removed a.out binftm support\n in commit eac616557050 (\"x86: Deprecate a.out support\").\n\n But I\u0027m not sure which came first: getdents() or ELF support, so let\u0027s\n pretend somebody might still have a working binary that uses the\n legacy readdir() case.. ]\n\nLink: https://lore.kernel.org/lkml/CAHk-\u003dwjbvzCAhAtvG0d81W5o0-KT5PPTHhfJ5ieDFq+bGtgOYg@mail.gmail.com/\nAcked-by: Al Viro \u003cviro@zeniv.linux.org.uk\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "88a5af943985fb43b4c9472b5abd9c0b9705533d", "tree": "f58881656d7d8b064b50ad3ca99072d7ee7920c2", "parents": [ "bdfd99e6d6bd690b47bd1d45dad218bf08be1dde", "f2764bd4f6a8dffaec3e220728385d9756b3c2cb" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 17 09:57:15 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 17 09:57:15 2021 -0700" }, "message": "Merge tag \u0027net-5.12-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net\n\nPull networking fixes from Jakub Kicinski:\n \"Networking fixes for 5.12-rc8, including fixes from netfilter, and\n bpf. BPF verifier changes stand out, otherwise things have slowed\n down.\n\n Current release - regressions:\n\n - gro: ensure frag0 meets IP header alignment\n\n - Revert \"net: stmmac: re-init rx buffers when mac resume back\"\n\n - ethernet: macb: fix the restore of cmp registers\n\n Previous releases - regressions:\n\n - ixgbe: Fix NULL pointer dereference in ethtool loopback test\n\n - ixgbe: fix unbalanced device enable/disable in suspend/resume\n\n - phy: marvell: fix detection of PHY on Topaz switches\n\n - make tcp_allowed_congestion_control readonly in non-init netns\n\n - xen-netback: Check for hotplug-status existence before watching\n\n Previous releases - always broken:\n\n - bpf: mitigate a speculative oob read of up to map value size by\n tightening the masking window\n\n - sctp: fix race condition in sctp_destroy_sock\n\n - sit, ip6_tunnel: Unregister catch-all devices\n\n - netfilter: nftables: clone set element expression template\n\n - netfilter: flowtable: fix NAT IPv6 offload mangling\n\n - net: geneve: check skb is large enough for IPv4/IPv6 header\n\n - netlink: don\u0027t call -\u003enetlink_bind with table lock held\"\n\n* tag \u0027net-5.12-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (52 commits)\n netlink: don\u0027t call -\u003enetlink_bind with table lock held\n MAINTAINERS: update my email\n bpf: Update selftests to reflect new error states\n bpf: Tighten speculative pointer arithmetic mask\n bpf: Move sanitize_val_alu out of op switch\n bpf: Refactor and streamline bounds check into helper\n bpf: Improve verifier error messages for users\n bpf: Rework ptr_limit into alu_limit and add common error path\n bpf: Ensure off_reg has no mixed signed bounds for all types\n bpf: Move off_reg into sanitize_ptr_alu\n bpf: Use correct permission flag for mixed signed bounds arithmetic\n ch_ktls: do not send snd_una update to TCB in middle\n ch_ktls: tcb close causes tls connection failure\n ch_ktls: fix device connection close\n ch_ktls: Fix kernel panic\n i40e: fix the panic when running bpf in xdpdrv mode\n net/mlx5e: fix ingress_ifindex check in mlx5e_flower_parse_meta\n net/mlx5e: Fix setting of RS FEC mode\n net/mlx5: Fix setting of devlink traps in switchdev mode\n Revert \"net: stmmac: re-init rx buffers when mac resume back\"\n ...\n" }, { "commit": "bdfd99e6d6bd690b47bd1d45dad218bf08be1dde", "tree": "3a73c1657b5a1392c9b34850b2f3844b794bf531", "parents": [ "7c22677407243d63df3aee1bb2f6d9aa12c01a24", "11d2498f1568a0f923dc8ef7621de15a9e89267f" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 17 09:40:44 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 17 09:40:44 2021 -0700" }, "message": "Merge tag \u0027libnvdimm-fixes-for-5.12-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm\n\nPull libnvdimm fixes from Dan Williams:\n \"The largest change is for a regression that landed during -rc1 for\n block-device read-only handling. Vaibhav found a new use for the\n ability (originally introduced by virtio_pmem) to call back to the\n platform to flush data, but also found an original bug in that\n implementation. Lastly, Arnd cleans up some compile warnings in dax.\n\n This has all appeared in -next with no reported issues.\n\n Summary:\n\n - Fix a regression of read-only handling in the pmem driver\n\n - Fix a compile warning\n\n - Fix support for platform cache flush commands on powerpc/papr\"\n\n* tag \u0027libnvdimm-fixes-for-5.12-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm:\n libnvdimm/region: Fix nvdimm_has_flush() to handle ND_REGION_ASYNC\n libnvdimm: Notify disk drivers to revalidate region read-only\n dax: avoid -Wempty-body warnings\n" }, { "commit": "7c22677407243d63df3aee1bb2f6d9aa12c01a24", "tree": "f92ad0274a97b10418fb4d5a1d811cd8264df473", "parents": [ "fdb5d6cab638a2881687b59b27817c74d8efac76", "fae8817ae804a682c6823ad1672438f39fc46c28" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 17 09:30:58 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 17 09:30:58 2021 -0700" }, "message": "Merge tag \u0027cxl-fixes-for-5.12-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxl\n\nPull CXL memory class fixes from Dan Williams:\n \"A collection of fixes for the CXL memory class driver introduced in\n this release cycle.\n\n The driver was primarily developed on a work-in-progress QEMU\n emulation of the interface and we have since found a couple places\n where it hid spec compliance bugs in the driver, or had a spec\n implementation bug itself.\n\n The biggest change here is replacing a percpu_ref with an rwsem to\n cleanup a couple bugs in the error unwind path during ioctl device\n init. Lastly there were some minor cleanups to not export the\n power-management sysfs-ABI for the ioctl device, use the proper sysfs\n helper for emitting values, and prevent subtle bugs as new\n administration commands are added to the supported list.\n\n The bulk of it has appeared in -next save for the top commit which was\n found today and validated on a fixed-up QEMU model.\n\n Summary:\n\n - Fix support for CXL memory devices with registers offset from the\n BAR base.\n\n - Fix the reporting of device capacity.\n\n - Fix the driver commands list definition to be disconnected from the\n UAPI command list.\n\n - Replace percpu_ref with rwsem to fix initialization error path.\n\n - Fix leaks in the driver initialization error path.\n\n - Drop the power/ directory from CXL device sysfs.\n\n - Use the recommended sysfs helper for attribute \u0027show\u0027\n implementations\"\n\n* tag \u0027cxl-fixes-for-5.12-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxl:\n cxl/mem: Fix memory device capacity probing\n cxl/mem: Fix register block offset calculation\n cxl/mem: Force array size of mem_commands[] to CXL_MEM_COMMAND_ID_MAX\n cxl/mem: Disable cxl device power management\n cxl/mem: Do not rely on device_add() side effects for dev_set_name() failures\n cxl/mem: Fix synchronization mechanism for device removal vs ioctl operations\n cxl/mem: Use sysfs_emit() for attribute show routines\n" }, { "commit": "fdb5d6cab638a2881687b59b27817c74d8efac76", "tree": "af26113cf18046a0debd0915653319f4da301873", "parents": [ "9cdbf6467424045617cd6e79dcaad06bb8efa31c", "c95c2d328cd051484bea161e66dfa715c02a7d7e" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 17 08:38:23 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 17 08:38:23 2021 -0700" }, "message": "Merge branch \u0027akpm\u0027 (patches from Andrew)\n\nMerge misc fixes from Andrew Morton:\n \"12 patches.\n\n Subsystems affected by this patch series: mm (documentation, kasan,\n and pagemap), csky, ia64, gcov, and lib\"\n\n* emailed patches from Andrew Morton \u003cakpm@linux-foundation.org\u003e:\n lib: remove \"expecting prototype\" kernel-doc warnings\n gcov: clang: fix clang-11+ build\n mm: ptdump: fix build failure\n mm/mapping_dirty_helpers: guard hugepage pud\u0027s usage\n ia64: tools: remove duplicate definition of ia64_mf() on ia64\n ia64: tools: remove inclusion of ia64-specific version of errno.h header\n ia64: fix discontig.c section mismatches\n ia64: remove duplicate entries in generic_defconfig\n csky: change a Kconfig symbol name to fix e1000 build error\n kasan: remove redundant config option\n kasan: fix hwasan build for gcc\n mm: eliminate \"expecting prototype\" kernel-doc warnings\n" }, { "commit": "fae8817ae804a682c6823ad1672438f39fc46c28", "tree": "2776b4624c017f4f28c581352686bebb266d0681", "parents": [ "b21bb4cd1102dd9e24a169d09cf4e6f3c8a46bcf" ], "author": { "name": "Dan Williams", "email": "dan.j.williams@intel.com", "time": "Fri Apr 16 17:43:30 2021 -0700" }, "committer": { "name": "Dan Williams", "email": "dan.j.williams@intel.com", "time": "Fri Apr 16 18:21:56 2021 -0700" }, "message": "cxl/mem: Fix memory device capacity probing\n\nThe CXL Identify Memory Device output payload emits capacity in 256MB\nunits. The driver is treating the capacity field as bytes. This was\nmissed because QEMU reports bytes when it should report bytes / 256MB.\n\nFixes: 8adaf747c9f0 (\"cxl/mem: Find device capabilities\")\nReviewed-by: Vishal Verma \u003cvishal.l.verma@intel.com\u003e\nCc: Ben Widawsky \u003cben.widawsky@intel.com\u003e\nLink: https://lore.kernel.org/r/161862021044.3259705.7008520073059739760.stgit@dwillia2-desk3.amr.corp.intel.com\nSigned-off-by: Dan Williams \u003cdan.j.williams@intel.com\u003e\n" }, { "commit": "f2764bd4f6a8dffaec3e220728385d9756b3c2cb", "tree": "23600181883a422fbe145e849744dd39246273db", "parents": [ "b02265429681c9c827c45978a61a9f00be5ea9aa" ], "author": { "name": "Florian Westphal", "email": "fw@strlen.de", "time": "Fri Apr 16 21:29:13 2021 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Apr 16 17:01:04 2021 -0700" }, "message": "netlink: don\u0027t call -\u003enetlink_bind with table lock held\n\nWhen I added support to allow generic netlink multicast groups to be\nrestricted to subscribers with CAP_NET_ADMIN I was unaware that a\ngenl_bind implementation already existed in the past.\n\nIt was reverted due to ABBA deadlock:\n\n1. -\u003enetlink_bind gets called with the table lock held.\n2. genetlink bind callback is invoked, it grabs the genl lock.\n\nBut when a new genl subsystem is (un)registered, these two locks are\ntaken in reverse order.\n\nOne solution would be to revert again and add a comment in genl\nreferring 1e82a62fec613, \"genetlink: remove genl_bind\").\n\nThis would need a second change in mptcp to not expose the raw token\nvalue anymore, e.g. by hashing the token with a secret key so userspace\ncan still associate subflow events with the correct mptcp connection.\n\nHowever, Paolo Abeni reminded me to double-check why the netlink table is\nlocked in the first place.\n\nI can\u0027t find one. netlink_bind() is already called without this lock\nwhen userspace joins a group via NETLINK_ADD_MEMBERSHIP setsockopt.\nSame holds for the netlink_unbind operation.\n\nDigging through the history, commit f773608026ee1\n(\"netlink: access nlk groups safely in netlink bind and getname\")\nexpanded the lock scope.\n\ncommit 3a20773beeeeade (\"net: netlink: cap max groups which will be considered in netlink_bind()\")\n... removed the nlk-\u003engroups access that the lock scope\nextension was all about.\n\nReduce the lock scope again and always call -\u003enetlink_bind without\nthe table lock.\n\nThe Fixes tag should be vs. the patch mentioned in the link below,\nbut that one got squash-merged into the patch that came earlier in the\nseries.\n\nFixes: 4d54cc32112d8d (\"mptcp: avoid lock_fast usage in accept path\")\nLink: https://lore.kernel.org/mptcp/20210213000001.379332-8-mathew.j.martineau@linux.intel.com/T/#u\nCc: Cong Wang \u003cxiyou.wangcong@gmail.com\u003e\nCc: Xin Long \u003clucien.xin@gmail.com\u003e\nCc: Johannes Berg \u003cjohannes.berg@intel.com\u003e\nCc: Sean Tranchetti \u003cstranche@codeaurora.org\u003e\nCc: Paolo Abeni \u003cpabeni@redhat.com\u003e\nCc: Pablo Neira Ayuso \u003cpablo@netfilter.org\u003e\nSigned-off-by: Florian Westphal \u003cfw@strlen.de\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "9cdbf6467424045617cd6e79dcaad06bb8efa31c", "tree": "7d9fd167cf16408c1dbb4e0f71b3bc60ad1b09c0", "parents": [ "151501160401e2dc669ea7dac2c599b53f220c33", "c7d95613c7d6e003969722a290397b8271bdad17" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 16:18:53 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 16:18:53 2021 -0700" }, "message": "Merge tag \u0027io_uring-5.12-2021-04-16\u0027 of git://git.kernel.dk/linux-block\n\nPull io_uring fix from Jens Axboe:\n \"Fix for a potential hang at exit with SQPOLL from Pavel\"\n\n* tag \u0027io_uring-5.12-2021-04-16\u0027 of git://git.kernel.dk/linux-block:\n io_uring: fix early sqd_list removal sqpoll hangs\n" }, { "commit": "c95c2d328cd051484bea161e66dfa715c02a7d7e", "tree": "a3e049406277ddfe017853d77c66823a198ebb6d", "parents": [ "04c53de57cb6435738961dace8b1b71d3ecd3c39" ], "author": { "name": "Randy Dunlap", "email": "rdunlap@infradead.org", "time": "Fri Apr 16 15:46:26 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 16:10:37 2021 -0700" }, "message": "lib: remove \"expecting prototype\" kernel-doc warnings\n\nFix various kernel-doc warnings in lib/ due to missing or erroneous\nfunction names.\n\nAdd kernel-doc for some function parameters that was missing. Use\nkernel-doc \"Return:\" notation in earlycpio.c.\n\nQuietens the following warnings:\n\n lib/earlycpio.c:61: warning: expecting prototype for cpio_data find_cpio_data(). Prototype was for find_cpio_data() instead\n\n lib/lru_cache.c:640: warning: expecting prototype for lc_dump(). Prototype was for lc_seq_dump_details() instead\n lru_cache.c:90: warning: Function parameter or member \u0027cache\u0027 not described in \u0027lc_create\u0027\n\n lib/parman.c:368: warning: expecting prototype for parman_item_del(). Prototype was for parman_item_remove() instead\n parman.c:309: warning: Excess function parameter \u0027prority\u0027 description in \u0027parman_prio_init\u0027\n\n lib/radix-tree.c:703: warning: expecting prototype for __radix_tree_insert(). Prototype was for radix_tree_insert() instead\n radix-tree.c:180: warning: Excess function parameter \u0027addr\u0027 description in \u0027radix_tree_find_next_bit\u0027\n radix-tree.c:180: warning: Excess function parameter \u0027size\u0027 description in \u0027radix_tree_find_next_bit\u0027\n radix-tree.c:931: warning: Function parameter or member \u0027iter\u0027 not described in \u0027radix_tree_iter_replace\u0027\n\nLink: https://lkml.kernel.org/r/20210411221756.15461-1-rdunlap@infradead.org\nSigned-off-by: Randy Dunlap \u003crdunlap@infradead.org\u003e\nCc: Philipp Reisner \u003cphilipp.reisner@linbit.com\u003e\nCc: Lars Ellenberg \u003clars.ellenberg@linbit.com\u003e\nCc: Jiri Pirko \u003cjiri@nvidia.com\u003e\nCc: Matthew Wilcox \u003cwilly@infradead.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "04c53de57cb6435738961dace8b1b71d3ecd3c39", "tree": "be53511525bf368b6ba5532d71ef349d995a6115", "parents": [ "458376913d86bed2fb781b4952eb6861675ef3be" ], "author": { "name": "Johannes Berg", "email": "johannes.berg@intel.com", "time": "Fri Apr 16 15:46:23 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 16:10:37 2021 -0700" }, "message": "gcov: clang: fix clang-11+ build\n\nWith clang-11+, the code is broken due to my kvmalloc() conversion\n(which predated the clang-11 support code) leaving one vmalloc() in\nplace. Fix that.\n\nLink: https://lkml.kernel.org/r/20210412214210.6e1ecca9cdc5.I24459763acf0591d5e6b31c7e3a59890d802f79c@changeid\nSigned-off-by: Johannes Berg \u003cjohannes.berg@intel.com\u003e\nReviewed-by: Nick Desaulniers \u003cndesaulniers@google.com\u003e\nTested-by: Nick Desaulniers \u003cndesaulniers@google.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "458376913d86bed2fb781b4952eb6861675ef3be", "tree": "cfe62267bb4b8dbb1104c71df165c4b6418e16af", "parents": [ "94036f4c884377bdf2da1ba7666c9599d6df0191" ], "author": { "name": "Christophe Leroy", "email": "christophe.leroy@csgroup.eu", "time": "Fri Apr 16 15:46:20 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 16:10:37 2021 -0700" }, "message": "mm: ptdump: fix build failure\n\nREAD_ONCE() cannot be used for reading PTEs. Use ptep_get() instead, to\navoid the following errors:\n\n CC mm/ptdump.o\n In file included from \u003ccommand-line\u003e:\n mm/ptdump.c: In function \u0027ptdump_pte_entry\u0027:\n include/linux/compiler_types.h:320:38: error: call to \u0027__compiletime_assert_207\u0027 declared with attribute error: Unsupported access size for {READ,WRITE}_ONCE().\n 320 | _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)\n | ^\n include/linux/compiler_types.h:301:4: note: in definition of macro \u0027__compiletime_assert\u0027\n 301 | prefix ## suffix(); \\\n | ^~~~~~\n include/linux/compiler_types.h:320:2: note: in expansion of macro \u0027_compiletime_assert\u0027\n 320 | _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)\n | ^~~~~~~~~~~~~~~~~~~\n include/asm-generic/rwonce.h:36:2: note: in expansion of macro \u0027compiletime_assert\u0027\n 36 | compiletime_assert(__native_word(t) || sizeof(t) \u003d\u003d sizeof(long long), \\\n | ^~~~~~~~~~~~~~~~~~\n include/asm-generic/rwonce.h:49:2: note: in expansion of macro \u0027compiletime_assert_rwonce_type\u0027\n 49 | compiletime_assert_rwonce_type(x); \\\n | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n mm/ptdump.c:114:14: note: in expansion of macro \u0027READ_ONCE\u0027\n 114 | pte_t val \u003d READ_ONCE(*pte);\n | ^~~~~~~~~\n make[2]: *** [mm/ptdump.o] Error 1\n\nSee commit 481e980a7c19 (\"mm: Allow arches to provide ptep_get()\") and\ncommit c0e1c8c22beb (\"powerpc/8xx: Provide ptep_get() with 16k pages\")\nfor details.\n\nLink: https://lkml.kernel.org/r/912b349e2bcaa88939904815ca0af945740c6bd4.1618478922.git.christophe.leroy@csgroup.eu\nFixes: 30d621f6723b (\"mm: add generic ptdump\")\nSigned-off-by: Christophe Leroy \u003cchristophe.leroy@csgroup.eu\u003e\nCc: Steven Price \u003csteven.price@arm.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "94036f4c884377bdf2da1ba7666c9599d6df0191", "tree": "6fbfd0aefe3b50004c678235c526afa15733d8ca", "parents": [ "f4bf09dc3aaa4b07cd15630f2023f68cb2668809" ], "author": { "name": "Zack Rusin", "email": "zackr@vmware.com", "time": "Fri Apr 16 15:46:18 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 16:10:37 2021 -0700" }, "message": "mm/mapping_dirty_helpers: guard hugepage pud\u0027s usage\n\nMapping dirty helpers have, so far, been only used on X86, but a port of\nvmwgfx to ARM64 exposed a problem which results in a compilation error\non ARM64 systems:\n\n mm/mapping_dirty_helpers.c: In function `wp_clean_pud_entry\u0027:\n mm/mapping_dirty_helpers.c:172:32: error: implicit declaration of function `pud_dirty\u0027; did you mean `pmd_dirty\u0027? [-Werror\u003dimplicit-function-declaration]\n\nThis is due to the fact that mapping_dirty_helpers code assumes that\npud_dirty is always defined, which is not the case for architectures\nthat don\u0027t define CONFIG_HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD.\n\nARM64 arch is a little inconsistent when it comes to PUD hugepage\nhelpers, e.g. it defines pud_young but not pud_dirty but regardless of\nthat the core kernel code shouldn\u0027t assume that any of the PUD hugepage\nhelpers are available unless CONFIG_HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD\nis defined. This prevents compilation errors whenever one of the\ndrivers is ported to new architectures.\n\nLink: https://lkml.kernel.org/r/20210409165151.694574-1-zackr@vmware.com\nSigned-off-by: Zack Rusin \u003czackr@vmware.com\u003e\nReviewed-by: Thomas Hellstrm (Intel) \u003cthomas_os@shipmail.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "f4bf09dc3aaa4b07cd15630f2023f68cb2668809", "tree": "23c13910ef66cc69a3aa5260c88b424c5347d0fc", "parents": [ "17786fea414393813b56e33a1a01b2dfa03c0915" ], "author": { "name": "John Paul Adrian Glaubitz", "email": "glaubitz@physik.fu-berlin.de", "time": "Fri Apr 16 15:46:15 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 16:10:37 2021 -0700" }, "message": "ia64: tools: remove duplicate definition of ia64_mf() on ia64\n\nThe ia64_mf() macro defined in tools/arch/ia64/include/asm/barrier.h is\nalready defined in \u003casm/gcc_intrin.h\u003e on ia64 which causes libbpf\nfailing to build:\n\n CC /usr/src/linux/tools/bpf/bpftool//libbpf/staticobjs/libbpf.o\n In file included from /usr/src/linux/tools/include/asm/barrier.h:24,\n from /usr/src/linux/tools/include/linux/ring_buffer.h:4,\n from libbpf.c:37:\n /usr/src/linux/tools/include/asm/../../arch/ia64/include/asm/barrier.h:43: error: \"ia64_mf\" redefined [-Werror]\n 43 | #define ia64_mf() asm volatile (\"mf\" ::: \"memory\")\n |\n In file included from /usr/include/ia64-linux-gnu/asm/intrinsics.h:20,\n from /usr/include/ia64-linux-gnu/asm/swab.h:11,\n from /usr/include/linux/swab.h:8,\n from /usr/include/linux/byteorder/little_endian.h:13,\n from /usr/include/ia64-linux-gnu/asm/byteorder.h:5,\n from /usr/src/linux/tools/include/uapi/linux/perf_event.h:20,\n from libbpf.c:36:\n /usr/include/ia64-linux-gnu/asm/gcc_intrin.h:382: note: this is the location of the previous definition\n 382 | #define ia64_mf() __asm__ volatile (\"mf\" ::: \"memory\")\n |\n cc1: all warnings being treated as errors\n\nThus, remove the definition from tools/arch/ia64/include/asm/barrier.h.\n\nSigned-off-by: John Paul Adrian Glaubitz \u003cglaubitz@physik.fu-berlin.de\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "17786fea414393813b56e33a1a01b2dfa03c0915", "tree": "e9cdc3bf2e9bcf5170e1e952cc2dbcfb24445b02", "parents": [ "e2af9da4f867a1a54f1252bf3abc1a5c63951778" ], "author": { "name": "John Paul Adrian Glaubitz", "email": "glaubitz () physik ! fu-berlin ! de", "time": "Fri Apr 16 15:46:12 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 16:10:37 2021 -0700" }, "message": "ia64: tools: remove inclusion of ia64-specific version of errno.h header\n\nThere is no longer an ia64-specific version of the errno.h header below\narch/ia64/include/uapi/asm/, so trying to build tools/bpf fails with:\n\n CC /usr/src/linux/tools/bpf/bpftool/btf_dumper.o\n In file included from /usr/src/linux/tools/include/linux/err.h:8,\n from btf_dumper.c:11:\n /usr/src/linux/tools/include/uapi/asm/errno.h:13:10: fatal error: ../../../arch/ia64/include/uapi/asm/errno.h: No such file or directory\n 13 | #include \"../../../arch/ia64/include/uapi/asm/errno.h\"\n | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n compilation terminated.\n\nThus, just remove the inclusion of the ia64-specific errno.h so that the\nbuild will use the generic errno.h header on this target which was used\nthere anyway as the ia64-specific errno.h was just a wrapper for the\ngeneric header.\n\nFixes: c25f867ddd00 (\"ia64: remove unneeded uapi asm-generic wrappers\")\nSigned-off-by: John Paul Adrian Glaubitz \u003cglaubitz@physik.fu-berlin.de\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "e2af9da4f867a1a54f1252bf3abc1a5c63951778", "tree": "bf88de9b74e16a5ac960f4048d756ed9b8c5d25f", "parents": [ "19d000d93303e05bd7b1326e3de9df05a41b25b5" ], "author": { "name": "Randy Dunlap", "email": "rdunlap@infradead.org", "time": "Fri Apr 16 15:46:09 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 16:10:37 2021 -0700" }, "message": "ia64: fix discontig.c section mismatches\n\nFix IA64 discontig.c Section mismatch warnings.\n\nWhen CONFIG_SPARSEMEM\u003dy and CONFIG_MEMORY_HOTPLUG\u003dy, the functions\ncomputer_pernodesize() and scatter_node_data() should not be marked as\n__meminit because they are needed after init, on any memory hotplug\nevent. Also, early_nr_cpus_node() is called by compute_pernodesize(),\nso early_nr_cpus_node() cannot be __meminit either.\n\n WARNING: modpost: vmlinux.o(.text.unlikely+0x1612): Section mismatch in reference from the function arch_alloc_nodedata() to the function .meminit.text:compute_pernodesize()\n The function arch_alloc_nodedata() references the function __meminit compute_pernodesize().\n This is often because arch_alloc_nodedata lacks a __meminit annotation or the annotation of compute_pernodesize is wrong.\n\n WARNING: modpost: vmlinux.o(.text.unlikely+0x1692): Section mismatch in reference from the function arch_refresh_nodedata() to the function .meminit.text:scatter_node_data()\n The function arch_refresh_nodedata() references the function __meminit scatter_node_data().\n This is often because arch_refresh_nodedata lacks a __meminit annotation or the annotation of scatter_node_data is wrong.\n\n WARNING: modpost: vmlinux.o(.text.unlikely+0x1502): Section mismatch in reference from the function compute_pernodesize() to the function .meminit.text:early_nr_cpus_node()\n The function compute_pernodesize() references the function __meminit early_nr_cpus_node().\n This is often because compute_pernodesize lacks a __meminit annotation or the annotation of early_nr_cpus_node is wrong.\n\nLink: https://lkml.kernel.org/r/20210411001201.3069-1-rdunlap@infradead.org\nSigned-off-by: Randy Dunlap \u003crdunlap@infradead.org\u003e\nCc: Mike Rapoport \u003crppt@kernel.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "19d000d93303e05bd7b1326e3de9df05a41b25b5", "tree": "43522ab07aa85ca27c22f538d8432d5905d5c054", "parents": [ "d199161653d612b8fb96ac51bfd5b2d2782ecef3" ], "author": { "name": "Randy Dunlap", "email": "rdunlap@infradead.org", "time": "Fri Apr 16 15:46:06 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 16:10:37 2021 -0700" }, "message": "ia64: remove duplicate entries in generic_defconfig\n\nFix ia64 generic_defconfig duplicate entries, as warned by:\n\n arch/ia64/configs/generic_defconfig: warning: override: reassigning to symbol ATA: \u003d\u003e 58\n arch/ia64/configs/generic_defconfig: warning: override: reassigning to symbol ATA_PIIX: \u003d\u003e 59\n\nThese 2 symbols still have the same value as in the removed lines.\n\nLink: https://lkml.kernel.org/r/20210411020255.18052-1-rdunlap@infradead.org\nFixes: c331649e6371 (\"ia64: Use libata instead of the legacy ide driver in defconfigs\")\nSigned-off-by: Randy Dunlap \u003crdunlap@infradead.org\u003e\nReported-by: Geert Uytterhoeven \u003cgeert@linux-m68k.org\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Tony Luck \u003ctony.luck@intel.com\u003e\nCc: Fenghua Yu \u003cfenghua.yu@intel.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "d199161653d612b8fb96ac51bfd5b2d2782ecef3", "tree": "a5dfcc0f7941e0d9de4c84a2fe40c89d12129a62", "parents": [ "02c587733c8161355a43e6e110c2e29bd0acff72" ], "author": { "name": "Randy Dunlap", "email": "rdunlap@infradead.org", "time": "Fri Apr 16 15:46:03 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 16:10:36 2021 -0700" }, "message": "csky: change a Kconfig symbol name to fix e1000 build error\n\ne1000\u0027s #define of CONFIG_RAM_BASE conflicts with a Kconfig symbol in\narch/csky/Kconfig.\n\nThe symbol in e1000 has been around longer, so change arch/csky/ to use\nDRAM_BASE instead of RAM_BASE to remove the conflict. (although e1000\nis also a 2-line change)\n\nLink: https://lkml.kernel.org/r/20210411055335.7111-1-rdunlap@infradead.org\nSigned-off-by: Randy Dunlap \u003crdunlap@infradead.org\u003e\nReported-by: kernel test robot \u003clkp@intel.com\u003e\nAcked-by: Guo Ren \u003cguoren@kernel.org\u003e\nCc: Jesse Brandeburg \u003cjesse.brandeburg@intel.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "02c587733c8161355a43e6e110c2e29bd0acff72", "tree": "d5b8a2132110e8691df9254d9b344bad866b0a8d", "parents": [ "5c595ac4c776c44b5c59de22ab43b3fe256d9fbb" ], "author": { "name": "Walter Wu", "email": "walter-zh.wu@mediatek.com", "time": "Fri Apr 16 15:46:00 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 16:10:36 2021 -0700" }, "message": "kasan: remove redundant config option\n\nCONFIG_KASAN_STACK and CONFIG_KASAN_STACK_ENABLE both enable KASAN stack\ninstrumentation, but we should only need one config, so that we remove\nCONFIG_KASAN_STACK_ENABLE and make CONFIG_KASAN_STACK workable. see [1].\n\nWhen enable KASAN stack instrumentation, then for gcc we could do no\nprompt and default value y, and for clang prompt and default value n.\n\nThis patch fixes the following compilation warning:\n\n include/linux/kasan.h:333:30: warning: \u0027CONFIG_KASAN_STACK\u0027 is not defined, evaluates to 0 [-Wundef]\n\n[akpm@linux-foundation.org: fix merge snafu]\n\nLink: https://bugzilla.kernel.org/show_bug.cgi?id\u003d210221 [1]\nLink: https://lkml.kernel.org/r/20210226012531.29231-1-walter-zh.wu@mediatek.com\nFixes: d9b571c885a8 (\"kasan: fix KASAN_STACK dependency for HW_TAGS\")\nSigned-off-by: Walter Wu \u003cwalter-zh.wu@mediatek.com\u003e\nSuggested-by: Dmitry Vyukov \u003cdvyukov@google.com\u003e\nReviewed-by: Nathan Chancellor \u003cnatechancellor@gmail.com\u003e\nAcked-by: Arnd Bergmann \u003carnd@arndb.de\u003e\nReviewed-by: Andrey Konovalov \u003candreyknvl@google.com\u003e\nCc: Andrey Ryabinin \u003cryabinin.a.a@gmail.com\u003e\nCc: Dmitry Vyukov \u003cdvyukov@google.com\u003e\nCc: Alexander Potapenko \u003cglider@google.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "5c595ac4c776c44b5c59de22ab43b3fe256d9fbb", "tree": "74b365db8f703503d42bb225cbe828693cbabcf0", "parents": [ "845be1cd34464620861b457b808e5fb2115d06b0" ], "author": { "name": "Arnd Bergmann", "email": "arnd@arndb.de", "time": "Fri Apr 16 15:45:57 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 16:10:36 2021 -0700" }, "message": "kasan: fix hwasan build for gcc\n\ngcc-11 adds support for -fsanitize\u003dkernel-hwaddress, so it becomes\npossible to enable CONFIG_KASAN_SW_TAGS.\n\nUnfortunately this fails to build at the moment, because the\ncorresponding command line arguments use llvm specific syntax.\n\nChange it to use the cc-param macro instead, which works on both clang\nand gcc.\n\n[elver@google.com: fixup for \"kasan: fix hwasan build for gcc\"]\n Link: https://lkml.kernel.org/r/YHQZVfVVLE/LDK2v@elver.google.com\n\nLink: https://lkml.kernel.org/r/20210323124112.1229772-1-arnd@kernel.org\nSigned-off-by: Arnd Bergmann \u003carnd@arndb.de\u003e\nSigned-off-by: Marco Elver \u003celver@google.com\u003e\nReviewed-by: Marco Elver \u003celver@google.com\u003e\nAcked-by: Andrey Konovalov \u003candreyknvl@gmail.com\u003e\nCc: Masahiro Yamada \u003cmasahiroy@kernel.org\u003e\nCc: Michal Marek \u003cmichal.lkml@markovi.net\u003e\nCc: Andrey Ryabinin \u003cryabinin.a.a@gmail.com\u003e\nCc: Nathan Chancellor \u003cnathan@kernel.org\u003e\nCc: Nick Desaulniers \u003cndesaulniers@google.com\u003e\nCc: Alexander Potapenko \u003cglider@google.com\u003e\nCc: Dmitry Vyukov \u003cdvyukov@google.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "845be1cd34464620861b457b808e5fb2115d06b0", "tree": "ff3989be9b642ccdad7eff01888ab9d37102263d", "parents": [ "06c2aac4014c38247256fe49c61b7f55890271e7" ], "author": { "name": "Randy Dunlap", "email": "rdunlap@infradead.org", "time": "Fri Apr 16 15:45:54 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 16:10:36 2021 -0700" }, "message": "mm: eliminate \"expecting prototype\" kernel-doc warnings\n\nFix stray kernel-doc warnings in mm/ due to mis-typed or missing function\nnames.\n\nQuietens these kernel-doc warnings:\n\n mm/mmu_gather.c:264: warning: expecting prototype for tlb_gather_mmu(). Prototype was for __tlb_gather_mmu() instead\n mm/oom_kill.c:180: warning: expecting prototype for Check whether unreclaimable slab amount is greater than(). Prototype was for should_dump_unreclaim_slab() instead\n mm/shuffle.c:155: warning: expecting prototype for shuffle_free_memory(). Prototype was for __shuffle_free_memory() instead\n\nLink: https://lkml.kernel.org/r/20210411210642.11362-1-rdunlap@infradead.org\nSigned-off-by: Randy Dunlap \u003crdunlap@infradead.org\u003e\nReviewed-by: Matthew Wilcox (Oracle) \u003cwilly@infradead.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "b02265429681c9c827c45978a61a9f00be5ea9aa", "tree": "172c9fac6ccc63b4c103cc23143ae87a8ea5e5f9", "parents": [ "6b389c16378a03fe71f3b1365b593ba41d2dd8ec", "d7a5091351756d0ae8e63134313c455624e36a13" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Apr 16 15:48:08 2021 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Apr 16 15:48:08 2021 -0700" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf\n\nDaniel Borkmann says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\npull-request: bpf 2021-04-17\n\nThe following pull-request contains BPF updates for your *net* tree.\n\nWe\u0027ve added 10 non-merge commits during the last 9 day(s) which contain\na total of 8 files changed, 175 insertions(+), 111 deletions(-).\n\nThe main changes are:\n\n1) Fix a potential NULL pointer dereference in libbpf\u0027s xsk\n umem handling, from Ciara Loftus.\n\n2) Mitigate a speculative oob read of up to map value size by\n tightening the masking window, from Daniel Borkmann.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "6b389c16378a03fe71f3b1365b593ba41d2dd8ec", "tree": "6cf58b60b23be4384f34268a32c20a1e6d6dfc56", "parents": [ "0e0704bb9ea00813dd3d5405516089fee286079b" ], "author": { "name": "Lijun Pan", "email": "lijunp213@gmail.com", "time": "Thu Apr 15 23:18:13 2021 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Apr 16 15:26:32 2021 -0700" }, "message": "MAINTAINERS: update my email\n\nUpdate my email and change myself to Reviewer.\n\nSigned-off-by: Lijun Pan \u003clijunp213@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "d7a5091351756d0ae8e63134313c455624e36a13", "tree": "4f9179a6cfdd07b6c82c179a57fe11cedcf50d74", "parents": [ "7fedb63a8307dda0ec3b8969a3b233a1dd7ea8e0" ], "author": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Wed Mar 24 14:52:31 2021 +0100" }, "committer": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Fri Apr 16 23:52:15 2021 +0200" }, "message": "bpf: Update selftests to reflect new error states\n\nUpdate various selftest error messages:\n\n * The \u0027Rx tried to sub from different maps, paths, or prohibited types\u0027\n is reworked into more specific/differentiated error messages for better\n guidance.\n\n * The change into \u0027value -4294967168 makes map_value pointer be out of\n bounds\u0027 is due to moving the mixed bounds check into the speculation\n handling and thus occuring slightly later than above mentioned sanity\n check.\n\n * The change into \u0027math between map_value pointer and register with\n unbounded min value\u0027 is similarly due to register sanity check coming\n before the mixed bounds check.\n\n * The case of \u0027map access: known scalar +\u003d value_ptr from different maps\u0027\n now loads fine given masks are the same from the different paths (despite\n max map value size being different).\n\nSigned-off-by: Daniel Borkmann \u003cdaniel@iogearbox.net\u003e\nReviewed-by: John Fastabend \u003cjohn.fastabend@gmail.com\u003e\nAcked-by: Alexei Starovoitov \u003cast@kernel.org\u003e\n" }, { "commit": "7fedb63a8307dda0ec3b8969a3b233a1dd7ea8e0", "tree": "4bcdfe16eb98bdef7b7b65356906b39efe462a54", "parents": [ "f528819334881fd622fdadeddb3f7edaed8b7c9b" ], "author": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Wed Mar 24 10:38:26 2021 +0100" }, "committer": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Fri Apr 16 23:52:01 2021 +0200" }, "message": "bpf: Tighten speculative pointer arithmetic mask\n\nThis work tightens the offset mask we use for unprivileged pointer arithmetic\nin order to mitigate a corner case reported by Piotr and Benedict where in\nthe speculative domain it is possible to advance, for example, the map value\npointer by up to value_size-1 out-of-bounds in order to leak kernel memory\nvia side-channel to user space.\n\nBefore this change, the computed ptr_limit for retrieve_ptr_limit() helper\nrepresents largest valid distance when moving pointer to the right or left\nwhich is then fed as aux-\u003ealu_limit to generate masking instructions against\nthe offset register. After the change, the derived aux-\u003ealu_limit represents\nthe largest potential value of the offset register which we mask against which\nis just a narrower subset of the former limit.\n\nFor minimal complexity, we call sanitize_ptr_alu() from 2 observation points\nin adjust_ptr_min_max_vals(), that is, before and after the simulated alu\noperation. In the first step, we retieve the alu_state and alu_limit before\nthe operation as well as we branch-off a verifier path and push it to the\nverification stack as we did before which checks the dst_reg under truncation,\nin other words, when the speculative domain would attempt to move the pointer\nout-of-bounds.\n\nIn the second step, we retrieve the new alu_limit and calculate the absolute\ndistance between both. Moreover, we commit the alu_state and final alu_limit\nvia update_alu_sanitation_state() to the env\u0027s instruction aux data, and bail\nout from there if there is a mismatch due to coming from different verification\npaths with different states.\n\nReported-by: Piotr Krysiuk \u003cpiotras@gmail.com\u003e\nReported-by: Benedict Schlueter \u003cbenedict.schlueter@rub.de\u003e\nSigned-off-by: Daniel Borkmann \u003cdaniel@iogearbox.net\u003e\nReviewed-by: John Fastabend \u003cjohn.fastabend@gmail.com\u003e\nAcked-by: Alexei Starovoitov \u003cast@kernel.org\u003e\nTested-by: Benedict Schlueter \u003cbenedict.schlueter@rub.de\u003e\n" }, { "commit": "f528819334881fd622fdadeddb3f7edaed8b7c9b", "tree": "151b9ecf65ed126be5b63b92bfc8852b2addc41f", "parents": [ "073815b756c51ba9d8384d924c5d1c03ca3d1ae4" ], "author": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Wed Mar 24 11:25:39 2021 +0100" }, "committer": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Fri Apr 16 23:51:57 2021 +0200" }, "message": "bpf: Move sanitize_val_alu out of op switch\n\nAdd a small sanitize_needed() helper function and move sanitize_val_alu()\nout of the main opcode switch. In upcoming work, we\u0027ll move sanitize_ptr_alu()\nas well out of its opcode switch so this helps to streamline both.\n\nSigned-off-by: Daniel Borkmann \u003cdaniel@iogearbox.net\u003e\nReviewed-by: John Fastabend \u003cjohn.fastabend@gmail.com\u003e\nAcked-by: Alexei Starovoitov \u003cast@kernel.org\u003e\n" }, { "commit": "073815b756c51ba9d8384d924c5d1c03ca3d1ae4", "tree": "618347c5c7ee8cb2850bd11502863fa005a5a1f6", "parents": [ "a6aaece00a57fa6f22575364b3903dfbccf5345d" ], "author": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Tue Mar 23 15:05:48 2021 +0100" }, "committer": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Fri Apr 16 23:51:54 2021 +0200" }, "message": "bpf: Refactor and streamline bounds check into helper\n\nMove the bounds check in adjust_ptr_min_max_vals() into a small helper named\nsanitize_check_bounds() in order to simplify the former a bit.\n\nSigned-off-by: Daniel Borkmann \u003cdaniel@iogearbox.net\u003e\nReviewed-by: John Fastabend \u003cjohn.fastabend@gmail.com\u003e\nAcked-by: Alexei Starovoitov \u003cast@kernel.org\u003e\n" }, { "commit": "a6aaece00a57fa6f22575364b3903dfbccf5345d", "tree": "d68b11e79877a5468a87cae5fba6ad20db20e914", "parents": [ "b658bbb844e28f1862867f37e8ca11a8e2aa94a3" ], "author": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Tue Mar 23 09:30:01 2021 +0100" }, "committer": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Fri Apr 16 23:51:48 2021 +0200" }, "message": "bpf: Improve verifier error messages for users\n\nConsolidate all error handling and provide more user-friendly error messages\nfrom sanitize_ptr_alu() and sanitize_val_alu().\n\nSigned-off-by: Daniel Borkmann \u003cdaniel@iogearbox.net\u003e\nReviewed-by: John Fastabend \u003cjohn.fastabend@gmail.com\u003e\nAcked-by: Alexei Starovoitov \u003cast@kernel.org\u003e\n" }, { "commit": "b658bbb844e28f1862867f37e8ca11a8e2aa94a3", "tree": "b342f07b0438c09019ccd2f5fac52808a9a7d81e", "parents": [ "24c109bb1537c12c02aeed2d51a347b4d6a9b76e" ], "author": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Tue Mar 23 09:04:10 2021 +0100" }, "committer": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Fri Apr 16 23:51:43 2021 +0200" }, "message": "bpf: Rework ptr_limit into alu_limit and add common error path\n\nSmall refactor with no semantic changes in order to consolidate the max\nptr_limit boundary check.\n\nSigned-off-by: Daniel Borkmann \u003cdaniel@iogearbox.net\u003e\nReviewed-by: John Fastabend \u003cjohn.fastabend@gmail.com\u003e\nAcked-by: Alexei Starovoitov \u003cast@kernel.org\u003e\n" }, { "commit": "24c109bb1537c12c02aeed2d51a347b4d6a9b76e", "tree": "17362854c4e4a2205a112ffbe9e7c4452a8174ed", "parents": [ "6f55b2f2a1178856c19bbce2f71449926e731914" ], "author": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Tue Mar 23 08:51:02 2021 +0100" }, "committer": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Fri Apr 16 23:51:39 2021 +0200" }, "message": "bpf: Ensure off_reg has no mixed signed bounds for all types\n\nThe mixed signed bounds check really belongs into retrieve_ptr_limit()\ninstead of outside of it in adjust_ptr_min_max_vals(). The reason is\nthat this check is not tied to PTR_TO_MAP_VALUE only, but to all pointer\ntypes that we handle in retrieve_ptr_limit() and given errors from the latter\npropagate back to adjust_ptr_min_max_vals() and lead to rejection of the\nprogram, it\u0027s a better place to reside to avoid anything slipping through\nfor future types. The reason why we must reject such off_reg is that we\notherwise would not be able to derive a mask, see details in 9d7eceede769\n(\"bpf: restrict unknown scalars of mixed signed bounds for unprivileged\").\n\nSigned-off-by: Daniel Borkmann \u003cdaniel@iogearbox.net\u003e\nReviewed-by: John Fastabend \u003cjohn.fastabend@gmail.com\u003e\nAcked-by: Alexei Starovoitov \u003cast@kernel.org\u003e\n" }, { "commit": "6f55b2f2a1178856c19bbce2f71449926e731914", "tree": "9e259148696b9f2331746ad16e7fec8f8c211ef6", "parents": [ "9601148392520e2e134936e76788fc2a6371e7be" ], "author": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Mon Mar 22 15:45:52 2021 +0100" }, "committer": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Fri Apr 16 23:51:36 2021 +0200" }, "message": "bpf: Move off_reg into sanitize_ptr_alu\n\nSmall refactor to drag off_reg into sanitize_ptr_alu(), so we later on can\nuse off_reg for generalizing some of the checks for all pointer types.\n\nSigned-off-by: Daniel Borkmann \u003cdaniel@iogearbox.net\u003e\nReviewed-by: John Fastabend \u003cjohn.fastabend@gmail.com\u003e\nAcked-by: Alexei Starovoitov \u003cast@kernel.org\u003e\n" }, { "commit": "9601148392520e2e134936e76788fc2a6371e7be", "tree": "4b12106c68f6229ce8292b94ad12e790e4ec756d", "parents": [ "afd0be7299533bb2e2b09104399d8a467ecbd2c5" ], "author": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Tue Mar 23 08:32:59 2021 +0100" }, "committer": { "name": "Daniel Borkmann", "email": "daniel@iogearbox.net", "time": "Fri Apr 16 23:51:21 2021 +0200" }, "message": "bpf: Use correct permission flag for mixed signed bounds arithmetic\n\nWe forbid adding unknown scalars with mixed signed bounds due to the\nspectre v1 masking mitigation. Hence this also needs bypass_spec_v1\nflag instead of allow_ptr_leaks.\n\nFixes: 2c78ee898d8f (\"bpf: Implement CAP_BPF\")\nSigned-off-by: Daniel Borkmann \u003cdaniel@iogearbox.net\u003e\nReviewed-by: John Fastabend \u003cjohn.fastabend@gmail.com\u003e\nAcked-by: Alexei Starovoitov \u003cast@kernel.org\u003e\n" }, { "commit": "151501160401e2dc669ea7dac2c599b53f220c33", "tree": "c8fbe121cc8b650f260c353320e61eb2e9ed320f", "parents": [ "06c2aac4014c38247256fe49c61b7f55890271e7", "7ae11635ec90072083503c6b6485cdffe46203b3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 10:05:42 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 10:05:42 2021 -0700" }, "message": "Merge tag \u0027riscv-for-linus-5.12-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux\n\nPull RISC-V fixes from Palmer Dabbelt:\n \"A handful of fixes:\n\n - a fix to properly select SPARSEMEM_STATIC on rv32\n\n - a few fixes to kprobes\n\n I don\u0027t generally like sending stuff this late, but these all seem\n pretty safe\"\n\n* tag \u0027riscv-for-linus-5.12-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux:\n riscv: keep interrupts disabled for BREAKPOINT exception\n riscv: kprobes/ftrace: Add recursion protection to the ftrace callback\n riscv: add do_page_fault and do_trap_break into the kprobes blacklist\n riscv: Fix spelling mistake \"SPARSEMEM\" to \"SPARSMEM\"\n" }, { "commit": "06c2aac4014c38247256fe49c61b7f55890271e7", "tree": "62c3eb8b69fb5086b530a9e04eccecd214da15f4", "parents": [ "2f7b98d1e55ccd34e4998bf5f321ec7b9d6b451b", "22315a2296f4c251fa92aec45fbbae37e9301b6c" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 09:45:30 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 09:45:30 2021 -0700" }, "message": "Merge tag \u0027arm64-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux\n\nPull arm64 fix from Catalin Marinas:\n \"Fix kernel compilation when using the LLVM integrated assembly.\n\n A recent commit (2decad92f473, \"arm64: mte: Ensure TIF_MTE_ASYNC_FAULT\n is set atomically\") broke the kernel build when using the LLVM\n integrated assembly (only noticeable with clang-12 as MTE is not\n supported by earlier versions and the code in question not compiled).\n The Fixes: tag in the commit refers to the original patch introducing\n subsections for the alternative code sequences\"\n\n* tag \u0027arm64-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:\n arm64: alternatives: Move length validation in alternative_{insn, endif}\n" }, { "commit": "2f7b98d1e55ccd34e4998bf5f321ec7b9d6b451b", "tree": "4eb72e7f08d9dfdcbd1aba64534deebd12150d23", "parents": [ "7e25f40eab52c57ff6772d27d2aef3640a3237d7", "4d2e1288372ccc5ac60290bc10cace49c9bfa6d0" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 07:49:04 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 16 07:49:04 2021 -0700" }, "message": "Merge tag \u0027drm-fixes-2021-04-16\u0027 of git://anongit.freedesktop.org/drm/drm\n\nPull drm fixes from Daniel Vetter:\n \"I pinged the usual suspects, only intel fixes pending\"\n\n* tag \u0027drm-fixes-2021-04-16\u0027 of git://anongit.freedesktop.org/drm/drm:\n drm/i915/display/vlv_dsi: Do not skip panel_pwr_cycle_delay when disabling the panel\n drm/i915: Don\u0027t zero out the Y plane\u0027s watermarks\n drm/i915/dpcd_bl: Don\u0027t try vesa interface unless specified by VBT\n" }, { "commit": "7ae11635ec90072083503c6b6485cdffe46203b3", "tree": "f7914021e422b50549e819decdb5992901ebe276", "parents": [ "e31be8d343e64e7ab17aef55c1d1b36dc504da67" ], "author": { "name": "Jisheng Zhang", "email": "jszhang@kernel.org", "time": "Tue Mar 30 02:16:24 2021 +0800" }, "committer": { "name": "Palmer Dabbelt", "email": "palmerdabbelt@google.com", "time": "Thu Apr 15 21:32:40 2021 -0700" }, "message": "riscv: keep interrupts disabled for BREAKPOINT exception\n\nCurrent riscv\u0027s kprobe handlers are run with both preemption and\ninterrupt enabled, this violates kprobe requirements. Fix this issue\nby keeping interrupts disabled for BREAKPOINT exception.\n\nFixes: c22b0bcb1dd0 (\"riscv: Add kprobes supported\")\nCc: stable@vger.kernel.org\nSigned-off-by: Jisheng Zhang \u003cjszhang@kernel.org\u003e\nReviewed-by: Masami Hiramatsu \u003cmhiramat@kernel.org\u003e\n[Palmer: add a comment]\nSigned-off-by: Palmer Dabbelt \u003cpalmerdabbelt@google.com\u003e\n" }, { "commit": "e31be8d343e64e7ab17aef55c1d1b36dc504da67", "tree": "ddaad7d4e24b5d2ff2f63069e0be6482a5d0af14", "parents": [ "2349a3b26e29b8d860466bafda2e02b4b87a9e40" ], "author": { "name": "Jisheng Zhang", "email": "jszhang@kernel.org", "time": "Tue Mar 30 02:14:40 2021 +0800" }, "committer": { "name": "Palmer Dabbelt", "email": "palmerdabbelt@google.com", "time": "Thu Apr 15 21:32:34 2021 -0700" }, "message": "riscv: kprobes/ftrace: Add recursion protection to the ftrace callback\n\nCurrently, the riscv\u0027s kprobes(powerred by ftrace) handler is\npreemptible. Futher check indicates we miss something similar as the\ncommit c536aa1c5b17 (\"kprobes/ftrace: Add recursion protection to the\nftrace callback\"), so do similar modifications as the commit does.\n\nFixes: 829adda597fe (\"riscv: Add KPROBES_ON_FTRACE supported\")\nCc: stable@vger.kernel.org\nSigned-off-by: Jisheng Zhang \u003cjszhang@kernel.org\u003e\nReviewed-by: Steven Rostedt (VMware) \u003crostedt@goodmis.org\u003e\nSigned-off-by: Palmer Dabbelt \u003cpalmerdabbelt@google.com\u003e\n" }, { "commit": "2349a3b26e29b8d860466bafda2e02b4b87a9e40", "tree": "e6d1e076576baf645d520c261f5d96e22b70e757", "parents": [ "199fc6b8dee7d6d50467a57e0dc7e3e1b7d59966" ], "author": { "name": "Jisheng Zhang", "email": "jszhang@kernel.org", "time": "Tue Mar 30 02:12:26 2021 +0800" }, "committer": { "name": "Palmer Dabbelt", "email": "palmerdabbelt@google.com", "time": "Thu Apr 15 21:32:28 2021 -0700" }, "message": "riscv: add do_page_fault and do_trap_break into the kprobes blacklist\n\nThese two functions are used to implement the kprobes feature so they\ncan\u0027t be kprobed.\n\nFixes: c22b0bcb1dd0 (\"riscv: Add kprobes supported\")\nCc: stable@vger.kernel.org\nSigned-off-by: Jisheng Zhang \u003cjszhang@kernel.org\u003e\nReviewed-by: Masami Hiramatsu \u003cmhiramat@kernel.org\u003e\nSigned-off-by: Palmer Dabbelt \u003cpalmerdabbelt@google.com\u003e\n" }, { "commit": "199fc6b8dee7d6d50467a57e0dc7e3e1b7d59966", "tree": "5b07f9f83694e01d3b718c2cc774deb5417cd1ff", "parents": [ "e49d033bddf5b565044e2abe4241353959bc9120" ], "author": { "name": "Kefeng Wang", "email": "wangkefeng.wang@huawei.com", "time": "Mon Mar 29 11:13:07 2021 +0800" }, "committer": { "name": "Palmer Dabbelt", "email": "palmerdabbelt@google.com", "time": "Thu Apr 15 21:32:10 2021 -0700" }, "message": "riscv: Fix spelling mistake \"SPARSEMEM\" to \"SPARSMEM\"\n\nThere is a spelling mistake when SPARSEMEM Kconfig copy.\n\nFixes: a5406a7ff56e (\"riscv: Correct SPARSEMEM configuration\")\nCc: stable@vger.kernel.org\nSigned-off-by: Kefeng Wang \u003cwangkefeng.wang@huawei.com\u003e\nSigned-off-by: Palmer Dabbelt \u003cpalmerdabbelt@google.com\u003e\n" }, { "commit": "b21bb4cd1102dd9e24a169d09cf4e6f3c8a46bcf", "tree": "05602b8c26ae0fd91964dbb1ddc55f39cbe9f89f", "parents": [ "392be0bda730df3c71241b2a16bbecac78ee627d" ], "author": { "name": "Ben Widawsky", "email": "ben.widawsky@intel.com", "time": "Thu Apr 15 16:26:08 2021 -0700" }, "committer": { "name": "Dan Williams", "email": "dan.j.williams@intel.com", "time": "Thu Apr 15 19:27:54 2021 -0700" }, "message": "cxl/mem: Fix register block offset calculation\n\nThe \"Register Offset Low\" register of a \"DVSEC Register Locator\"\ncontains the 64K aligned offset for the registers along with the BAR\nindicator and an id. The implementation was treating the \"Register Block\nOffset Low\" field a value rather than as a pre-aligned component of the\n64-bit offset. So, just mask, don\u0027t mask and shift (FIELD_GET).\n\nThe user visible result of this bug is that the driver fails to bind to\nthe device after none of the required blocks are found.\n\nThis was missed earlier because the primary development done in the QEMU\nenvironment only uses 0 offsets, i.e. 0 shifted is still 0.\n\nFixes: 8adaf747c9f0 (\"cxl/mem: Find device capabilities\")\nReported-by: Vishal Verma \u003cvishal.l.verma@intel.com\u003e\nSigned-off-by: Ben Widawsky \u003cben.widawsky@intel.com\u003e\nLink: https://lore.kernel.org/r/20210415232610.603273-1-ben.widawsky@intel.com\nSigned-off-by: Dan Williams \u003cdan.j.williams@intel.com\u003e\n" }, { "commit": "0e0704bb9ea00813dd3d5405516089fee286079b", "tree": "b8e2407301f122c03f50d9e6fd584e24c58f3c0c", "parents": [ "61d773586eba5c102bcb2cccdd88239227ce8b26", "e8a4155567b3c903f49cbf89b8017e9cc22c4fe4" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Apr 15 16:55:49 2021 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Apr 15 16:55:49 2021 -0700" }, "message": "Merge branch \u0027ch_tlss-fixes\u0027\n\nVinay Kumar Yadav says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nchelsio/ch_ktls: chelsio inline tls driver bug fixes\n\nThis series of patches fix following bugs in Chelsio inline tls driver.\nPatch1: kernel panic.\nPatch2: connection close issue.\nPatch3: tcb close call issue.\nPatch4: unnecessary snd_una update.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "e8a4155567b3c903f49cbf89b8017e9cc22c4fe4", "tree": "b8e2407301f122c03f50d9e6fd584e24c58f3c0c", "parents": [ "21d8c25e3f4b9052a471ced8f47b531956eb9963" ], "author": { "name": "Vinay Kumar Yadav", "email": "vinay.yadav@chelsio.com", "time": "Thu Apr 15 13:17:48 2021 +0530" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Apr 15 16:55:49 2021 -0700" }, "message": "ch_ktls: do not send snd_una update to TCB in middle\n\nsnd_una update should not be done when the same skb is being\nsent out.chcr_short_record_handler() sends it again even\nthough SND_UNA update is already sent for the skb in\nchcr_ktls_xmit(), which causes mismatch in un-acked\nTCP seq number, later causes problem in sending out\ncomplete record.\n\nFixes: 429765a149f1 (\"chcr: handle partial end part of a record\")\nSigned-off-by: Vinay Kumar Yadav \u003cvinay.yadav@chelsio.com\u003e\nSigned-off-by: Rohit Maheshwari \u003crohitm@chelsio.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "21d8c25e3f4b9052a471ced8f47b531956eb9963", "tree": "8ee86f2b12d8f192e33c1dd9c6f0ffff181516cc", "parents": [ "bc16efd2430652f894ae34b1de5eccc3bf0d2810" ], "author": { "name": "Vinay Kumar Yadav", "email": "vinay.yadav@chelsio.com", "time": "Thu Apr 15 13:17:47 2021 +0530" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Apr 15 16:55:49 2021 -0700" }, "message": "ch_ktls: tcb close causes tls connection failure\n\nHW doesn\u0027t need marking TCB closed. This TCB state change\nsometimes causes problem to the new connection which gets\nthe same tid.\n\nFixes: 34aba2c45024 (\"cxgb4/chcr : Register to tls add and del callback\")\nSigned-off-by: Vinay Kumar Yadav \u003cvinay.yadav@chelsio.com\u003e\nSigned-off-by: Rohit Maheshwari \u003crohitm@chelsio.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "bc16efd2430652f894ae34b1de5eccc3bf0d2810", "tree": "8934371d618bc64d5f065a2726e5471cd17c41fd", "parents": [ "1a73e427b824133940c2dd95ebe26b6dce1cbf10" ], "author": { "name": "Vinay Kumar Yadav", "email": "vinay.yadav@chelsio.com", "time": "Thu Apr 15 13:17:46 2021 +0530" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Apr 15 16:55:49 2021 -0700" }, "message": "ch_ktls: fix device connection close\n\nWhen sge queue is full and chcr_ktls_xmit_wr_complete()\nreturns failure, skb is not freed if it is not the last tls record in\nthis skb, causes refcount never gets freed and tls_dev_del()\nnever gets called on this connection.\n\nFixes: 5a4b9fe7fece (\"cxgb4/chcr: complete record tx handling\")\nSigned-off-by: Vinay Kumar Yadav \u003cvinay.yadav@chelsio.com\u003e\nSigned-off-by: Rohit Maheshwari \u003crohitm@chelsio.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "1a73e427b824133940c2dd95ebe26b6dce1cbf10", "tree": "5fac6ce9c7ac4bf88a55f8bd7442f10edb9462a7", "parents": [ "61d773586eba5c102bcb2cccdd88239227ce8b26" ], "author": { "name": "Vinay Kumar Yadav", "email": "vinay.yadav@chelsio.com", "time": "Thu Apr 15 13:17:45 2021 +0530" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Apr 15 16:55:49 2021 -0700" }, "message": "ch_ktls: Fix kernel panic\n\nTaking page refcount is not ideal and causes kernel panic\nsometimes. It\u0027s better to take tx_ctx lock for the complete\nskb transmit, to avoid page cleanup if ACK received in middle.\n\nFixes: 5a4b9fe7fece (\"cxgb4/chcr: complete record tx handling\")\nSigned-off-by: Vinay Kumar Yadav \u003cvinay.yadav@chelsio.com\u003e\nSigned-off-by: Rohit Maheshwari \u003crohitm@chelsio.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "61d773586eba5c102bcb2cccdd88239227ce8b26", "tree": "113fcd128e130286d7e866d878746933dd1028e3", "parents": [ "4e39a072a6a0fc422ba7da5e4336bdc295d70211", "e3e0f9b279705154b951d579dc3d8b7041710e24" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Apr 15 16:43:29 2021 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Apr 15 16:43:29 2021 -0700" }, "message": "Merge tag \u0027mlx5-fixes-2021-04-14\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 2021-04-14\n\nThis series provides 3 small fixes to mlx5 driver.\nPlease pull and let me know if there is any problem.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4e39a072a6a0fc422ba7da5e4336bdc295d70211", "tree": "074c5695a825b1d7b6dd9a7a99dbbdbfe5a4f6c4", "parents": [ "9e196e4398869e4dfe3f49f7f3e0935e07b1d1b9" ], "author": { "name": "Jason Xing", "email": "xingwanli@kuaishou.com", "time": "Wed Apr 14 10:34:28 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Apr 15 14:37:35 2021 -0700" }, "message": "i40e: fix the panic when running bpf in xdpdrv mode\n\nFix this panic by adding more rules to calculate the value of @rss_size_max\nwhich could be used in allocating the queues when bpf is loaded, which,\nhowever, could cause the failure and then trigger the NULL pointer of\nvsi-\u003erx_rings. Prio to this fix, the machine doesn\u0027t care about how many\ncpus are online and then allocates 256 queues on the machine with 32 cpus\nonline actually.\n\nOnce the load of bpf begins, the log will go like this \"failed to get\ntracking for 256 queues for VSI 0 err -12\" and this \"setup of MAIN VSI\nfailed\".\n\nThus, I attach the key information of the crash-log here.\n\nBUG: unable to handle kernel NULL pointer dereference at\n0000000000000000\nRIP: 0010:i40e_xdp+0xdd/0x1b0 [i40e]\nCall Trace:\n[2160294.717292] ? i40e_reconfig_rss_queues+0x170/0x170 [i40e]\n[2160294.717666] dev_xdp_install+0x4f/0x70\n[2160294.718036] dev_change_xdp_fd+0x11f/0x230\n[2160294.718380] ? dev_disable_lro+0xe0/0xe0\n[2160294.718705] do_setlink+0xac7/0xe70\n[2160294.719035] ? __nla_parse+0xed/0x120\n[2160294.719365] rtnl_newlink+0x73b/0x860\n\nFixes: 41c445ff0f48 (\"i40e: main driver core\")\nCo-developed-by: Shujin Li \u003clishujin@kuaishou.com\u003e\nSigned-off-by: Shujin Li \u003clishujin@kuaishou.com\u003e\nSigned-off-by: Jason Xing \u003cxingwanli@kuaishou.com\u003e\nReviewed-by: Jesse Brandeburg \u003cjesse.brandeburg@intel.com\u003e\nAcked-by: Jesper Dangaard Brouer \u003cbrouer@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "39930213e7779b9c4257499972b8afb8858f1a2d", "tree": "39feba35d1d635c23526d99c5bc634110d77ee54", "parents": [ "d434405aaab7d0ebc516b68a8fc4100922d7f5ef" ], "author": { "name": "Marek Behún", "email": "kabel@kernel.org", "time": "Thu Apr 08 04:00:00 2021 +0200" }, "committer": { "name": "Wolfram Sang", "email": "wsa@kernel.org", "time": "Thu Apr 15 22:13:19 2021 +0200" }, "message": "i2c: mv64xxx: Fix random system lock caused by runtime PM\n\nI noticed a weird bug with this driver on Marvell CN9130 Customer\nReference Board.\n\nSometime after boot, the system locks with the following message:\n [104.071363] i2c i2c-0: mv64xxx: I2C bus locked, block: 1, time_left: 0\n\nThe system does not respond afterwards, only warns about RCU stalls.\n\nThis first appeared with commit e5c02cf54154 (\"i2c: mv64xxx: Add runtime\nPM support\").\n\nWith further experimentation I discovered that adding a delay into\nmv64xxx_i2c_hw_init() fixes this issue. This function is called before\nevery xfer, due to how runtime PM works in this driver. It seems that in\norder to work correctly, a delay is needed after the bus is reset in\nthis function.\n\nSince there already is a known erratum with this controller needing a\ndelay, I assume that this is just another place this needs to be\napplied. Therefore I apply the delay only if errata_delay is true.\n\nSigned-off-by: Marek Behún \u003ckabel@kernel.org\u003e\nAcked-by: Gregory CLEMENT \u003cgregory.clement@bootlin.com\u003e\nReviewed-by: Samuel Holland \u003csamuel@sholland.org\u003e\nSigned-off-by: Wolfram Sang \u003cwsa@kernel.org\u003e\n" }, { "commit": "7e25f40eab52c57ff6772d27d2aef3640a3237d7", "tree": "d666f40cc2789bf1f9fb9922b019d834500b9bf2", "parents": [ "33f0d9d94a0ef0814d23320c2536c4135d230114", "6998a8800d73116187aad542391ce3b2dd0f9e30" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Apr 15 10:53:39 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Apr 15 10:53:39 2021 -0700" }, "message": "Merge tag \u0027acpi-5.12-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm\n\nPull ACPI fix from Rafael Wysocki:\n \"Restore the initrd-based ACPI table override functionality broken by\n one of the recent fixes\"\n\n* tag \u0027acpi-5.12-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:\n ACPI: x86: Call acpi_boot_table_init() after acpi_table_upgrade()\n" }, { "commit": "33f0d9d94a0ef0814d23320c2536c4135d230114", "tree": "3d5c7298356b3993f4da297e25bfb9d23665e1c8", "parents": [ "e7e3a53b30d6e6f54eef81400ddfe8b32224b77f", "23cf00ddd2e1aacf1873e43f5e0c519c120daf7a" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Apr 15 10:48:51 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Apr 15 10:48:51 2021 -0700" }, "message": "Merge tag \u0027gpio-fixes-for-v5.12-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux\n\nPull gpio fix from Bartosz Golaszewski:\n \"A single fix for an older problem with the sysfs interface: do not\n allow exporting GPIO lines which were marked invalid by the driver\"\n\n* tag \u0027gpio-fixes-for-v5.12-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux:\n gpio: sysfs: Obey valid_mask\n" }, { "commit": "e7e3a53b30d6e6f54eef81400ddfe8b32224b77f", "tree": "8436a1e017b7957abd79ae8fade9c70392c65ca8", "parents": [ "1df01322f00a0aedd4a589597ce9c0b680ae6068", "276559d8d02c2709281578976ca2f53bc62063d4" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Apr 15 10:43:18 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Apr 15 10:43:18 2021 -0700" }, "message": "Merge branch \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/hid/hid\n\nPull HID fixes from Jiri Kosina:\n \"The changes are all device/driver specific fixes:\n\n - EV_KEY and EV_ABS regression fix for Wacom from Ping Cheng\n\n - BIOS-specific quirk to fix some of the AMD_SFH-based systems, from\n Hans de Goede\n\n - other small error handling fixes and device ID additions\"\n\n* \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/hid/hid:\n HID: wacom: set EV_KEY and EV_ABS only for non-HID_GENERIC type of devices\n AMD_SFH: Add DMI quirk table for BIOS-es which don\u0027t set the activestatus bits\n AMD_SFH: Add sensor_mask module parameter\n AMD_SFH: Removed unused activecontrolstatus member from the amd_mp2_dev struct\n HID: wacom: Assign boolean values to a bool variable\n HID cp2112: fix support for multiple gpiochips\n HID: alps: fix error return code in alps_input_configured()\n HID: asus: Add support for 2021 ASUS N-Key keyboard\n HID: google: add don USB id\n" }, { "commit": "22315a2296f4c251fa92aec45fbbae37e9301b6c", "tree": "d6be89f7456097ad3afed89d35a419ddefef27fe", "parents": [ "738fa58ee1328481d1d7889e7c430b3401c571b9" ], "author": { "name": "Nathan Chancellor", "email": "nathan@kernel.org", "time": "Tue Apr 13 17:08:04 2021 -0700" }, "committer": { "name": "Catalin Marinas", "email": "catalin.marinas@arm.com", "time": "Thu Apr 15 18:33:25 2021 +0100" }, "message": "arm64: alternatives: Move length validation in alternative_{insn, endif}\n\nAfter commit 2decad92f473 (\"arm64: mte: Ensure TIF_MTE_ASYNC_FAULT is\nset atomically\"), LLVM\u0027s integrated assembler fails to build entry.S:\n\n\u003cinstantiation\u003e:5:7: error: expected assembly-time absolute expression\n .org . - (664b-663b) + (662b-661b)\n ^\n\u003cinstantiation\u003e:6:7: error: expected assembly-time absolute expression\n .org . - (662b-661b) + (664b-663b)\n ^\n\nThe root cause is LLVM\u0027s assembler has a one-pass design, meaning it\ncannot figure out these instruction lengths when the .org directive is\noutside of the subsection that they are in, which was changed by the\n.arch_extension directive added in the above commit.\n\nApply the same fix from commit 966a0acce2fc (\"arm64/alternatives: move\nlength validation inside the subsection\") to the alternative_endif\nmacro, shuffling the .org directives so that the length validation\nhappen will always happen in the same subsections. alternative_insn has\nnot shown any issue yet but it appears that it could have the same issue\nin the future so just preemptively change it.\n\nFixes: f7b93d42945c (\"arm64/alternatives: use subsections for replacement sequences\")\nCc: \u003cstable@vger.kernel.org\u003e # 5.8.x\nLink: https://github.com/ClangBuiltLinux/linux/issues/1347\nSigned-off-by: Nathan Chancellor \u003cnathan@kernel.org\u003e\nReviewed-by: Sami Tolvanen \u003csamitolvanen@google.com\u003e\nTested-by: Sami Tolvanen \u003csamitolvanen@google.com\u003e\nReviewed-by: Nick Desaulniers \u003cndesaulniers@google.com\u003e\nTested-by: Nick Desaulniers \u003cndesaulniers@google.com\u003e\nLink: https://lore.kernel.org/r/20210414000803.662534-1-nathan@kernel.org\nSigned-off-by: Catalin Marinas \u003ccatalin.marinas@arm.com\u003e\n" }, { "commit": "1df01322f00a0aedd4a589597ce9c0b680ae6068", "tree": "17d3e06070b9b42531c712b8dc4bc3077a6911b6", "parents": [ "7f75285ca572eaabc028cf78c6ab5473d0d160be", "56cfe6f820a6315291eb5a1b82bb49633b993d3b" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Apr 15 10:23:44 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Apr 15 10:23:44 2021 -0700" }, "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 \"Just a few driver fixes here\"\n\n* \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:\n Input: elants_i2c - drop zero-checking of ABS_MT_TOUCH_MAJOR resolution\n Input: elants_i2c - fix division by zero if firmware reports zero phys size\n Input: nspire-keypad - enable interrupts only when opened\n Input: i8042 - fix Pegatron C15B ID entry\n Input: n64joy - fix return value check in n64joy_probe()\n Input: s6sy761 - fix coordinate read bit shift\n" }, { "commit": "4d2e1288372ccc5ac60290bc10cace49c9bfa6d0", "tree": "81be747a8a55513e81fcaf015a7f4099122dba20", "parents": [ "d434405aaab7d0ebc516b68a8fc4100922d7f5ef", "aee6f25e9c911323aa89a200e1bb160c1613ed3d" ], "author": { "name": "Daniel Vetter", "email": "daniel.vetter@ffwll.ch", "time": "Thu Apr 15 15:24:17 2021 +0200" }, "committer": { "name": "Daniel Vetter", "email": "daniel.vetter@ffwll.ch", "time": "Thu Apr 15 15:24:17 2021 +0200" }, "message": "Merge tag \u0027drm-intel-fixes-2021-04-15\u0027 of git://anongit.freedesktop.org/drm/drm-intel into drm-fixes\n\nDisplay panel \u0026 power related fixes:\n\n- Backlight fix (Lyude)\n- Display watermark fix (Ville)\n- VLV panel power fix (Hans)\n\nSigned-off-by: Daniel Vetter \u003cdaniel.vetter@ffwll.ch\u003e\n\nFrom: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/YHg4nz/ndzDRmPjd@intel.com\n" }, { "commit": "e3e0f9b279705154b951d579dc3d8b7041710e24", "tree": "5effa7f5b973b4ffa4568690c6dc94dd623898d4", "parents": [ "7a320c9db3e73fb6c4f9a331087df9df18767221" ], "author": { "name": "wenxu", "email": "wenxu@ucloud.cn", "time": "Fri Apr 09 13:33:48 2021 +0800" }, "committer": { "name": "Saeed Mahameed", "email": "saeedm@nvidia.com", "time": "Wed Apr 14 16:13:00 2021 -0700" }, "message": "net/mlx5e: fix ingress_ifindex check in mlx5e_flower_parse_meta\n\nIn the nft_offload there is the mate flow_dissector with no\ningress_ifindex but with ingress_iftype that only be used\nin the software. So if the mask of ingress_ifindex in meta is\n0, this meta check should be bypass.\n\nFixes: 6d65bc64e232 (\"net/mlx5e: Add mlx5e_flower_parse_meta support\")\nSigned-off-by: wenxu \u003cwenxu@ucloud.cn\u003e\nAcked-by: Pablo Neira Ayuso \u003cpablo@netfilter.org\u003e\nSigned-off-by: Saeed Mahameed \u003csaeedm@nvidia.com\u003e\n" }, { "commit": "7a320c9db3e73fb6c4f9a331087df9df18767221", "tree": "e6d97004dd8e7f304b403c57b760b457fb976799", "parents": [ "41bafb31dcd58d834bdffa5db703f94fd2cec727" ], "author": { "name": "Aya Levin", "email": "ayal@nvidia.com", "time": "Sun Apr 11 09:33:12 2021 +0300" }, "committer": { "name": "Saeed Mahameed", "email": "saeedm@nvidia.com", "time": "Wed Apr 14 16:12:57 2021 -0700" }, "message": "net/mlx5e: Fix setting of RS FEC mode\n\nChange register setting from bit number to bit mask.\n\nFixes: b5ede32d3329 (\"net/mlx5e: Add support for FEC modes based on 50G per lane links\")\nSigned-off-by: Aya Levin \u003cayal@nvidia.com\u003e\nReviewed-by: Eran Ben Elisha \u003ceranbe@nvidia.com\u003e\nSigned-off-by: Saeed Mahameed \u003csaeedm@nvidia.com\u003e\n" }, { "commit": "41bafb31dcd58d834bdffa5db703f94fd2cec727", "tree": "3b27494be9ecef60e6941e0aa2ea7b459a00e402", "parents": [ "2afeec08ab5c86ae21952151f726bfe184f6b23d" ], "author": { "name": "Aya Levin", "email": "ayal@nvidia.com", "time": "Mon Apr 12 17:50:08 2021 +0300" }, "committer": { "name": "Saeed Mahameed", "email": "saeedm@nvidia.com", "time": "Wed Apr 14 16:12:54 2021 -0700" }, "message": "net/mlx5: Fix setting of devlink traps in switchdev mode\n\nPrevent setting of devlink traps on the uplink while in switchdev mode.\nIn this mode, it is the SW switch responsibility to handle both packets\nwith a mismatch in destination MAC or VLAN ID. Therefore, there are no\nflow steering tables to trap undesirable packets and driver crashes upon\nsetting a trap.\n\nFixes: 241dc159391f (\"net/mlx5: Notify on trap action by blocking event\")\nSigned-off-by: Aya Levin \u003cayal@nvidia.com\u003e\nReviewed-by: Moshe Shemesh \u003cmoshe@nvidia.com\u003e\nReviewed-by: Roi Dayan \u003croid@nvidia.com\u003e\nReviewed-by: Tariq Toukan \u003ctariqt@nvidia.com\u003e\nSigned-off-by: Saeed Mahameed \u003csaeedm@nvidia.com\u003e\n" }, { "commit": "9e196e4398869e4dfe3f49f7f3e0935e07b1d1b9", "tree": "092fb849798495be664823eed179bc381e905cf8", "parents": [ "00423969d806d7169d16fa6314c570a472ca26c9", "ef963ae427aa4669905e0a96b3bd9d44dc85db32" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Apr 14 14:12:12 2021 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Apr 14 14:12:12 2021 -0700" }, "message": "Merge branch \u002710GbE\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue\n\nTony Nguyen says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nIntel Wired LAN Driver Updates 2021-04-14\n\nThis series contains updates to ixgbe and ice drivers.\n\nAlex Duyck fixes a NULL pointer dereference for ixgbe.\n\nYongxin Liu fixes an unbalanced enable/disable which was causing a call\ntrace with suspend for ixgbe.\n\nColin King fixes a potential infinite loop for ice.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "00423969d806d7169d16fa6314c570a472ca26c9", "tree": "43400feb26786cb59af67b1607c50ac3fe3518f0", "parents": [ "416dcc5ce9d2a810477171c62ffa061a98f87367" ], "author": { "name": "Thierry Reding", "email": "treding@nvidia.com", "time": "Wed Apr 14 17:10:07 2021 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Apr 14 14:06:17 2021 -0700" }, "message": "Revert \"net: stmmac: re-init rx buffers when mac resume back\"\n\nThis reverts commit 9c63faaa931e443e7abbbee9de0169f1d4710546, which\nintroduces a suspend/resume regression on Jetson TX2 boards that can be\nreproduced every time. Given that the issue that this was supposed to\nfix only occurs very sporadically the safest course of action is to\nrevert before v5.12 and then we can have another go at fixing the more\nrare issue in the next release (and perhaps backport it if necessary).\n\nThe root cause of the observed problem seems to be that when the system\nis suspended, some packets are still in transit. When the descriptors\nfor these buffers are cleared on resume, the descriptors become invalid\nand cause a fatal bus error.\n\nLink: https://lore.kernel.org/r/708edb92-a5df-ecc4-3126-5ab36707e275@nvidia.com/\nReported-by: Jonathan Hunter \u003cjonathanh@nvidia.com\u003e\nSigned-off-by: Thierry Reding \u003ctreding@nvidia.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "416dcc5ce9d2a810477171c62ffa061a98f87367", "tree": "ea87260a5b7d1ee185d784aea406ca59cd8b4655", "parents": [ "a714e27ea8bdee2b238748029d31472d0a65b611" ], "author": { "name": "Wan Jiabing", "email": "wanjiabing@vivo.com", "time": "Wed Apr 14 19:31:48 2021 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Apr 14 14:02:56 2021 -0700" }, "message": "cavium/liquidio: Fix duplicate argument\n\nFix the following coccicheck warning:\n\n./drivers/net/ethernet/cavium/liquidio/cn66xx_regs.h:413:6-28:\nduplicated argument to \u0026 or |\n\nThe CN6XXX_INTR_M1UPB0_ERR here is duplicate.\nHere should be CN6XXX_INTR_M1UNB0_ERR.\n\nSigned-off-by: Wan Jiabing \u003cwanjiabing@vivo.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "a714e27ea8bdee2b238748029d31472d0a65b611", "tree": "34dd498261b52591abdfc91a15380607d2fb6a56", "parents": [ "2e1534f395e73152e2051332034bff61a56a8368" ], "author": { "name": "Claudiu Beznea", "email": "claudiu.beznea@microchip.com", "time": "Wed Apr 14 14:20:29 2021 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Apr 14 13:57:17 2021 -0700" }, "message": "net: macb: fix the restore of cmp registers\n\nCommit a14d273ba159 (\"net: macb: restore cmp registers on resume path\")\nintroduces the restore of CMP registers on resume path. In case the IP\ndoesn\u0027t support type 2 screeners (zero on DCFG8 register) the\nstruct macb::rx_fs_list::list is not initialized and thus the\nlist_for_each_entry(item, \u0026bp-\u003erx_fs_list.list, list) loop introduced in\ncommit a14d273ba159 (\"net: macb: restore cmp registers on resume path\")\nwill access an uninitialized list leading to crash. Thus, initialize\nthe struct macb::rx_fs_list::list without taking into account if the\nIP supports type 2 screeners or not.\n\nFixes: a14d273ba159 (\"net: macb: restore cmp registers on resume path\")\nSigned-off-by: Claudiu Beznea \u003cclaudiu.beznea@microchip.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "2ef4fb92363c44e8a6f93fd0877b6a7dee6f874d", "tree": "c0b3475babab37484f784507a2d5fb163b53e169", "parents": [ "68ce556bd1643498080af310d4544f46f3c4f3df" ], "author": { "name": "Zack Rusin", "email": "zackr@vmware.com", "time": "Mon Mar 22 13:04:11 2021 -0400" }, "committer": { "name": "Zack Rusin", "email": "zackr@vmware.com", "time": "Wed Apr 14 16:41:31 2021 -0400" }, "message": "drm/vmwgfx: Make sure bo\u0027s are unpinned before putting them back\n\nDuring cotable resize we pin the backup buffer to make sure the\ntrylock doesn\u0027t fail. We were never unpinning the backup buffer\nresulting in every subsequent cotable resize trying to release a\npinned bo. After we copy the old backup to the new we can release\nthe pin.\nMob\u0027s are always pinned so we just have to make sure we unpin\nthem before releasing them.\n\nReviewed-by: Thomas Hellström (Intel) \u003cthomas_os@shipmail.org\u003e\nFixes: d1a73c641afd (\"drm/vmwgfx: Make sure we unpin no longer needed buffers\")\nLink: https://patchwork.freedesktop.org/patch/msgid/20210413205938.788366-1-zackr@vmware.com\nSigned-off-by: Zack Rusin \u003czackr@vmware.com\u003e\n" }, { "commit": "68ce556bd1643498080af310d4544f46f3c4f3df", "tree": "52ea753cb7e88b322484db5e05c2501980c134ea", "parents": [ "ab4d9913632b1e5ffcf3365783e98718b3c83c7f" ], "author": { "name": "Zack Rusin", "email": "zackr@vmware.com", "time": "Mon Mar 22 12:54:57 2021 -0400" }, "committer": { "name": "Zack Rusin", "email": "zackr@vmware.com", "time": "Wed Apr 14 16:41:17 2021 -0400" }, "message": "drm/vmwgfx: Fix the lockdep breakage\n\nThomas has noticed that the lockdep was broken in vmwgfx. It\nwas broken during the pci initialization rework. This fixes\nthe breakage by making sure we initialize the locking code\nbefore doing anything else. This was independently spotted\nand fixed by Tetsuo Handa as well.\n\nReviewed-by: Martin Krastev \u003ckrastevm@vmware.com\u003e\nReviewed-by: Roland Scheidegger \u003csroland@vmware.com\u003e\nCc: Tetsuo Handa \u003cpenguin-kernel@I-love.SAKURA.ne.jp\u003e\nCc: dri-devel@lists.freedesktop.org\nSigned-off-by: Zack Rusin \u003czackr@vmware.com\u003e\nFixes: 8772c0bb58bbf98a (\"drm/vmwgfx: Cleanup pci resource allocation\")\nLink: https://patchwork.freedesktop.org/patch/msgid/20210408172245.673785-1-zackr@vmware.com\n" }, { "commit": "ab4d9913632b1e5ffcf3365783e98718b3c83c7f", "tree": "8cceefe98f43f4ba5cd1f9de06ef9cc3495c84ec", "parents": [ "d434405aaab7d0ebc516b68a8fc4100922d7f5ef" ], "author": { "name": "Zack Rusin", "email": "zackr@vmware.com", "time": "Thu Jan 14 18:38:16 2021 -0500" }, "committer": { "name": "Zack Rusin", "email": "zackr@vmware.com", "time": "Wed Apr 14 16:41:02 2021 -0400" }, "message": "drm/vmwgfx: Make sure we unpin no longer needed buffers\n\nWe were not correctly unpinning no longer needed buffers. In particular\nvmw_buffer_object, which is internally often pinned on creation wasn\u0027t\nunpinned on destruction and none of the internal MOB buffers were\nunpinned before being put back. Technically this existed for a\nlong time but commit 57fcd550eb15 (\"drm/ttm: Warn on pinning without\nholding a reference\") introduced a WARN_ON which was filling up the\nkernel logs rather quickly.\n\nQuite frankly internal usage of vmw_buffer_object and in general\npinning needs to be refactored in vmwgfx but for now this makes\nit work.\n\nSigned-off-by: Zack Rusin \u003czackr@vmware.com\u003e\nReviewed-by: Martin Krastev \u003ckrastevm@vmware.com\u003e\nReviewed-by: Roland Scheidegger \u003csroland@vmware.com\u003e\nFixes: 57fcd550eb15 (\"drm/ttm: Warn on pinning without holding a reference\")\nLink: https://patchwork.freedesktop.org/patch/414984/?series\u003d86052\u0026rev\u003d1\nCc: Huang Rui \u003cray.huang@amd.com\u003e\nCc: Christian König \u003cchristian.koenig@amd.com\u003e\nCc: Daniel Vetter \u003cdaniel.vetter@intel.com\u003e\nCc: Christian Koenig \u003cchristian.koenig@amd.com\u003e\nCc: dri-devel@lists.freedesktop.org\n" }, { "commit": "7f75285ca572eaabc028cf78c6ab5473d0d160be", "tree": "d67da3ceae1fd84d2a616ec6445331c7de71bbf3", "parents": [ "8e2e0104ef78d49a51db9acfd24eaf6d52dc779e", "8ca7cab82bda4eb0b8064befeeeaa38106cac637" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Apr 14 13:23:54 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Apr 14 13:23:54 2021 -0700" }, "message": "Merge tag \u0027for-5.12/dm-fixes-3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm\n\nPull device mapper fix from Mike Snitzer:\n \"Fix DM verity target FEC support\u0027s RS roots IO to always be aligned.\n\n This fixes a previous stable@ fix that overcorrected for a different\n configuration that also resulted in misaligned roots IO\"\n\n* tag \u0027for-5.12/dm-fixes-3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm:\n dm verity fec: fix misaligned RS roots IO\n" }, { "commit": "2e1534f395e73152e2051332034bff61a56a8368", "tree": "1f15e1c8f42ba0ec74587fa2423cb38f1889b59f", "parents": [ "292ecd9f5a94dd29d09fe03b5b669cb20b44f19e" ], "author": { "name": "Nicolas Dichtel", "email": "nicolas.dichtel@6wind.com", "time": "Wed Apr 14 12:03:25 2021 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Apr 14 13:13:53 2021 -0700" }, "message": "vrf: fix a comment about loopback device\n\nThis is a leftover of the below commit.\n\nFixes: 4f04256c983a (\"net: vrf: Drop local rtable and rt6_info\")\nSigned-off-by: Nicolas Dichtel \u003cnicolas.dichtel@6wind.com\u003e\nAcked-by: David Ahern \u003cdsahern@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "292ecd9f5a94dd29d09fe03b5b669cb20b44f19e", "tree": "2a8142fb6e8dce00a07abf227d072286069b32e4", "parents": [ "49963eed2c239365c12418d07ca9ba621dc69313" ], "author": { "name": "Nicolas Dichtel", "email": "nicolas.dichtel@6wind.com", "time": "Wed Apr 14 12:00:27 2021 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Apr 14 13:13:15 2021 -0700" }, "message": "doc: move seg6_flowlabel to seg6-sysctl.rst\n\nLet\u0027s have all seg6 sysctl at the same place.\n\nFixes: a6dc6670cd7e (\"ipv6: sr: Add documentation for seg_flowlabel sysctl\")\nSigned-off-by: Nicolas Dichtel \u003cnicolas.dichtel@6wind.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "49963eed2c239365c12418d07ca9ba621dc69313", "tree": "f73e7e1f057656825a663bcb144d309bec691558", "parents": [ "453a77894efa4d9b6ef9644d74b9419c47ac427c", "7c451f3ef676c805a4b77a743a01a5c21a250a73" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Apr 14 13:10:58 2021 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Apr 14 13:10:58 2021 -0700" }, "message": "Merge branch \u0027ibmvnic-napi-fixes\u0027\n\nLijun Pan says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nibmvnic: correctly call NAPI APIs\n\nThis series correct some misuse of NAPI APIs in the driver.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "7c451f3ef676c805a4b77a743a01a5c21a250a73", "tree": "f73e7e1f057656825a663bcb144d309bec691558", "parents": [ "d3a6abccbd272aea7dc2c6f984bb5a2c11278e44" ], "author": { "name": "Lijun Pan", "email": "lijunp213@gmail.com", "time": "Wed Apr 14 02:46:16 2021 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Apr 14 13:10:58 2021 -0700" }, "message": "ibmvnic: remove duplicate napi_schedule call in open function\n\nRemove the unnecessary napi_schedule() call in __ibmvnic_open() since\ninterrupt_rx() calls napi_schedule_prep/__napi_schedule during every\nreceive interrupt.\n\nFixes: ed651a10875f (\"ibmvnic: Updated reset handling\")\nSigned-off-by: Lijun Pan \u003clijunp213@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "d3a6abccbd272aea7dc2c6f984bb5a2c11278e44", "tree": "a6e673e81e72c513c1acb9aff8394c9eb8b58808", "parents": [ "0775ebc4cf8554bdcd2c212669a0868ab68df5c0" ], "author": { "name": "Lijun Pan", "email": "lijunp213@gmail.com", "time": "Wed Apr 14 02:46:15 2021 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Apr 14 13:10:58 2021 -0700" }, "message": "ibmvnic: remove duplicate napi_schedule call in do_reset function\n\nDuring adapter reset, do_reset/do_hard_reset calls ibmvnic_open(),\nwhich will calls napi_schedule if previous state is VNIC_CLOSED\n(i.e, the reset case, and \"ifconfig down\" case). So there is no need\nfor do_reset to call napi_schedule again at the end of the function\nthough napi_schedule will neglect the request if napi is already\nscheduled.\n\nFixes: ed651a10875f (\"ibmvnic: Updated reset handling\")\nSigned-off-by: Lijun Pan \u003clijunp213@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "0775ebc4cf8554bdcd2c212669a0868ab68df5c0", "tree": "d08a30c883146b03332e2878b1b269d0922deb4c", "parents": [ "453a77894efa4d9b6ef9644d74b9419c47ac427c" ], "author": { "name": "Lijun Pan", "email": "lijunp213@gmail.com", "time": "Wed Apr 14 02:46:14 2021 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Apr 14 13:10:58 2021 -0700" }, "message": "ibmvnic: avoid calling napi_disable() twice\n\n__ibmvnic_open calls napi_disable without checking whether NAPI polling\nhas already been disabled or not. This could cause napi_disable\nbeing called twice, which could generate deadlock. For example,\nthe first napi_disable will spin until NAPI_STATE_SCHED is cleared\nby napi_complete_done, then set it again.\nWhen napi_disable is called the second time, it will loop infinitely\nbecause no dev-\u003epoll will be running to clear NAPI_STATE_SCHED.\n\nTo prevent above scenario from happening, call ibmvnic_napi_disable()\nwhich checks if napi is disabled or not before calling napi_disable.\n\nFixes: bfc32f297337 (\"ibmvnic: Move resource initialization to its own routine\")\nSuggested-by: Thomas Falcon \u003ctlfalcon@linux.ibm.com\u003e\nSigned-off-by: Lijun Pan \u003clijunp213@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "453a77894efa4d9b6ef9644d74b9419c47ac427c", "tree": "09ffd7241b52c8227bb671d3d974bc1bba453ced", "parents": [ "16756d3e77ad58cd07e36cbed724aa13ae5a0278" ], "author": { "name": "Heiner Kallweit", "email": "hkallweit1@gmail.com", "time": "Wed Apr 14 10:47:10 2021 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Apr 14 13:05:40 2021 -0700" }, "message": "r8169: don\u0027t advertise pause in jumbo mode\n\nIt has been reported [0] that using pause frames in jumbo mode impacts\nperformance. There\u0027s no available chip documentation, but vendor\ndrivers r8168 and r8125 don\u0027t advertise pause in jumbo mode. So let\u0027s\ndo the same, according to Roman it fixes the issue.\n\n[0] https://bugzilla.kernel.org/show_bug.cgi?id\u003d212617\n\nFixes: 9cf9b84cc701 (\"r8169: make use of phy_set_asym_pause\")\nReported-by: Roman Mamedov \u003crm+bko@romanrm.net\u003e\nTested-by: Roman Mamedov \u003crm+bko@romanrm.net\u003e\nSigned-off-by: Heiner Kallweit \u003chkallweit1@gmail.com\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "16756d3e77ad58cd07e36cbed724aa13ae5a0278", "tree": "c9ba1c42ab61d04ca305e747e0d86d562272753a", "parents": [ "2afeec08ab5c86ae21952151f726bfe184f6b23d" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Apr 13 20:46:14 2021 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Apr 14 13:03:06 2021 -0700" }, "message": "ethtool: pause: make sure we init driver stats\n\nThe intention was for pause statistics to not be reported\nwhen driver does not have the relevant callback (only\nreport an empty netlink nest). What happens currently\nwe report all 0s instead. Make sure statistics are\ninitialized to \"not set\" (which is -1) so the dumping\ncode skips them.\n\nFixes: 9a27a33027f2 (\"ethtool: add standard pause stats\")\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "c7d95613c7d6e003969722a290397b8271bdad17", "tree": "5eddfc6c45f59cea3351e2b0e3ffc3adb5cb2d1e", "parents": [ "c60eb049f4a19ddddcd3ee97a9c79ab8066a6a03" ], "author": { "name": "Pavel Begunkov", "email": "asml.silence@gmail.com", "time": "Tue Apr 13 11:43:00 2021 +0100" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Wed Apr 14 13:07:27 2021 -0600" }, "message": "io_uring: fix early sqd_list removal sqpoll hangs\n\n[ 245.463317] INFO: task iou-sqp-1374:1377 blocked for more than 122 seconds.\n[ 245.463334] task:iou-sqp-1374 state:D flags:0x00004000\n[ 245.463345] Call Trace:\n[ 245.463352] __schedule+0x36b/0x950\n[ 245.463376] schedule+0x68/0xe0\n[ 245.463385] __io_uring_cancel+0xfb/0x1a0\n[ 245.463407] do_exit+0xc0/0xb40\n[ 245.463423] io_sq_thread+0x49b/0x710\n[ 245.463445] ret_from_fork+0x22/0x30\n\nIt happens when sqpoll forgot to run park_task_work and goes to exit,\nthen exiting user may remove ctx from sqd_list, and so corresponding\nio_sq_thread() -\u003e io_uring_cancel_sqpoll() won\u0027t be executed. Hopefully\nit just stucks in do_exit() in this case.\n\nFixes: dbe1bdbb39db (\"io_uring: handle signals for IO threads like a normal thread\")\nReported-by: Joakim Hassila \u003cjoj@mac.com\u003e\nSigned-off-by: Pavel Begunkov \u003casml.silence@gmail.com\u003e\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "8ca7cab82bda4eb0b8064befeeeaa38106cac637", "tree": "85e7c10725c382e9ba342690e457badff7268daa", "parents": [ "d434405aaab7d0ebc516b68a8fc4100922d7f5ef" ], "author": { "name": "Jaegeuk Kim", "email": "jaegeuk@google.com", "time": "Wed Apr 14 08:28:28 2021 -0700" }, "committer": { "name": "Mike Snitzer", "email": "snitzer@redhat.com", "time": "Wed Apr 14 14:28:29 2021 -0400" }, "message": "dm verity fec: fix misaligned RS roots IO\n\ncommit df7b59ba9245 (\"dm verity: fix FEC for RS roots unaligned to\nblock size\") introduced the possibility for misaligned roots IO\nrelative to the underlying device\u0027s logical block size. E.g. Android\u0027s\ndefault RS roots\u003d2 results in dm_bufio-\u003eblock_size\u003d1024, which causes\nthe following EIO if the logical block size of the device is 4096,\ngiven v-\u003edata_dev_block_bits\u003d12:\n\nE sd 0 : 0:0:0: [sda] tag#30 request not aligned to the logical block size\nE blk_update_request: I/O error, dev sda, sector 10368424 op 0x0:(READ) flags 0x0 phys_seg 1 prio class 0\nE device-mapper: verity-fec: 254:8: FEC 9244672: parity read failed (block 18056): -5\n\nFix this by onlu using f-\u003eroots for dm_bufio blocksize IFF it is\naligned to v-\u003edata_dev_block_bits.\n\nFixes: df7b59ba9245 (\"dm verity: fix FEC for RS roots unaligned to block size\")\nCc: stable@vger.kernel.org\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk@google.com\u003e\nSigned-off-by: Mike Snitzer \u003csnitzer@redhat.com\u003e\n" }, { "commit": "8e2e0104ef78d49a51db9acfd24eaf6d52dc779e", "tree": "29b05dea088e01e85097cc3c4996678f1b181ada", "parents": [ "ec97a7296ad1ca3ccb8bca1e72739cb8262686f1", "a994eddb947ea9ebb7b14d9a1267001699f0a136" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Apr 14 10:55:56 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Apr 14 10:55:56 2021 -0700" }, "message": "Merge tag \u0027s390-5.12-7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux\n\nPull s390 fixes from Heiko Carstens:\n\n - setup stack backchain properly in external and i/o interrupt handler\n to fix stack unwinding. This broke when converting to generic entry\n\n - save caller address of psw_idle to get a sane stacktrace\n\n* tag \u0027s390-5.12-7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:\n s390/entry: save the caller of psw_idle\n s390/entry: avoid setting up backchain in ext|io handlers\n" }, { "commit": "ec97a7296ad1ca3ccb8bca1e72739cb8262686f1", "tree": "7af8b62caabde85cf07d21b942c61694b30940b4", "parents": [ "c17a3066b4c1acdf36fa307faaa391f558ac0420", "738fa58ee1328481d1d7889e7c430b3401c571b9" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Apr 14 10:36:03 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Apr 14 10:36:03 2021 -0700" }, "message": "Merge tag \u0027arm64-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux\n\nPull arm64 fixes from Will Deacon:\n\n - Fix incorrect asm constraint for load_unaligned_zeropad() fixup\n\n - Fix thread flag update when setting TIF_MTE_ASYNC_FAULT\n\n - Fix restored irq state when handling fault on kprobe\n\n* tag \u0027arm64-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:\n arm64: kprobes: Restore local irqflag if kprobes is cancelled\n arm64: mte: Ensure TIF_MTE_ASYNC_FAULT is set atomically\n arm64: fix inline asm in load_unaligned_zeropad()\n" }, { "commit": "c17a3066b4c1acdf36fa307faaa391f558ac0420", "tree": "82007641c1aee2011b1cbfa132a8e41fe1fecddb", "parents": [ "e70b911acc1687100c31e550251715dbdac96a12", "ea9aadc06a9f10ad20a90edc0a484f1147d88a7a" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Apr 14 09:36:54 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Apr 14 09:36:54 2021 -0700" }, "message": "Merge tag \u0027dmaengine-fix-5.12\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine\n\nPull dmaengine fixes from Vinod Koul:\n \"A couple of dmaengine driver fixes for:\n\n - race and descriptor issue for xilinx driver\n\n - fix interrupt handling, wq state \u0026 cleanup, field sizes for\n completion, msix permissions for idxd driver\n\n - runtime pm fix for tegra driver\n\n - double free fix in dma_async_device_register\"\n\n* tag \u0027dmaengine-fix-5.12\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine:\n dmaengine: idxd: fix wq cleanup of WQCFG registers\n dmaengine: idxd: clear MSIX permission entry on shutdown\n dmaengine: plx_dma: add a missing put_device() on error path\n dmaengine: tegra20: Fix runtime PM imbalance on error\n dmaengine: Fix a double free in dma_async_device_register\n dmaengine: dw: Make it dependent to HAS_IOMEM\n dmaengine: idxd: fix wq size store permission state\n dmaengine: idxd: fix opcap sysfs attribute output\n dmaengine: idxd: fix delta_rec and crc size field for completion record\n dmaengine: idxd: Fix clobbering of SWERR overflow bit on writeback\n dmaengine: xilinx: dpdma: Fix race condition in done IRQ\n dmaengine: xilinx: dpdma: Fix descriptor issuing on video group\n" }, { "commit": "e70b911acc1687100c31e550251715dbdac96a12", "tree": "9705f1140d063393a06569fb8d99ddaa578bdcf7", "parents": [ "2558258d78873998b8cd81ce7661dc68541b8b51", "909290786ea335366e21d7f1ed5812b90f2f0a92" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Apr 14 09:10:54 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Apr 14 09:10:54 2021 -0700" }, "message": "Merge tag \u0027vfio-v5.12-rc8\u0027 of git://github.com/awilliam/linux-vfio\n\nPull VFIO fix from Alex Williamson:\n \"Verify mmap region within range (Christian A. Ehrhardt)\"\n\n* tag \u0027vfio-v5.12-rc8\u0027 of git://github.com/awilliam/linux-vfio:\n vfio/pci: Add missing range check in vfio_pci_mmap\n" }, { "commit": "2558258d78873998b8cd81ce7661dc68541b8b51", "tree": "ac3bba36709063467350e63b1d847286c8a91848", "parents": [ "50987beca096a7ed4f453a6da245fd6a2fadedeb", "04c4f2ee3f68c9a4bf1653d15f1a9a435ae33f7a" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Apr 14 08:50:46 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Apr 14 08:50:46 2021 -0700" }, "message": "Merge tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/virt/kvm/kvm\n\nPull kvm fix from Paolo Bonzini:\n \"Fix for a possible out-of-bounds access\"\n\n* tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/virt/kvm/kvm:\n KVM: VMX: Don\u0027t use vcpu-\u003erun-\u003einternal.ndata as an array index\n" }, { "commit": "50987beca096a7ed4f453a6da245fd6a2fadedeb", "tree": "4dd912eeded6cc698d02850050b04dad450d77c1", "parents": [ "eebe426d32e1a10ac7c35f8ffab5f818c32a2454", "8db403b9631331ef1d5e302cdf353c48849ca9d5" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Apr 13 18:40:00 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Apr 13 18:40:00 2021 -0700" }, "message": "Merge tag \u0027trace-v5.12-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace\n\nPull tracing fix from Steven Rostedt:\n \"Fix a memory link in dyn_event_release().\n\n An error path exited the function before freeing the allocated \u0027argv\u0027\n variable\"\n\n* tag \u0027trace-v5.12-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:\n tracing/dynevent: Fix a memory leak in an error handling path\n" }, { "commit": "2afeec08ab5c86ae21952151f726bfe184f6b23d", "tree": "68d755c22e5a1ebea73397b4e1d958f88adb16a2", "parents": [ "38ec4944b593fd90c5ef42aaaa53e66ae5769d04" ], "author": { "name": "Michael Brown", "email": "mbrown@fensystems.co.uk", "time": "Tue Apr 13 16:25:12 2021 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Apr 13 15:24:14 2021 -0700" }, "message": "xen-netback: Check for hotplug-status existence before watching\n\nThe logic in connect() is currently written with the assumption that\nxenbus_watch_pathfmt() will return an error for a node that does not\nexist. This assumption is incorrect: xenstore does allow a watch to\nbe registered for a nonexistent node (and will send notifications\nshould the node be subsequently created).\n\nAs of commit 1f2565780 (\"xen-netback: remove \u0027hotplug-status\u0027 once it\nhas served its purpose\"), this leads to a failure when a domU\ntransitions into XenbusStateConnected more than once. On the first\ndomU transition into Connected state, the \"hotplug-status\" node will\nbe deleted by the hotplug_status_changed() callback in dom0. On the\nsecond or subsequent domU transition into Connected state, the\nhotplug_status_changed() callback will therefore never be invoked, and\nso the backend will remain stuck in InitWait.\n\nThis failure prevents scenarios such as reloading the xen-netfront\nmodule within a domU, or booting a domU via iPXE. There is\nunfortunately no way for the domU to work around this dom0 bug.\n\nFix by explicitly checking for existence of the \"hotplug-status\" node,\nthereby creating the behaviour that was previously assumed to exist.\n\nSigned-off-by: Michael Brown \u003cmbrown@fensystems.co.uk\u003e\nReviewed-by: Paul Durrant \u003cpaul@xen.org\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "04c4f2ee3f68c9a4bf1653d15f1a9a435ae33f7a", "tree": "43993d21b858cc8b24c9f6cb909f2dd837fb3bcb", "parents": [ "315f02c60d9425b38eb8ad7f21b8a35e40db23f9" ], "author": { "name": "Reiji Watanabe", "email": "reijiw@google.com", "time": "Tue Apr 13 15:47:40 2021 +0000" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Tue Apr 13 18:23:41 2021 -0400" }, "message": "KVM: VMX: Don\u0027t use vcpu-\u003erun-\u003einternal.ndata as an array index\n\n__vmx_handle_exit() uses vcpu-\u003erun-\u003einternal.ndata as an index for\nan array access. Since vcpu-\u003erun is (can be) mapped to a user address\nspace with a writer permission, the \u0027ndata\u0027 could be updated by the\nuser process at anytime (the user process can set it to outside the\nbounds of the array).\nSo, it is not safe that __vmx_handle_exit() uses the \u0027ndata\u0027 that way.\n\nFixes: 1aa561b1a4c0 (\"kvm: x86: Add \"last CPU\" to some KVM_EXIT information\")\nSigned-off-by: Reiji Watanabe \u003creijiw@google.com\u003e\nReviewed-by: Jim Mattson \u003cjmattson@google.com\u003e\nMessage-Id: \u003c20210413154739.490299-1-reijiw@google.com\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "38ec4944b593fd90c5ef42aaaa53e66ae5769d04", "tree": "ada126e3929b87d7b29b858478f5b8f5bf70475c", "parents": [ "b166a20b07382b8bc1dcee2a448715c9c2c81b5b" ], "author": { "name": "Eric Dumazet", "email": "edumazet@google.com", "time": "Tue Apr 13 05:41:35 2021 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Apr 13 15:09:31 2021 -0700" }, "message": "gro: ensure frag0 meets IP header alignment\n\nAfter commit 0f6925b3e8da (\"virtio_net: Do not pull payload in skb-\u003ehead\")\nGuenter Roeck reported one failure in his tests using sh architecture.\n\nAfter much debugging, we have been able to spot silent unaligned accesses\nin inet_gro_receive()\n\nThe issue at hand is that upper networking stacks assume their header\nis word-aligned. Low level drivers are supposed to reserve NET_IP_ALIGN\nbytes before the Ethernet header to make that happen.\n\nThis patch hardens skb_gro_reset_offset() to not allow frag0 fast-path\nif the fragment is not properly aligned.\n\nSome arches like x86, arm64 and powerpc do not care and define NET_IP_ALIGN\nas 0, this extra check will be a NOP for them.\n\nNote that if frag0 is not used, GRO will call pskb_may_pull()\nas many times as needed to pull network and transport headers.\n\nFixes: 0f6925b3e8da (\"virtio_net: Do not pull payload in skb-\u003ehead\")\nFixes: 78a478d0efd9 (\"gro: Inline skb_gro_header and cache frag0 virtual address\")\nSigned-off-by: Eric Dumazet \u003cedumazet@google.com\u003e\nReported-by: Guenter Roeck \u003clinux@roeck-us.net\u003e\nCc: Xuan Zhuo \u003cxuanzhuo@linux.alibaba.com\u003e\nCc: \"Michael S. Tsirkin\" \u003cmst@redhat.com\u003e\nCc: Jason Wang \u003cjasowang@redhat.com\u003e\nAcked-by: Michael S. Tsirkin \u003cmst@redhat.com\u003e\nTested-by: Guenter Roeck \u003clinux@roeck-us.net\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "b166a20b07382b8bc1dcee2a448715c9c2c81b5b", "tree": "ddc605b5d010d800b84225d1113abb26641e95a4", "parents": [ "ca09bf7bb109a37a7ff05f230bb3fa3627e6625f" ], "author": { "name": "Or Cohen", "email": "orcohen@paloaltonetworks.com", "time": "Tue Apr 13 21:10:31 2021 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Apr 13 14:59:46 2021 -0700" }, "message": "net/sctp: fix race condition in sctp_destroy_sock\n\nIf sctp_destroy_sock is called without sock_net(sk)-\u003esctp.addr_wq_lock\nheld and sp-\u003edo_auto_asconf is true, then an element is removed\nfrom the auto_asconf_splist without any proper locking.\n\nThis can happen in the following functions:\n1. In sctp_accept, if sctp_sock_migrate fails.\n2. In inet_create or inet6_create, if there is a bpf program\n attached to BPF_CGROUP_INET_SOCK_CREATE which denies\n creation of the sctp socket.\n\nThe bug is fixed by acquiring addr_wq_lock in sctp_destroy_sock\ninstead of sctp_close.\n\nThis addresses CVE-2021-23133.\n\nReported-by: Or Cohen \u003corcohen@paloaltonetworks.com\u003e\nReviewed-by: Xin Long \u003clucien.xin@gmail.com\u003e\nFixes: 610236587600 (\"bpf: Add new cgroup attach type to enable sock modifications\")\nSigned-off-by: Or Cohen \u003corcohen@paloaltonetworks.com\u003e\nAcked-by: Marcelo Ricardo Leitner \u003cmarcelo.leitner@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "ca09bf7bb109a37a7ff05f230bb3fa3627e6625f", "tree": "85aa36b590287ba1ea051550077daccb53becddc", "parents": [ "97684f0970f6e112926de631fdd98d9693c7e5c1" ], "author": { "name": "Lijun Pan", "email": "lijunp213@gmail.com", "time": "Tue Apr 13 03:33:25 2021 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Apr 13 14:49:30 2021 -0700" }, "message": "ibmvnic: correctly use dev_consume/free_skb_irq\n\nIt is more correct to use dev_kfree_skb_irq when packets are dropped,\nand to use dev_consume_skb_irq when packets are consumed.\n\nFixes: 0d973388185d (\"ibmvnic: Introduce xmit_more support using batched subCRQ hcalls\")\nSuggested-by: Thomas Falcon \u003ctlfalcon@linux.ibm.com\u003e\nSigned-off-by: Lijun Pan \u003clijunp213@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "97684f0970f6e112926de631fdd98d9693c7e5c1", "tree": "9ad8cbca70d7d7f89635fd6f78dac56f1e2b7c5f", "parents": [ "61aaa1aa9a8f83dbdc763c6939568952a2a30c90" ], "author": { "name": "Jonathon Reinhart", "email": "jonathon.reinhart@gmail.com", "time": "Tue Apr 13 03:08:48 2021 -0400" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Apr 13 14:42:51 2021 -0700" }, "message": "net: Make tcp_allowed_congestion_control readonly in non-init netns\n\nCurrently, tcp_allowed_congestion_control is global and writable;\nwriting to it in any net namespace will leak into all other net\nnamespaces.\n\ntcp_available_congestion_control and tcp_allowed_congestion_control are\nthe only sysctls in ipv4_net_table (the per-netns sysctl table) with a\nNULL data pointer; their handlers (proc_tcp_available_congestion_control\nand proc_allowed_congestion_control) have no other way of referencing a\nstruct net. Thus, they operate globally.\n\nBecause ipv4_net_table does not use designated initializers, there is no\neasy way to fix up this one \"bad\" table entry. However, the data pointer\nupdating logic shouldn\u0027t be applied to NULL pointers anyway, so we\ninstead force these entries to be read-only.\n\nThese sysctls used to exist in ipv4_table (init-net only), but they were\nmoved to the per-net ipv4_net_table, presumably without realizing that\ntcp_allowed_congestion_control was writable and thus introduced a leak.\n\nBecause the intent of that commit was only to know (i.e. read) \"which\ncongestion algorithms are available or allowed\", this read-only solution\nshould be sufficient.\n\nThe logic added in recent commit\n31c4d2f160eb: (\"net: Ensure net namespace isolation of sysctls\")\ndoes not and cannot check for NULL data pointers, because\nother table entries (e.g. /proc/sys/net/netfilter/nf_log/) have\n.data\u003dNULL but use other methods (.extra2) to access the struct net.\n\nFixes: 9cb8e048e5d9 (\"net/ipv4/sysctl: show tcp_{allowed, available}_congestion_control in non-initial netns\")\nSigned-off-by: Jonathon Reinhart \u003cjonathon.reinhart@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" } ], "next": "61aaa1aa9a8f83dbdc763c6939568952a2a30c90" }