)]}' { "log": [ { "commit": "62fb9874f5da54fdb243003b386128037319b219", "tree": "e07e8b0138ae8ec5bdf138cf11f538c7ddc3dfb0", "parents": [ "b4b27b9eed8ebdbf9f3046197d29d733c8c944f3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Jun 27 15:21:11 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Jun 27 15:21:11 2021 -0700" }, "message": "Linux 5.13\n" }, { "commit": "b4b27b9eed8ebdbf9f3046197d29d733c8c944f3", "tree": "836d7997effa9b977233dedc5f5f7c05685c111b", "parents": [ "625acffd7ae2c52898d249e6c5c39f348db0d8df" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Jun 27 13:32:54 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Jun 27 13:32:54 2021 -0700" }, "message": "Revert \"signal: Allow tasks to cache one sigqueue struct\"\n\nThis reverts commits 4bad58ebc8bc4f20d89cff95417c9b4674769709 (and\n399f8dd9a866e107639eabd3c1979cd526ca3a98, which tried to fix it).\n\nI do not believe these are correct, and I\u0027m about to release 5.13, so am\nreverting them out of an abundance of caution.\n\nThe locking is odd, and appears broken.\n\nOn the allocation side (in __sigqueue_alloc()), the locking is somewhat\nstraightforward: it depends on sighand-\u003esiglock. Since one caller\ndoesn\u0027t hold that lock, it further then tests \u0027sigqueue_flags\u0027 to avoid\nthe case with no locks held.\n\nOn the freeing side (in sigqueue_cache_or_free()), there is no locking\nat all, and the logic instead depends on \u0027current\u0027 being a single\nthread, and not able to race with itself.\n\nTo make things more exciting, there\u0027s also the data race between freeing\na signal and allocating one, which is handled by using WRITE_ONCE() and\nREAD_ONCE(), and being mutually exclusive wrt the initial state (ie\nfreeing will only free if the old state was NULL, while allocating will\nobviously only use the value if it was non-NULL, so only one or the\nother will actually act on the value).\n\nHowever, while the free-\u003ealloc paths do seem mutually exclusive thanks\nto just the data value dependency, it\u0027s not clear what the memory\nordering constraints are on it. Could writes from the previous\nallocation possibly be delayed and seen by the new allocation later,\ncausing logical inconsistencies?\n\nSo it\u0027s all very exciting and unusual.\n\nAnd in particular, it seems that the freeing side is incorrect in\ndepending on \"current\" being single-threaded. Yes, \u0027current\u0027 is a\nsingle thread, but in the presense of asynchronous events even a single\nthread can have data races.\n\nAnd such asynchronous events can and do happen, with interrupts causing\nsignals to be flushed and thus free\u0027d (for example - sending a\nSIGCONT/SIGSTOP can happen from interrupt context, and can flush\npreviously queued process control signals).\n\nSo regardless of all the other questions about the memory ordering and\nlocking for this new cached allocation, the sigqueue_cache_or_free()\nassumptions seem to be fundamentally incorrect.\n\nIt may be that people will show me the errors of my ways, and tell me\nwhy this is all safe after all. We can reinstate it if so. But my\ncurrent belief is that the WRITE_ONCE() that sets the cached entry needs\nto be a smp_store_release(), and the READ_ONCE() that finds a cached\nentry needs to be a smp_load_acquire() to handle memory ordering\ncorrectly.\n\nAnd the sequence in sigqueue_cache_or_free() would need to either use a\nlock or at least be interrupt-safe some way (perhaps by using something\nlike the percpu \u0027cmpxchg\u0027: it doesn\u0027t need to be SMP-safe, but like the\npercpu operations it needs to be interrupt-safe).\n\nFixes: 399f8dd9a866 (\"signal: Prevent sigqueue caching after task got released\")\nFixes: 4bad58ebc8bc (\"signal: Allow tasks to cache one sigqueue struct\")\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nCc: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nCc: Oleg Nesterov \u003coleg@redhat.com\u003e\nCc: Christian Brauner \u003cchristian.brauner@ubuntu.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "625acffd7ae2c52898d249e6c5c39f348db0d8df", "tree": "af448a062513dd41e0a223af859ed081ea301300", "parents": [ "b7050b242430f3170e0b57f5f55136e44cb8dc66", "67147e96a332b56c7206238162771d82467f86c0" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Jun 26 09:50:10 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Jun 26 09:50:10 2021 -0700" }, "message": "Merge tag \u0027s390-5.13-5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux\n\nPull s390 fixes from Vasily Gorbik:\n\n - Fix a couple of late pt_regs flags handling findings of conversion to\n generic entry.\n\n - Fix potential register clobbering in stack switch helper.\n\n - Fix thread/group masks for offline cpus.\n\n - Fix cleanup of mdev resources when remove callback is invoked in\n vfio-ap code.\n\n* tag \u0027s390-5.13-5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:\n s390/stack: fix possible register corruption with stack switch helper\n s390/topology: clear thread/group maps for offline cpus\n s390/vfio-ap: clean up mdev resources when remove callback invoked\n s390: clear pt_regs::flags on irq entry\n s390: fix system call restart with multiple signals\n" }, { "commit": "b7050b242430f3170e0b57f5f55136e44cb8dc66", "tree": "7b5aaa6a0bc5ef817e2cf98522ca723c6a9d726c", "parents": [ "e2f527b58e8115dae15ae344215accdd7a42e5ba", "67e2996f72c71ebe4ac2fcbcf77e54479bb7aa11" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 19:06:24 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 19:06:24 2021 -0700" }, "message": "Merge tag \u0027pinctrl-v5.13-3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl\n\nPull pin control fixes from Linus Walleij:\n \"Two last-minute fixes:\n\n - Put an fwnode in the errorpath in the SGPIO driver\n\n - Fix the number of GPIO lines per bank in the STM32 driver\"\n\n* tag \u0027pinctrl-v5.13-3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl:\n pinctrl: stm32: fix the reported number of GPIO lines per bank\n pinctrl: microchip-sgpio: Put fwnode in error case during -\u003eprobe()\n" }, { "commit": "e2f527b58e8115dae15ae344215accdd7a42e5ba", "tree": "667ce3013c9e5fe3a20e66fd00ee5343068f5564", "parents": [ "7ce32ac6fb2fc73584b567c73ae0c47528954ec6", "d1b7f92035c6fb42529ada531e2cbf3534544c82" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 15:59:14 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 15:59:14 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 small fixes, both in upper layer drivers (scsi disk and cdrom).\n\n The sd one is fixing a commit changing revalidation that came from the\n block tree a while ago (5.10) and the sr one adds handling of a\n condition we didn\u0027t previously handle for manually removed media\"\n\n* tag \u0027scsi-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:\n scsi: sd: Call sd_revalidate_disk() for ioctl(BLKRRPART)\n scsi: sr: Return appropriate error code when disk is ejected\n" }, { "commit": "7ce32ac6fb2fc73584b567c73ae0c47528954ec6", "tree": "e10dd353d9c646109f7481dad1023496142f8f10", "parents": [ "808e9df477757955a9644ca323010339be0c40ee", "72a461adbe88acf6a8cc5dba7720cf94d7056154" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 11:05:03 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 11:05:03 2021 -0700" }, "message": "Merge branch \u0027akpm\u0027 (patches from Andrew)\n\nMerge misc fixes from Andrew Morton:\n \"24 patches, based on 4a09d388f2ab382f217a764e6a152b3f614246f6.\n\n Subsystems affected by this patch series: mm (thp, vmalloc, hugetlb,\n memory-failure, and pagealloc), nilfs2, kthread, MAINTAINERS, and\n mailmap\"\n\n* emailed patches from Andrew Morton \u003cakpm@linux-foundation.org\u003e: (24 commits)\n mailmap: add Marek\u0027s other e-mail address and identity without diacritics\n MAINTAINERS: fix Marek\u0027s identity again\n mm/page_alloc: do bulk array bounds check after checking populated elements\n mm/page_alloc: __alloc_pages_bulk(): do bounds check before accessing array\n mm/hwpoison: do not lock page again when me_huge_page() successfully recovers\n mm,hwpoison: return -EHWPOISON to denote that the page has already been poisoned\n mm/memory-failure: use a mutex to avoid memory_failure() races\n mm, futex: fix shared futex pgoff on shmem huge page\n kthread: prevent deadlock when kthread_mod_delayed_work() races with kthread_cancel_delayed_work_sync()\n kthread_worker: split code for canceling the delayed work timer\n mm/vmalloc: unbreak kasan vmalloc support\n KVM: s390: prepare for hugepage vmalloc\n mm/vmalloc: add vmalloc_no_huge\n nilfs2: fix memory leak in nilfs_sysfs_delete_device_group\n mm/thp: another PVMW_SYNC fix in page_vma_mapped_walk()\n mm/thp: fix page_vma_mapped_walk() if THP mapped by ptes\n mm: page_vma_mapped_walk(): get vma_address_end() earlier\n mm: page_vma_mapped_walk(): use goto instead of while (1)\n mm: page_vma_mapped_walk(): add a level of indentation\n mm: page_vma_mapped_walk(): crossing page table boundary\n ...\n" }, { "commit": "808e9df477757955a9644ca323010339be0c40ee", "tree": "3788dfa060a6d7be79d0a4a3d12697fa7b79b8ad", "parents": [ "55fcd4493da5ac8a0f7a0b3b5ae8448aee2041bb" ], "author": { "name": "Gleb Fotengauer-Malinovskiy", "email": "glebfm@altlinux.org", "time": "Fri Jun 25 20:36:55 2021 +0300" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 10:53:26 2021 -0700" }, "message": "userfaultfd: uapi: fix UFFDIO_CONTINUE ioctl request definition\n\nThis ioctl request reads from uffdio_continue structure written by\nuserspace which justifies _IOC_WRITE flag. It also writes back to that\nstructure which justifies _IOC_READ flag.\n\nSee NOTEs in include/uapi/asm-generic/ioctl.h for more information.\n\nFixes: f619147104c8 (\"userfaultfd: add UFFDIO_CONTINUE ioctl\")\nSigned-off-by: Gleb Fotengauer-Malinovskiy \u003cglebfm@altlinux.org\u003e\nAcked-by: Peter Xu \u003cpeterx@redhat.com\u003e\nReviewed-by: Axel Rasmussen \u003caxelrasmussen@google.com\u003e\nReviewed-by: Dmitry V. Levin \u003cldv@altlinux.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "55fcd4493da5ac8a0f7a0b3b5ae8448aee2041bb", "tree": "16c2f3af439c874cc107318e62a95cf49f117ab9", "parents": [ "7764c62f9848cd4585801019168a6272240ab4d3", "4ca070ef0dd885616ef294d269a9bf8e3b258e1a" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 10:44:03 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 10:44:03 2021 -0700" }, "message": "Merge branch \u0027i2c/for-current\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux\n\nPull i2c fixes from Wolfram Sang:\n \"Three more driver bugfixes and an annotation fix for the core\"\n\n* \u0027i2c/for-current\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux:\n i2c: robotfuzz-osif: fix control-request directions\n i2c: dev: Add __user annotation\n i2c: cp2615: check for allocation failure in cp2615_i2c_recv()\n i2c: i801: Ensure that SMBHSTSTS_INUSE_STS is cleared when leaving i801_access\n" }, { "commit": "7764c62f9848cd4585801019168a6272240ab4d3", "tree": "e59157a9ef12f91ff7e1af88330521a23ce09742", "parents": [ "b960e0147451915b5d4cd208b7abd3b07ceaf1a2", "5dca69e26fe97f17d4a6cbd6872103c868577b14" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 10:30:28 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 10:30:28 2021 -0700" }, "message": "Merge tag \u0027devprop-5.13-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm\n\nPull device properties framework fix from Rafael Wysocki:\n \"Fix a NULL pointer dereference introduced by a recent commit and\n occurring when device_remove_software_node() is used with a device\n that has never been registered (Heikki Krogerus)\"\n\n* tag \u0027devprop-5.13-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:\n software node: Handle software node injection to an existing device properly\n" }, { "commit": "b960e0147451915b5d4cd208b7abd3b07ceaf1a2", "tree": "13a1541e033fb5ca3989f1cff330759290236b0a", "parents": [ "616a99dd146a799d0cac43f884a3a46571bd2796", "3de218ff39b9e3f0d453fe3154f12a174de44b25" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 10:19:01 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 10:19:01 2021 -0700" }, "message": "Merge tag \u0027for-linus-5.13b-rc8-tag\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip\n\nPull xen fix from Juergen Gross:\n \"A fix for a regression introduced in 5.12: when migrating an irq\n related to a Xen user event to another cpu, a race might result\n in a WARN() triggering\"\n\n* tag \u0027for-linus-5.13b-rc8-tag\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:\n xen/events: reset active flag for lateeoi events later\n" }, { "commit": "616a99dd146a799d0cac43f884a3a46571bd2796", "tree": "a359ed8eebce395d55e9f199674c1680c3e35298", "parents": [ "94ca94bbbb5f50180ecaac31566dfe5ed44c7994", "f8be156be163a052a067306417cd0ff679068c97" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 10:15:35 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 10:15:35 2021 -0700" }, "message": "Merge tag \u0027for-linus-urgent\u0027 of git://git.kernel.org/pub/scm/virt/kvm/kvm\n\nPull kvm fixes from Paolo Bonzini:\n \"A selftests fix for ARM, and the fix for page reference count\n underflow. This is a very small fix that was provided by Nick Piggin\n and tested by myself\"\n\n* tag \u0027for-linus-urgent\u0027 of git://git.kernel.org/pub/scm/virt/kvm/kvm:\n KVM: do not allow mapping valid but non-reference-counted pages\n KVM: selftests: Fix mapping length truncation in m{,un}map()\n" }, { "commit": "94ca94bbbb5f50180ecaac31566dfe5ed44c7994", "tree": "6be0cccd19471d0fd9c431cf1958d9b5aff5cdcf", "parents": [ "edf54d9d0ae0a230199a58e57b46c2d4b37a1462", "f9dfb5e390fab2df9f7944bb91e7705aba14cd26" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 10:00:25 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 10:00:25 2021 -0700" }, "message": "Merge tag \u0027x86_urgent_for_v5.13\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull x86 fixes from Borislav Petkov:\n \"Two more urgent FPU fixes:\n\n - prevent unprivileged userspace from reinitializing supervisor\n states\n\n - prepare init_fpstate, which is the buffer used when initializing\n FPU state, properly in case the skip-writing-state-components\n XSAVE* variants are used\"\n\n* tag \u0027x86_urgent_for_v5.13\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n x86/fpu: Make init_fpstate correct with optimized XSAVE\n x86/fpu: Preserve supervisor states in sanitize_restored_user_xstate()\n" }, { "commit": "edf54d9d0ae0a230199a58e57b46c2d4b37a1462", "tree": "170ca35357bfe07143c08d4a130c4829d0ccc2ff", "parents": [ "9e736cf7d6f0dac63855ba74c94b85898485ba7a", "03af4c7bad8ca59143bca488b90b3775d10d7f94" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 09:50:30 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 09:50:30 2021 -0700" }, "message": "Merge tag \u0027ceph-for-5.13-rc8\u0027 of https://github.com/ceph/ceph-client\n\nPull ceph fixes from Ilya Dryomov:\n \"Two regression fixes from the merge window: one in the auth code\n affecting old clusters and one in the filesystem for proper\n propagation of MDS request errors.\n\n Also included a locking fix for async creates, marked for stable\"\n\n* tag \u0027ceph-for-5.13-rc8\u0027 of https://github.com/ceph/ceph-client:\n libceph: set global_id as soon as we get an auth ticket\n libceph: don\u0027t pass result into ac-\u003eops-\u003ehandle_reply()\n ceph: fix error handling in ceph_atomic_open and ceph_lookup\n ceph: must hold snap_rwsem when filling inode for async create\n" }, { "commit": "9e736cf7d6f0dac63855ba74c94b85898485ba7a", "tree": "22a4528cbd9de0be750e49cfb114f9000819156e", "parents": [ "c13e3021331ed7736996fe61d6f26983ac3b84cc", "827a746f405d25f79560c7868474aec5aee174e1" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 09:41:29 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 09:41:29 2021 -0700" }, "message": "Merge tag \u0027netfs-fixes-20210621\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs\n\nPull netfs fixes from David Howells:\n \"This contains patches to fix netfs_write_begin() and afs_write_end()\n in the following ways:\n\n (1) In netfs_write_begin(), extract the decision about whether to skip\n a page out to its own helper and have that clear around the region\n to be written, but not clear that region. This requires the\n filesystem to patch it up afterwards if the hole doesn\u0027t get\n completely filled.\n\n (2) Use offset_in_thp() in (1) rather than manually calculating the\n offset into the page.\n\n (3) Due to (1), afs_write_end() now needs to handle short data write\n into the page by generic_perform_write(). I\u0027ve adopted an\n analogous approach to ceph of just returning 0 in this case and\n letting the caller go round again.\n\n It also adds a note that (in the future) the len parameter may extend\n beyond the page allocated. This is because the page allocation is\n deferred to write_begin() and that gets to decide what size of THP to\n allocate.\"\n\nJeff Layton points out:\n \"The netfs fix in particular fixes a data corruption bug in cephfs\"\n\n* tag \u0027netfs-fixes-20210621\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs:\n netfs: fix test for whether we can skip read when writing beyond EOF\n afs: Fix afs_write_end() to handle short writes\n" }, { "commit": "c13e3021331ed7736996fe61d6f26983ac3b84cc", "tree": "b993660f580fd8ee8207bd2a2eb237f1bc99c6d5", "parents": [ "e41fc7c8e275ddb292556698c7b60a1bf1199920", "c6414e1a2bd26b0071e2b9d6034621f705dfd4c0" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 09:32:57 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 09:32:57 2021 -0700" }, "message": "Merge tag \u0027gpio-fixes-for-v5.13\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux\n\nPull gpio fixes from Bartosz Golaszewski:\n\n - fix wake-up interrupt support on gpio-mxc\n\n - zero the padding bytes in a structure passed to user-space in the\n GPIO character device\n\n - require HAS_IOPORT_MAP in two drivers that need it to fix a Kbuild\n issue\n\n* tag \u0027gpio-fixes-for-v5.13\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux:\n gpio: AMD8111 and TQMX86 require HAS_IOPORT_MAP\n gpiolib: cdev: zero padding during conversion to gpioline_info_changed\n gpio: mxc: Fix disabled interrupt wake-up support\n" }, { "commit": "e41fc7c8e275ddb292556698c7b60a1bf1199920", "tree": "c0b2cb36762f596d3ccdeda0ec266d2d1c5a09ad", "parents": [ "44db63d1ad8d71c6932cbe007eb41f31c434d140", "5c6d4f97267f02f47acea8a652265348ec12de51" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 09:20:22 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Jun 25 09:20:22 2021 -0700" }, "message": "Merge tag \u0027sound-5.13-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound\n\nPull sound fixes from Takashi Iwai:\n \"Two small changes have been cherry-picked as a last material for 5.13:\n a coverage after UMN revert action and a stale MAINTAINERS entry fix\"\n\n* tag \u0027sound-5.13-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:\n MAINTAINERS: remove Timur Tabi from Freescale SOC sound drivers\n ASoC: rt5645: Avoid upgrading static warnings to errors\n" }, { "commit": "c6414e1a2bd26b0071e2b9d6034621f705dfd4c0", "tree": "c72ced9e7f88b30a93205268f7095129b6647a22", "parents": [ "cb8f63b8cbf39845244f3ccae43bb7e63bd70543" ], "author": { "name": "Johannes Berg", "email": "johannes.berg@intel.com", "time": "Fri Jun 25 10:37:34 2021 +0200" }, "committer": { "name": "Bartosz Golaszewski", "email": "bgolaszewski@baylibre.com", "time": "Fri Jun 25 12:13:53 2021 +0200" }, "message": "gpio: AMD8111 and TQMX86 require HAS_IOPORT_MAP\n\nBoth of these drivers use ioport_map(), so they need to\ndepend on HAS_IOPORT_MAP. Otherwise, they cannot be built\neven with COMPILE_TEST on architectures without an ioport\nimplementation, such as ARCH\u003dum.\n\nReported-by: kernel test robot \u003clkp@intel.com\u003e\nSigned-off-by: Johannes Berg \u003cjohannes.berg@intel.com\u003e\nSigned-off-by: Bartosz Golaszewski \u003cbgolaszewski@baylibre.com\u003e\n" }, { "commit": "72a461adbe88acf6a8cc5dba7720cf94d7056154", "tree": "8d1e53ca1d9bb5d34fab9ac0ad88f8f030a8fbe0", "parents": [ "ee924d3ddd4561b7e6671bd431ff55bb9a24c47c" ], "author": { "name": "Marek Behún", "email": "kabel@kernel.org", "time": "Thu Jun 24 18:40:13 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:54 2021 -0700" }, "message": "mailmap: add Marek\u0027s other e-mail address and identity without diacritics\n\nSome of my commits were sent with identities\n Marek Behun \u003cmarek.behun@nic.cz\u003e\n Marek Behún \u003cmarek.behun@nic.cz\u003e\nwhile the correct one is\n Marek Behún \u003ckabel@kernel.org\u003e\n\nPut this into mailmap so that git shortlog prints all my commits under\none identity.\n\nLink: https://lkml.kernel.org/r/20210616113624.19351-2-kabel@kernel.org\nSigned-off-by: Marek Behún \u003ckabel@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": "ee924d3ddd4561b7e6671bd431ff55bb9a24c47c", "tree": "652c3eec53f6b262201d5e635fc20ed0bc6e4d28", "parents": [ "b3b64ebd38225d8032b5db42938d969b602040c2" ], "author": { "name": "Marek Behún", "email": "kabel@kernel.org", "time": "Thu Jun 24 18:40:10 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:54 2021 -0700" }, "message": "MAINTAINERS: fix Marek\u0027s identity again\n\nFix my name to use diacritics, since MAINTAINERS supports it.\n\nFix my e-mail address in MAINTAINERS\u0027 marvell10g PHY driver description,\nI accidentally put my other e-mail address here.\n\nLink: https://lkml.kernel.org/r/20210616113624.19351-1-kabel@kernel.org\nSigned-off-by: Marek Behún \u003ckabel@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": "b3b64ebd38225d8032b5db42938d969b602040c2", "tree": "c5002be80ee46f6ee3e49e1c34a3ed5c3dc2458a", "parents": [ "b08e50dd64489e3997029d204f761cb57a3762d2" ], "author": { "name": "Mel Gorman", "email": "mgorman@techsingularity.net", "time": "Thu Jun 24 18:40:07 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:54 2021 -0700" }, "message": "mm/page_alloc: do bulk array bounds check after checking populated elements\n\nDan Carpenter reported the following\n\n The patch 0f87d9d30f21: \"mm/page_alloc: add an array-based interface\n to the bulk page allocator\" from Apr 29, 2021, leads to the following\n static checker warning:\n\n mm/page_alloc.c:5338 __alloc_pages_bulk()\n warn: potentially one past the end of array \u0027page_array[nr_populated]\u0027\n\nThe problem can occur if an array is passed in that is fully populated.\nThat potentially ends up allocating a single page and storing it past\nthe end of the array. This patch returns 0 if the array is fully\npopulated.\n\nLink: https://lkml.kernel.org/r/20210618125102.GU30378@techsingularity.net\nFixes: 0f87d9d30f21 (\"mm/page_alloc: add an array-based interface to the bulk page allocator\")\nSigned-off-by: Mel Gorman \u003cmgorman@techsinguliarity.net\u003e\nReported-by: Dan Carpenter \u003cdan.carpenter@oracle.com\u003e\nCc: Jesper Dangaard Brouer \u003cbrouer@redhat.com\u003e\nCc: Vlastimil Babka \u003cvbabka@suse.cz\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "b08e50dd64489e3997029d204f761cb57a3762d2", "tree": "2e3cb23ccd9b949e687584e7c3f581ca06feb318", "parents": [ "ea6d0630100b285f059d0a8d8e86f38a46407536" ], "author": { "name": "Rasmus Villemoes", "email": "linux@rasmusvillemoes.dk", "time": "Thu Jun 24 18:40:04 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:54 2021 -0700" }, "message": "mm/page_alloc: __alloc_pages_bulk(): do bounds check before accessing array\n\nIn the event that somebody would call this with an already fully\npopulated page_array, the last loop iteration would do an access beyond\nthe end of page_array.\n\nIt\u0027s of course extremely unlikely that would ever be done, but this\ntriggers my internal static analyzer. Also, if it really is not\nsupposed to be invoked this way (i.e., with no NULL entries in\npage_array), the nr_populated\u003cnr_pages check could simply be removed\ninstead.\n\nLink: https://lkml.kernel.org/r/20210507064504.1712559-1-linux@rasmusvillemoes.dk\nFixes: 0f87d9d30f21 (\"mm/page_alloc: add an array-based interface to the bulk page allocator\")\nSigned-off-by: Rasmus Villemoes \u003clinux@rasmusvillemoes.dk\u003e\nAcked-by: Mel Gorman \u003cmgorman@techsingularity.net\u003e\nReviewed-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "ea6d0630100b285f059d0a8d8e86f38a46407536", "tree": "d7acb0884336336bb3d0031c00f7a43a716bce1b", "parents": [ "47af12bae17f99b5e77f8651cb7f3e1877610acf" ], "author": { "name": "Naoya Horiguchi", "email": "naoya.horiguchi@nec.com", "time": "Thu Jun 24 18:40:01 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:54 2021 -0700" }, "message": "mm/hwpoison: do not lock page again when me_huge_page() successfully recovers\n\nCurrently me_huge_page() temporary unlocks page to perform some actions\nthen locks it again later. My testcase (which calls hard-offline on\nsome tail page in a hugetlb, then accesses the address of the hugetlb\nrange) showed that page allocation code detects this page lock on buddy\npage and printed out \"BUG: Bad page state\" message.\n\ncheck_new_page_bad() does not consider a page with __PG_HWPOISON as bad\npage, so this flag works as kind of filter, but this filtering doesn\u0027t\nwork in this case because the \"bad page\" is not the actual hwpoisoned\npage. So stop locking page again. Actions to be taken depend on the\npage type of the error, so page unlocking should be done in -\u003eaction()\ncallbacks. So let\u0027s make it assumed and change all existing callbacks\nthat way.\n\nLink: https://lkml.kernel.org/r/20210609072029.74645-1-nao.horiguchi@gmail.com\nFixes: commit 78bb920344b8 (\"mm: hwpoison: dissolve in-use hugepage in unrecoverable memory error\")\nSigned-off-by: Naoya Horiguchi \u003cnaoya.horiguchi@nec.com\u003e\nCc: Oscar Salvador \u003cosalvador@suse.de\u003e\nCc: Michal Hocko \u003cmhocko@suse.com\u003e\nCc: Tony Luck \u003ctony.luck@intel.com\u003e\nCc: \"Aneesh Kumar K.V\" \u003caneesh.kumar@linux.vnet.ibm.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": "47af12bae17f99b5e77f8651cb7f3e1877610acf", "tree": "1761a4b5ad4d3b07822e5b75972bcbcae67162e8", "parents": [ "171936ddaf97e6f4e1264f4128bb5cf15691339c" ], "author": { "name": "Aili Yao", "email": "yaoaili@kingsoft.com", "time": "Thu Jun 24 18:39:58 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:54 2021 -0700" }, "message": "mm,hwpoison: return -EHWPOISON to denote that the page has already been poisoned\n\nWhen memory_failure() is called with MF_ACTION_REQUIRED on the page that\nhas already been hwpoisoned, memory_failure() could fail to send SIGBUS\nto the affected process, which results in infinite loop of MCEs.\n\nCurrently memory_failure() returns 0 if it\u0027s called for already\nhwpoisoned page, then the caller, kill_me_maybe(), could return without\nsending SIGBUS to current process. An action required MCE is raised\nwhen the current process accesses to the broken memory, so no SIGBUS\nmeans that the current process continues to run and access to the error\npage again soon, so running into MCE loop.\n\nThis issue can arise for example in the following scenarios:\n\n - Two or more threads access to the poisoned page concurrently. If\n local MCE is enabled, MCE handler independently handles the MCE\n events. So there\u0027s a race among MCE events, and the second or latter\n threads fall into the situation in question.\n\n - If there was a precedent memory error event and memory_failure() for\n the event failed to unmap the error page for some reason, the\n subsequent memory access to the error page triggers the MCE loop\n situation.\n\nTo fix the issue, make memory_failure() return an error code when the\nerror page has already been hwpoisoned. This allows memory error\nhandler to control how it sends signals to userspace. And make sure\nthat any process touching a hwpoisoned page should get a SIGBUS even in\n\"already hwpoisoned\" path of memory_failure() as is done in page fault\npath.\n\nLink: https://lkml.kernel.org/r/20210521030156.2612074-3-nao.horiguchi@gmail.com\nSigned-off-by: Aili Yao \u003cyaoaili@kingsoft.com\u003e\nSigned-off-by: Naoya Horiguchi \u003cnaoya.horiguchi@nec.com\u003e\nReviewed-by: Oscar Salvador \u003cosalvador@suse.de\u003e\nCc: Andy Lutomirski \u003cluto@kernel.org\u003e\nCc: Borislav Petkov \u003cbp@alien8.de\u003e\nCc: Borislav Petkov \u003cbp@suse.de\u003e\nCc: David Hildenbrand \u003cdavid@redhat.com\u003e\nCc: Jue Wang \u003cjuew@google.com\u003e\nCc: Tony Luck \u003ctony.luck@intel.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": "171936ddaf97e6f4e1264f4128bb5cf15691339c", "tree": "e0bb11aba675fa17176613cf9ba82f0ffff9a630", "parents": [ "fe19bd3dae3d15d2fbfdb3de8839a6ea0fe94264" ], "author": { "name": "Tony Luck", "email": "tony.luck@intel.com", "time": "Thu Jun 24 18:39:55 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:54 2021 -0700" }, "message": "mm/memory-failure: use a mutex to avoid memory_failure() races\n\nPatch series \"mm,hwpoison: fix sending SIGBUS for Action Required MCE\", v5.\n\nI wrote this patchset to materialize what I think is the current\nallowable solution mentioned by the previous discussion [1]. I simply\nborrowed Tony\u0027s mutex patch and Aili\u0027s return code patch, then I queued\nanother one to find error virtual address in the best effort manner. I\nknow that this is not a perfect solution, but should work for some\ntypical case.\n\n[1]: https://lore.kernel.org/linux-mm/20210331192540.2141052f@alex-virtual-machine/\n\nThis patch (of 2):\n\nThere can be races when multiple CPUs consume poison from the same page.\nThe first into memory_failure() atomically sets the HWPoison page flag\nand begins hunting for tasks that map this page. Eventually it\ninvalidates those mappings and may send a SIGBUS to the affected tasks.\n\nBut while all that work is going on, other CPUs see a \"success\" return\ncode from memory_failure() and so they believe the error has been\nhandled and continue executing.\n\nFix by wrapping most of the internal parts of memory_failure() in a\nmutex.\n\n[akpm@linux-foundation.org: make mf_mutex local to memory_failure()]\n\nLink: https://lkml.kernel.org/r/20210521030156.2612074-1-nao.horiguchi@gmail.com\nLink: https://lkml.kernel.org/r/20210521030156.2612074-2-nao.horiguchi@gmail.com\nSigned-off-by: Tony Luck \u003ctony.luck@intel.com\u003e\nSigned-off-by: Naoya Horiguchi \u003cnaoya.horiguchi@nec.com\u003e\nReviewed-by: Borislav Petkov \u003cbp@suse.de\u003e\nReviewed-by: Oscar Salvador \u003cosalvador@suse.de\u003e\nCc: Aili Yao \u003cyaoaili@kingsoft.com\u003e\nCc: Andy Lutomirski \u003cluto@kernel.org\u003e\nCc: Borislav Petkov \u003cbp@alien8.de\u003e\nCc: David Hildenbrand \u003cdavid@redhat.com\u003e\nCc: Jue Wang \u003cjuew@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": "fe19bd3dae3d15d2fbfdb3de8839a6ea0fe94264", "tree": "fab0cfb0af57acaa2ece48ca74d5cf2964b10862", "parents": [ "5fa54346caf67b4b1b10b1f390316ae466da4d53" ], "author": { "name": "Hugh Dickins", "email": "hughd@google.com", "time": "Thu Jun 24 18:39:52 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:54 2021 -0700" }, "message": "mm, futex: fix shared futex pgoff on shmem huge page\n\nIf more than one futex is placed on a shmem huge page, it can happen\nthat waking the second wakes the first instead, and leaves the second\nwaiting: the key\u0027s shared.pgoff is wrong.\n\nWhen 3.11 commit 13d60f4b6ab5 (\"futex: Take hugepages into account when\ngenerating futex_key\"), the only shared huge pages came from hugetlbfs,\nand the code added to deal with its exceptional page-\u003eindex was put into\nhugetlb source. Then that was missed when 4.8 added shmem huge pages.\n\npage_to_pgoff() is what others use for this nowadays: except that, as\ncurrently written, it gives the right answer on hugetlbfs head, but\nnonsense on hugetlbfs tails. Fix that by calling hugetlbfs-specific\nhugetlb_basepage_index() on PageHuge tails as well as on head.\n\nYes, it\u0027s unconventional to declare hugetlb_basepage_index() there in\npagemap.h, rather than in hugetlb.h; but I do not expect anything but\npage_to_pgoff() ever to need it.\n\n[akpm@linux-foundation.org: give hugetlb_basepage_index() prototype the correct scope]\n\nLink: https://lkml.kernel.org/r/b17d946b-d09-326e-b42a-52884c36df32@google.com\nFixes: 800d8c63b2e9 (\"shmem: add huge pages support\")\nReported-by: Neel Natu \u003cneelnatu@google.com\u003e\nSigned-off-by: Hugh Dickins \u003chughd@google.com\u003e\nReviewed-by: Matthew Wilcox (Oracle) \u003cwilly@infradead.org\u003e\nAcked-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nCc: \"Kirill A. Shutemov\" \u003ckirill.shutemov@linux.intel.com\u003e\nCc: Zhang Yi \u003cwetpzy@gmail.com\u003e\nCc: Mel Gorman \u003cmgorman@techsingularity.net\u003e\nCc: Mike Kravetz \u003cmike.kravetz@oracle.com\u003e\nCc: Ingo Molnar \u003cmingo@redhat.com\u003e\nCc: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nCc: Darren Hart \u003cdvhart@infradead.org\u003e\nCc: Davidlohr Bueso \u003cdave@stgolabs.net\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": "5fa54346caf67b4b1b10b1f390316ae466da4d53", "tree": "46795b4f8028c01ee808f2b96f08e4a6cbe89bda", "parents": [ "34b3d5344719d14fd2185b2d9459b3abcb8cf9d8" ], "author": { "name": "Petr Mladek", "email": "pmladek@suse.com", "time": "Thu Jun 24 18:39:48 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:54 2021 -0700" }, "message": "kthread: prevent deadlock when kthread_mod_delayed_work() races with kthread_cancel_delayed_work_sync()\n\nThe system might hang with the following backtrace:\n\n\tschedule+0x80/0x100\n\tschedule_timeout+0x48/0x138\n\twait_for_common+0xa4/0x134\n\twait_for_completion+0x1c/0x2c\n\tkthread_flush_work+0x114/0x1cc\n\tkthread_cancel_work_sync.llvm.16514401384283632983+0xe8/0x144\n\tkthread_cancel_delayed_work_sync+0x18/0x2c\n\txxxx_pm_notify+0xb0/0xd8\n\tblocking_notifier_call_chain_robust+0x80/0x194\n\tpm_notifier_call_chain_robust+0x28/0x4c\n\tsuspend_prepare+0x40/0x260\n\tenter_state+0x80/0x3f4\n\tpm_suspend+0x60/0xdc\n\tstate_store+0x108/0x144\n\tkobj_attr_store+0x38/0x88\n\tsysfs_kf_write+0x64/0xc0\n\tkernfs_fop_write_iter+0x108/0x1d0\n\tvfs_write+0x2f4/0x368\n\tksys_write+0x7c/0xec\n\nIt is caused by the following race between kthread_mod_delayed_work()\nand kthread_cancel_delayed_work_sync():\n\nCPU0\t\t\t\tCPU1\n\nContext: Thread A\t\tContext: Thread B\n\nkthread_mod_delayed_work()\n spin_lock()\n __kthread_cancel_work()\n spin_unlock()\n del_timer_sync()\n\t\t\t\tkthread_cancel_delayed_work_sync()\n\t\t\t\t spin_lock()\n\t\t\t\t __kthread_cancel_work()\n\t\t\t\t spin_unlock()\n\t\t\t\t del_timer_sync()\n\t\t\t\t spin_lock()\n\n\t\t\t\t work-\u003ecanceling++\n\t\t\t\t spin_unlock\n spin_lock()\n queue_delayed_work()\n // dwork is put into the worker-\u003edelayed_work_list\n\n spin_unlock()\n\n\t\t\t\t kthread_flush_work()\n // flush_work is put at the tail of the dwork\n\n\t\t\t\t wait_for_completion()\n\nContext: IRQ\n\n kthread_delayed_work_timer_fn()\n spin_lock()\n list_del_init(\u0026work-\u003enode);\n spin_unlock()\n\nBANG: flush_work is not longer linked and will never get proceed.\n\nThe problem is that kthread_mod_delayed_work() checks work-\u003ecanceling\nflag before canceling the timer.\n\nA simple solution is to (re)check work-\u003ecanceling after\n__kthread_cancel_work(). But then it is not clear what should be\nreturned when __kthread_cancel_work() removed the work from the queue\n(list) and it can\u0027t queue it again with the new @delay.\n\nThe return value might be used for reference counting. The caller has\nto know whether a new work has been queued or an existing one was\nreplaced.\n\nThe proper solution is that kthread_mod_delayed_work() will remove the\nwork from the queue (list) _only_ when work-\u003ecanceling is not set. The\nflag must be checked after the timer is stopped and the remaining\noperations can be done under worker-\u003elock.\n\nNote that kthread_mod_delayed_work() could remove the timer and then\nbail out. It is fine. The other canceling caller needs to cancel the\ntimer as well. The important thing is that the queue (list)\nmanipulation is done atomically under worker-\u003elock.\n\nLink: https://lkml.kernel.org/r/20210610133051.15337-3-pmladek@suse.com\nFixes: 9a6b06c8d9a220860468a (\"kthread: allow to modify delayed kthread work\")\nSigned-off-by: Petr Mladek \u003cpmladek@suse.com\u003e\nReported-by: Martin Liu \u003cliumartin@google.com\u003e\nCc: \u003cjenhaochen@google.com\u003e\nCc: Minchan Kim \u003cminchan@google.com\u003e\nCc: Nathan Chancellor \u003cnathan@kernel.org\u003e\nCc: Nick Desaulniers \u003cndesaulniers@google.com\u003e\nCc: Oleg Nesterov \u003coleg@redhat.com\u003e\nCc: Tejun Heo \u003ctj@kernel.org\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": "34b3d5344719d14fd2185b2d9459b3abcb8cf9d8", "tree": "db3a54c835903c8ad922472a6e0a901bf5d3a742", "parents": [ "7ca3027b726be681c8e6292b5a81ebcde7581710" ], "author": { "name": "Petr Mladek", "email": "pmladek@suse.com", "time": "Thu Jun 24 18:39:45 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:54 2021 -0700" }, "message": "kthread_worker: split code for canceling the delayed work timer\n\nPatch series \"kthread_worker: Fix race between kthread_mod_delayed_work()\nand kthread_cancel_delayed_work_sync()\".\n\nThis patchset fixes the race between kthread_mod_delayed_work() and\nkthread_cancel_delayed_work_sync() including proper return value\nhandling.\n\nThis patch (of 2):\n\nSimple code refactoring as a preparation step for fixing a race between\nkthread_mod_delayed_work() and kthread_cancel_delayed_work_sync().\n\nIt does not modify the existing behavior.\n\nLink: https://lkml.kernel.org/r/20210610133051.15337-2-pmladek@suse.com\nSigned-off-by: Petr Mladek \u003cpmladek@suse.com\u003e\nCc: \u003cjenhaochen@google.com\u003e\nCc: Martin Liu \u003cliumartin@google.com\u003e\nCc: Minchan Kim \u003cminchan@google.com\u003e\nCc: Nathan Chancellor \u003cnathan@kernel.org\u003e\nCc: Nick Desaulniers \u003cndesaulniers@google.com\u003e\nCc: Oleg Nesterov \u003coleg@redhat.com\u003e\nCc: Tejun Heo \u003ctj@kernel.org\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": "7ca3027b726be681c8e6292b5a81ebcde7581710", "tree": "4834aca0a45837341108be0fd034cc4711173264", "parents": [ "185cca24e977411495d57ec71e43350b69c08e63" ], "author": { "name": "Daniel Axtens", "email": "dja@axtens.net", "time": "Thu Jun 24 18:39:42 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:54 2021 -0700" }, "message": "mm/vmalloc: unbreak kasan vmalloc support\n\nIn commit 121e6f3258fe (\"mm/vmalloc: hugepage vmalloc mappings\"),\n__vmalloc_node_range was changed such that __get_vm_area_node was no\nlonger called with the requested/real size of the vmalloc allocation,\nbut rather with a rounded-up size.\n\nThis means that __get_vm_area_node called kasan_unpoision_vmalloc() with\na rounded up size rather than the real size. This led to it allowing\naccess to too much memory and so missing vmalloc OOBs and failing the\nkasan kunit tests.\n\nPass the real size and the desired shift into __get_vm_area_node. This\nallows it to round up the size for the underlying allocators while still\nunpoisioning the correct quantity of shadow memory.\n\nAdjust the other call-sites to pass in PAGE_SHIFT for the shift value.\n\nLink: https://lkml.kernel.org/r/20210617081330.98629-1-dja@axtens.net\nLink: https://bugzilla.kernel.org/show_bug.cgi?id\u003d213335\nFixes: 121e6f3258fe (\"mm/vmalloc: hugepage vmalloc mappings\")\nSigned-off-by: Daniel Axtens \u003cdja@axtens.net\u003e\nTested-by: David Gow \u003cdavidgow@google.com\u003e\nReviewed-by: Nicholas Piggin \u003cnpiggin@gmail.com\u003e\nReviewed-by: Uladzislau Rezki (Sony) \u003curezki@gmail.com\u003e\nTested-by: Andrey Konovalov \u003candreyknvl@gmail.com\u003e\nAcked-by: Andrey Konovalov \u003candreyknvl@gmail.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": "185cca24e977411495d57ec71e43350b69c08e63", "tree": "c5f48289b08a62cc6e4e51b51a2e27ad9f3e6dfb", "parents": [ "15a64f5a8870b5610b616a4aa753262dfaa5d76e" ], "author": { "name": "Claudio Imbrenda", "email": "imbrenda@linux.ibm.com", "time": "Thu Jun 24 18:39:39 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:54 2021 -0700" }, "message": "KVM: s390: prepare for hugepage vmalloc\n\nThe Create Secure Configuration Ultravisor Call does not support using\nlarge pages for the virtual memory area. This is a hardware limitation.\n\nThis patch replaces the vzalloc call with an almost equivalent call to\nthe newly introduced vmalloc_no_huge function, which guarantees that\nonly small pages will be used for the backing.\n\nThe new call will not clear the allocated memory, but that has never\nbeen an actual requirement.\n\nLink: https://lkml.kernel.org/r/20210614132357.10202-3-imbrenda@linux.ibm.com\nFixes: 121e6f3258fe3 (\"mm/vmalloc: hugepage vmalloc mappings\")\nSigned-off-by: Claudio Imbrenda \u003cimbrenda@linux.ibm.com\u003e\nReviewed-by: Janosch Frank \u003cfrankja@linux.ibm.com\u003e\nAcked-by: Christian Borntraeger \u003cborntraeger@de.ibm.com\u003e\nAcked-by: Nicholas Piggin \u003cnpiggin@gmail.com\u003e\nReviewed-by: David Hildenbrand \u003cdavid@redhat.com\u003e\nCc: Nicholas Piggin \u003cnpiggin@gmail.com\u003e\nCc: Uladzislau Rezki (Sony) \u003curezki@gmail.com\u003e\nCc: Catalin Marinas \u003ccatalin.marinas@arm.com\u003e\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nCc: Ingo Molnar \u003cmingo@redhat.com\u003e\nCc: David Rientjes \u003crientjes@google.com\u003e\nCc: Christoph Hellwig \u003chch@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": "15a64f5a8870b5610b616a4aa753262dfaa5d76e", "tree": "bc3b8e5b0d343720387a9ddbfd19095d1b265cdd", "parents": [ "8fd0c1b0647a6bda4067ee0cd61e8395954b6f28" ], "author": { "name": "Claudio Imbrenda", "email": "imbrenda@linux.ibm.com", "time": "Thu Jun 24 18:39:36 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:53 2021 -0700" }, "message": "mm/vmalloc: add vmalloc_no_huge\n\nPatch series \"mm: add vmalloc_no_huge and use it\", v4.\n\nAdd vmalloc_no_huge() and export it, so modules can allocate memory with\nsmall pages.\n\nUse the newly added vmalloc_no_huge() in KVM on s390 to get around a\nhardware limitation.\n\nThis patch (of 2):\n\nCommit 121e6f3258fe3 (\"mm/vmalloc: hugepage vmalloc mappings\") added\nsupport for hugepage vmalloc mappings, it also added the flag\nVM_NO_HUGE_VMAP for __vmalloc_node_range to request the allocation to be\nperformed with 0-order non-huge pages.\n\nThis flag is not accessible when calling vmalloc, the only option is to\ncall directly __vmalloc_node_range, which is not exported.\n\nThis means that a module can\u0027t vmalloc memory with small pages.\n\nCase in point: KVM on s390x needs to vmalloc a large area, and it needs\nto be mapped with non-huge pages, because of a hardware limitation.\n\nThis patch adds the function vmalloc_no_huge, which works like vmalloc,\nbut it is guaranteed to always back the mapping using small pages. This\nnew function is exported, therefore it is usable by modules.\n\n[akpm@linux-foundation.org: whitespace fixes, per Christoph]\n\nLink: https://lkml.kernel.org/r/20210614132357.10202-1-imbrenda@linux.ibm.com\nLink: https://lkml.kernel.org/r/20210614132357.10202-2-imbrenda@linux.ibm.com\nFixes: 121e6f3258fe3 (\"mm/vmalloc: hugepage vmalloc mappings\")\nSigned-off-by: Claudio Imbrenda \u003cimbrenda@linux.ibm.com\u003e\nReviewed-by: Uladzislau Rezki (Sony) \u003curezki@gmail.com\u003e\nAcked-by: Nicholas Piggin \u003cnpiggin@gmail.com\u003e\nReviewed-by: David Hildenbrand \u003cdavid@redhat.com\u003e\nAcked-by: David Rientjes \u003crientjes@google.com\u003e\nCc: Uladzislau Rezki (Sony) \u003curezki@gmail.com\u003e\nCc: Catalin Marinas \u003ccatalin.marinas@arm.com\u003e\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nCc: Ingo Molnar \u003cmingo@redhat.com\u003e\nCc: Christoph Hellwig \u003chch@infradead.org\u003e\nCc: Cornelia Huck \u003ccohuck@redhat.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "8fd0c1b0647a6bda4067ee0cd61e8395954b6f28", "tree": "353a4b6404639449297a2af806b5cfd90b3ba8b1", "parents": [ "a7a69d8ba88d8dcee7ef00e91d413a4bd003a814" ], "author": { "name": "Pavel Skripkin", "email": "paskripkin@gmail.com", "time": "Thu Jun 24 18:39:33 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:53 2021 -0700" }, "message": "nilfs2: fix memory leak in nilfs_sysfs_delete_device_group\n\nMy local syzbot instance hit memory leak in nilfs2. The problem was in\nmissing kobject_put() in nilfs_sysfs_delete_device_group().\n\nkobject_del() does not call kobject_cleanup() for passed kobject and it\nleads to leaking duped kobject name if kobject_put() was not called.\n\nFail log:\n\n BUG: memory leak\n unreferenced object 0xffff8880596171e0 (size 8):\n comm \"syz-executor379\", pid 8381, jiffies 4294980258 (age 21.100s)\n hex dump (first 8 bytes):\n 6c 6f 6f 70 30 00 00 00 loop0...\n backtrace:\n kstrdup+0x36/0x70 mm/util.c:60\n kstrdup_const+0x53/0x80 mm/util.c:83\n kvasprintf_const+0x108/0x190 lib/kasprintf.c:48\n kobject_set_name_vargs+0x56/0x150 lib/kobject.c:289\n kobject_add_varg lib/kobject.c:384 [inline]\n kobject_init_and_add+0xc9/0x160 lib/kobject.c:473\n nilfs_sysfs_create_device_group+0x150/0x800 fs/nilfs2/sysfs.c:999\n init_nilfs+0xe26/0x12b0 fs/nilfs2/the_nilfs.c:637\n\nLink: https://lkml.kernel.org/r/20210612140559.20022-1-paskripkin@gmail.com\nFixes: da7141fb78db (\"nilfs2: add /sys/fs/nilfs2/\u003cdevice\u003e group\")\nSigned-off-by: Pavel Skripkin \u003cpaskripkin@gmail.com\u003e\nAcked-by: Ryusuke Konishi \u003ckonishi.ryusuke@gmail.com\u003e\nCc: Michael L. Semon \u003cmlsemon35@gmail.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "a7a69d8ba88d8dcee7ef00e91d413a4bd003a814", "tree": "31e5fe7533a654e098d37c62bd508bca14dc49af", "parents": [ "a9a7504d9beaf395481faa91e70e2fd08f7a3dde" ], "author": { "name": "Hugh Dickins", "email": "hughd@google.com", "time": "Thu Jun 24 18:39:30 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:53 2021 -0700" }, "message": "mm/thp: another PVMW_SYNC fix in page_vma_mapped_walk()\n\nAha! Shouldn\u0027t that quick scan over pte_none()s make sure that it holds\nptlock in the PVMW_SYNC case? That too might have been responsible for\nBUGs or WARNs in split_huge_page_to_list() or its unmap_page(), though\nI\u0027ve never seen any.\n\nLink: https://lkml.kernel.org/r/1bdf384c-8137-a149-2a1e-475a4791c3c@google.com\nLink: https://lore.kernel.org/linux-mm/20210412180659.B9E3.409509F4@e16-tech.com/\nFixes: ace71a19cec5 (\"mm: introduce page_vma_mapped_walk()\")\nSigned-off-by: Hugh Dickins \u003chughd@google.com\u003e\nAcked-by: Kirill A. Shutemov \u003ckirill.shutemov@linux.intel.com\u003e\nTested-by: Wang Yugui \u003cwangyugui@e16-tech.com\u003e\nCc: Alistair Popple \u003capopple@nvidia.com\u003e\nCc: Matthew Wilcox \u003cwilly@infradead.org\u003e\nCc: Peter Xu \u003cpeterx@redhat.com\u003e\nCc: Ralph Campbell \u003crcampbell@nvidia.com\u003e\nCc: Will Deacon \u003cwill@kernel.org\u003e\nCc: Yang Shi \u003cshy828301@gmail.com\u003e\nCc: Zi Yan \u003cziy@nvidia.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": "a9a7504d9beaf395481faa91e70e2fd08f7a3dde", "tree": "02b96e9bf0891ded5ff4592f097b76c2f01ac05f", "parents": [ "a765c417d876cc635f628365ec9aa6f09470069a" ], "author": { "name": "Hugh Dickins", "email": "hughd@google.com", "time": "Thu Jun 24 18:39:26 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:53 2021 -0700" }, "message": "mm/thp: fix page_vma_mapped_walk() if THP mapped by ptes\n\nRunning certain tests with a DEBUG_VM kernel would crash within hours,\non the total_mapcount BUG() in split_huge_page_to_list(), while trying\nto free up some memory by punching a hole in a shmem huge page: split\u0027s\ntry_to_unmap() was unable to find all the mappings of the page (which,\non a !DEBUG_VM kernel, would then keep the huge page pinned in memory).\n\nCrash dumps showed two tail pages of a shmem huge page remained mapped\nby pte: ptes in a non-huge-aligned vma of a gVisor process, at the end\nof a long unmapped range; and no page table had yet been allocated for\nthe head of the huge page to be mapped into.\n\nAlthough designed to handle these odd misaligned huge-page-mapped-by-pte\ncases, page_vma_mapped_walk() falls short by returning false prematurely\nwhen !pmd_present or !pud_present or !p4d_present or !pgd_present: there\nare cases when a huge page may span the boundary, with ptes present in\nthe next.\n\nRestructure page_vma_mapped_walk() as a loop to continue in these cases,\nwhile keeping its layout much as before. Add a step_forward() helper to\nadvance pvmw-\u003eaddress across those boundaries: originally I tried to use\nmm\u0027s standard p?d_addr_end() macros, but hit the same crash 512 times\nless often: because of the way redundant levels are folded together, but\nfolded differently in different configurations, it was just too\ndifficult to use them correctly; and step_forward() is simpler anyway.\n\nLink: https://lkml.kernel.org/r/fedb8632-1798-de42-f39e-873551d5bc81@google.com\nFixes: ace71a19cec5 (\"mm: introduce page_vma_mapped_walk()\")\nSigned-off-by: Hugh Dickins \u003chughd@google.com\u003e\nAcked-by: Kirill A. Shutemov \u003ckirill.shutemov@linux.intel.com\u003e\nCc: Alistair Popple \u003capopple@nvidia.com\u003e\nCc: Matthew Wilcox \u003cwilly@infradead.org\u003e\nCc: Peter Xu \u003cpeterx@redhat.com\u003e\nCc: Ralph Campbell \u003crcampbell@nvidia.com\u003e\nCc: Wang Yugui \u003cwangyugui@e16-tech.com\u003e\nCc: Will Deacon \u003cwill@kernel.org\u003e\nCc: Yang Shi \u003cshy828301@gmail.com\u003e\nCc: Zi Yan \u003cziy@nvidia.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": "a765c417d876cc635f628365ec9aa6f09470069a", "tree": "7b213a4aac5f5b8c8c5b80189007049054764586", "parents": [ "474466301dfd8b39a10c01db740645f3f7ae9a28" ], "author": { "name": "Hugh Dickins", "email": "hughd@google.com", "time": "Thu Jun 24 18:39:23 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:53 2021 -0700" }, "message": "mm: page_vma_mapped_walk(): get vma_address_end() earlier\n\npage_vma_mapped_walk() cleanup: get THP\u0027s vma_address_end() at the\nstart, rather than later at next_pte.\n\nIt\u0027s a little unnecessary overhead on the first call, but makes for a\nsimpler loop in the following commit.\n\nLink: https://lkml.kernel.org/r/4542b34d-862f-7cb4-bb22-e0df6ce830a2@google.com\nSigned-off-by: Hugh Dickins \u003chughd@google.com\u003e\nAcked-by: Kirill A. Shutemov \u003ckirill.shutemov@linux.intel.com\u003e\nCc: Alistair Popple \u003capopple@nvidia.com\u003e\nCc: Matthew Wilcox \u003cwilly@infradead.org\u003e\nCc: Peter Xu \u003cpeterx@redhat.com\u003e\nCc: Ralph Campbell \u003crcampbell@nvidia.com\u003e\nCc: Wang Yugui \u003cwangyugui@e16-tech.com\u003e\nCc: Will Deacon \u003cwill@kernel.org\u003e\nCc: Yang Shi \u003cshy828301@gmail.com\u003e\nCc: Zi Yan \u003cziy@nvidia.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": "474466301dfd8b39a10c01db740645f3f7ae9a28", "tree": "a10d26aa505e881bda6ef59d84c05393ec2e3da3", "parents": [ "b3807a91aca7d21c05d5790612e49969117a72b9" ], "author": { "name": "Hugh Dickins", "email": "hughd@google.com", "time": "Thu Jun 24 18:39:20 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:53 2021 -0700" }, "message": "mm: page_vma_mapped_walk(): use goto instead of while (1)\n\npage_vma_mapped_walk() cleanup: add a label this_pte, matching next_pte,\nand use \"goto this_pte\", in place of the \"while (1)\" loop at the end.\n\nLink: https://lkml.kernel.org/r/a52b234a-851-3616-2525-f42736e8934@google.com\nSigned-off-by: Hugh Dickins \u003chughd@google.com\u003e\nAcked-by: Kirill A. Shutemov \u003ckirill.shutemov@linux.intel.com\u003e\nCc: Alistair Popple \u003capopple@nvidia.com\u003e\nCc: Matthew Wilcox \u003cwilly@infradead.org\u003e\nCc: Peter Xu \u003cpeterx@redhat.com\u003e\nCc: Ralph Campbell \u003crcampbell@nvidia.com\u003e\nCc: Wang Yugui \u003cwangyugui@e16-tech.com\u003e\nCc: Will Deacon \u003cwill@kernel.org\u003e\nCc: Yang Shi \u003cshy828301@gmail.com\u003e\nCc: Zi Yan \u003cziy@nvidia.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": "b3807a91aca7d21c05d5790612e49969117a72b9", "tree": "ba8326f7981d59409b384440a50777cc53ff3c23", "parents": [ "448282487483d6fa5b2eeeafaa0acc681e544a9c" ], "author": { "name": "Hugh Dickins", "email": "hughd@google.com", "time": "Thu Jun 24 18:39:17 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:53 2021 -0700" }, "message": "mm: page_vma_mapped_walk(): add a level of indentation\n\npage_vma_mapped_walk() cleanup: add a level of indentation to much of\nthe body, making no functional change in this commit, but reducing the\nlater diff when this is all converted to a loop.\n\n[hughd@google.com: : page_vma_mapped_walk(): add a level of indentation fix]\n Link: https://lkml.kernel.org/r/7f817555-3ce1-c785-e438-87d8efdcaf26@google.com\n\nLink: https://lkml.kernel.org/r/efde211-f3e2-fe54-977-ef481419e7f3@google.com\nSigned-off-by: Hugh Dickins \u003chughd@google.com\u003e\nAcked-by: Kirill A. Shutemov \u003ckirill.shutemov@linux.intel.com\u003e\nCc: Alistair Popple \u003capopple@nvidia.com\u003e\nCc: Matthew Wilcox \u003cwilly@infradead.org\u003e\nCc: Peter Xu \u003cpeterx@redhat.com\u003e\nCc: Ralph Campbell \u003crcampbell@nvidia.com\u003e\nCc: Wang Yugui \u003cwangyugui@e16-tech.com\u003e\nCc: Will Deacon \u003cwill@kernel.org\u003e\nCc: Yang Shi \u003cshy828301@gmail.com\u003e\nCc: Zi Yan \u003cziy@nvidia.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": "448282487483d6fa5b2eeeafaa0acc681e544a9c", "tree": "23097bf7a0c1de1b5ce8e8035cf11365e78d32c5", "parents": [ "e2e1d4076c77b3671cf8ce702535ae7dee3acf89" ], "author": { "name": "Hugh Dickins", "email": "hughd@google.com", "time": "Thu Jun 24 18:39:14 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:53 2021 -0700" }, "message": "mm: page_vma_mapped_walk(): crossing page table boundary\n\npage_vma_mapped_walk() cleanup: adjust the test for crossing page table\nboundary - I believe pvmw-\u003eaddress is always page-aligned, but nothing\nelse here assumed that; and remember to reset pvmw-\u003epte to NULL after\nunmapping the page table, though I never saw any bug from that.\n\nLink: https://lkml.kernel.org/r/799b3f9c-2a9e-dfef-5d89-26e9f76fd97@google.com\nSigned-off-by: Hugh Dickins \u003chughd@google.com\u003e\nAcked-by: Kirill A. Shutemov \u003ckirill.shutemov@linux.intel.com\u003e\nCc: Alistair Popple \u003capopple@nvidia.com\u003e\nCc: Matthew Wilcox \u003cwilly@infradead.org\u003e\nCc: Peter Xu \u003cpeterx@redhat.com\u003e\nCc: Ralph Campbell \u003crcampbell@nvidia.com\u003e\nCc: Wang Yugui \u003cwangyugui@e16-tech.com\u003e\nCc: Will Deacon \u003cwill@kernel.org\u003e\nCc: Yang Shi \u003cshy828301@gmail.com\u003e\nCc: Zi Yan \u003cziy@nvidia.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": "e2e1d4076c77b3671cf8ce702535ae7dee3acf89", "tree": "a61555cd33ce8d29a3e3f36301acda6a0cac8336", "parents": [ "3306d3119ceacc43ea8b141a73e21fea68eec30c" ], "author": { "name": "Hugh Dickins", "email": "hughd@google.com", "time": "Thu Jun 24 18:39:10 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:53 2021 -0700" }, "message": "mm: page_vma_mapped_walk(): prettify PVMW_MIGRATION block\n\npage_vma_mapped_walk() cleanup: rearrange the !pmd_present() block to\nfollow the same \"return not_found, return not_found, return true\"\npattern as the block above it (note: returning not_found there is never\npremature, since existence or prior existence of huge pmd guarantees\ngood alignment).\n\nLink: https://lkml.kernel.org/r/378c8650-1488-2edf-9647-32a53cf2e21@google.com\nSigned-off-by: Hugh Dickins \u003chughd@google.com\u003e\nAcked-by: Kirill A. Shutemov \u003ckirill.shutemov@linux.intel.com\u003e\nReviewed-by: Peter Xu \u003cpeterx@redhat.com\u003e\nCc: Alistair Popple \u003capopple@nvidia.com\u003e\nCc: Matthew Wilcox \u003cwilly@infradead.org\u003e\nCc: Ralph Campbell \u003crcampbell@nvidia.com\u003e\nCc: Wang Yugui \u003cwangyugui@e16-tech.com\u003e\nCc: Will Deacon \u003cwill@kernel.org\u003e\nCc: Yang Shi \u003cshy828301@gmail.com\u003e\nCc: Zi Yan \u003cziy@nvidia.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": "3306d3119ceacc43ea8b141a73e21fea68eec30c", "tree": "3f3860084e8821788f151bf6baad91c91605b30d", "parents": [ "6d0fd5987657cb0c9756ce684e3a74c0f6351728" ], "author": { "name": "Hugh Dickins", "email": "hughd@google.com", "time": "Thu Jun 24 18:39:07 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:53 2021 -0700" }, "message": "mm: page_vma_mapped_walk(): use pmde for *pvmw-\u003epmd\n\npage_vma_mapped_walk() cleanup: re-evaluate pmde after taking lock, then\nuse it in subsequent tests, instead of repeatedly dereferencing pointer.\n\nLink: https://lkml.kernel.org/r/53fbc9d-891e-46b2-cb4b-468c3b19238e@google.com\nSigned-off-by: Hugh Dickins \u003chughd@google.com\u003e\nAcked-by: Kirill A. Shutemov \u003ckirill.shutemov@linux.intel.com\u003e\nReviewed-by: Peter Xu \u003cpeterx@redhat.com\u003e\nCc: Alistair Popple \u003capopple@nvidia.com\u003e\nCc: Matthew Wilcox \u003cwilly@infradead.org\u003e\nCc: Ralph Campbell \u003crcampbell@nvidia.com\u003e\nCc: Wang Yugui \u003cwangyugui@e16-tech.com\u003e\nCc: Will Deacon \u003cwill@kernel.org\u003e\nCc: Yang Shi \u003cshy828301@gmail.com\u003e\nCc: Zi Yan \u003cziy@nvidia.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": "6d0fd5987657cb0c9756ce684e3a74c0f6351728", "tree": "8e4d1b9e7d8bd6c69d805f5bac1427892d4d9f51", "parents": [ "f003c03bd29e6f46fef1b9a8e8d636ac732286d5" ], "author": { "name": "Hugh Dickins", "email": "hughd@google.com", "time": "Thu Jun 24 18:39:04 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:53 2021 -0700" }, "message": "mm: page_vma_mapped_walk(): settle PageHuge on entry\n\npage_vma_mapped_walk() cleanup: get the hugetlbfs PageHuge case out of\nthe way at the start, so no need to worry about it later.\n\nLink: https://lkml.kernel.org/r/e31a483c-6d73-a6bb-26c5-43c3b880a2@google.com\nSigned-off-by: Hugh Dickins \u003chughd@google.com\u003e\nAcked-by: Kirill A. Shutemov \u003ckirill.shutemov@linux.intel.com\u003e\nReviewed-by: Peter Xu \u003cpeterx@redhat.com\u003e\nCc: Alistair Popple \u003capopple@nvidia.com\u003e\nCc: \"Kirill A. Shutemov\" \u003ckirill.shutemov@linux.intel.com\u003e\nCc: Matthew Wilcox \u003cwilly@infradead.org\u003e\nCc: Ralph Campbell \u003crcampbell@nvidia.com\u003e\nCc: Wang Yugui \u003cwangyugui@e16-tech.com\u003e\nCc: Will Deacon \u003cwill@kernel.org\u003e\nCc: Yang Shi \u003cshy828301@gmail.com\u003e\nCc: Zi Yan \u003cziy@nvidia.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": "f003c03bd29e6f46fef1b9a8e8d636ac732286d5", "tree": "7aa1d51b913ab504aed32b373887382912396da0", "parents": [ "4a09d388f2ab382f217a764e6a152b3f614246f6" ], "author": { "name": "Hugh Dickins", "email": "hughd@google.com", "time": "Thu Jun 24 18:39:01 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 19:40:53 2021 -0700" }, "message": "mm: page_vma_mapped_walk(): use page for pvmw-\u003epage\n\nPatch series \"mm: page_vma_mapped_walk() cleanup and THP fixes\".\n\nI\u0027ve marked all of these for stable: many are merely cleanups, but I\nthink they are much better before the main fix than after.\n\nThis patch (of 11):\n\npage_vma_mapped_walk() cleanup: sometimes the local copy of pvwm-\u003epage\nwas used, sometimes pvmw-\u003epage itself: use the local copy \"page\"\nthroughout.\n\nLink: https://lkml.kernel.org/r/589b358c-febc-c88e-d4c2-7834b37fa7bf@google.com\nLink: https://lkml.kernel.org/r/88e67645-f467-c279-bf5e-af4b5c6b13eb@google.com\nSigned-off-by: Hugh Dickins \u003chughd@google.com\u003e\nReviewed-by: Alistair Popple \u003capopple@nvidia.com\u003e\nAcked-by: Kirill A. Shutemov \u003ckirill.shutemov@linux.intel.com\u003e\nReviewed-by: Peter Xu \u003cpeterx@redhat.com\u003e\nCc: Yang Shi \u003cshy828301@gmail.com\u003e\nCc: Wang Yugui \u003cwangyugui@e16-tech.com\u003e\nCc: Matthew Wilcox \u003cwilly@infradead.org\u003e\nCc: Ralph Campbell \u003crcampbell@nvidia.com\u003e\nCc: Zi Yan \u003cziy@nvidia.com\u003e\nCc: Will Deacon \u003cwill@kernel.org\u003e\nCc: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "44db63d1ad8d71c6932cbe007eb41f31c434d140", "tree": "0cd88b5095c19f18789a639a5c5d2031a2a67cf2", "parents": [ "4a09d388f2ab382f217a764e6a152b3f614246f6", "5e0e7a407675d9f50f1d840214beaec42293b79b" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 13:27:07 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 13:27:07 2021 -0700" }, "message": "Merge tag \u0027drm-fixes-2021-06-25\u0027 of git://anongit.freedesktop.org/drm/drm\n\nPull drm fixes from Dave Airlie:\n \"This is a bit bigger than I\u0027d like at this stage, and I guess last\n week was extra quiet, but it\u0027s mostly one fix across three drivers to\n wait for buffer move pinning to complete.\n\n There was one locking change that got reverted so it\u0027s just noise.\n\n Otherwise the amdgpu/nouveau changes are for known regressions, and\n otherwise it\u0027s just misc changes in kmb/atmel/vc4 drivers.\n\n Summary:\n\n core:\n - auth locking change + brown paper bag revert\n\n radeon/nouveau/amdgpu/ttm:\n - wait for BO to be pinned after moving it (same fix in three\n drivers)\n\n amdgpu:\n - Revert GFX9/10 doorbell fixes, we just end up trading one bug for\n another\n - Potential memory corruption fix in framebuffer handling\n\n nouveau:\n - fix regression checking dma addresses\n\n kmb:\n - error return fix\n\n atmel-hlcdc:\n - fix kernel warnings at boot\n - enable async flips\n\n vc4:\n - fix CPU hang due to power management\"\n\n* tag \u0027drm-fixes-2021-06-25\u0027 of git://anongit.freedesktop.org/drm/drm:\n drm/nouveau: fix dma_address check for CPU/GPU sync\n drm/kmb: Fix error return code in kmb_hw_init()\n drm/amdgpu: wait for moving fence after pinning\n drm/radeon: wait for moving fence after pinning\n drm/nouveau: wait for moving fence after pinning v2\n Revert \"drm: add a locked version of drm_is_current_master\"\n Revert \"drm/amdgpu/gfx9: fix the doorbell missing when in CGPG issue.\"\n Revert \"drm/amdgpu/gfx10: enlarge CP_MEC_DOORBELL_RANGE_UPPER to cover full doorbell.\"\n drm/amdgpu: Call drm_framebuffer_init last for framebuffer init\n drm: add a locked version of drm_is_current_master\n drm/atmel-hlcdc: Allow async page flips\n drm/panel: ld9040: reference spi_device_id table\n drm: atmel_hlcdc: Enable the crtc vblank prior to crtc usage.\n drm/vc4: hdmi: Make sure the controller is powered in detect\n drm/vc4: hdmi: Move the HSM clock enable to runtime_pm\n" }, { "commit": "4ca070ef0dd885616ef294d269a9bf8e3b258e1a", "tree": "31cf0d52c78ee2a6e7069c31e6051e8577c366ff", "parents": [ "3265a7e6b41bae8608e7e91ac6798de5e5564164" ], "author": { "name": "Johan Hovold", "email": "johan@kernel.org", "time": "Mon May 24 11:09:12 2021 +0200" }, "committer": { "name": "Wolfram Sang", "email": "wsa@kernel.org", "time": "Thu Jun 24 22:08:00 2021 +0200" }, "message": "i2c: robotfuzz-osif: fix control-request directions\n\nThe direction of the pipe argument must match the request-type direction\nbit or control requests may fail depending on the host-controller-driver\nimplementation.\n\nControl transfers without a data stage are treated as OUT requests by\nthe USB stack and should be using usb_sndctrlpipe(). Failing to do so\nwill now trigger a warning.\n\nFix the OSIFI2C_SET_BIT_RATE and OSIFI2C_STOP requests which erroneously\nused the osif_usb_read() helper and set the IN direction bit.\n\nReported-by: syzbot+9d7dadd15b8819d73f41@syzkaller.appspotmail.com\nFixes: 83e53a8f120f (\"i2c: Add bus driver for for OSIF USB i2c device.\")\nCc: stable@vger.kernel.org # 3.14\nSigned-off-by: Johan Hovold \u003cjohan@kernel.org\u003e\nSigned-off-by: Wolfram Sang \u003cwsa@kernel.org\u003e\n" }, { "commit": "5e0e7a407675d9f50f1d840214beaec42293b79b", "tree": "3da7ab3c8916fd47c616b3ab450522d567dcefd8", "parents": [ "efea0c12a47bc41d22dbe6499a5fd8a5a856b792", "d330099115597bbc238d6758a4930e72b49ea9ba" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Jun 25 05:44:32 2021 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Jun 25 06:05:13 2021 +1000" }, "message": "Merge tag \u0027drm-misc-fixes-2021-06-24\u0027 of git://anongit.freedesktop.org/drm/drm-misc into drm-fixes\n\nA DMA address check for nouveau, an error code return fix for kmb, fixes\nto wait for a moving fence after pinning the BO for amdgpu, nouveau and\nradeon, a crtc and async page flip fix for atmel-hlcdc and a cpu hang\nfix for vc4.\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n\nFrom: Maxime Ripard \u003cmaxime@cerno.tech\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20210624190353.wyizoil3wqrrxz5d@gilmour\n" }, { "commit": "3265a7e6b41bae8608e7e91ac6798de5e5564164", "tree": "b27f825fa9063d3a45c72a75adde8297f1887b5a", "parents": [ "2269583753d2b8fdd3c861a516ff0cdbfcf4ef0b" ], "author": { "name": "Andreas Hecht", "email": "andreas.e.hecht@gmail.com", "time": "Thu Jun 24 17:25:35 2021 +0200" }, "committer": { "name": "Wolfram Sang", "email": "wsa@kernel.org", "time": "Thu Jun 24 21:47:43 2021 +0200" }, "message": "i2c: dev: Add __user annotation\n\nFix Sparse warnings:\ndrivers/i2c/i2c-dev.c:546:19: warning: incorrect type in assignment (different address spaces)\ndrivers/i2c/i2c-dev.c:549:53: warning: incorrect type in argument 2 (different address spaces)\n\ncompat_ptr() returns a pointer tagged __user which gets assigned to a\npointer missing the __user annotation. The same pointer is passed to\ncopy_from_user() as an argument where it is expected to have the __user\nannotation. Fix both by adding the __user annotation to the pointer.\n\nFixes: 7d5cb45655f2 (\"i2c compat ioctls: move to -\u003ecompat_ioctl()\")\nSigned-off-by: Andreas Hecht \u003candreas.e.hecht@gmail.com\u003e\nSigned-off-by: Wolfram Sang \u003cwsa@kernel.org\u003e\n" }, { "commit": "03af4c7bad8ca59143bca488b90b3775d10d7f94", "tree": "acc80188616ede6bd9488eb29601f41a1e0c0d44", "parents": [ "3c0d0894320cc517fda657c69939cd0313d0b4e2" ], "author": { "name": "Ilya Dryomov", "email": "idryomov@gmail.com", "time": "Mon Jun 21 12:17:40 2021 +0200" }, "committer": { "name": "Ilya Dryomov", "email": "idryomov@gmail.com", "time": "Thu Jun 24 21:03:17 2021 +0200" }, "message": "libceph: set global_id as soon as we get an auth ticket\n\nCommit 61ca49a9105f (\"libceph: don\u0027t set global_id until we get an\nauth ticket\") delayed the setting of global_id too much. It is set\nonly after all tickets are received, but in pre-nautilus clusters an\nauth ticket and the service tickets are obtained in separate steps\n(for a total of three MAuth replies). When the service tickets are\nrequested, global_id is used to build an authorizer; if global_id is\nstill 0 we never get them and fail to establish the session.\n\nMoving the setting of global_id into protocol implementations. This\nway global_id can be set exactly when an auth ticket is received, not\nsooner nor later.\n\nFixes: 61ca49a9105f (\"libceph: don\u0027t set global_id until we get an auth ticket\")\nSigned-off-by: Ilya Dryomov \u003cidryomov@gmail.com\u003e\nReviewed-by: Jeff Layton \u003cjlayton@kernel.org\u003e\n" }, { "commit": "3c0d0894320cc517fda657c69939cd0313d0b4e2", "tree": "461605c1079ae14df53653e4e424357456bb222c", "parents": [ "7a971e2c0767b6fc9a77c4108eceff0509c61cdb" ], "author": { "name": "Ilya Dryomov", "email": "idryomov@gmail.com", "time": "Mon Jun 21 11:53:38 2021 +0200" }, "committer": { "name": "Ilya Dryomov", "email": "idryomov@gmail.com", "time": "Thu Jun 24 21:03:16 2021 +0200" }, "message": "libceph: don\u0027t pass result into ac-\u003eops-\u003ehandle_reply()\n\nThere is no result to pass in msgr2 case because authentication\nfailures are reported through auth_bad_method frame and in MAuth\ncase an error is returned immediately.\n\nSigned-off-by: Ilya Dryomov \u003cidryomov@gmail.com\u003e\nReviewed-by: Jeff Layton \u003cjlayton@kernel.org\u003e\n" }, { "commit": "4a09d388f2ab382f217a764e6a152b3f614246f6", "tree": "d50a529dc55ab374663e73b94441f2c6bfae3bbc", "parents": [ "7749b0337b4e92d83f7e04b86434dcf4fe531377", "103a5348c22c3fca8b96c735a9e353b8a0801842" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 10:53:05 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 10:53:05 2021 -0700" }, "message": "Merge tag \u0027mmc-v5.13-rc6\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc\n\nPull MMC fix from Ulf Hansson:\n \"Use memcpy_to/fromio for dram-access-quirk in the meson-gx host\n driver\"\n\n* tag \u0027mmc-v5.13-rc6\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc:\n mmc: meson-gx: use memcpy_to/fromio for dram-access-quirk\n" }, { "commit": "7749b0337b4e92d83f7e04b86434dcf4fe531377", "tree": "9ad9cd57ce020a2168f215c5e46296bb9f6ef0c7", "parents": [ "666751701b6e4b6b6ebc82186434806fa8a09cf3", "399f8dd9a866e107639eabd3c1979cd526ca3a98" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 09:06:19 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 09:06:19 2021 -0700" }, "message": "Merge tag \u0027core-urgent-2021-06-24\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull sigqueue cache fix from Ingo Molnar:\n \"Fix a memory leak in the recently introduced sigqueue cache\"\n\n* tag \u0027core-urgent-2021-06-24\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n signal: Prevent sigqueue caching after task got released\n" }, { "commit": "666751701b6e4b6b6ebc82186434806fa8a09cf3", "tree": "6cc11146f899a41e0bc80f4217c28d2871a4a239", "parents": [ "df5011000496355d8508d56cd3ce45b1196b8b43", "fdaba61ef8a268d4136d0a113d153f7a89eb9984" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 08:58:23 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 08:58:23 2021 -0700" }, "message": "Merge tag \u0027sched-urgent-2021-06-24\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull scheduler fix from Ingo Molnar:\n \"A last minute cgroup bandwidth scheduling fix for a recently\n introduced logic fail which triggered a kernel warning by LTP\u0027s\n cfs_bandwidth01 test\"\n\n* tag \u0027sched-urgent-2021-06-24\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n sched/fair: Ensure that the CFS parent is added after unthrottling\n" }, { "commit": "df5011000496355d8508d56cd3ce45b1196b8b43", "tree": "e8b469c9425664a69775ff1eff2a011747e38013", "parents": [ "c0e457851fffd90eac14ad2528dfea3994945c28", "7f049fbdd57f6ea71dc741d903c19c73b2f70950" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 08:55:12 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 08:55:12 2021 -0700" }, "message": "Merge tag \u0027perf-urgent-2021-06-24\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull x86 perf fix from Ingo Molnar:\n \"An LBR buffer fix for code that probably only worked accidentally\"\n\n* tag \u0027perf-urgent-2021-06-24\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n perf/x86/intel/lbr: Zero the xstate buffer on allocation\n" }, { "commit": "f8be156be163a052a067306417cd0ff679068c97", "tree": "2f291bec641f75e1f350177da6179e8316699d71", "parents": [ "309505dd56854c1f9744c9a2b8aa40d897002bca" ], "author": { "name": "Nicholas Piggin", "email": "npiggin@gmail.com", "time": "Thu Jun 24 08:29:04 2021 -0400" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Jun 24 11:55:11 2021 -0400" }, "message": "KVM: do not allow mapping valid but non-reference-counted pages\n\nIt\u0027s possible to create a region which maps valid but non-refcounted\npages (e.g., tail pages of non-compound higher order allocations). These\nhost pages can then be returned by gfn_to_page, gfn_to_pfn, etc., family\nof APIs, which take a reference to the page, which takes it from 0 to 1.\nWhen the reference is dropped, this will free the page incorrectly.\n\nFix this by only taking a reference on valid pages if it was non-zero,\nwhich indicates it is participating in normal refcounting (and can be\nreleased with put_page).\n\nThis addresses CVE-2021-22543.\n\nSigned-off-by: Nicholas Piggin \u003cnpiggin@gmail.com\u003e\nTested-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "c0e457851fffd90eac14ad2528dfea3994945c28", "tree": "5735fee4d28530258d1fa9b83352300bc74e95bd", "parents": [ "7426cedc7dad67bf3c71ea6cc29ab7822e1a453f", "49faa77759b211fff344898edc23bb780707fff5" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 08:47:33 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jun 24 08:47:33 2021 -0700" }, "message": "Merge tag \u0027objtool-urgent-2021-06-24\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull objtool fixes from Ingo Molnar:\n \"Address a number of objtool warnings that got reported.\n\n No change in behavior intended, but code generation might be impacted\n by commit 1f008d46f124 (\"x86: Always inline task_size_max()\")\"\n\n* tag \u0027objtool-urgent-2021-06-24\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n locking/lockdep: Improve noinstr vs errors\n x86: Always inline task_size_max()\n x86/xen: Fix noinstr fail in exc_xen_unknown_trap()\n x86/xen: Fix noinstr fail in xen_pv_evtchn_do_upcall()\n x86/entry: Fix noinstr fail in __do_fast_syscall_32()\n objtool/x86: Ignore __x86_indirect_alt_* symbols\n" }, { "commit": "d330099115597bbc238d6758a4930e72b49ea9ba", "tree": "b9ebabc0c73b1c2c57d1c0e08c323fc8b498e444", "parents": [ "6fd8f323b3e4e5290d02174559308669507c00dd" ], "author": { "name": "Christian König", "email": "christian.koenig@amd.com", "time": "Fri Jun 11 14:34:50 2021 +0200" }, "committer": { "name": "Christian König", "email": "christian.koenig@amd.com", "time": "Thu Jun 24 15:40:44 2021 +0200" }, "message": "drm/nouveau: fix dma_address check for CPU/GPU sync\n\nAGP for example doesn\u0027t have a dma_address array.\n\nSigned-off-by: Christian König \u003cchristian.koenig@amd.com\u003e\nAcked-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20210614110517.1624-1-christian.koenig@amd.com\n" }, { "commit": "3de218ff39b9e3f0d453fe3154f12a174de44b25", "tree": "992878585c4a8ec453556fa77989796fe510849d", "parents": [ "107866a8eb0b664675a260f1ba0655010fac1e08" ], "author": { "name": "Juergen Gross", "email": "jgross@suse.com", "time": "Wed Jun 23 15:09:13 2021 +0200" }, "committer": { "name": "Juergen Gross", "email": "jgross@suse.com", "time": "Thu Jun 24 12:52:36 2021 +0200" }, "message": "xen/events: reset active flag for lateeoi events later\n\nIn order to avoid a race condition for user events when changing\ncpu affinity reset the active flag only when EOI-ing the event.\n\nThis is working fine as all user events are lateeoi events. Note that\nlateeoi_ack_mask_dynirq() is not modified as there is no explicit call\nto xen_irq_lateeoi() expected later.\n\nCc: stable@vger.kernel.org\nReported-by: Julien Grall \u003cjulien@xen.org\u003e\nFixes: b6622798bc50b62 (\"xen/events: avoid handling the same event on two cpus at the same time\")\nTested-by: Julien Grall \u003cjulien@xen.org\u003e\nSigned-off-by: Juergen Gross \u003cjgross@suse.com\u003e\nReviewed-by: Boris Ostrovsky \u003cboris.ostrvsky@oracle.com\u003e\nLink: https://lore.kernel.org/r/20210623130913.9405-1-jgross@suse.com\nSigned-off-by: Juergen Gross \u003cjgross@suse.com\u003e\n" }, { "commit": "5c6d4f97267f02f47acea8a652265348ec12de51", "tree": "ee90576e11fcc867657e15100c9b4e32dc43220d", "parents": [ "10043bb6af4230c57aeabaee02e6a42302f18d0f" ], "author": { "name": "Timur Tabi", "email": "timur@kernel.org", "time": "Sun Jun 20 11:01:35 2021 -0500" }, "committer": { "name": "Takashi Iwai", "email": "tiwai@suse.de", "time": "Thu Jun 24 12:23:45 2021 +0200" }, "message": "MAINTAINERS: remove Timur Tabi from Freescale SOC sound drivers\n\nI haven\u0027t touched these drivers in seven years, and none of the\npatches sent to me these days affect code that I wrote. The\nother maintainers are doing a very good job without me.\n\nSigned-off-by: Timur Tabi \u003ctimur@kernel.org\u003e\nReviewed-by: Fabio Estevam \u003cfestevam@gmail.com\u003e\nLink: https://lore.kernel.org/r/20210620160135.28651-1-timur@kernel.org\nSigned-off-by: Mark Brown \u003cbroonie@kernel.org\u003e\n(cherry picked from commit 50b1ce617d66d04f1f9006e51793e6cffcdec6ea)\nSigned-off-by: Takashi Iwai \u003ctiwai@suse.de\u003e\n" }, { "commit": "10043bb6af4230c57aeabaee02e6a42302f18d0f", "tree": "bd10a03dc46051b4c54084bca254b1227884e0d1", "parents": [ "13311e74253fe64329390df80bed3f07314ddd61" ], "author": { "name": "Mark Brown", "email": "broonie@kernel.org", "time": "Tue Jun 08 17:07:13 2021 +0100" }, "committer": { "name": "Takashi Iwai", "email": "tiwai@suse.de", "time": "Thu Jun 24 12:22:27 2021 +0200" }, "message": "ASoC: rt5645: Avoid upgrading static warnings to errors\n\nOne of the fixes reverted as part of the UMN fallout was actually fine,\nhowever rather than undoing the revert the process that handled all this\nstuff resulted in a patch which attempted to add extra error checks\ninstead. Unfortunately this new change wasn\u0027t really based on a good\nunderstanding of the subsystem APIs and bypassed the usual patch flow\nwithout ensuring it was reviewed by people with subsystem knowledge and\nwas merged as a fix rather than during the merge window.\n\nThe effect of the new fix is to upgrade what were previously warnings on\nstatic data in the code to hard errors on that data. If this actually\nhappens then it would break existing systems, if it doesn\u0027t happen then\nthe change has no effect so this was not a safe change to apply as a fix\nto the release candidates. Since the new code has not been tested and\ndoesn\u0027t in practice improve error handling revert it instead, and also\ndrop the original revert since the original fix was fine. This takes\nthe driver back to what it was in -rc1.\n\nFixes: 5e70b8e22b64e (\"ASoC: rt5645: add error checking to rt5645_probe function\")\nFixes: 1e0ce84215dbf (\"Revert \"ASoC: rt5645: fix a NULL pointer dereference\")\nSigned-off-by: Mark Brown \u003cbroonie@kernel.org\u003e\nCc: Greg Kroah-Hartman \u003cgregkh@linuxfoundation.org\u003e\nCc: Phillip Potter \u003cphil@philpotter.co.uk\u003e\nReviewed-by: Greg Kroah-Hartman \u003cgregkh@linuxfoundation.org\u003e\nLink: https://lore.kernel.org/r/20210608160713.21040-1-broonie@kernel.org\nSigned-off-by: Mark Brown \u003cbroonie@kernel.org\u003e\n(cherry picked from commit 916cccb5078eee57fce131c5fe18e417545083e2)\nSigned-off-by: Takashi Iwai \u003ctiwai@suse.de\u003e\n" }, { "commit": "309505dd56854c1f9744c9a2b8aa40d897002bca", "tree": "89495e398fdf35b8c3714ee042b3c08a85c32131", "parents": [ "13311e74253fe64329390df80bed3f07314ddd61" ], "author": { "name": "Zenghui Yu", "email": "yuzenghui@huawei.com", "time": "Thu Jun 24 15:09:31 2021 +0800" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Jun 24 04:04:38 2021 -0400" }, "message": "KVM: selftests: Fix mapping length truncation in m{,un}map()\n\nmax_mem_slots is now declared as uint32_t. The result of (0x200000 * 32767)\nis unexpectedly truncated to be 0xffe00000, whilst we actually need to\nallocate about, 63GB. Cast max_mem_slots to size_t in both mmap() and\nmunmap() to fix the length truncation.\n\nWe\u0027ll otherwise see the failure on arm64 thanks to the access_ok() checking\nin __kvm_set_memory_region(), as the unmapped VA happen to go beyond the\ntask\u0027s allowed address space.\n\n # ./set_memory_region_test\nAllowed number of memory slots: 32767\nAdding slots 0..32766, each memory region with 2048K size\n\u003d\u003d\u003d\u003d Test Assertion Failure \u003d\u003d\u003d\u003d\n set_memory_region_test.c:391: ret \u003d\u003d 0\n pid\u003d94861 tid\u003d94861 errno\u003d22 - Invalid argument\n 1\t0x00000000004015a7: test_add_max_memory_regions at set_memory_region_test.c:389\n 2\t (inlined by) main at set_memory_region_test.c:426\n 3\t0x0000ffffb8e67bdf: ?? ??:0\n 4\t0x00000000004016db: _start at :?\n KVM_SET_USER_MEMORY_REGION IOCTL failed,\n rc: -1 errno: 22 slot: 2615\n\nFixes: 3bf0fcd75434 (\"KVM: selftests: Speed up set_memory_region_test\")\nSigned-off-by: Zenghui Yu \u003cyuzenghui@huawei.com\u003e\nMessage-Id: \u003c20210624070931.565-1-yuzenghui@huawei.com\u003e\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "efea0c12a47bc41d22dbe6499a5fd8a5a856b792", "tree": "a208d16e8368b95d461f9adcb3e146c6610e0705", "parents": [ "13311e74253fe64329390df80bed3f07314ddd61", "ee5468b9f1d3bf48082eed351dace14598e8ca39" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Thu Jun 24 17:06:12 2021 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Thu Jun 24 17:06:12 2021 +1000" }, "message": "Merge tag \u0027amd-drm-fixes-5.13-2021-06-21\u0027 of https://gitlab.freedesktop.org/agd5f/linux into drm-fixes\n\namd-drm-fixes-5.13-2021-06-21:\n\namdgpu:\n- Revert GFX9, 10 doorbell fixes, we just\n end up trading one bug for another\n- Potential memory corruption fix in framebuffer handling\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\nFrom: Alex Deucher \u003calexander.deucher@amd.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20210621214132.4004-1-alexander.deucher@amd.com\n" }, { "commit": "7f049fbdd57f6ea71dc741d903c19c73b2f70950", "tree": "5b11e726fd368def34a736cfa73511ac95ded45e", "parents": [ "13311e74253fe64329390df80bed3f07314ddd61" ], "author": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Fri Jun 11 15:03:16 2021 +0200" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Thu Jun 24 08:49:03 2021 +0200" }, "message": "perf/x86/intel/lbr: Zero the xstate buffer on allocation\n\nXRSTORS requires a valid xstate buffer to work correctly. XSAVES does not\nguarantee to write a fully valid buffer according to the SDM:\n\n \"XSAVES does not write to any parts of the XSAVE header other than the\n XSTATE_BV and XCOMP_BV fields.\"\n\nXRSTORS triggers a #GP:\n\n \"If bytes 63:16 of the XSAVE header are not all zero.\"\n\nIt\u0027s dubious at best how this can work at all when the buffer is not zeroed\nbefore use.\n\nAllocate the buffers with __GFP_ZERO to prevent XRSTORS failure.\n\nFixes: ce711ea3cab9 (\"perf/x86/intel/lbr: Support XSAVES/XRSTORS for LBR context switch\")\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\nLink: https://lore.kernel.org/r/87wnr0wo2z.ffs@nanos.tec.linutronix.de\n" }, { "commit": "7426cedc7dad67bf3c71ea6cc29ab7822e1a453f", "tree": "985bc301971c0a48e1b39cdcf6d452deb58ae8de", "parents": [ "7266f2030eb0e17c442aa6cf527263828eed79bd", "f422316c8e9d3c4aff3c56549dfb44a677d02f14" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jun 23 11:29:15 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jun 23 11:29:15 2021 -0700" }, "message": "Merge tag \u0027spi-fix-v5.13-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi\n\nPull spi fixes from Mark Brown:\n \"A couple of small, driver specific fixes that arrived in the past few\n weeks\"\n\n* tag \u0027spi-fix-v5.13-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi:\n spi: spi-nxp-fspi: move the register operation after the clock enable\n spi: tegra20-slink: Ensure SPI controller reset is deasserted\n" }, { "commit": "5dca69e26fe97f17d4a6cbd6872103c868577b14", "tree": "e6638b0f32b8b2ab911010141e923ab7349ef43c", "parents": [ "13311e74253fe64329390df80bed3f07314ddd61" ], "author": { "name": "Heikki Krogerus", "email": "heikki.krogerus@linux.intel.com", "time": "Wed Jun 23 16:14:21 2021 +0300" }, "committer": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Wed Jun 23 19:34:58 2021 +0200" }, "message": "software node: Handle software node injection to an existing device properly\n\nThe function software_node_notify() - the function that creates\nand removes the symlinks between the node and the device - was\ncalled unconditionally in device_add_software_node() and\ndevice_remove_software_node(), but it needs to be called in\nthose functions only in the special case where the node is\nadded to a device that has already been registered.\n\nThis fixes NULL pointer dereference that happens if\ndevice_remove_software_node() is used with device that was\nnever registered.\n\nFixes: b622b24519f5 (\"software node: Allow node addition to already existing device\")\nReported-and-tested-by: Dominik Brodowski \u003clinux@dominikbrodowski.net\u003e\nReviewed-by: Andy Shevchenko \u003candriy.shevchenko@linux.intel.com\u003e\nSigned-off-by: Heikki Krogerus \u003cheikki.krogerus@linux.intel.com\u003e\nSigned-off-by: Rafael J. Wysocki \u003crafael.j.wysocki@intel.com\u003e\n" }, { "commit": "7266f2030eb0e17c442aa6cf527263828eed79bd", "tree": "48dc945416c7b76c9085e720920c10798d3617f4", "parents": [ "8fd2ed1c01dd96396c39dfa203c54705c6b6e03b", "4d6035f9bf4ea12776322746a216e856dfe46698" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jun 23 09:40:55 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jun 23 09:40:55 2021 -0700" }, "message": "Merge tag \u0027pm-5.13-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm\n\nPull power management fix from Rafael Wysocki:\n \"Revert a recent PCI power management commit that causes initialization\n issues to appear on some systems\"\n\n* tag \u0027pm-5.13-rc8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:\n Revert \"PCI: PM: Do not read power state in pci_enable_device_flags()\"\n" }, { "commit": "8fd2ed1c01dd96396c39dfa203c54705c6b6e03b", "tree": "588e5d9bd8750b1f392173e3822fb8d1dc57f631", "parents": [ "0c18f29aae7ce3dadd26d8ee3505d07cc982df75", "5f89468e2f060031cd89fd4287298e0eaf246bf6" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jun 23 09:04:07 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jun 23 09:04:07 2021 -0700" }, "message": "Merge branch \u0027stable/for-linus-5.14\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/swiotlb\n\nPull swiotlb fix from Konrad Rzeszutek Wilk:\n \"A fix for the regression for the DMA operations where the offset was\n ignored and corruptions would appear.\n\n Going forward there will be a cleanups to make the offset and\n alignment logic more clearer and better test-cases to help with this\"\n\n* \u0027stable/for-linus-5.14\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/swiotlb:\n swiotlb: manipulate orig_addr when tlb_addr has offset\n" }, { "commit": "d1b7f92035c6fb42529ada531e2cbf3534544c82", "tree": "15b7cb92667c77c0999097fb538f29b7a752116d", "parents": [ "7dd753ca59d6c8cc09aa1ed24f7657524803c7f3" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Thu Jun 17 13:55:04 2021 +0200" }, "committer": { "name": "Martin K. Petersen", "email": "martin.petersen@oracle.com", "time": "Tue Jun 22 22:02:20 2021 -0400" }, "message": "scsi: sd: Call sd_revalidate_disk() for ioctl(BLKRRPART)\n\nWhile the disk state has nothing to do with partitions, BLKRRPART is used\nto force a full revalidate after things like a disk format for historical\nreasons. Restore that behavior.\n\nLink: https://lore.kernel.org/r/20210617115504.1732350-1-hch@lst.de\nFixes: 471bd0af544b (\"sd: use bdev_check_media_change\")\nReported-by: Xiang Chen \u003cchenxiang66@hisilicon.com\u003e\nTested-by: Xiang Chen \u003cchenxiang66@hisilicon.com\u003e\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Martin K. Petersen \u003cmartin.petersen@oracle.com\u003e\n" }, { "commit": "0c18f29aae7ce3dadd26d8ee3505d07cc982df75", "tree": "51bdd30e0767c17e9f617eda0b6440ac1085fe39", "parents": [ "a96bfed64c8986d6404e553f18203cae1f5ac7e6" ], "author": { "name": "Mimi Zohar", "email": "zohar@linux.ibm.com", "time": "Tue Jun 22 13:36:41 2021 +0200" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Jun 22 11:13:19 2021 -0700" }, "message": "module: limit enabling module.sig_enforce\n\nIrrespective as to whether CONFIG_MODULE_SIG is configured, specifying\n\"module.sig_enforce\u003d1\" on the boot command line sets \"sig_enforce\".\nOnly allow \"sig_enforce\" to be set when CONFIG_MODULE_SIG is configured.\n\nThis patch makes the presence of /sys/module/module/parameters/sig_enforce\ndependent on CONFIG_MODULE_SIG\u003dy.\n\nFixes: fda784e50aac (\"module: export module signature enforcement status\")\nReported-by: Nayna Jain \u003cnayna@linux.ibm.com\u003e\nTested-by: Mimi Zohar \u003czohar@linux.ibm.com\u003e\nTested-by: Jessica Yu \u003cjeyu@kernel.org\u003e\nSigned-off-by: Mimi Zohar \u003czohar@linux.ibm.com\u003e\nSigned-off-by: Jessica Yu \u003cjeyu@kernel.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "6fd8f323b3e4e5290d02174559308669507c00dd", "tree": "24967c045deb3276a15366532859744a25883700", "parents": [ "8ddf5b9bb479570a3825d70fecfb9399bc15700c" ], "author": { "name": "Zhen Lei", "email": "thunder.leizhen@huawei.com", "time": "Thu May 13 21:46:38 2021 +0800" }, "committer": { "name": "Anitha Chrisanthus", "email": "anitha.chrisanthus@intel.com", "time": "Tue Jun 22 10:33:49 2021 -0700" }, "message": "drm/kmb: Fix error return code in kmb_hw_init()\n\nWhen the call to platform_get_irq() to obtain the IRQ of the lcd fails, the\nreturned error code should be propagated. However, we currently do not\nexplicitly assign this error code to \u0027ret\u0027. As a result, 0 was incorrectly\nreturned.\n\nFixes: 7f7b96a8a0a1 (\"drm/kmb: Add support for KeemBay Display\")\nReported-by: Hulk Robot \u003chulkci@huawei.com\u003e\nSigned-off-by: Zhen Lei \u003cthunder.leizhen@huawei.com\u003e\nSigned-off-by: Anitha Chrisanthus \u003canitha.chrisanthus@intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20210513134639.6541-1-thunder.leizhen@huawei.com\n" }, { "commit": "4d6035f9bf4ea12776322746a216e856dfe46698", "tree": "2f9bc24631f713b9901f1d3520824c504f32b6d2", "parents": [ "13311e74253fe64329390df80bed3f07314ddd61" ], "author": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Tue Jun 22 17:35:18 2021 +0200" }, "committer": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Tue Jun 22 17:35:18 2021 +0200" }, "message": "Revert \"PCI: PM: Do not read power state in pci_enable_device_flags()\"\n\nRevert commit 4514d991d992 (\"PCI: PM: Do not read power state in\npci_enable_device_flags()\") that is reported to cause PCI device\ninitialization issues on some systems.\n\nBugLink: https://bugzilla.kernel.org/show_bug.cgi?id\u003d213481\nLink: https://lore.kernel.org/linux-acpi/YNDoGICcg0V8HhpQ@eldamar.lan\nReported-by: Michael \u003cphyre@rogers.com\u003e\nReported-by: Salvatore Bonaccorso \u003ccarnil@debian.org\u003e\nFixes: 4514d991d992 (\"PCI: PM: Do not read power state in pci_enable_device_flags()\")\nSigned-off-by: Rafael J. Wysocki \u003crafael.j.wysocki@intel.com\u003e\n" }, { "commit": "399f8dd9a866e107639eabd3c1979cd526ca3a98", "tree": "5e9cfbb7c122d0216fbbcadf9dcd5f4ee32864f3", "parents": [ "13311e74253fe64329390df80bed3f07314ddd61" ], "author": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Tue Jun 22 01:08:30 2021 +0200" }, "committer": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Tue Jun 22 15:55:41 2021 +0200" }, "message": "signal: Prevent sigqueue caching after task got released\n\nsyzbot reported a memory leak related to sigqueue caching.\n\nThe assumption that a task cannot cache a sigqueue after the signal handler\nhas been dropped and exit_task_sigqueue_cache() has been invoked turns out\nto be wrong.\n\nSuch a task can still invoke release_task(other_task), which cleans up the\nsignals of \u0027other_task\u0027 and ends up in sigqueue_cache_or_free(), which in\nturn will cache the signal because task-\u003esigqueue_cache is NULL. That\u0027s\nobviously bogus because nothing will free the cached signal of that task\nanymore, so the cached item is leaked.\n\nThis happens when e.g. the last non-leader thread exits and reaps the\nzombie leader.\n\nPrevent this by setting tsk::sigqueue_cache to an error pointer value in\nexit_task_sigqueue_cache() which forces any subsequent invocation of\nsigqueue_cache_or_free() from that task to hand the sigqueue back to the\nkmemcache.\n\nAdd comments to all relevant places.\n\nFixes: 4bad58ebc8bc (\"signal: Allow tasks to cache one sigqueue struct\")\nReported-by: syzbot+0bac5fec63d4f399ba98@syzkaller.appspotmail.com\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nReviewed-by: Oleg Nesterov \u003coleg@redhat.com\u003e\nAcked-by: Christian Brauner \u003cchristian.brauner@ubuntu.com\u003e\nLink: https://lore.kernel.org/r/878s32g6j5.ffs@nanos.tec.linutronix.de\n\n" }, { "commit": "8ddf5b9bb479570a3825d70fecfb9399bc15700c", "tree": "30c88932a8e180187fb2be054c18484123a7e68a", "parents": [ "4b41726aae563273bb4b4a9462ba51ce4d372f78" ], "author": { "name": "Christian König", "email": "christian.koenig@amd.com", "time": "Mon Jun 21 14:29:14 2021 +0200" }, "committer": { "name": "Christian König", "email": "christian.koenig@amd.com", "time": "Tue Jun 22 15:29:03 2021 +0200" }, "message": "drm/amdgpu: wait for moving fence after pinning\n\nWe actually need to wait for the moving fence after pinning\nthe BO to make sure that the pin is completed.\n\nSigned-off-by: Christian König \u003cchristian.koenig@amd.com\u003e\nReviewed-by: Daniel Vetter \u003cdaniel.vetter@ffwll.ch\u003e\nReferences: https://lore.kernel.org/dri-devel/20210621151758.2347474-1-daniel.vetter@ffwll.ch/\nCC: stable@kernel.org\nLink: https://patchwork.freedesktop.org/patch/msgid/20210622114506.106349-3-christian.koenig@amd.com\n" }, { "commit": "4b41726aae563273bb4b4a9462ba51ce4d372f78", "tree": "dfe60fa25bf52b880b1070922a6a97eec9475a4a", "parents": [ "17b11f71795abdce46f62a808f906857e525cea8" ], "author": { "name": "Christian König", "email": "christian.koenig@amd.com", "time": "Mon Jun 21 13:43:05 2021 +0200" }, "committer": { "name": "Christian König", "email": "christian.koenig@amd.com", "time": "Tue Jun 22 15:29:03 2021 +0200" }, "message": "drm/radeon: wait for moving fence after pinning\n\nWe actually need to wait for the moving fence after pinning\nthe BO to make sure that the pin is completed.\n\nSigned-off-by: Christian König \u003cchristian.koenig@amd.com\u003e\nReviewed-by: Daniel Vetter \u003cdaniel.vetter@ffwll.ch\u003e\nReferences: https://lore.kernel.org/dri-devel/20210621151758.2347474-1-daniel.vetter@ffwll.ch/\nCC: stable@kernel.org\nLink: https://patchwork.freedesktop.org/patch/msgid/20210622114506.106349-2-christian.koenig@amd.com\n" }, { "commit": "17b11f71795abdce46f62a808f906857e525cea8", "tree": "a149a8332643a93cb5d2fb43eb8b7096d5376615", "parents": [ "f54b3ca7ea1e5e02f481cf4ca54568e57bd66086" ], "author": { "name": "Christian König", "email": "christian.koenig@amd.com", "time": "Mon Jun 21 13:36:35 2021 +0200" }, "committer": { "name": "Christian König", "email": "christian.koenig@amd.com", "time": "Tue Jun 22 15:29:03 2021 +0200" }, "message": "drm/nouveau: wait for moving fence after pinning v2\n\nWe actually need to wait for the moving fence after pinning\nthe BO to make sure that the pin is completed.\n\nv2: grab the lock while waiting\n\nSigned-off-by: Christian König \u003cchristian.koenig@amd.com\u003e\nReviewed-by: Daniel Vetter \u003cdaniel.vetter@ffwll.ch\u003e\nReferences: https://lore.kernel.org/dri-devel/20210621151758.2347474-1-daniel.vetter@ffwll.ch/\nCC: stable@kernel.org\nLink: https://patchwork.freedesktop.org/patch/msgid/20210622114506.106349-1-christian.koenig@amd.com\n" }, { "commit": "fdaba61ef8a268d4136d0a113d153f7a89eb9984", "tree": "d2fd026e595d2b5eabbe15aadd3dc6382f55d240", "parents": [ "a96bfed64c8986d6404e553f18203cae1f5ac7e6" ], "author": { "name": "Rik van Riel", "email": "riel@surriel.com", "time": "Mon Jun 21 19:43:30 2021 +0200" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Tue Jun 22 14:06:57 2021 +0200" }, "message": "sched/fair: Ensure that the CFS parent is added after unthrottling\n\nEnsure that a CFS parent will be in the list whenever one of its children is also\nin the list.\n\nA warning on rq-\u003etmp_alone_branch !\u003d \u0026rq-\u003eleaf_cfs_rq_list has been\nreported while running LTP test cfs_bandwidth01.\n\nOdin Ugedal found the root cause:\n\n\t$ tree /sys/fs/cgroup/ltp/ -d --charset\u003dascii\n\t/sys/fs/cgroup/ltp/\n\t|-- drain\n\t`-- test-6851\n\t `-- level2\n\t\t|-- level3a\n\t\t| |-- worker1\n\t\t| `-- worker2\n\t\t`-- level3b\n\t\t `-- worker3\n\nTimeline (ish):\n- worker3 gets throttled\n- level3b is decayed, since it has no more load\n- level2 get throttled\n- worker3 get unthrottled\n- level2 get unthrottled\n - worker3 is added to list\n - level3b is not added to list, since nr_running\u003d\u003d0 and is decayed\n\n [ Vincent Guittot: Rebased and updated to fix for the reported warning. ]\n\nFixes: a7b359fc6a37 (\"sched/fair: Correctly insert cfs_rq\u0027s to list on unthrottle\")\nReported-by: Sachin Sant \u003csachinp@linux.vnet.ibm.com\u003e\nSuggested-by: Vincent Guittot \u003cvincent.guittot@linaro.org\u003e\nSigned-off-by: Rik van Riel \u003criel@surriel.com\u003e\nSigned-off-by: Vincent Guittot \u003cvincent.guittot@linaro.org\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\nTested-by: Sachin Sant \u003csachinp@linux.vnet.ibm.com\u003e\nAcked-by: Odin Ugedal \u003codin@uged.al\u003e\nLink: https://lore.kernel.org/r/20210621174330.11258-1-vincent.guittot@linaro.org\n" }, { "commit": "49faa77759b211fff344898edc23bb780707fff5", "tree": "9f36b8c797c3f312f08e12634ec2cdbafb6f0a95", "parents": [ "1f008d46f1243899d27fd034ab5c41985bd16cee" ], "author": { "name": "Peter Zijlstra", "email": "peterz@infradead.org", "time": "Mon Jun 21 13:12:38 2021 +0200" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Tue Jun 22 13:56:43 2021 +0200" }, "message": "locking/lockdep: Improve noinstr vs errors\n\nBetter handle the failure paths.\n\n vmlinux.o: warning: objtool: debug_locks_off()+0x23: call to console_verbose() leaves .noinstr.text section\n vmlinux.o: warning: objtool: debug_locks_off()+0x19: call to __kasan_check_write() leaves .noinstr.text section\n\n debug_locks_off+0x19/0x40:\n instrument_atomic_write at include/linux/instrumented.h:86\n (inlined by) __debug_locks_off at include/linux/debug_locks.h:17\n (inlined by) debug_locks_off at lib/debug_locks.c:41\n\nFixes: 6eebad1ad303 (\"lockdep: __always_inline more for noinstr\")\nSigned-off-by: Peter Zijlstra (Intel) \u003cpeterz@infradead.org\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\nLink: https://lore.kernel.org/r/20210621120120.784404944@infradead.org\n" }, { "commit": "1f008d46f1243899d27fd034ab5c41985bd16cee", "tree": "f60e7bbc53dcb2120240a044e48f083b562a1ca5", "parents": [ "4c9c26f1e67648f41f28f8c997c5c9467a3dbbe4" ], "author": { "name": "Peter Zijlstra", "email": "peterz@infradead.org", "time": "Mon Jun 21 13:12:37 2021 +0200" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Tue Jun 22 13:56:43 2021 +0200" }, "message": "x86: Always inline task_size_max()\n\nFix:\n\n vmlinux.o: warning: objtool: handle_bug()+0x10: call to task_size_max() leaves .noinstr.text section\n\nWhen #UD isn\u0027t a BUG, we shouldn\u0027t violate noinstr (we\u0027ll still\nprobably die, but that\u0027s another story).\n\nFixes: 025768a966a3 (\"x86/cpu: Use alternative to generate the TASK_SIZE_MAX constant\")\nSigned-off-by: Peter Zijlstra (Intel) \u003cpeterz@infradead.org\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\nLink: https://lore.kernel.org/r/20210621120120.682468274@infradead.org\n" }, { "commit": "4c9c26f1e67648f41f28f8c997c5c9467a3dbbe4", "tree": "91bdb7426a226e2e160b218710b1e54b75c6d1fe", "parents": [ "84e60065df9ef03759115a7e48c04bbc0d292165" ], "author": { "name": "Peter Zijlstra", "email": "peterz@infradead.org", "time": "Mon Jun 21 13:12:36 2021 +0200" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Tue Jun 22 13:56:42 2021 +0200" }, "message": "x86/xen: Fix noinstr fail in exc_xen_unknown_trap()\n\nFix:\n\n vmlinux.o: warning: objtool: exc_xen_unknown_trap()+0x7: call to printk() leaves .noinstr.text section\n\nFixes: 2e92493637a0 (\"x86/xen: avoid warning in Xen pv guest with CONFIG_AMD_MEM_ENCRYPT enabled\")\nSigned-off-by: Peter Zijlstra (Intel) \u003cpeterz@infradead.org\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\nLink: https://lore.kernel.org/r/20210621120120.606560778@infradead.org\n" }, { "commit": "84e60065df9ef03759115a7e48c04bbc0d292165", "tree": "cd95ee4a9c76577a954d5317742d092da0a66f86", "parents": [ "240001d4e3041832e8a2654adc3ccf1683132b92" ], "author": { "name": "Peter Zijlstra", "email": "peterz@infradead.org", "time": "Mon Jun 21 13:12:35 2021 +0200" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Tue Jun 22 13:56:42 2021 +0200" }, "message": "x86/xen: Fix noinstr fail in xen_pv_evtchn_do_upcall()\n\nFix:\n\n vmlinux.o: warning: objtool: xen_pv_evtchn_do_upcall()+0x23: call to irq_enter_rcu() leaves .noinstr.text section\n\nFixes: 359f01d1816f (\"x86/entry: Use run_sysvec_on_irqstack_cond() for XEN upcall\")\nSigned-off-by: Peter Zijlstra (Intel) \u003cpeterz@infradead.org\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\nLink: https://lore.kernel.org/r/20210621120120.532960208@infradead.org\n" }, { "commit": "240001d4e3041832e8a2654adc3ccf1683132b92", "tree": "6430ecdfbff6af5c2c4dbffaa0e137971df990f1", "parents": [ "31197d3a0f1caeb60fb01f6755e28347e4f44037" ], "author": { "name": "Peter Zijlstra", "email": "peterz@infradead.org", "time": "Mon Jun 21 13:12:34 2021 +0200" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Tue Jun 22 13:56:42 2021 +0200" }, "message": "x86/entry: Fix noinstr fail in __do_fast_syscall_32()\n\nFix:\n\n vmlinux.o: warning: objtool: __do_fast_syscall_32()+0xf5: call to trace_hardirqs_off() leaves .noinstr.text section\n\nFixes: 5d5675df792f (\"x86/entry: Fix entry/exit mismatch on failed fast 32-bit syscalls\")\nSigned-off-by: Peter Zijlstra (Intel) \u003cpeterz@infradead.org\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\nLink: https://lore.kernel.org/r/20210621120120.467898710@infradead.org\n" }, { "commit": "7a971e2c0767b6fc9a77c4108eceff0509c61cdb", "tree": "df30eb87a1fcd07e2ddae903ce7617cee17766d1", "parents": [ "27171ae6a0fdc75571e5bf3d0961631a1e4fb765" ], "author": { "name": "Jeff Layton", "email": "jlayton@kernel.org", "time": "Wed Jun 02 12:46:07 2021 -0400" }, "committer": { "name": "Ilya Dryomov", "email": "idryomov@gmail.com", "time": "Tue Jun 22 13:53:28 2021 +0200" }, "message": "ceph: fix error handling in ceph_atomic_open and ceph_lookup\n\nCommit aa60cfc3f7ee broke the error handling in these functions such\nthat they don\u0027t handle non-ENOENT errors from ceph_mdsc_do_request\nproperly.\n\nMove the checking of -ENOENT out of ceph_handle_snapdir and into the\ncallers, and if we get a different error, return it immediately.\n\nFixes: aa60cfc3f7ee (\"ceph: don\u0027t use d_add in ceph_handle_snapdir\")\nSigned-off-by: Jeff Layton \u003cjlayton@kernel.org\u003e\nReviewed-by: Ilya Dryomov \u003cidryomov@gmail.com\u003e\nSigned-off-by: Ilya Dryomov \u003cidryomov@gmail.com\u003e\n" }, { "commit": "27171ae6a0fdc75571e5bf3d0961631a1e4fb765", "tree": "110d0dc1ca33573d9a58637cabb45b0bdd474550", "parents": [ "13311e74253fe64329390df80bed3f07314ddd61" ], "author": { "name": "Jeff Layton", "email": "jlayton@kernel.org", "time": "Tue Jun 01 09:40:25 2021 -0400" }, "committer": { "name": "Ilya Dryomov", "email": "idryomov@gmail.com", "time": "Tue Jun 22 13:53:28 2021 +0200" }, "message": "ceph: must hold snap_rwsem when filling inode for async create\n\n...and add a lockdep assertion for it to ceph_fill_inode().\n\nCc: stable@vger.kernel.org # v5.7+\nFixes: 9a8d03ca2e2c3 (\"ceph: attempt to do async create when possible\")\nSigned-off-by: Jeff Layton \u003cjlayton@kernel.org\u003e\nReviewed-by: Ilya Dryomov \u003cidryomov@gmail.com\u003e\nSigned-off-by: Ilya Dryomov \u003cidryomov@gmail.com\u003e\n" }, { "commit": "f9dfb5e390fab2df9f7944bb91e7705aba14cd26", "tree": "803faf70167a03486e5bbb1cabb6662fb8b5b87c", "parents": [ "9301982c424a003c0095bf157154a85bf5322bd0" ], "author": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Fri Jun 18 16:18:25 2021 +0200" }, "committer": { "name": "Borislav Petkov", "email": "bp@suse.de", "time": "Tue Jun 22 11:06:21 2021 +0200" }, "message": "x86/fpu: Make init_fpstate correct with optimized XSAVE\n\nThe XSAVE init code initializes all enabled and supported components with\nXRSTOR(S) to init state. Then it XSAVEs the state of the components back\ninto init_fpstate which is used in several places to fill in the init state\nof components.\n\nThis works correctly with XSAVE, but not with XSAVEOPT and XSAVES because\nthose use the init optimization and skip writing state of components which\nare in init state. So init_fpstate.xsave still contains all zeroes after\nthis operation.\n\nThere are two ways to solve that:\n\n 1) Use XSAVE unconditionally, but that requires to reshuffle the buffer when\n XSAVES is enabled because XSAVES uses compacted format.\n\n 2) Save the components which are known to have a non-zero init state by other\n means.\n\nLooking deeper, #2 is the right thing to do because all components the\nkernel supports have all-zeroes init state except the legacy features (FP,\nSSE). Those cannot be hard coded because the states are not identical on all\nCPUs, but they can be saved with FXSAVE which avoids all conditionals.\n\nUse FXSAVE to save the legacy FP/SSE components in init_fpstate along with\na BUILD_BUG_ON() which reminds developers to validate that a newly added\ncomponent has all zeroes init state. As a bonus remove the now unused\ncopy_xregs_to_kernel_booting() crutch.\n\nThe XSAVE and reshuffle method can still be implemented in the unlikely\ncase that components are added which have a non-zero init state and no\nother means to save them. For now, FXSAVE is just simple and good enough.\n\n [ bp: Fix a typo or two in the text. ]\n\nFixes: 6bad06b76892 (\"x86, xsave: Use xsaveopt in context-switch path when supported\")\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nSigned-off-by: Borislav Petkov \u003cbp@suse.de\u003e\nReviewed-by: Borislav Petkov \u003cbp@suse.de\u003e\nCc: stable@vger.kernel.org\nLink: https://lkml.kernel.org/r/20210618143444.587311343@linutronix.de\n" }, { "commit": "9301982c424a003c0095bf157154a85bf5322bd0", "tree": "f96f40f5010ecdb6218364ddaff74f24b2f0852e", "parents": [ "13311e74253fe64329390df80bed3f07314ddd61" ], "author": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Fri Jun 18 16:18:24 2021 +0200" }, "committer": { "name": "Borislav Petkov", "email": "bp@suse.de", "time": "Tue Jun 22 10:51:23 2021 +0200" }, "message": "x86/fpu: Preserve supervisor states in sanitize_restored_user_xstate()\n\nsanitize_restored_user_xstate() preserves the supervisor states only\nwhen the fx_only argument is zero, which allows unprivileged user space\nto put supervisor states back into init state.\n\nPreserve them unconditionally.\n\n [ bp: Fix a typo or two in the text. ]\n\nFixes: 5d6b6a6f9b5c (\"x86/fpu/xstate: Update sanitize_restored_xstate() for supervisor xstates\")\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nSigned-off-by: Borislav Petkov \u003cbp@suse.de\u003e\nCc: stable@vger.kernel.org\nLink: https://lkml.kernel.org/r/20210618143444.438635017@linutronix.de\n" }, { "commit": "f54b3ca7ea1e5e02f481cf4ca54568e57bd66086", "tree": "80663197832fc55ba4d5602403df93e40d705839", "parents": [ "1815d9c86e3090477fbde066ff314a7e9721ee0f" ], "author": { "name": "Daniel Vetter", "email": "daniel.vetter@ffwll.ch", "time": "Tue Jun 22 09:54:09 2021 +0200" }, "committer": { "name": "Daniel Vetter", "email": "daniel.vetter@ffwll.ch", "time": "Tue Jun 22 10:41:55 2021 +0200" }, "message": "Revert \"drm: add a locked version of drm_is_current_master\"\n\nThis reverts commit 1815d9c86e3090477fbde066ff314a7e9721ee0f.\n\nUnfortunately this inverts the locking hierarchy, so back to the\ndrawing board. Full lockdep splat below:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nWARNING: possible circular locking dependency detected\n5.13.0-rc7-CI-CI_DRM_10254+ #1 Not tainted\n------------------------------------------------------\nkms_frontbuffer/1087 is trying to acquire lock:\nffff88810dcd01a8 (\u0026dev-\u003emaster_mutex){+.+.}-{3:3}, at: drm_is_current_master+0x1b/0x40\nbut task is already holding lock:\nffff88810dcd0488 (\u0026dev-\u003emode_config.mutex){+.+.}-{3:3}, at: drm_mode_getconnector+0x1c6/0x4a0\nwhich lock already depends on the new lock.\nthe existing dependency chain (in reverse order) is:\n-\u003e #2 (\u0026dev-\u003emode_config.mutex){+.+.}-{3:3}:\n __mutex_lock+0xab/0x970\n drm_client_modeset_probe+0x22e/0xca0\n __drm_fb_helper_initial_config_and_unlock+0x42/0x540\n intel_fbdev_initial_config+0xf/0x20 [i915]\n async_run_entry_fn+0x28/0x130\n process_one_work+0x26d/0x5c0\n worker_thread+0x37/0x380\n kthread+0x144/0x170\n ret_from_fork+0x1f/0x30\n-\u003e #1 (\u0026client-\u003emodeset_mutex){+.+.}-{3:3}:\n __mutex_lock+0xab/0x970\n drm_client_modeset_commit_locked+0x1c/0x180\n drm_client_modeset_commit+0x1c/0x40\n __drm_fb_helper_restore_fbdev_mode_unlocked+0x88/0xb0\n drm_fb_helper_set_par+0x34/0x40\n intel_fbdev_set_par+0x11/0x40 [i915]\n fbcon_init+0x270/0x4f0\n visual_init+0xc6/0x130\n do_bind_con_driver+0x1e5/0x2d0\n do_take_over_console+0x10e/0x180\n do_fbcon_takeover+0x53/0xb0\n register_framebuffer+0x22d/0x310\n __drm_fb_helper_initial_config_and_unlock+0x36c/0x540\n intel_fbdev_initial_config+0xf/0x20 [i915]\n async_run_entry_fn+0x28/0x130\n process_one_work+0x26d/0x5c0\n worker_thread+0x37/0x380\n kthread+0x144/0x170\n ret_from_fork+0x1f/0x30\n-\u003e #0 (\u0026dev-\u003emaster_mutex){+.+.}-{3:3}:\n __lock_acquire+0x151e/0x2590\n lock_acquire+0xd1/0x3d0\n __mutex_lock+0xab/0x970\n drm_is_current_master+0x1b/0x40\n drm_mode_getconnector+0x37e/0x4a0\n drm_ioctl_kernel+0xa8/0xf0\n drm_ioctl+0x1e8/0x390\n __x64_sys_ioctl+0x6a/0xa0\n do_syscall_64+0x39/0xb0\n entry_SYSCALL_64_after_hwframe+0x44/0xae\nother info that might help us debug this:\nChain exists of: \u0026dev-\u003emaster_mutex --\u003e \u0026client-\u003emodeset_mutex --\u003e \u0026dev-\u003emode_config.mutex\n Possible unsafe locking scenario:\n CPU0 CPU1\n ---- ----\n lock(\u0026dev-\u003emode_config.mutex);\n lock(\u0026client-\u003emodeset_mutex);\n lock(\u0026dev-\u003emode_config.mutex);\n lock(\u0026dev-\u003emaster_mutex);\n*** DEADLOCK ***\n1 lock held by kms_frontbuffer/1087:\n #0: ffff88810dcd0488 (\u0026dev-\u003emode_config.mutex){+.+.}-{3:3}, at: drm_mode_getconnector+0x1c6/0x4a0\nstack backtrace:\nCPU: 7 PID: 1087 Comm: kms_frontbuffer Not tainted 5.13.0-rc7-CI-CI_DRM_10254+ #1\nHardware name: Intel Corporation Ice Lake Client Platform/IceLake U DDR4 SODIMM PD RVP TLC, BIOS ICLSFWR1.R00.3234.A01.1906141750 06/14/2019\nCall Trace:\n dump_stack+0x7f/0xad\n check_noncircular+0x12e/0x150\n __lock_acquire+0x151e/0x2590\n lock_acquire+0xd1/0x3d0\n __mutex_lock+0xab/0x970\n drm_is_current_master+0x1b/0x40\n drm_mode_getconnector+0x37e/0x4a0\n drm_ioctl_kernel+0xa8/0xf0\n drm_ioctl+0x1e8/0x390\n __x64_sys_ioctl+0x6a/0xa0\n do_syscall_64+0x39/0xb0\n entry_SYSCALL_64_after_hwframe+0x44/0xae\n\nNote that this broke the intel-gfx CI pretty much across the board\nbecause it has to reboot machines after it hits a lockdep splat.\n\nTestcase: igt/debugfs_test/read_all_entries\nAcked-by: Petri Latvala \u003cpetri.latvala@intel.com\u003e\nFixes: 1815d9c86e30 (\"drm: add a locked version of drm_is_current_master\")\nCc: Desmond Cheong Zhi Xi \u003cdesmondcheongzx@gmail.com\u003e\nCc: Emil Velikov \u003cemil.l.velikov@gmail.com\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Daniel Vetter \u003cdaniel.vetter@intel.com\u003e\nCc: Maarten Lankhorst \u003cmaarten.lankhorst@linux.intel.com\u003e\nCc: Maxime Ripard \u003cmripard@kernel.org\u003e\nCc: Thomas Zimmermann \u003ctzimmermann@suse.de\u003e\nCc: David Airlie \u003cairlied@linux.ie\u003e\nCc: Daniel Vetter \u003cdaniel@ffwll.ch\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20210622075409.2673805-1-daniel.vetter@ffwll.ch\n" }, { "commit": "cb8f63b8cbf39845244f3ccae43bb7e63bd70543", "tree": "1b6267bdc7e0e75bdb730ed75276077e7c458941", "parents": [ "3093e6cca3ba7d47848068cb256c489675125181" ], "author": { "name": "Gabriel Knezek", "email": "gabeknez@linux.microsoft.com", "time": "Mon Jun 21 15:28:59 2021 -0700" }, "committer": { "name": "Bartosz Golaszewski", "email": "bgolaszewski@baylibre.com", "time": "Tue Jun 22 09:54:13 2021 +0200" }, "message": "gpiolib: cdev: zero padding during conversion to gpioline_info_changed\n\nWhen userspace requests a GPIO v1 line info changed event,\nlineinfo_watch_read() populates and returns the gpioline_info_changed\nstructure. It contains 5 words of padding at the end which are not\ninitialized before being returned to userspace.\n\nZero the structure in gpio_v2_line_info_change_to_v1() before populating\nits contents.\n\nFixes: aad955842d1c (\"gpiolib: cdev: support GPIO_V2_GET_LINEINFO_IOCTL and GPIO_V2_GET_LINEINFO_WATCH_IOCTL\")\nSigned-off-by: Gabriel Knezek \u003cgabeknez@linux.microsoft.com\u003e\nReviewed-by: Kent Gibson \u003cwarthog618@gmail.com\u003e\nSigned-off-by: Bartosz Golaszewski \u003cbgolaszewski@baylibre.com\u003e\n" }, { "commit": "ee5468b9f1d3bf48082eed351dace14598e8ca39", "tree": "a208d16e8368b95d461f9adcb3e146c6610e0705", "parents": [ "baacf52a473b24e10322b67757ddb92ab8d86717" ], "author": { "name": "Yifan Zhang", "email": "yifan1.zhang@amd.com", "time": "Sat Jun 19 11:40:54 2021 +0800" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Mon Jun 21 17:22:52 2021 -0400" }, "message": "Revert \"drm/amdgpu/gfx9: fix the doorbell missing when in CGPG issue.\"\n\nThis reverts commit 4cbbe34807938e6e494e535a68d5ff64edac3f20.\n\nReason for revert: side effect of enlarging CP_MEC_DOORBELL_RANGE may\ncause some APUs fail to enter gfxoff in certain user cases.\n\nSigned-off-by: Yifan Zhang \u003cyifan1.zhang@amd.com\u003e\nAcked-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\nCc: stable@vger.kernel.org\n" }, { "commit": "baacf52a473b24e10322b67757ddb92ab8d86717", "tree": "614e907aa88d958014da320a278b82f720e97888", "parents": [ "4c6a23188e26339fd3dbc78e6ce6fe0fc4009553" ], "author": { "name": "Yifan Zhang", "email": "yifan1.zhang@amd.com", "time": "Sat Jun 19 11:39:43 2021 +0800" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Mon Jun 21 17:22:06 2021 -0400" }, "message": "Revert \"drm/amdgpu/gfx10: enlarge CP_MEC_DOORBELL_RANGE_UPPER to cover full doorbell.\"\n\nThis reverts commit 1c0b0efd148d5b24c4932ddb3fa03c8edd6097b3.\n\nReason for revert: Side effect of enlarging CP_MEC_DOORBELL_RANGE may\ncause some APUs fail to enter gfxoff in certain user cases.\n\nSigned-off-by: Yifan Zhang \u003cyifan1.zhang@amd.com\u003e\nAcked-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\nCc: stable@vger.kernel.org\n" }, { "commit": "4c6a23188e26339fd3dbc78e6ce6fe0fc4009553", "tree": "62eafbe8c6c9a99394fc024b47fc27a2790bf580", "parents": [ "13311e74253fe64329390df80bed3f07314ddd61" ], "author": { "name": "Michel Dänzer", "email": "mdaenzer@redhat.com", "time": "Wed Jun 16 12:46:51 2021 +0200" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Mon Jun 21 17:21:49 2021 -0400" }, "message": "drm/amdgpu: Call drm_framebuffer_init last for framebuffer init\n\nOnce drm_framebuffer_init has returned 0, the framebuffer is hooked up\nto the reference counting machinery and can no longer be destroyed with\na simple kfree. Therefore, it must be called last.\n\nIf drm_framebuffer_init returns 0 but its caller then returns non-0,\nthere will likely be memory corruption fireworks down the road.\nThe following lead me to this fix:\n\n[ 12.891228] kernel BUG at lib/list_debug.c:25!\n[...]\n[ 12.891263] RIP: 0010:__list_add_valid+0x4b/0x70\n[...]\n[ 12.891324] Call Trace:\n[ 12.891330] drm_framebuffer_init+0xb5/0x100 [drm]\n[ 12.891378] amdgpu_display_gem_fb_verify_and_init+0x47/0x120 [amdgpu]\n[ 12.891592] ? amdgpu_display_user_framebuffer_create+0x10d/0x1f0 [amdgpu]\n[ 12.891794] amdgpu_display_user_framebuffer_create+0x126/0x1f0 [amdgpu]\n[ 12.891995] drm_internal_framebuffer_create+0x378/0x3f0 [drm]\n[ 12.892036] ? drm_internal_framebuffer_create+0x3f0/0x3f0 [drm]\n[ 12.892075] drm_mode_addfb2+0x34/0xd0 [drm]\n[ 12.892115] ? drm_internal_framebuffer_create+0x3f0/0x3f0 [drm]\n[ 12.892153] drm_ioctl_kernel+0xe2/0x150 [drm]\n[ 12.892193] drm_ioctl+0x3da/0x460 [drm]\n[ 12.892232] ? drm_internal_framebuffer_create+0x3f0/0x3f0 [drm]\n[ 12.892274] amdgpu_drm_ioctl+0x43/0x80 [amdgpu]\n[ 12.892475] __se_sys_ioctl+0x72/0xc0\n[ 12.892483] do_syscall_64+0x33/0x40\n[ 12.892491] entry_SYSCALL_64_after_hwframe+0x44/0xae\n\nFixes: f258907fdd835e \"drm/amdgpu: Verify bo size can fit framebuffer size on init.\"\nSigned-off-by: Michel Dänzer \u003cmdaenzer@redhat.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "827a746f405d25f79560c7868474aec5aee174e1", "tree": "020798c6aa0d9eee41ce18b91e8f2b4f93ae0c7f", "parents": [ "66e9c6a86b800f60b1e1ea1ff7271f9e6ed1fa96" ], "author": { "name": "Jeff Layton", "email": "jlayton@kernel.org", "time": "Sun Jun 13 19:33:45 2021 -0400" }, "committer": { "name": "David Howells", "email": "dhowells@redhat.com", "time": "Mon Jun 21 21:24:07 2021 +0100" }, "message": "netfs: fix test for whether we can skip read when writing beyond EOF\n\nIt\u0027s not sufficient to skip reading when the pos is beyond the EOF.\nThere may be data at the head of the page that we need to fill in\nbefore the write.\n\nAdd a new helper function that corrects and clarifies the logic of\nwhen we can skip reads, and have it only zero out the part of the page\nthat won\u0027t have data copied in for the write.\n\nFinally, don\u0027t set the page Uptodate after zeroing. It\u0027s not up to date\nsince the write data won\u0027t have been copied in yet.\n\n[DH made the following changes:\n\n - Prefixed the new function with \"netfs_\".\n\n - Don\u0027t call zero_user_segments() for a full-page write.\n\n - Altered the beyond-last-page check to avoid a DIV instruction and got\n rid of then-redundant zero-length file check.\n]\n\nFixes: e1b1240c1ff5f (\"netfs: Add write_begin helper\")\nReported-by: Andrew W Elble \u003caweits@rit.edu\u003e\nSigned-off-by: Jeff Layton \u003cjlayton@kernel.org\u003e\nSigned-off-by: David Howells \u003cdhowells@redhat.com\u003e\nReviewed-by: Matthew Wilcox (Oracle) \u003cwilly@infradead.org\u003e\ncc: ceph-devel@vger.kernel.org\nLink: https://lore.kernel.org/r/20210613233345.113565-1-jlayton@kernel.org/\nLink: https://lore.kernel.org/r/162367683365.460125.4467036947364047314.stgit@warthog.procyon.org.uk/ # v1\nLink: https://lore.kernel.org/r/162391826758.1173366.11794946719301590013.stgit@warthog.procyon.org.uk/ # v2\n" }, { "commit": "66e9c6a86b800f60b1e1ea1ff7271f9e6ed1fa96", "tree": "9d12875d5d73cb8cfbc86c773e35e5fa150a7ff4", "parents": [ "009c9aa5be652675a06d5211e1640e02bbb1c33d" ], "author": { "name": "David Howells", "email": "dhowells@redhat.com", "time": "Mon Jun 14 14:13:41 2021 +0100" }, "committer": { "name": "David Howells", "email": "dhowells@redhat.com", "time": "Mon Jun 21 21:23:36 2021 +0100" }, "message": "afs: Fix afs_write_end() to handle short writes\n\nFix afs_write_end() to correctly handle a short copy into the intended\nwrite region of the page. Two things are necessary:\n\n (1) If the page is not up to date, then we should just return 0\n (ie. indicating a zero-length copy). The loop in\n generic_perform_write() will go around again, possibly breaking up the\n iterator into discrete chunks[1].\n\n This is analogous to commit b9de313cf05fe08fa59efaf19756ec5283af672a\n for ceph.\n\n (2) The page should not have been set uptodate if it wasn\u0027t completely set\n up by netfs_write_begin() (this will be fixed in the next patch), so\n we need to set uptodate here in such a case.\n\nAlso remove the assertion that was checking that the page was set uptodate\nsince it\u0027s now set uptodate if it wasn\u0027t already a few lines above. The\nassertion was from when uptodate was set elsewhere.\n\nChanges:\nv3: Remove the handling of len exceeding the end of the page.\n\nFixes: 3003bbd0697b (\"afs: Use the netfs_write_begin() helper\")\nReported-by: Jeff Layton \u003cjlayton@kernel.org\u003e\nSigned-off-by: David Howells \u003cdhowells@redhat.com\u003e\nAcked-by: Jeff Layton \u003cjlayton@kernel.org\u003e\nReviewed-by: Matthew Wilcox (Oracle) \u003cwilly@infradead.org\u003e\ncc: Al Viro \u003cviro@zeniv.linux.org.uk\u003e\ncc: linux-afs@lists.infradead.org\nLink: https://lore.kernel.org/r/YMwVp268KTzTf8cN@zeniv-ca.linux.org.uk/ [1]\nLink: https://lore.kernel.org/r/162367682522.460125.5652091227576721609.stgit@warthog.procyon.org.uk/ # v1\nLink: https://lore.kernel.org/r/162391825688.1173366.3437507255136307904.stgit@warthog.procyon.org.uk/ # v2\n" }, { "commit": "3093e6cca3ba7d47848068cb256c489675125181", "tree": "76031bc3b8b0175995a07de459862d8a04344f98", "parents": [ "dbec64b11c65d74f31427e2b9d5746fbf17bf840" ], "author": { "name": "Loic Poulain", "email": "loic.poulain@linaro.org", "time": "Thu Jun 17 15:54:13 2021 +0200" }, "committer": { "name": "Bartosz Golaszewski", "email": "bgolaszewski@baylibre.com", "time": "Mon Jun 21 20:43:52 2021 +0200" }, "message": "gpio: mxc: Fix disabled interrupt wake-up support\n\nA disabled/masked interrupt marked as wakeup source must be re-enable\nand unmasked in order to be able to wake-up the host. That can be done\nby flaging the irqchip with IRQCHIP_ENABLE_WAKEUP_ON_SUSPEND.\n\nNote: It \u0027sometimes\u0027 works without that change, but only thanks to the\nlazy generic interrupt disabling (keeping interrupt unmasked).\n\nReported-by: Michal Koziel \u003cmichal.koziel@emlogic.no\u003e\nSigned-off-by: Loic Poulain \u003cloic.poulain@linaro.org\u003e\nReviewed-by: Linus Walleij \u003clinus.walleij@linaro.org\u003e\nSigned-off-by: Bartosz Golaszewski \u003cbgolaszewski@baylibre.com\u003e\n" }, { "commit": "a96bfed64c8986d6404e553f18203cae1f5ac7e6", "tree": "64147aaa297861ab2552a1c3807212ea0d229db8", "parents": [ "13311e74253fe64329390df80bed3f07314ddd61", "dad7b9896a5dbac5da8275d5a6147c65c81fb5f2" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Jun 21 09:49:48 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Jun 21 09:49:48 2021 -0700" }, "message": "Merge tag \u0027for-linus\u0027 of git://git.armlinux.org.uk/~rmk/linux-arm\n\nPull ARM fix from Russell King:\n\n - fix gcc 10 compiler regression with cpu_init()\n\n* tag \u0027for-linus\u0027 of git://git.armlinux.org.uk/~rmk/linux-arm:\n ARM: 9081/1: fix gcc-10 thumb2-kernel regression\n" }, { "commit": "1815d9c86e3090477fbde066ff314a7e9721ee0f", "tree": "b8b020baa41abd8fe9f046745ed84fbeaeb26984", "parents": [ "e541845ae0858616c52dd97df4bf91568c7a7a1b" ], "author": { "name": "Desmond Cheong Zhi Xi", "email": "desmondcheongzx@gmail.com", "time": "Sun Jun 20 19:03:26 2021 +0800" }, "committer": { "name": "Daniel Vetter", "email": "daniel.vetter@ffwll.ch", "time": "Mon Jun 21 17:42:28 2021 +0200" }, "message": "drm: add a locked version of drm_is_current_master\n\nWhile checking the master status of the DRM file in\ndrm_is_current_master(), the device\u0027s master mutex should be\nheld. Without the mutex, the pointer fpriv-\u003emaster may be freed\nconcurrently by another process calling drm_setmaster_ioctl(). This\ncould lead to use-after-free errors when the pointer is subsequently\ndereferenced in drm_lease_owner().\n\nThe callers of drm_is_current_master() from drm_auth.c hold the\ndevice\u0027s master mutex, but external callers do not. Hence, we implement\ndrm_is_current_master_locked() to be used within drm_auth.c, and\nmodify drm_is_current_master() to grab the device\u0027s master mutex\nbefore checking the master status.\n\nReported-by: Daniel Vetter \u003cdaniel.vetter@ffwll.ch\u003e\nSigned-off-by: Desmond Cheong Zhi Xi \u003cdesmondcheongzx@gmail.com\u003e\nReviewed-by: Emil Velikov \u003cemil.l.velikov@gmail.com\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Daniel Vetter \u003cdaniel.vetter@ffwll.ch\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20210620110327.4964-2-desmondcheongzx@gmail.com\n" }, { "commit": "31197d3a0f1caeb60fb01f6755e28347e4f44037", "tree": "0df86c920d5b315fa3c2a97a2965bdef23633253", "parents": [ "13311e74253fe64329390df80bed3f07314ddd61" ], "author": { "name": "Peter Zijlstra", "email": "peterz@infradead.org", "time": "Mon Jun 21 16:13:55 2021 +0200" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Mon Jun 21 17:26:57 2021 +0200" }, "message": "objtool/x86: Ignore __x86_indirect_alt_* symbols\n\nBecause the __x86_indirect_alt* symbols are just that, objtool will\ntry and validate them as regular symbols, instead of the alternative\nreplacements that they are.\n\nThis goes sideways for FRAME_POINTER\u003dy builds; which generate a fair\namount of warnings.\n\nFixes: 9bc0bb50727c (\"objtool/x86: Rewrite retpoline thunk calls\")\nSigned-off-by: Peter Zijlstra (Intel) \u003cpeterz@infradead.org\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\nLink: https://lore.kernel.org/r/YNCgxwLBiK9wclYJ@hirez.programming.kicks-ass.net\n" }, { "commit": "5f89468e2f060031cd89fd4287298e0eaf246bf6", "tree": "1be0fd73073ad09ce92ec7ce1f3c6328679551eb", "parents": [ "dfc06b389a4f54e78c03abecd5b42ab6ea8d492a" ], "author": { "name": "Bumyong Lee", "email": "bumyong.lee@samsung.com", "time": "Mon May 10 18:10:04 2021 +0900" }, "committer": { "name": "Konrad Rzeszutek Wilk", "email": "konrad@kernel.org", "time": "Mon Jun 21 08:59:02 2021 -0400" }, "message": "swiotlb: manipulate orig_addr when tlb_addr has offset\n\nin case of driver wants to sync part of ranges with offset,\nswiotlb_tbl_sync_single() copies from orig_addr base to tlb_addr with\noffset and ends up with data mismatch.\n\nIt was removed from\n\"swiotlb: don\u0027t modify orig_addr in swiotlb_tbl_sync_single\",\nbut said logic has to be added back in.\n\nFrom Linus\u0027s email:\n\"That commit which the removed the offset calculation entirely, because the old\n\n (unsigned long)tlb_addr \u0026 (IO_TLB_SIZE - 1)\n\nwas wrong, but instead of removing it, I think it should have just\nfixed it to be\n\n (tlb_addr - mem-\u003estart) \u0026 (IO_TLB_SIZE - 1);\n\ninstead. That way the slot offset always matches the slot index calculation.\"\n\n(Unfortunatly that broke NVMe).\n\nThe use-case that drivers are hitting is as follow:\n\n1. Get dma_addr_t from dma_map_single()\n\ndma_addr_t tlb_addr \u003d dma_map_single(dev, vaddr, vsize, DMA_TO_DEVICE);\n\n |\u003c---------------vsize-------------\u003e|\n +-----------------------------------+\n | | original buffer\n +-----------------------------------+\n vaddr\n\n swiotlb_align_offset\n |\u003c-----\u003e|\u003c---------------vsize-------------\u003e|\n +-------+-----------------------------------+\n | | | swiotlb buffer\n +-------+-----------------------------------+\n tlb_addr\n\n2. Do something\n3. Sync dma_addr_t through dma_sync_single_for_device(..)\n\ndma_sync_single_for_device(dev, tlb_addr + offset, size, DMA_TO_DEVICE);\n\n Error case.\n Copy data to original buffer but it is from base addr (instead of\n base addr + offset) in original buffer:\n\n swiotlb_align_offset\n |\u003c-----\u003e|\u003c- offset -\u003e|\u003c- size -\u003e|\n +-------+-----------------------------------+\n | | |##########| | swiotlb buffer\n +-------+-----------------------------------+\n tlb_addr\n\n |\u003c- size -\u003e|\n +-----------------------------------+\n |##########| | original buffer\n +-----------------------------------+\n vaddr\n\nThe fix is to copy the data to the original buffer and take into\naccount the offset, like so:\n\n swiotlb_align_offset\n |\u003c-----\u003e|\u003c- offset -\u003e|\u003c- size -\u003e|\n +-------+-----------------------------------+\n | | |##########| | swiotlb buffer\n +-------+-----------------------------------+\n tlb_addr\n\n |\u003c- offset -\u003e|\u003c- size -\u003e|\n +-----------------------------------+\n | |##########| | original buffer\n +-----------------------------------+\n vaddr\n\n[One fix which was Linus\u0027s that made more sense to as it created a\nsymmetry would break NVMe. The reason for that is the:\n unsigned int offset \u003d (tlb_addr - mem-\u003estart) \u0026 (IO_TLB_SIZE - 1);\n\nwould come up with the proper offset, but it would lose the\nalignment (which this patch contains).]\n\nFixes: 16fc3cef33a0 (\"swiotlb: don\u0027t modify orig_addr in swiotlb_tbl_sync_single\")\nSigned-off-by: Bumyong Lee \u003cbumyong.lee@samsung.com\u003e\nSigned-off-by: Chanho Park \u003cchanho61.park@samsung.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nReported-by: Dominique MARTINET \u003cdominique.martinet@atmark-techno.com\u003e\nReported-by: Horia Geantă \u003choria.geanta@nxp.com\u003e\nTested-by: Horia Geantă \u003choria.geanta@nxp.com\u003e\nCC: stable@vger.kernel.org\nSigned-off-by: Konrad Rzeszutek Wilk \u003ckonrad.wilk@oracle.com\u003e\n" }, { "commit": "67147e96a332b56c7206238162771d82467f86c0", "tree": "8641c17c4af3f413489ec76c23a1cd4b32892f1a", "parents": [ "9e3d62d55bf455d4f9fdf2ede5c8756410c64102" ], "author": { "name": "Heiko Carstens", "email": "hca@linux.ibm.com", "time": "Fri Jun 18 16:58:47 2021 +0200" }, "committer": { "name": "Vasily Gorbik", "email": "gor@linux.ibm.com", "time": "Mon Jun 21 11:19:18 2021 +0200" }, "message": "s390/stack: fix possible register corruption with stack switch helper\n\nThe CALL_ON_STACK macro is used to call a C function from inline\nassembly, and therefore must consider the C ABI, which says that only\nregisters 6-13, and 15 are non-volatile (restored by the called\nfunction).\n\nThe inline assembly incorrectly marks all registers used to pass\nparameters to the called function as read-only input operands, instead\nof operands that are read and written to. This might result in\nregister corruption depending on usage, compiler, and compile options.\n\nFix this by marking all operands used to pass parameters as read/write\noperands. To keep the code simple even register 6, if used, is marked\nas read-write operand.\n\nFixes: ff340d2472ec (\"s390: add stack switch helper\")\nCc: \u003cstable@kernel.org\u003e # 4.20\nReviewed-by: Vasily Gorbik \u003cgor@linux.ibm.com\u003e\nSigned-off-by: Heiko Carstens \u003chca@linux.ibm.com\u003e\nSigned-off-by: Vasily Gorbik \u003cgor@linux.ibm.com\u003e\n" }, { "commit": "9e3d62d55bf455d4f9fdf2ede5c8756410c64102", "tree": "aaf22e51576329f329311b7b9d077da19e5143d3", "parents": [ "8c0795d2a0f50e2b131f5b2a8c2795939a94058e" ], "author": { "name": "Sven Schnelle", "email": "svens@linux.ibm.com", "time": "Tue Jun 15 15:05:22 2021 +0200" }, "committer": { "name": "Vasily Gorbik", "email": "gor@linux.ibm.com", "time": "Mon Jun 21 11:19:18 2021 +0200" }, "message": "s390/topology: clear thread/group maps for offline cpus\n\nThe current code doesn\u0027t clear the thread/group maps for offline\nCPUs. This may cause kernel crashes like the one bewlow in common\ncode that assumes if a CPU has sibblings it is online.\n\nUnable to handle kernel pointer dereference in virtual kernel address space\n\nCall Trace:\n [\u003c000000013a4b8c3c\u003e] blk_mq_map_swqueue+0x10c/0x388\n([\u003c000000013a4b8bcc\u003e] blk_mq_map_swqueue+0x9c/0x388)\n [\u003c000000013a4b9300\u003e] blk_mq_init_allocated_queue+0x448/0x478\n [\u003c000000013a4b9416\u003e] blk_mq_init_queue+0x4e/0x90\n [\u003c000003ff8019d3e6\u003e] loop_add+0x106/0x278 [loop]\n [\u003c000003ff801b8148\u003e] loop_init+0x148/0x1000 [loop]\n [\u003c0000000139de4924\u003e] do_one_initcall+0x3c/0x1e0\n [\u003c0000000139ef449a\u003e] do_init_module+0x6a/0x2a0\n [\u003c0000000139ef61bc\u003e] __do_sys_finit_module+0xa4/0xc0\n [\u003c0000000139de9e6e\u003e] do_syscall+0x7e/0xd0\n [\u003c000000013a8e0aec\u003e] __do_syscall+0xbc/0x110\n [\u003c000000013a8ee2e8\u003e] system_call+0x78/0xa0\n\nFixes: 52aeda7accb6 (\"s390/topology: remove offline CPUs from CPU topology masks\")\nCc: \u003cstable@kernel.org\u003e # 5.7+\nReported-by: Marius Hillenbrand \u003cmhillen@linux.ibm.com\u003e\nSigned-off-by: Sven Schnelle \u003csvens@linux.ibm.com\u003e\nReviewed-by: Heiko Carstens \u003chca@linux.ibm.com\u003e\nSigned-off-by: Vasily Gorbik \u003cgor@linux.ibm.com\u003e\n" }, { "commit": "8c0795d2a0f50e2b131f5b2a8c2795939a94058e", "tree": "46dec4cc07c8753010cea30a6c83b94564b2134e", "parents": [ "ca1f4d702d534387aa1f16379edb3b03cdb6ceda" ], "author": { "name": "Tony Krowiak", "email": "akrowiak@linux.ibm.com", "time": "Wed Jun 09 18:46:32 2021 -0400" }, "committer": { "name": "Vasily Gorbik", "email": "gor@linux.ibm.com", "time": "Mon Jun 21 11:19:18 2021 +0200" }, "message": "s390/vfio-ap: clean up mdev resources when remove callback invoked\n\nThe mdev remove callback for the vfio_ap device driver bails out with\n-EBUSY if the mdev is in use by a KVM guest (i.e., the KVM pointer in the\nstruct ap_matrix_mdev is not NULL). The intended purpose was\nto prevent the mdev from being removed while in use. There are two\nproblems with this scenario:\n\n1. Returning a non-zero return code from the remove callback does not\n prevent the removal of the mdev.\n\n2. The KVM pointer in the struct ap_matrix_mdev will always be NULL because\n the remove callback will not get invoked until the mdev fd is closed.\n When the mdev fd is closed, the mdev release callback is invoked and\n clears the KVM pointer from the struct ap_matrix_mdev.\n\nLet\u0027s go ahead and remove the check for KVM in the remove callback and\nallow the cleanup of mdev resources to proceed.\n\nSigned-off-by: Tony Krowiak \u003cakrowiak@linux.ibm.com\u003e\nReviewed-by: Jason Gunthorpe \u003cjgg@nvidia.com\u003e\nLink: https://lore.kernel.org/r/20210609224634.575156-2-akrowiak@linux.ibm.com\nSigned-off-by: Vasily Gorbik \u003cgor@linux.ibm.com\u003e\n" }, { "commit": "ca1f4d702d534387aa1f16379edb3b03cdb6ceda", "tree": "38203bd018acae7d0386ae45d7e3803a04bd626e", "parents": [ "fc66127dc3396338f287c3b494dfbf102547e770" ], "author": { "name": "Sven Schnelle", "email": "svens@linux.ibm.com", "time": "Fri Jun 11 16:08:18 2021 +0200" }, "committer": { "name": "Vasily Gorbik", "email": "gor@linux.ibm.com", "time": "Mon Jun 21 11:19:18 2021 +0200" }, "message": "s390: clear pt_regs::flags on irq entry\n\nThe current irq entry code doesn\u0027t initialize pt_regs::flags. On exit to\nuser mode arch_do_signal_or_restart() tests whether PIF_SYSCALL is set,\nwhich might yield wrong results.\n\nFix this by clearing pt_regs::flags in the entry.S irq handler\ncode.\n\nReported-by: Heiko Carstens \u003chca@linux.ibm.com\u003e\nSigned-off-by: Sven Schnelle \u003csvens@linux.ibm.com\u003e\nReviewed-by: Heiko Carstens \u003chca@linux.ibm.com\u003e\nFixes: 56e62a737028 (\"s390: convert to generic entry\")\nCc: \u003cstable@vger.kernel.org\u003e # 5.12\nSigned-off-by: Vasily Gorbik \u003cgor@linux.ibm.com\u003e\n" }, { "commit": "fc66127dc3396338f287c3b494dfbf102547e770", "tree": "5c18d8af2e864a4251bb9f33983c18cd53ffcfea", "parents": [ "13311e74253fe64329390df80bed3f07314ddd61" ], "author": { "name": "Sven Schnelle", "email": "svens@linux.ibm.com", "time": "Fri Jun 11 10:27:51 2021 +0200" }, "committer": { "name": "Vasily Gorbik", "email": "gor@linux.ibm.com", "time": "Mon Jun 21 11:19:18 2021 +0200" }, "message": "s390: fix system call restart with multiple signals\n\nglibc complained with \"The futex facility returned an unexpected error\ncode.\". It turned out that the futex syscall returned -ERESTARTSYS because\na signal is pending. arch_do_signal_or_restart() restored the syscall\nparameters (nameley regs-\u003egprs[2]) and set PIF_SYSCALL_RESTART. When\nanother signal is made pending later in the exit loop\narch_do_signal_or_restart() is called again. This function clears\nPIF_SYSCALL_RESTART and checks the return code which is set in\nregs-\u003egprs[2]. However, regs-\u003egprs[2] was restored in the previous run\nand no longer contains -ERESTARTSYS, so PIF_SYSCALL_RESTART isn\u0027t set\nagain and the syscall is skipped.\n\nFix this by not clearing PIF_SYSCALL_RESTART - it is already cleared in\n__do_syscall() when the syscall is restarted.\n\nReported-by: Bjoern Walk \u003cbwalk@linux.ibm.com\u003e\nSigned-off-by: Sven Schnelle \u003csvens@linux.ibm.com\u003e\nReviewed-by: Heiko Carstens \u003chca@linux.ibm.com\u003e\nFixes: 56e62a737028 (\"s390: convert to generic entry\")\nCc: \u003cstable@vger.kernel.org\u003e # 5.12\nSigned-off-by: Vasily Gorbik \u003cgor@linux.ibm.com\u003e\n" }, { "commit": "13311e74253fe64329390df80bed3f07314ddd61", "tree": "6d436fd3a5be105a44524317846392003b49aebe", "parents": [ "cba5e97280f53ec7feb656fcdf0ec00a5c6dd539" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Jun 20 15:03:15 2021 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Jun 20 15:03:15 2021 -0700" }, "message": "Linux 5.13-rc7\n" } ], "next": "2269583753d2b8fdd3c861a516ff0cdbfcf4ef0b" }