)]}' { "log": [ { "commit": "2c58d548f5706d085c4b009f6abb945220460632", "tree": "b263f432d3ad778b6970ebbb842bd0a577e88cdc", "parents": [ "4c4f7c19b3c721aed418bc97907b411608c5c6a0" ], "author": { "name": "Eric Biggers", "email": "ebiggers@google.com", "time": "Wed Apr 10 13:21:15 2019 -0700" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Apr 17 12:43:29 2019 -0400" }, "message": "fscrypt: cache decrypted symlink target in -\u003ei_link\n\nPath lookups that traverse encrypted symlink(s) are very slow because\neach encrypted symlink needs to be decrypted each time it\u0027s followed.\nThis also involves dropping out of rcu-walk mode.\n\nMake encrypted symlinks faster by caching the decrypted symlink target\nin -\u003ei_link. The first call to fscrypt_get_symlink() sets it. Then,\nthe existing VFS path lookup code uses the non-NULL -\u003ei_link to take the\nfast path where -\u003eget_link() isn\u0027t called, and lookups in rcu-walk mode\nremain in rcu-walk mode.\n\nAlso set -\u003ei_link immediately when a new encrypted symlink is created.\n\nTo safely free the symlink target after an RCU grace period has elapsed,\nintroduce a new function fscrypt_free_inode(), and make the relevant\nfilesystems call it just before actually freeing the inode.\n\nCc: Al Viro \u003cviro@zeniv.linux.org.uk\u003e\nSigned-off-by: Eric Biggers \u003cebiggers@google.com\u003e\nSigned-off-by: Theodore Ts\u0027o \u003ctytso@mit.edu\u003e\n" }, { "commit": "4c4f7c19b3c721aed418bc97907b411608c5c6a0", "tree": "f277f0ab8ee56d45ad6ae93950645c405ac7f49c", "parents": [ "b01531db6cec2aa330dbc91bfbfaaef4a0d387a4" ], "author": { "name": "Eric Biggers", "email": "ebiggers@google.com", "time": "Wed Apr 10 13:21:14 2019 -0700" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Apr 17 12:43:14 2019 -0400" }, "message": "vfs: use READ_ONCE() to access -\u003ei_link\n\nUse \u0027READ_ONCE(inode-\u003ei_link)\u0027 to explicitly support filesystems caching\nthe symlink target in -\u003ei_link later if it was unavailable at iget()\ntime, or wasn\u0027t easily available. I\u0027ll be doing this in fscrypt, to\nimprove the performance of encrypted symlinks on ext4, f2fs, and ubifs.\n\n-\u003ei_link will start NULL and may later be set to a non-NULL value by a\nsmp_store_release() or cmpxchg_release(). READ_ONCE() is needed on the\nread side. smp_load_acquire() is unnecessary because only a data\ndependency barrier is required. (Thanks to Al for pointing this out.)\n\nAcked-by: Al Viro \u003cviro@zeniv.linux.org.uk\u003e\nSigned-off-by: Eric Biggers \u003cebiggers@google.com\u003e\nSigned-off-by: Theodore Ts\u0027o \u003ctytso@mit.edu\u003e\n" }, { "commit": "b01531db6cec2aa330dbc91bfbfaaef4a0d387a4", "tree": "215e407f512e5b48d4f3269c3ab717f3080fcf51", "parents": [ "d456a33f041af4b54f3ce495a86d00c246165032" ], "author": { "name": "Eric Biggers", "email": "ebiggers@google.com", "time": "Wed Mar 20 11:39:13 2019 -0700" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Apr 17 10:07:51 2019 -0400" }, "message": "fscrypt: fix race where -\u003elookup() marks plaintext dentry as ciphertext\n\n-\u003elookup() in an encrypted directory begins as follows:\n\n1. fscrypt_prepare_lookup():\n a. Try to load the directory\u0027s encryption key.\n b. If the key is unavailable, mark the dentry as a ciphertext name\n via d_flags.\n2. fscrypt_setup_filename():\n a. Try to load the directory\u0027s encryption key.\n b. If the key is available, encrypt the name (treated as a plaintext\n name) to get the on-disk name. Otherwise decode the name\n (treated as a ciphertext name) to get the on-disk name.\n\nBut if the key is concurrently added, it may be found at (2a) but not at\n(1a). In this case, the dentry will be wrongly marked as a ciphertext\nname even though it was actually treated as plaintext.\n\nThis will cause the dentry to be wrongly invalidated on the next lookup,\npotentially causing problems. For example, if the racy -\u003elookup() was\npart of sys_mount(), then the new mount will be detached when anything\ntries to access it. This is despite the mountpoint having a plaintext\npath, which should remain valid now that the key was added.\n\nOf course, this is only possible if there\u0027s a userspace race. Still,\nthe additional kernel-side race is confusing and unexpected.\n\nClose the kernel-side race by changing fscrypt_prepare_lookup() to also\nset the on-disk filename (step 2b), consistent with the d_flags update.\n\nFixes: 28b4c263961c (\"ext4 crypto: revalidate dentry after adding or removing the key\")\nSigned-off-by: Eric Biggers \u003cebiggers@google.com\u003e\nSigned-off-by: Theodore Ts\u0027o \u003ctytso@mit.edu\u003e\n" }, { "commit": "d456a33f041af4b54f3ce495a86d00c246165032", "tree": "5ff534692506594efb7fa3ec0f8e2f34b9808d8b", "parents": [ "0bf3d5c1604ecbbd4e49e9f5b3c79152b87adb0d" ], "author": { "name": "Eric Biggers", "email": "ebiggers@google.com", "time": "Wed Mar 20 11:39:12 2019 -0700" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Apr 17 10:06:32 2019 -0400" }, "message": "fscrypt: only set dentry_operations on ciphertext dentries\n\nPlaintext dentries are always valid, so only set fscrypt_d_ops on\nciphertext dentries.\n\nBesides marginally improved performance, this allows overlayfs to use an\nfscrypt-encrypted upperdir, provided that all the following are true:\n\n (1) The fscrypt encryption key is placed in the keyring before\n\tmounting overlayfs, and remains while the overlayfs is mounted.\n\n (2) The overlayfs workdir uses the same encryption policy.\n\n (3) No dentries for the ciphertext names of subdirectories have been\n\tcreated in the upperdir or workdir yet. (Since otherwise\n\td_splice_alias() will reuse the old dentry with -\u003ed_op set.)\n\nOne potential use case is using an ephemeral encryption key to encrypt\nall files created or changed by a container, so that they can be\nsecurely erased (\"crypto-shredded\") after the container stops.\n\nSigned-off-by: Eric Biggers \u003cebiggers@google.com\u003e\nSigned-off-by: Theodore Ts\u0027o \u003ctytso@mit.edu\u003e\n" }, { "commit": "0bf3d5c1604ecbbd4e49e9f5b3c79152b87adb0d", "tree": "97a5f5cedec9b310490df829e03997036869c9fa", "parents": [ "968dd6d0c6d6b6a989c6ddb9e2584a031b83e7b5" ], "author": { "name": "Eric Biggers", "email": "ebiggers@google.com", "time": "Wed Mar 20 11:39:11 2019 -0700" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Apr 17 10:05:51 2019 -0400" }, "message": "fs, fscrypt: clear DCACHE_ENCRYPTED_NAME when unaliasing directory\n\nMake __d_move() clear DCACHE_ENCRYPTED_NAME on the source dentry. This\nis needed for when d_splice_alias() moves a directory\u0027s encrypted alias\nto its decrypted alias as a result of the encryption key being added.\n\nOtherwise, the decrypted alias will incorrectly be invalidated on the\nnext lookup, causing problems such as unmounting a mount the user just\nmount()ed there.\n\nNote that we don\u0027t have to support arbitrary moves of this flag because\nfscrypt doesn\u0027t allow dentries with DCACHE_ENCRYPTED_NAME to be the\nsource or target of a rename().\n\nFixes: 28b4c263961c (\"ext4 crypto: revalidate dentry after adding or removing the key\")\nReported-by: Sarthak Kukreti \u003csarthakkukreti@chromium.org\u003e\nSigned-off-by: Eric Biggers \u003cebiggers@google.com\u003e\nSigned-off-by: Theodore Ts\u0027o \u003ctytso@mit.edu\u003e\n" }, { "commit": "968dd6d0c6d6b6a989c6ddb9e2584a031b83e7b5", "tree": "747e028b00d4453e8de1b8959b739517e025efb0", "parents": [ "6cc248684d3d23bbd073ae2fa73d3416c0558909" ], "author": { "name": "Eric Biggers", "email": "ebiggers@google.com", "time": "Wed Mar 20 11:39:10 2019 -0700" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Apr 17 09:51:20 2019 -0400" }, "message": "fscrypt: fix race allowing rename() and link() of ciphertext dentries\n\nClose some race conditions where fscrypt allowed rename() and link() on\nciphertext dentries that had been looked up just prior to the key being\nconcurrently added. It\u0027s better to return -ENOKEY in this case.\n\nThis avoids doing the nonsensical thing of encrypting the names a second\ntime when searching for the actual on-disk dir entries. It also\nguarantees that DCACHE_ENCRYPTED_NAME dentries are never rename()d, so\nthe dcache won\u0027t have support all possible combinations of moving\nDCACHE_ENCRYPTED_NAME around during __d_move().\n\nSigned-off-by: Eric Biggers \u003cebiggers@google.com\u003e\nSigned-off-by: Theodore Ts\u0027o \u003ctytso@mit.edu\u003e\n" }, { "commit": "6cc248684d3d23bbd073ae2fa73d3416c0558909", "tree": "1fe0b6a608ccdae37511795fd20857e45435622d", "parents": [ "e37a784d8b6a1e726de5ddc7b4809c086a08db09" ], "author": { "name": "Eric Biggers", "email": "ebiggers@google.com", "time": "Wed Mar 20 11:39:09 2019 -0700" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Apr 17 09:48:46 2019 -0400" }, "message": "fscrypt: clean up and improve dentry revalidation\n\nMake various improvements to fscrypt dentry revalidation:\n\n- Don\u0027t try to handle the case where the per-directory key is removed,\n as this can\u0027t happen without the inode (and dentries) being evicted.\n\n- Flag ciphertext dentries rather than plaintext dentries, since it\u0027s\n ciphertext dentries that need the special handling.\n\n- Avoid doing unnecessary work for non-ciphertext dentries.\n\n- When revalidating ciphertext dentries, try to set up the directory\u0027s\n i_crypt_info to make sure the key is really still absent, rather than\n invalidating all negative dentries as the previous code did. An old\n comment suggested we can\u0027t do this for locking reasons, but AFAICT\n this comment was outdated and it actually works fine.\n\nSigned-off-by: Eric Biggers \u003cebiggers@google.com\u003e\nSigned-off-by: Theodore Ts\u0027o \u003ctytso@mit.edu\u003e\n" }, { "commit": "e37a784d8b6a1e726de5ddc7b4809c086a08db09", "tree": "965d40cec69107aba4324f72db93d896b12d0737", "parents": [ "ff5d3a97075c65731a46453d36e75b9cf925e165" ], "author": { "name": "Eric Biggers", "email": "ebiggers@google.com", "time": "Thu Apr 11 14:32:15 2019 -0700" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Apr 16 18:57:09 2019 -0400" }, "message": "fscrypt: use READ_ONCE() to access -\u003ei_crypt_info\n\n-\u003ei_crypt_info starts out NULL and may later be locklessly set to a\nnon-NULL value by the cmpxchg() in fscrypt_get_encryption_info().\n\nBut -\u003ei_crypt_info is used directly, which technically is incorrect.\nIt\u0027s a data race, and it doesn\u0027t include the data dependency barrier\nneeded to safely dereference the pointer on at least one architecture.\n\nFix this by using READ_ONCE() instead. Note: we don\u0027t need to use\nsmp_load_acquire(), since dereferencing the pointer only requires a data\ndependency barrier, which is already included in READ_ONCE(). We also\ndon\u0027t need READ_ONCE() in places where -\u003ei_crypt_info is unconditionally\ndereferenced, since it must have already been checked.\n\nAlso downgrade the cmpxchg() to cmpxchg_release(), since RELEASE\nsemantics are sufficient on the write side.\n\nSigned-off-by: Eric Biggers \u003cebiggers@google.com\u003e\nSigned-off-by: Theodore Ts\u0027o \u003ctytso@mit.edu\u003e\n" }, { "commit": "ff5d3a97075c65731a46453d36e75b9cf925e165", "tree": "8f45dad94b8d4aa47315eba32133a15f9e91c1cb", "parents": [ "cd0265fcd2eae9004c68ef2123a9dac0dc5a666a" ], "author": { "name": "Eric Biggers", "email": "ebiggers@google.com", "time": "Fri Mar 15 14:16:32 2019 -0700" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Apr 16 18:44:44 2019 -0400" }, "message": "fscrypt: remove WARN_ON_ONCE() when decryption fails\n\nIf decrypting a block fails, fscrypt did a WARN_ON_ONCE(). But WARN is\nmeant for kernel bugs, which this isn\u0027t; this could be hit by fuzzers\nusing fault injection, for example. Also, there is already a proper\nwarning message logged in fscrypt_do_page_crypto(), so the WARN doesn\u0027t\nadd much.\n\nJust remove the unnessary WARN.\n\nSigned-off-by: Eric Biggers \u003cebiggers@google.com\u003e\nSigned-off-by: Theodore Ts\u0027o \u003ctytso@mit.edu\u003e\n" }, { "commit": "cd0265fcd2eae9004c68ef2123a9dac0dc5a666a", "tree": "3d0ecb2c81a38689f21b9f056938cbdaaf6156ed", "parents": [ "dc4060a5dc2557e6b5aa813bf5b73677299d62d2" ], "author": { "name": "Eric Biggers", "email": "ebiggers@google.com", "time": "Mon Mar 18 10:23:33 2019 -0700" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Apr 16 18:37:25 2019 -0400" }, "message": "fscrypt: drop inode argument from fscrypt_get_ctx()\n\nThe only reason the inode is being passed to fscrypt_get_ctx() is to\nverify that the encryption key is available. However, all callers\nalready ensure this because if we get as far as trying to do I/O to an\nencrypted file without the key, there\u0027s already a bug.\n\nTherefore, remove this unnecessary argument.\n\nSigned-off-by: Eric Biggers \u003cebiggers@google.com\u003e\nSigned-off-by: Theodore Ts\u0027o \u003ctytso@mit.edu\u003e\n" }, { "commit": "dc4060a5dc2557e6b5aa813bf5b73677299d62d2", "tree": "83850b3f700cdacac6edaeb80df2fc7c4ba87a8e", "parents": [ "6b3a707736301c2128ca85ce85fb13f60b5e350a" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Apr 14 15:17:41 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Apr 14 15:17:41 2019 -0700" }, "message": "Linux 5.1-rc5\n" }, { "commit": "6b3a707736301c2128ca85ce85fb13f60b5e350a", "tree": "2bf1892cf29121150adece8d1221ecd513a4e792", "parents": [ "4443f8e6ac7755cd775c70d08be8042dc2f936cb", "15fab63e1e57be9fdb5eec1bbc5916e9825e9acb" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Apr 14 15:09:40 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Apr 14 15:09:40 2019 -0700" }, "message": "Merge branch \u0027page-refs\u0027 (page ref overflow)\n\nMerge page ref overflow branch.\n\nJann Horn reported that he can overflow the page ref count with\nsufficient memory (and a filesystem that is intentionally extremely\nslow).\n\nAdmittedly it\u0027s not exactly easy. To have more than four billion\nreferences to a page requires a minimum of 32GB of kernel memory just\nfor the pointers to the pages, much less any metadata to keep track of\nthose pointers. Jann needed a total of 140GB of memory and a specially\ncrafted filesystem that leaves all reads pending (in order to not ever\nfree the page references and just keep adding more).\n\nStill, we have a fairly straightforward way to limit the two obvious\nuser-controllable sources of page references: direct-IO like page\nreferences gotten through get_user_pages(), and the splice pipe page\nduplication. So let\u0027s just do that.\n\n* branch page-refs:\n fs: prevent page refcount overflow in pipe_buf_get\n mm: prevent get_user_pages() from overflowing page refcount\n mm: add \u0027try_get_page()\u0027 helper function\n mm: make page ref count overflow check tighter and more explicit\n" }, { "commit": "15fab63e1e57be9fdb5eec1bbc5916e9825e9acb", "tree": "1353971310401012cf4b108a375e525065d8bec0", "parents": [ "8fde12ca79aff9b5ba951fce1a2641901b8d8e64" ], "author": { "name": "Matthew Wilcox", "email": "willy@infradead.org", "time": "Fri Apr 05 14:02:10 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Apr 14 10:00:04 2019 -0700" }, "message": "fs: prevent page refcount overflow in pipe_buf_get\n\nChange pipe_buf_get() to return a bool indicating whether it succeeded\nin raising the refcount of the page (if the thing in the pipe is a page).\nThis removes another mechanism for overflowing the page refcount. All\ncallers converted to handle a failure.\n\nReported-by: Jann Horn \u003cjannh@google.com\u003e\nSigned-off-by: Matthew Wilcox \u003cwilly@infradead.org\u003e\nCc: stable@kernel.org\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "8fde12ca79aff9b5ba951fce1a2641901b8d8e64", "tree": "23ee93c180e690ccd12257fb677fa9b40e1be53a", "parents": [ "88b1a17dfc3ed7728316478fae0f5ad508f50397" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Apr 11 10:49:19 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Apr 14 10:00:04 2019 -0700" }, "message": "mm: prevent get_user_pages() from overflowing page refcount\n\nIf the page refcount wraps around past zero, it will be freed while\nthere are still four billion references to it. One of the possible\navenues for an attacker to try to make this happen is by doing direct IO\non a page multiple times. This patch makes get_user_pages() refuse to\ntake a new page reference if there are already more than two billion\nreferences to the page.\n\nReported-by: Jann Horn \u003cjannh@google.com\u003e\nAcked-by: Matthew Wilcox \u003cwilly@infradead.org\u003e\nCc: stable@kernel.org\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "88b1a17dfc3ed7728316478fae0f5ad508f50397", "tree": "dd019e0d5979c6aa1ceb93b5f36ea16dca86afab", "parents": [ "f958d7b528b1b40c44cfda5eabe2d82760d868c3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Apr 11 10:14:59 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Apr 14 10:00:04 2019 -0700" }, "message": "mm: add \u0027try_get_page()\u0027 helper function\n\nThis is the same as the traditional \u0027get_page()\u0027 function, but instead\nof unconditionally incrementing the reference count of the page, it only\ndoes so if the count was \"safe\". It returns whether the reference count\nwas incremented (and is marked __must_check, since the caller obviously\nhas to be aware of it).\n\nAlso like \u0027get_page()\u0027, you can\u0027t use this function unless you already\nhad a reference to the page. The intent is that you can use this\nexactly like get_page(), but in situations where you want to limit the\nmaximum reference count.\n\nThe code currently does an unconditional WARN_ON_ONCE() if we ever hit\nthe reference count issues (either zero or negative), as a notification\nthat the conditional non-increment actually happened.\n\nNOTE! The count access for the \"safety\" check is inherently racy, but\nthat doesn\u0027t matter since the buffer we use is basically half the range\nof the reference count (ie we look at the sign of the count).\n\nAcked-by: Matthew Wilcox \u003cwilly@infradead.org\u003e\nCc: Jann Horn \u003cjannh@google.com\u003e\nCc: stable@kernel.org\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "f958d7b528b1b40c44cfda5eabe2d82760d868c3", "tree": "4064635cb1fdb10668227084cb68c2c4be2fb0f8", "parents": [ "1c163f4c7b3f621efff9b28a47abb36f7378d783" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Apr 11 10:06:20 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Apr 14 10:00:04 2019 -0700" }, "message": "mm: make page ref count overflow check tighter and more explicit\n\nWe have a VM_BUG_ON() to check that the page reference count doesn\u0027t\nunderflow (or get close to overflow) by checking the sign of the count.\n\nThat\u0027s all fine, but we actually want to allow people to use a \"get page\nref unless it\u0027s already very high\" helper function, and we want that one\nto use the sign of the page ref (without triggering this VM_BUG_ON).\n\nChange the VM_BUG_ON to only check for small underflows (or _very_ close\nto overflowing), and ignore overflows which have strayed into negative\nterritory.\n\nAcked-by: Matthew Wilcox \u003cwilly@infradead.org\u003e\nCc: Jann Horn \u003cjannh@google.com\u003e\nCc: stable@kernel.org\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "4443f8e6ac7755cd775c70d08be8042dc2f936cb", "tree": "7adb7c79800f27cd327cbafa792d9a847b804937", "parents": [ "b60bc0665e6af8c55b946b67ea8cb235823bb74e", "a89afe58f1a74aac768a5eb77af95ef4ee15beaa" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 13 16:23:16 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 13 16:23:16 2019 -0700" }, "message": "Merge tag \u0027for-linus-20190412\u0027 of git://git.kernel.dk/linux-block\n\nPull block fixes from Jens Axboe:\n \"Set of fixes that should go into this round. This pull is larger than\n I\u0027d like at this time, but there\u0027s really no specific reason for that.\n Some are fixes for issues that went into this merge window, others are\n not. Anyway, this contains:\n\n - Hardware queue limiting for virtio-blk/scsi (Dongli)\n\n - Multi-page bvec fixes for lightnvm pblk\n\n - Multi-bio dio error fix (Jason)\n\n - Remove the cache hint from the io_uring tool side, since we didn\u0027t\n move forward with that (me)\n\n - Make io_uring SETUP_SQPOLL root restricted (me)\n\n - Fix leak of page in error handling for pc requests (Jérôme)\n\n - Fix BFQ regression introduced in this merge window (Paolo)\n\n - Fix break logic for bio segment iteration (Ming)\n\n - Fix NVMe cancel request error handling (Ming)\n\n - NVMe pull request with two fixes (Christoph):\n - fix the initial CSN for nvme-fc (James)\n - handle log page offsets properly in the target (Keith)\"\n\n* tag \u0027for-linus-20190412\u0027 of git://git.kernel.dk/linux-block:\n block: fix the return errno for direct IO\n nvmet: fix discover log page when offsets are used\n nvme-fc: correct csn initialization and increments on error\n block: do not leak memory in bio_copy_user_iov()\n lightnvm: pblk: fix crash in pblk_end_partial_read due to multipage bvecs\n nvme: cancel request synchronously\n blk-mq: introduce blk_mq_complete_request_sync()\n scsi: virtio_scsi: limit number of hw queues by nr_cpu_ids\n virtio-blk: limit number of hw queues by nr_cpu_ids\n block, bfq: fix use after free in bfq_bfqq_expire\n io_uring: restrict IORING_SETUP_SQPOLL to root\n tools/io_uring: remove IOCQE_FLAG_CACHEHIT\n block: don\u0027t use for-inside-for in bio_for_each_segment_all\n" }, { "commit": "b60bc0665e6af8c55b946b67ea8cb235823bb74e", "tree": "ea0e6914750fae6e3d12d3c76de2bf74f10d10d3", "parents": [ "87af0c3813570193c6708f6fddbf85178a5bab7e", "af6b61d7ef58099c82d854395a0e002be6bd036c" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 13 14:47:06 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 13 14:47:06 2019 -0700" }, "message": "Merge tag \u0027nfs-for-5.1-4\u0027 of git://git.linux-nfs.org/projects/trondmy/linux-nfs\n\nPull NFS client bugfixes from Trond Myklebust:\n \"Highlights include:\n\n Stable fix:\n\n - Fix a deadlock in close() due to incorrect draining of RDMA queues\n\n Bugfixes:\n\n - Revert \"SUNRPC: Micro-optimise when the task is known not to be\n sleeping\" as it is causing stack overflows\n\n - Fix a regression where NFSv4 getacl and fs_locations stopped\n working\n\n - Forbid setting AF_INET6 to \"struct sockaddr_in\"-\u003esin_family.\n\n - Fix xfstests failures due to incorrect copy_file_range() return\n values\"\n\n* tag \u0027nfs-for-5.1-4\u0027 of git://git.linux-nfs.org/projects/trondmy/linux-nfs:\n Revert \"SUNRPC: Micro-optimise when the task is known not to be sleeping\"\n NFSv4.1 fix incorrect return value in copy_file_range\n xprtrdma: Fix helper that drains the transport\n NFS: Fix handling of reply page vector\n NFS: Forbid setting AF_INET6 to \"struct sockaddr_in\"-\u003esin_family.\n" }, { "commit": "87af0c3813570193c6708f6fddbf85178a5bab7e", "tree": "21c2cbd1d325431edba79eeb83a6757e2abf5af5", "parents": [ "09bad0df3974ef9e682447f48bb7fd55c48513f3", "5c2442fd78998af60e13aba506d103f7f43f8701" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 13 14:37:49 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 13 14:37:49 2019 -0700" }, "message": "Merge tag \u0027scsi-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi\n\nPull SCSI fix from James Bottomley:\n \"One obvious fix for a ciostor data corruption on error bug\"\n\n* tag \u0027scsi-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:\n scsi: csiostor: fix missing data copy in csio_scsi_err_handler()\n" }, { "commit": "09bad0df3974ef9e682447f48bb7fd55c48513f3", "tree": "6779933d61f3f15eef4451d99e372bf5bff91ee6", "parents": [ "a3b84248628df52c39c8724fb69da1ea0f2c0dc7", "f89b9e1be7da8bb0aac667a0206a00975cefe6d3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 13 14:33:56 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 13 14:33:56 2019 -0700" }, "message": "Merge tag \u0027clk-fixes-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux\n\nPull clk fixes from Stephen Boyd:\n \"Here\u0027s more than a handful of clk driver fixes for changes that came\n in during the merge window:\n\n - Fix the AT91 sama5d2 programmable clk prescaler formula\n\n - A bunch of Amlogic meson clk driver fixes for the VPU clks\n\n - A DMI quirk for Intel\u0027s Bay Trail SoC\u0027s driver to properly mark pmc\n clks as critical only when really needed\n\n - Stop overwriting CLK_SET_RATE_PARENT flag in mediatek\u0027s clk gate\n implementation\n\n - Use the right structure to test for a frequency table in i.MX\u0027s\n PLL_1416x driver\"\n\n* tag \u0027clk-fixes-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux:\n clk: imx: Fix PLL_1416X not rounding rates\n clk: mediatek: fix clk-gate flag setting\n platform/x86: pmc_atom: Drop __initconst on dmi table\n clk: x86: Add system specific quirk to mark clocks as critical\n clk: meson: vid-pll-div: remove warning and return 0 on invalid config\n clk: meson: pll: fix rounding and setting a rate that matches precisely\n clk: meson-g12a: fix VPU clock parents\n clk: meson: g12a: fix VPU clock muxes mask\n clk: meson-gxbb: round the vdec dividers to closest\n clk: at91: fix programmable clock for sama5d2\n" }, { "commit": "a3b84248628df52c39c8724fb69da1ea0f2c0dc7", "tree": "da664789ca89f072f33c5248dc67b1305cfb72a6", "parents": [ "cf60528f8ab805ba03e21ce4ce6ab11647cede9b", "3943af9d01e94330d0cfac6fccdbc829aad50c92" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 13 14:29:21 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 13 14:29:21 2019 -0700" }, "message": "Merge tag \u0027pci-v5.1-fixes-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci\n\nPull PCI fixes from Bjorn Helgaas:\n\n - Add a DMA alias quirk for another Marvell SATA device (Andre\n Przywara)\n\n - Fix a pciehp regression that broke safe removal of devices (Sergey\n Miroshnichenko)\n\n* tag \u0027pci-v5.1-fixes-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci:\n PCI: pciehp: Ignore Link State Changes after powering off a slot\n PCI: Add function 1 DMA alias quirk for Marvell 9170 SATA controller\n" }, { "commit": "cf60528f8ab805ba03e21ce4ce6ab11647cede9b", "tree": "a26903dd378be043fcf530529fce63adec0fbf36", "parents": [ "5ded88718aef7e92a9806f6ff4b89c7f2a4f1570", "cf7cf6977f531acd5dfe55250d0ee8cbbb6f1ae8" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 13 09:03:09 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 13 09:03:09 2019 -0700" }, "message": "Merge tag \u0027powerpc-5.1-5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux\n\nPull powerpc fixes from Michael Ellerman:\n \"A minor build fix for 64-bit FLATMEM configs.\n\n A fix for a boot failure on 32-bit powermacs.\n\n My commit to fix CLOCK_MONOTONIC across Y2038 broke the 32-bit VDSO on\n 64-bit kernels, ie. compat mode, which is only used on big endian.\n\n The rewrite of the SLB code we merged in 4.20 missed the fact that the\n 0x380 exception is also used with the Radix MMU to report out of range\n accesses. This could lead to an oops if userspace tried to read from\n addresses outside the user or kernel range.\n\n Thanks to: Aneesh Kumar K.V, Christophe Leroy, Larry Finger, Nicholas\n Piggin\"\n\n* tag \u0027powerpc-5.1-5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:\n powerpc/mm: Define MAX_PHYSMEM_BITS for all 64-bit configs\n powerpc/64s/radix: Fix radix segment exception handling\n powerpc/vdso32: fix CLOCK_MONOTONIC on PPC64\n powerpc/32: Fix early boot failure with RTAS built-in\n" }, { "commit": "5ded88718aef7e92a9806f6ff4b89c7f2a4f1570", "tree": "a6fcf73123205891b848221eb63b10e051500d65", "parents": [ "6d0a598489ca687e1ebac37eef546526eca41347", "045afc24124d80c6998d9c770844c67912083506" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 13 08:57:00 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Apr 13 08:57:00 2019 -0700" }, "message": "Merge tag \u0027arm64-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux\n\nPull arm64 fixes from Will Deacon:\n \"The main thing is a fix to our FUTEX_WAKE_OP implementation which was\n unbelievably broken, but did actually work for the one scenario that\n GLIBC used to use.\n\n Summary:\n\n - Fix stack unwinding so we ignore user stacks\n\n - Fix ftrace module PLT trampoline initialisation checks\n\n - Fix terminally broken implementation of FUTEX_WAKE_OP atomics\"\n\n* tag \u0027arm64-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:\n arm64: futex: Fix FUTEX_WAKE_OP atomic ops with non-zero result value\n arm64: backtrace: Don\u0027t bother trying to unwind the userspace stack\n arm64/ftrace: fix inadvertent BUG() in trampoline check\n" }, { "commit": "6d0a598489ca687e1ebac37eef546526eca41347", "tree": "8b32e83f59095c09278dfa29cd621ccc9e747883", "parents": [ "122c215bfae884f10a189e6754d9603a06b981c3", "5b77e95dd7790ff6c8fbf1cd8d0104ebed818a03" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 20:54:40 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 20:54:40 2019 -0700" }, "message": "Merge branch \u0027x86-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull x86 fixes from Ingo Molnar:\n \"Fix typos in user-visible resctrl parameters, and also fix assembly\n constraint bugs that might result in miscompilation\"\n\n* \u0027x86-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n x86/asm: Use stricter assembly constraints in bitops\n x86/resctrl: Fix typos in the mba_sc mount option\n" }, { "commit": "122c215bfae884f10a189e6754d9603a06b981c3", "tree": "5402cb8e5c3acb668b8810cf5c8d669a119327b4", "parents": [ "5e6f1fee60a3d80582146835ac01d9808748434f", "07d7e12091f4ab869cc6a4bb276399057e73b0b3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 20:52:28 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 20:52:28 2019 -0700" }, "message": "Merge branch \u0027timers-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull timer fix from Ingo Molnar:\n \"Fix the alarm_timer_remaining() return value\"\n\n* \u0027timers-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n alarmtimer: Return correct remaining time\n" }, { "commit": "5e6f1fee60a3d80582146835ac01d9808748434f", "tree": "49452a45442283463705a17408b66c7650c7e405", "parents": [ "73fdb2c908c64a32e11c72c029d636f556859c0d", "0e9f02450da07fc7b1346c8c32c771555173e397" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 20:50:43 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 20:50:43 2019 -0700" }, "message": "Merge branch \u0027sched-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull scheduler fix from Ingo Molnar:\n \"Fix a NULL pointer dereference crash in certain environments\"\n\n* \u0027sched-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n sched/fair: Do not re-read -\u003eh_load_next during hierarchical load calculation\n" }, { "commit": "73fdb2c908c64a32e11c72c029d636f556859c0d", "tree": "da6f262c6538de2eb1f39937816d3134f5330117", "parents": [ "26e2b81977bb1ad70ff9b2acb4d4cb13c23facfd", "1d54ad944074010609562da5c89e4f5df2f4e5db" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 20:42:30 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 20:42:30 2019 -0700" }, "message": "Merge branch \u0027perf-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull perf fixes from Ingo Molnar:\n \"Six kernel side fixes: three related to NMI handling on AMD systems, a\n race fix, a kexec initialization fix and a PEBS sampling fix\"\n\n* \u0027perf-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n perf/core: Fix perf_event_disable_inatomic() race\n x86/perf/amd: Remove need to check \"running\" bit in NMI handler\n x86/perf/amd: Resolve NMI latency issues for active PMCs\n x86/perf/amd: Resolve race condition when disabling PMC\n perf/x86/intel: Initialize TFA MSR\n perf/x86/intel: Fix handling of wakeup_events for multi-entry PEBS\n" }, { "commit": "26e2b81977bb1ad70ff9b2acb4d4cb13c23facfd", "tree": "cdc210a8dfd71d231d92d07ffe0439e19d3322c0", "parents": [ "6a022984c3feda8cc7d2d8c028b429715d0af531", "90c1cba2b3b3851c151229f61801919b2904d437" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 20:31:08 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 20:31:08 2019 -0700" }, "message": "Merge branch \u0027locking-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull locking fix from Ingo Molnar:\n \"Fixes a crash when accessing /proc/lockdep\"\n\n* \u0027locking-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n locking/lockdep: Zap lock classes even with lock debugging disabled\n" }, { "commit": "6a022984c3feda8cc7d2d8c028b429715d0af531", "tree": "62f00e615799204d3279887ae1bfe1a41c499052", "parents": [ "54c63a75581f4b8d5d4b5660424f926510cfd98c", "325aa19598e410672175ed50982f902d4e3f31c5" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 20:21:59 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 20:21:59 2019 -0700" }, "message": "Merge branch \u0027irq-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull irq fixes from Ingo Molnar:\n \"Two genirq fixes, plus an irqchip driver error handling fix\"\n\n* \u0027irq-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n genirq: Respect IRQCHIP_SKIP_SET_WAKE in irq_chip_set_wake_parent()\n genirq: Initialize request_mutex if CONFIG_SPARSE_IRQ\u003dn\n irqchip/irq-ls1x: Missing error code in ls1x_intc_of_init()\n" }, { "commit": "54c63a75581f4b8d5d4b5660424f926510cfd98c", "tree": "62f1528eeb5d2142e5496beeb709408a7fca3a71", "parents": [ "8ee15f3248660f85102a47410079d408615723d4", "4fa5ecda2bf96be7464eb406df8aba9d89260227" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 20:13:13 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 20:13:13 2019 -0700" }, "message": "Merge branch \u0027core-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip\n\nPull core fixes from Ingo Molnar:\n \"Fix an objtool warning plus fix a u64_to_user_ptr() macro expansion\n bug\"\n\n* \u0027core-urgent-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:\n objtool: Add rewind_stack_do_exit() to the noreturn list\n linux/kernel.h: Use parentheses around argument in u64_to_user_ptr()\n" }, { "commit": "f89b9e1be7da8bb0aac667a0206a00975cefe6d3", "tree": "a052db2d577c89547a1188d072fe8eca197e6027", "parents": [ "b3cf181c65c4d49f86b67b399fe7203ecac730a9" ], "author": { "name": "Leonard Crestez", "email": "leonard.crestez@nxp.com", "time": "Fri Apr 12 14:10:03 2019 +0000" }, "committer": { "name": "Stephen Boyd", "email": "sboyd@kernel.org", "time": "Fri Apr 12 14:21:43 2019 -0700" }, "message": "clk: imx: Fix PLL_1416X not rounding rates\n\nCode which initializes the \"clk_init_data.ops\" checks pll-\u003erate_table\nbefore that field is ever assigned to so it always picks\n\"clk_pll1416x_min_ops\".\n\nThis breaks dynamic rate rounding for features such as cpufreq.\n\nFix by checking pll_clk-\u003erate_table instead, here pll_clk refers to\nthe constant initialization data coming from per-soc clk driver.\n\nSigned-off-by: Leonard Crestez \u003cleonard.crestez@nxp.com\u003e\nFixes: 8646d4dcc7fb (\"clk: imx: Add PLLs driver for imx8mm soc\")\nSigned-off-by: Stephen Boyd \u003csboyd@kernel.org\u003e\n" }, { "commit": "b3cf181c65c4d49f86b67b399fe7203ecac730a9", "tree": "3bb5b4d2afaad2686353e38e2b6a7157db0d3e52", "parents": [ "b995dcca7cf12f208cfd95fd9d5768dca7cccec7" ], "author": { "name": "Weiyi Lu", "email": "weiyi.lu@mediatek.com", "time": "Fri Apr 12 11:30:27 2019 +0800" }, "committer": { "name": "Stephen Boyd", "email": "sboyd@kernel.org", "time": "Fri Apr 12 09:41:49 2019 -0700" }, "message": "clk: mediatek: fix clk-gate flag setting\n\nCLK_SET_RATE_PARENT would be dropped.\nMerge two flag setting together to correct the error.\n\nFixes: 5a1cc4c27ad2 (\"clk: mediatek: Add flags to mtk_gate\")\nCc: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Weiyi Lu \u003cweiyi.lu@mediatek.com\u003e\nReviewed-by: Matthias Brugger \u003cmatthias.bgg@gmail.com\u003e\nSigned-off-by: Stephen Boyd \u003csboyd@kernel.org\u003e\n" }, { "commit": "8ee15f3248660f85102a47410079d408615723d4", "tree": "96e57bb60585f81f2bc99bf112aed8e782d12f08", "parents": [ "4876191cbe2a4702b0067d4156e1d8648ac36801", "8c5165430c0194df92369162d1c7f53f8672baa5" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 08:25:16 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 08:25:16 2019 -0700" }, "message": "Merge tag \u0027dma-mapping-5.1-1\u0027 of git://git.infradead.org/users/hch/dma-mapping\n\nPull dma-mapping fixes from Christoph Hellwig:\n \"Fix a sparc64 sun4v_pci regression introduced in this merged window,\n and a dma-debug stracktrace regression from the big refactor last\n merge window\"\n\n* tag \u0027dma-mapping-5.1-1\u0027 of git://git.infradead.org/users/hch/dma-mapping:\n dma-debug: only skip one stackframe entry\n sparc64/pci_sun4v: fix ATU checks for large DMA masks\n" }, { "commit": "4876191cbe2a4702b0067d4156e1d8648ac36801", "tree": "0bbcb392495738947b71b6dbf31b673a1f0d090f", "parents": [ "8e72d95d993de920a81cd71027d3bb9c22cd0b4e", "3c677d206210f53a4be972211066c0f1cd47fe12" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 08:21:15 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 08:21:15 2019 -0700" }, "message": "Merge tag \u0027iommu-fix-v5.1-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu\n\nPull IOMMU fix from Joerg Roedel:\n \"Fix an AMD IOMMU issue where the driver didn\u0027t correctly setup the\n exclusion range in the hardware registers, resulting in exclusion\n ranges being one page too big.\n\n This can cause data corruption of the address of that last page is\n used by DMA operations\"\n\n* tag \u0027iommu-fix-v5.1-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu:\n iommu/amd: Set exclusion range correctly\n" }, { "commit": "8e72d95d993de920a81cd71027d3bb9c22cd0b4e", "tree": "4fad4a7c2f79b8dc6753ccd64fc35e57ac0488bc", "parents": [ "ea951a943f8d046fcb77e7d259b3c45eb2017ab0", "f16628d6e8c6616b071ffe775908b95e07404cab" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 08:18:37 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 08:18:37 2019 -0700" }, "message": "Merge tag \u0027clang-format-for-linus-v5.1-rc5\u0027 of git://github.com/ojeda/linux\n\nPull clang-format update from Miguel Ojeda:\n \"The usual roughly-per-release .clang-format macro list update\"\n\n* tag \u0027clang-format-for-linus-v5.1-rc5\u0027 of git://github.com/ojeda/linux:\n clang-format: Update with the latest for_each macro list\n" }, { "commit": "ea951a943f8d046fcb77e7d259b3c45eb2017ab0", "tree": "ceec47c2ab84215c864c0ee0efa47d7da5f46d93", "parents": [ "372686e60c84b01efefd4fed3ec00123d9effd0c", "5c41ea6d52003b5bc77c2a82fd5ca7d480237d89" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 08:16:40 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 08:16:40 2019 -0700" }, "message": "Merge tag \u0027mmc-v5.1-rc2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc\n\nPull MMC host fixes from Ulf Hansson:\n\n - alcor: Stabilize data write requests\n\n - sdhci-omap: Fix command error path during tuning\n\n* tag \u0027mmc-v5.1-rc2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc:\n mmc: sdhci-omap: Don\u0027t finish_mrq() on a command error during tuning\n mmc: alcor: don\u0027t write data before command has completed\n" }, { "commit": "372686e60c84b01efefd4fed3ec00123d9effd0c", "tree": "4e81d8bf001033f64aaf3ea878b0e71057f453d9", "parents": [ "f2a7346955e56cacdc2f4ec6165537cc6fe5689a", "9b0dcd0e5a27958b57e3e390f63c098d63a055da" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 08:11:59 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 08:11:59 2019 -0700" }, "message": "Merge tag \u0027sound-5.1-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound\n\nPull sound fixes from Takashi Iwai:\n \"Well, this one became unpleasantly larger than previous pull requests,\n but it\u0027s a kind of usual pattern: now it contains a collection of ASoC\n fixes, and nothing to worry too much.\n\n The fixes for ASoC core (DAPM, DPCM, topology) are all small and just\n covering corner cases. The rest changes are driver-specific, many of\n which are for x86 platforms and new drivers like STM32, in addition to\n the usual fixups for HD-audio\"\n\n* tag \u0027sound-5.1-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound: (66 commits)\n ASoC: wcd9335: Fix missing regmap requirement\n ALSA: hda: Fix racy display power access\n ASoC: pcm: fix error handling when try_module_get() fails.\n ASoC: stm32: sai: fix master clock management\n ASoC: Intel: kbl: fix wrong number of channels\n ALSA: hda - Add two more machines to the power_save_blacklist\n ASoC: pcm: update module refcount if module_get_upon_open is set\n ASoC: core: conditionally increase module refcount on component open\n ASoC: stm32: fix sai driver name initialisation\n ASoC: topology: Use the correct dobj to free enum control values and texts\n ALSA: seq: Fix OOB-reads from strlcpy\n ASoC: intel: skylake: add remove() callback for component driver\n ASoC: cs35l35: Disable regulators on driver removal\n ALSA: xen-front: Do not use stream buffer size before it is set\n ASoC: rockchip: pdm: change dma burst to 8\n ASoC: rockchip: pdm: fix regmap_ops hang issue\n ASoC: simple-card: don\u0027t select DPCM via simple-audio-card\n ASoC: audio-graph-card: don\u0027t select DPCM via audio-graph-card\n ASoC: tlv320aic32x4: Change author\u0027s name\n ALSA: hda/realtek - Add quirk for Tuxedo XC 1509\n ...\n" }, { "commit": "f2a7346955e56cacdc2f4ec6165537cc6fe5689a", "tree": "18efb8077dfa79edaf08d75748bace9ee1c1c3df", "parents": [ "58890f31f98db230e708905a7f0bd700756ebbed", "c5781ffbbd4f742a58263458145fe7f0ac01d9e0" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 08:07:46 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 08:07:46 2019 -0700" }, "message": "Merge tag \u0027acpi-5.1-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm\n\nPull ACPI fix from Rafael Wysocki:\n \"Fix an ACPICA issue introduced during the 4.20 development cycle and\n causing some systems to crash because of leftover operation region\n data still maintained after the operation region in question has gone\n away (Erik Schmauss)\"\n\n* tag \u0027acpi-5.1-rc5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:\n ACPICA: Namespace: remove address node from global list after method termination\n" }, { "commit": "58890f31f98db230e708905a7f0bd700756ebbed", "tree": "cbb93284d6b90cb9db5cbd2e1efec4d8a34a520d", "parents": [ "2d06b235815e6bd20395f3db9ada786a6f7a876e", "788f07ebe040922c89e38f609b7890ba52a09db2" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 08:04:01 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Apr 12 08:04:01 2019 -0700" }, "message": "Merge tag \u0027drm-fixes-2019-04-12\u0027 of git://anongit.freedesktop.org/drm/drm\n\nPull drm fixes from Dave Airlie:\n \"Fixes across the driver spectrum this week, the mediatek fbdev support\n might be a bit late for this round, but I looked over it and it\u0027s not\n very large and seems like a useful feature for them.\n\n Otherwise the main thing is a regression fix for i915 5.0 bug that\n caused black screens on a bunch of Dell XPS 15s I think, I know at\n least Fedora is waiting for this to land, and the udl fix is also for\n a regression since 5.0 where unplugging the device would end badly.\n\n core:\n - make atomic hooks optional\n\n i915:\n - Revert a 5.0 regression where some eDP panels stopped working\n - DSI related fixes for platforms up to IceLake\n - GVT (regression fix, warning fix, use-after free fix)\n\n amdgpu:\n - Cursor fixes\n - missing PCI ID fix for KFD\n - XGMI fix\n - shadow buffer handling after reset fix\n\n udl:\n - fix unplugging device crashes.\n\n mediatek:\n - stabilise MT2701 HDMI support\n - fbdev support\n\n tegra:\n - fix for build regression in rc1.\n\n sun4i:\n - Allwinner A6 max freq improvements\n - null ptr deref fix\n\n dw-hdmi:\n - SCDC configuration improvements\n\n omap:\n - CEC clock management policy fix\"\n\n* tag \u0027drm-fixes-2019-04-12\u0027 of git://anongit.freedesktop.org/drm/drm: (32 commits)\n gpu: host1x: Fix compile error when IOMMU API is not available\n drm/i915/gvt: Roundup fb-\u003eheight into tile\u0027s height at calucation fb-\u003esize\n drm/i915/dp: revert back to max link rate and lane count on eDP\n drm/i915/icl: Fix port disable sequence for mipi-dsi\n drm/i915/icl: Ungate ddi clocks before IO enable\n drm/mediatek: no change parent rate in round_rate() for MT2701 hdmi phy\n drm/mediatek: using new factor for tvdpll for MT2701 hdmi phy\n drm/mediatek: remove flag CLK_SET_RATE_PARENT for MT2701 hdmi phy\n drm/mediatek: make implementation of recalc_rate() for MT2701 hdmi phy\n drm/mediatek: fix the rate and divder of hdmi phy for MT2701\n drm/mediatek: fix possible object reference leak\n drm/i915: Get power refs in encoder-\u003eget_power_domains()\n drm/i915: Fix pipe_bpp readout for BXT/GLK DSI\n drm/amd/display: Fix negative cursor pos programming (v2)\n drm/sun4i: tcon top: Fix NULL/invalid pointer dereference in sun8i_tcon_top_un/bind\n drm/udl: add a release method and delay modeset teardown\n drm/i915/gvt: Prevent use-after-free in ppgtt_free_all_spt()\n drm/i915/gvt: Annotate iomem usage\n drm/sun4i: DW HDMI: Lower max. supported rate for H6\n Revert \"Documentation/gpu/meson: Remove link to meson_canvas.c\"\n ...\n" }, { "commit": "045afc24124d80c6998d9c770844c67912083506", "tree": "8e02bd1946e0704bff6f2e96ec758eb8f61b60cc", "parents": [ "1e6f5440a6814d28c32d347f338bfef68bc3e69d" ], "author": { "name": "Will Deacon", "email": "will.deacon@arm.com", "time": "Mon Apr 08 12:45:09 2019 +0100" }, "committer": { "name": "Will Deacon", "email": "will.deacon@arm.com", "time": "Fri Apr 12 15:04:33 2019 +0100" }, "message": "arm64: futex: Fix FUTEX_WAKE_OP atomic ops with non-zero result value\n\nRather embarrassingly, our futex() FUTEX_WAKE_OP implementation doesn\u0027t\nexplicitly set the return value on the non-faulting path and instead\nleaves it holding the result of the underlying atomic operation. This\nmeans that any FUTEX_WAKE_OP atomic operation which computes a non-zero\nvalue will be reported as having failed. Regrettably, I wrote the buggy\ncode back in 2011 and it was upstreamed as part of the initial arm64\nsupport in 2012.\n\nThe reasons we appear to get away with this are:\n\n 1. FUTEX_WAKE_OP is rarely used and therefore doesn\u0027t appear to get\n exercised by futex() test applications\n\n 2. If the result of the atomic operation is zero, the system call\n behaves correctly\n\n 3. Prior to version 2.25, the only operation used by GLIBC set the\n futex to zero, and therefore worked as expected. From 2.25 onwards,\n FUTEX_WAKE_OP is not used by GLIBC at all.\n\nFix the implementation by ensuring that the return value is either 0\nto indicate that the atomic operation completed successfully, or -EFAULT\nif we encountered a fault when accessing the user mapping.\n\nCc: \u003cstable@kernel.org\u003e\nFixes: 6170a97460db (\"arm64: Atomic operations\")\nSigned-off-by: Will Deacon \u003cwill.deacon@arm.com\u003e\n" }, { "commit": "3c677d206210f53a4be972211066c0f1cd47fe12", "tree": "0784753e89b9847d082df06b83e6a915b0644d5d", "parents": [ "15ade5d2e7775667cf191cf2f94327a4889f8b9d" ], "author": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Fri Apr 12 12:50:31 2019 +0200" }, "committer": { "name": "Joerg Roedel", "email": "jroedel@suse.de", "time": "Fri Apr 12 12:59:45 2019 +0200" }, "message": "iommu/amd: Set exclusion range correctly\n\nThe exlcusion range limit register needs to contain the\nbase-address of the last page that is part of the range, as\nbits 0-11 of this register are treated as 0xfff by the\nhardware for comparisons.\n\nSo correctly set the exclusion range in the hardware to the\nlast page which is _in_ the range.\n\nFixes: b2026aa2dce44 (\u0027x86, AMD IOMMU: add functions for programming IOMMU MMIO space\u0027)\nSigned-off-by: Joerg Roedel \u003cjroedel@suse.de\u003e\n" }, { "commit": "f16628d6e8c6616b071ffe775908b95e07404cab", "tree": "5b281a6270feca006fb6256a12243a8160612d23", "parents": [ "8c2ffd9174779014c3fe1f96d9dc3641d9175f00" ], "author": { "name": "Miguel Ojeda", "email": "miguel.ojeda.sandonis@gmail.com", "time": "Sat Mar 30 09:20:16 2019 +0100" }, "committer": { "name": "Miguel Ojeda", "email": "miguel.ojeda.sandonis@gmail.com", "time": "Fri Apr 12 12:49:54 2019 +0200" }, "message": "clang-format: Update with the latest for_each macro list\n\nRe-run the shell fragment that generated the original list now that\nthere are two dozens of new entries after v5.1\u0027s merge window.\n\nSigned-off-by: Miguel Ojeda \u003cmiguel.ojeda.sandonis@gmail.com\u003e\n" }, { "commit": "1d54ad944074010609562da5c89e4f5df2f4e5db", "tree": "1840b09597a2d00ad131bbf9c94931f6123248a1", "parents": [ "3966c3feca3fd10b2935caa0b4a08c7dd59469e5" ], "author": { "name": "Peter Zijlstra", "email": "peterz@infradead.org", "time": "Thu Apr 04 15:03:00 2019 +0200" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Fri Apr 12 08:55:55 2019 +0200" }, "message": "perf/core: Fix perf_event_disable_inatomic() race\n\nThomas-Mich Richter reported he triggered a WARN()ing from event_function_local()\non his s390. The problem boils down to:\n\n\tCPU-A\t\t\t\tCPU-B\n\n\tperf_event_overflow()\n\t perf_event_disable_inatomic()\n\t @pending_disable \u003d 1\n\t irq_work_queue();\n\n\tsched-out\n\t event_sched_out()\n\t @pending_disable \u003d 0\n\n\t\t\t\t\tsched-in\n\t\t\t\t\tperf_event_overflow()\n\t\t\t\t\t perf_event_disable_inatomic()\n\t\t\t\t\t @pending_disable \u003d 1;\n\t\t\t\t\t irq_work_queue(); // FAILS\n\n\tirq_work_run()\n\t perf_pending_event()\n\t if (@pending_disable)\n\t perf_event_disable_local(); // WHOOPS\n\nThe problem exists in generic, but s390 is particularly sensitive\nbecause it doesn\u0027t implement arch_irq_work_raise(), nor does it call\nirq_work_run() from it\u0027s PMU interrupt handler (nor would that be\nsufficient in this case, because s390 also generates\nperf_event_overflow() from pmu::stop). Add to that the fact that s390\nis a virtual architecture and (virtual) CPU-A can stall long enough\nfor the above race to happen, even if it would self-IPI.\n\nAdding a irq_work_sync() to event_sched_in() would work for all hardare\nPMUs that properly use irq_work_run() but fails for software PMUs.\n\nInstead encode the CPU number in @pending_disable, such that we can\ntell which CPU requested the disable. This then allows us to detect\nthe above scenario and even redirect the IPI to make up for the failed\nqueue.\n\nReported-by: Thomas-Mich Richter \u003ctmricht@linux.ibm.com\u003e\nTested-by: Thomas Richter \u003ctmricht@linux.ibm.com\u003e\nSigned-off-by: Peter Zijlstra (Intel) \u003cpeterz@infradead.org\u003e\nAcked-by: Mark Rutland \u003cmark.rutland@arm.com\u003e\nCc: Alexander Shishkin \u003calexander.shishkin@linux.intel.com\u003e\nCc: Arnaldo Carvalho de Melo \u003cacme@redhat.com\u003e\nCc: Heiko Carstens \u003cheiko.carstens@de.ibm.com\u003e\nCc: Hendrik Brueckner \u003cbrueckner@linux.ibm.com\u003e\nCc: Jiri Olsa \u003cjolsa@redhat.com\u003e\nCc: Kees Cook \u003ckeescook@chromium.org\u003e\nCc: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nCc: Martin Schwidefsky \u003cschwidefsky@de.ibm.com\u003e\nCc: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\n" }, { "commit": "788f07ebe040922c89e38f609b7890ba52a09db2", "tree": "37b96117c5788c996f68467ca64b10ac3ce56a70", "parents": [ "ffb5d6fe140d2ef79c2a91861fc18a4a7a1f510a", "3f5f5d534bd40b666cf37bbeeb48bfe6c2efc1e0" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Apr 12 13:39:22 2019 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Apr 12 13:39:32 2019 +1000" }, "message": "Merge tag \u0027drm-intel-fixes-2019-04-11\u0027 of git://anongit.freedesktop.org/drm/drm-intel into drm-fixes\n\n- Revert back to max link rate and lane count on eDP.\n- DSI related fixes for all platforms including Ice Lake.\n- GVT Fixes including one vGPU display plane size regression fix,\none for preventing use-after-free in ppgtt shadow free function,\nand another warning fix for iomem access annotation.\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n\nFrom: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20190411235832.GA6476@intel.com\n" }, { "commit": "a89afe58f1a74aac768a5eb77af95ef4ee15beaa", "tree": "f4ca4eb4bc4793646997856ccae8f03be3cd4f41", "parents": [ "c0c14e935032abd6ee6828840a83df1de49d4838" ], "author": { "name": "Jason Yan", "email": "yanaijie@huawei.com", "time": "Fri Apr 12 10:09:16 2019 +0800" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Thu Apr 11 21:22:21 2019 -0600" }, "message": "block: fix the return errno for direct IO\n\nIf the last bio returned is not dio-\u003ebio, the status of the bio will\nnot assigned to dio-\u003ebio if it is error. This will cause the whole IO\nstatus wrong.\n\n ksoftirqd/21-117 [021] ..s. 4017.966090: 8,0 C N 4883648 [0]\n \u003cidle\u003e-0 [018] ..s. 4017.970888: 8,0 C WS 4924800 + 1024 [0]\n \u003cidle\u003e-0 [018] ..s. 4017.970909: 8,0 D WS 4935424 + 1024 [\u003cidle\u003e]\n \u003cidle\u003e-0 [018] ..s. 4017.970924: 8,0 D WS 4936448 + 321 [\u003cidle\u003e]\n ksoftirqd/21-117 [021] ..s. 4017.995033: 8,0 C R 4883648 + 336 [65475]\n ksoftirqd/21-117 [021] d.s. 4018.001988: myprobe1: (blkdev_bio_end_io+0x0/0x168) bi_status\u003d7\n ksoftirqd/21-117 [021] d.s. 4018.001992: myprobe: (aio_complete_rw+0x0/0x148) x0\u003d0xffff802f2595ad80 res\u003d0x12a000 res2\u003d0x0\n\nWe always have to assign bio-\u003ebi_status to dio-\u003ebio.bi_status because we\nwill only check dio-\u003ebio.bi_status when we return the whole IO to\nthe upper layer.\n\nFixes: 542ff7bf18c6 (\"block: new direct I/O implementation\")\nCc: stable@vger.kernel.org\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nCc: Jens Axboe \u003caxboe@kernel.dk\u003e\nReviewed-by: Ming Lei \u003cming.lei@redhat.com\u003e\nSigned-off-by: Jason Yan \u003cyanaijie@huawei.com\u003e\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "2d06b235815e6bd20395f3db9ada786a6f7a876e", "tree": "38dd49f6dabc62a8daa8c59243f148c955a7bc56", "parents": [ "582549e3fbe137eb6ce9be591aca25c2222a36b4", "272e5326c7837697882ce3162029ba893059b616" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Apr 11 14:19:02 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Apr 11 14:19:02 2019 -0700" }, "message": "Merge tag \u0027for-5.1-rc4-tag\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux\n\nPull btrfs fixes from David Sterba:\n\n - fix parsing of compression algorithm when set as a inode property,\n this could end up with eg. \u0027zst\u0027 or \u0027zli\u0027 in the value\n\n - don\u0027t allow trim on a filesystem with unreplayed log, this could\n cause data loss if there are pending updates to the block groups that\n would not be subject to trim after replay\n\n* tag \u0027for-5.1-rc4-tag\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:\n btrfs: prop: fix vanished compression property after failed set\n btrfs: prop: fix zstd compression parameter validation\n Btrfs: do not allow trimming when a fs is mounted with the nologreplay option\n" }, { "commit": "ffb5d6fe140d2ef79c2a91861fc18a4a7a1f510a", "tree": "6640724c3148343cf21190e1032cdf9856c85d73", "parents": [ "d47703d43ecaa9189d70fb5d151a6883cc44afd3", "1a07a94b47b1f528f39c3e6187b5eaf02efe44ea" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Apr 12 06:55:20 2019 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Apr 12 06:55:25 2019 +1000" }, "message": "Merge tag \u0027drm-misc-fixes-2019-04-11\u0027 of git://anongit.freedesktop.org/drm/drm-misc into drm-fixes\n\n - core: Make atomic_enable and disable optional for CRTC\n - dw-hdmi: Lower max frequency for the Allwinner H6, SCDC configuration\n improvements for older controller versions\n - omap: a fix for the CEC clock management policy\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n\nFrom: Maxime Ripard \u003cmaxime.ripard@bootlin.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20190411151658.orm46ccd5zmrw27l@flea\n" }, { "commit": "af6b61d7ef58099c82d854395a0e002be6bd036c", "tree": "870be0f84564dbbc5ae82534bb6d98a2054ce865", "parents": [ "0769663b4f580566ef6cdf366f3073dbe8022c39" ], "author": { "name": "Trond Myklebust", "email": "trond.myklebust@hammerspace.com", "time": "Thu Apr 11 15:16:52 2019 -0400" }, "committer": { "name": "Trond Myklebust", "email": "trond.myklebust@hammerspace.com", "time": "Thu Apr 11 15:41:14 2019 -0400" }, "message": "Revert \"SUNRPC: Micro-optimise when the task is known not to be sleeping\"\n\nThis reverts commit 009a82f6437490c262584d65a14094a818bcb747.\n\nThe ability to optimise here relies on compiler being able to optimise\naway tail calls to avoid stack overflows. Unfortunately, we are seeing\nreports of problems, so let\u0027s just revert.\n\nReported-by: Daniel Mack \u003cdaniel@zonque.org\u003e\nSigned-off-by: Trond Myklebust \u003ctrond.myklebust@hammerspace.com\u003e\n" }, { "commit": "0769663b4f580566ef6cdf366f3073dbe8022c39", "tree": "96eb99500c11d736c336f960b3e12bbdeb132e51", "parents": [ "e1ede312f17e96a9c5cda9aaa1cdcf442c1a5da8" ], "author": { "name": "Olga Kornievskaia", "email": "kolga@netapp.com", "time": "Thu Apr 11 14:34:18 2019 -0400" }, "committer": { "name": "Trond Myklebust", "email": "trond.myklebust@hammerspace.com", "time": "Thu Apr 11 15:23:48 2019 -0400" }, "message": "NFSv4.1 fix incorrect return value in copy_file_range\n\nAccording to the NFSv4.2 spec if the input and output file is the\nsame file, operation should fail with EINVAL. However, linux\ncopy_file_range() system call has no such restrictions. Therefore,\nin such case let\u0027s return EOPNOTSUPP and allow VFS to fallback\nto doing do_splice_direct(). Also when copy_file_range is called\non an NFSv4.0 or 4.1 mount (ie., a server that doesn\u0027t support\nCOPY functionality), we also need to return EOPNOTSUPP and\nfallback to a regular copy.\n\nFixes xfstest generic/075, generic/091, generic/112, generic/263\nfor all NFSv4.x versions.\n\nSigned-off-by: Olga Kornievskaia \u003ckolga@netapp.com\u003e\nSigned-off-by: Trond Myklebust \u003ctrond.myklebust@hammerspace.com\u003e\n" }, { "commit": "e1ede312f17e96a9c5cda9aaa1cdcf442c1a5da8", "tree": "5b392c34b573400b3077934c78c7b575a5bec41d", "parents": [ "29e7ca715f2a0b6c0a99b1aec1b0956d1f271955" ], "author": { "name": "Chuck Lever", "email": "chuck.lever@oracle.com", "time": "Tue Apr 09 17:04:09 2019 -0400" }, "committer": { "name": "Trond Myklebust", "email": "trond.myklebust@hammerspace.com", "time": "Thu Apr 11 15:23:48 2019 -0400" }, "message": "xprtrdma: Fix helper that drains the transport\n\nWe want to drain only the RQ first. Otherwise the transport can\ndeadlock on -\u003eclose if there are outstanding Send completions.\n\nFixes: 6d2d0ee27c7a (\"xprtrdma: Replace rpcrdma_receive_wq ... \")\nSigned-off-by: Chuck Lever \u003cchuck.lever@oracle.com\u003e\nCc: stable@vger.kernel.org # v5.0+\nSigned-off-by: Trond Myklebust \u003ctrond.myklebust@hammerspace.com\u003e\n" }, { "commit": "29e7ca715f2a0b6c0a99b1aec1b0956d1f271955", "tree": "adcd49b88fe18a38fc112545f22b0c3a0559a4a5", "parents": [ "7c2bd9a39845bfb6d72ddb55ce737650271f6f96" ], "author": { "name": "Chuck Lever", "email": "chuck.lever@oracle.com", "time": "Tue Apr 09 10:44:16 2019 -0400" }, "committer": { "name": "Trond Myklebust", "email": "trond.myklebust@hammerspace.com", "time": "Thu Apr 11 15:23:48 2019 -0400" }, "message": "NFS: Fix handling of reply page vector\n\nNFSv4 GETACL and FS_LOCATIONS requests stopped working in v5.1-rc.\n\nThese two need the extra padding to be added directly to the reply\nlength.\n\nReported-by: Olga Kornievskaia \u003caglo@umich.edu\u003e\nFixes: 02ef04e432ba (\"NFS: Account for XDR pad of buf-\u003epages\")\nSigned-off-by: Chuck Lever \u003cchuck.lever@oracle.com\u003e\nTested-by: Olga Kornievskaia \u003caglo@umich.edu\u003e\nSigned-off-by: Trond Myklebust \u003ctrond.myklebust@hammerspace.com\u003e\n" }, { "commit": "7c2bd9a39845bfb6d72ddb55ce737650271f6f96", "tree": "1418f0b2012665a16921443a17856c53ee3eb18b", "parents": [ "582549e3fbe137eb6ce9be591aca25c2222a36b4" ], "author": { "name": "Tetsuo Handa", "email": "penguin-kernel@I-love.SAKURA.ne.jp", "time": "Sat Mar 30 10:21:07 2019 +0900" }, "committer": { "name": "Trond Myklebust", "email": "trond.myklebust@hammerspace.com", "time": "Thu Apr 11 15:23:48 2019 -0400" }, "message": "NFS: Forbid setting AF_INET6 to \"struct sockaddr_in\"-\u003esin_family.\n\nsyzbot is reporting uninitialized value at rpc_sockaddr2uaddr() [1]. This\nis because syzbot is setting AF_INET6 to \"struct sockaddr_in\"-\u003esin_family\n(which is embedded into user-visible \"struct nfs_mount_data\" structure)\ndespite nfs23_validate_mount_data() cannot pass sizeof(struct sockaddr_in6)\nbytes of AF_INET6 address to rpc_sockaddr2uaddr().\n\nSince \"struct nfs_mount_data\" structure is user-visible, we can\u0027t change\n\"struct nfs_mount_data\" to use \"struct sockaddr_storage\". Therefore,\nassuming that everybody is using AF_INET family when passing address via\n\"struct nfs_mount_data\"-\u003eaddr, reject if its sin_family is not AF_INET.\n\n[1] https://syzkaller.appspot.com/bug?id\u003d599993614e7cbbf66bc2656a919ab2a95fb5d75c\n\nReported-by: syzbot \u003csyzbot+047a11c361b872896a4f@syzkaller.appspotmail.com\u003e\nSigned-off-by: Tetsuo Handa \u003cpenguin-kernel@I-love.SAKURA.ne.jp\u003e\nSigned-off-by: Trond Myklebust \u003ctrond.myklebust@hammerspace.com\u003e\n" }, { "commit": "8c5165430c0194df92369162d1c7f53f8672baa5", "tree": "739255d87917d08f3da0df4004e9d8d26fd8e139", "parents": [ "2a29e9f6b9b499f1fc5f4a48220dc3f4428499f9" ], "author": { "name": "Scott Wood", "email": "swood@redhat.com", "time": "Wed Apr 10 16:59:25 2019 -0500" }, "committer": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Thu Apr 11 19:29:50 2019 +0200" }, "message": "dma-debug: only skip one stackframe entry\n\nWith skip set to 1, I get a traceback like this:\n\n[ 106.867637] DMA-API: Mapped at:\n[ 106.870784] afu_dma_map_region+0x2cd/0x4f0 [dfl_afu]\n[ 106.875839] afu_ioctl+0x258/0x380 [dfl_afu]\n[ 106.880108] do_vfs_ioctl+0xa9/0x720\n[ 106.883688] ksys_ioctl+0x60/0x90\n[ 106.887007] __x64_sys_ioctl+0x16/0x20\n\nWith the previous value of 2, afu_dma_map_region was being omitted. I\nsuspect that the code paths have simply changed since the value of 2 was\nchosen a decade ago, but it\u0027s also possible that it varies based on which\nmapping function was used, compiler inlining choices, etc. In any case,\nit\u0027s best to err on the side of skipping less.\n\nSigned-off-by: Scott Wood \u003cswood@redhat.com\u003e\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\n" }, { "commit": "b995dcca7cf12f208cfd95fd9d5768dca7cccec7", "tree": "644a2ba1a6fefbe99fb1cb3a14dc01ffdf73d2a3", "parents": [ "7c2e07130090ae001a97a6b65597830d6815e93e" ], "author": { "name": "Stephen Boyd", "email": "sboyd@kernel.org", "time": "Thu Apr 11 10:22:43 2019 -0700" }, "committer": { "name": "Stephen Boyd", "email": "sboyd@kernel.org", "time": "Thu Apr 11 10:25:37 2019 -0700" }, "message": "platform/x86: pmc_atom: Drop __initconst on dmi table\n\nIt\u0027s used by probe and that isn\u0027t an init function. Drop this so that we\ndon\u0027t get a section mismatch.\n\nReported-by: kbuild test robot \u003clkp@intel.com\u003e\nCc: David Müller \u003cdave.mueller@gmx.ch\u003e\nCc: Hans de Goede \u003chdegoede@redhat.com\u003e\nCc: Andy Shevchenko \u003candy.shevchenko@gmail.com\u003e\nFixes: 7c2e07130090 (\"clk: x86: Add system specific quirk to mark clocks as critical\")\nSigned-off-by: Stephen Boyd \u003csboyd@kernel.org\u003e\n" }, { "commit": "3f5f5d534bd40b666cf37bbeeb48bfe6c2efc1e0", "tree": "8fd610851e27192504e9cac44509d5aefab59f6d", "parents": [ "21635d7311734d2d1b177f8a95e2f9386174b76d", "cd7879f79f83aec4bb13f0f823f323911dc5397b" ], "author": { "name": "Rodrigo Vivi", "email": "rodrigo.vivi@intel.com", "time": "Thu Apr 11 09:18:13 2019 -0700" }, "committer": { "name": "Rodrigo Vivi", "email": "rodrigo.vivi@intel.com", "time": "Thu Apr 11 09:18:14 2019 -0700" }, "message": "Merge tag \u0027gvt-fixes-2019-04-11\u0027 of https://github.com/intel/gvt-linux into drm-intel-fixes\n\ngvt-fixes-2019-04-11\n\n- Fix sparse warning on iomem usage (Chris)\n- Prevent use-after-free for ppgtt shadow table free (Chris)\n- Fix display plane size regression for tiled surface (Xiong)\n\nSigned-off-by: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\nFrom: Zhenyu Wang \u003czhenyuw@linux.intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20190411064910.GF17995@zhen-hp.sh.intel.com\n" }, { "commit": "c0c14e935032abd6ee6828840a83df1de49d4838", "tree": "6e5d46c670e1ed35f4cddadfec3f7b8c563d23cb", "parents": [ "a3761c3c91209b58b6f33bf69dd8bb8ec0c9d925", "d808b7f759b50acf0784ce6230ffa63e12ef465d" ], "author": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Thu Apr 11 09:36:41 2019 -0600" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Thu Apr 11 09:36:41 2019 -0600" }, "message": "Merge branch \u0027nvme-5.1\u0027 of git://git.infradead.org/nvme into for-linus\n\nPull NVMe fixes from Christoph:\n\n\"Two nvme fixes for 5.1 - fixing the initial CSN for nvme-fc, and handle\n log page offsets properly in the target.\"\n\n* \u0027nvme-5.1\u0027 of git://git.infradead.org/nvme:\n nvmet: fix discover log page when offsets are used\n nvme-fc: correct csn initialization and increments on error\n" }, { "commit": "d808b7f759b50acf0784ce6230ffa63e12ef465d", "tree": "6e5d46c670e1ed35f4cddadfec3f7b8c563d23cb", "parents": [ "67f471b6ed3b09033c4ac77ea03f92afdb1989fe" ], "author": { "name": "Keith Busch", "email": "keith.busch@intel.com", "time": "Tue Apr 09 10:03:59 2019 -0600" }, "committer": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Thu Apr 11 17:28:30 2019 +0200" }, "message": "nvmet: fix discover log page when offsets are used\n\nThe nvme target hadn\u0027t been taking the Get Log Page offset parameter\ninto consideration, and so has been returning corrupted log pages when\noffsets are used. Since many tools, including nvme-cli, split the log\nrequest to 4k, we\u0027ve been breaking discovery log responses when more\nthan 3 subsystems exist.\n\nFix the returned data by internally generating the entire discovery\nlog page and copying only the requested bytes into the user buffer. The\ncommand log page offset type has been modified to a native __le64 to\nmake it easier to extract the value from a command.\n\nSigned-off-by: Keith Busch \u003ckeith.busch@intel.com\u003e\nTested-by: Minwoo Im \u003cminwoo.im@samsung.com\u003e\nReviewed-by: Chaitanya Kulkarni \u003cchaitanya.kulkarni@wdc.com\u003e\nReviewed-by: Hannes Reinecke \u003chare@suse.com\u003e\nReviewed-by: James Smart \u003cjames.smart@broadcom.com\u003e\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\n" }, { "commit": "67f471b6ed3b09033c4ac77ea03f92afdb1989fe", "tree": "19f2e0412e291eb83411fd57775cf879689e2ba0", "parents": [ "a3761c3c91209b58b6f33bf69dd8bb8ec0c9d925" ], "author": { "name": "James Smart", "email": "jsmart2021@gmail.com", "time": "Mon Apr 08 11:15:19 2019 -0700" }, "committer": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Thu Apr 11 17:28:30 2019 +0200" }, "message": "nvme-fc: correct csn initialization and increments on error\n\nThis patch fixes a long-standing bug that initialized the FC-NVME\ncmnd iu CSN value to 1. Early FC-NVME specs had the connection starting\nwith CSN\u003d1. By the time the spec reached approval, the language had\nchanged to state a connection should start with CSN\u003d0. This patch\ncorrects the initialization value for FC-NVME connections.\n\nAdditionally, in reviewing the transport, the CSN value is assigned to\nthe new IU early in the start routine. It\u0027s possible that a later dma\nmap request may fail, causing the command to never be sent to the\ncontroller. Change the location of the assignment so that it is\nimmediately prior to calling the lldd. Add a comment block to explain\nthe impacts if the lldd were to additionally fail sending the command.\n\nSigned-off-by: Dick Kennedy \u003cdick.kennedy@broadcom.com\u003e\nSigned-off-by: James Smart \u003cjsmart2021@gmail.com\u003e\nReviewed-by: Ewan D. Milne \u003cemilne@redhat.com\u003e\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\n" }, { "commit": "9b0dcd0e5a27958b57e3e390f63c098d63a055da", "tree": "de778d683f121d3062df316994e9c4cf195eb12c", "parents": [ "d7a181da2dfa3190487c446042ba01e07d851c74", "ac71317e6be01812cc0c54d8be6d3c1139c8380b" ], "author": { "name": "Takashi Iwai", "email": "tiwai@suse.de", "time": "Thu Apr 11 14:36:30 2019 +0200" }, "committer": { "name": "Takashi Iwai", "email": "tiwai@suse.de", "time": "Thu Apr 11 14:36:30 2019 +0200" }, "message": "Merge tag \u0027asoc-fix-v5.1-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-linus\n\nASoC: Fixes for v5.1\n\nA few core fixes along with the driver specific ones, mainly fixing\nsmall issues that only affect x86 platforms for various reasons (their\nunusual machine enumeration mechanisms mainly, plus a fix for error\nhandling in topology).\n\nThere\u0027s some of the driver fixes that look larger than they are, like\nthe hdmi-codec changes which resulted in an indentation change, and most\nof the other large changes are for new drivers like the STM32 changes.\n" }, { "commit": "5c41ea6d52003b5bc77c2a82fd5ca7d480237d89", "tree": "728ed8df779df5c676d2e5308379ee50c6119935", "parents": [ "157c99c5a2956a9ab1ae12de0136a2d8a1b1a307" ], "author": { "name": "Faiz Abbas", "email": "faiz_abbas@ti.com", "time": "Thu Apr 11 14:29:37 2019 +0530" }, "committer": { "name": "Ulf Hansson", "email": "ulf.hansson@linaro.org", "time": "Thu Apr 11 12:40:32 2019 +0200" }, "message": "mmc: sdhci-omap: Don\u0027t finish_mrq() on a command error during tuning\n\ncommit 5b0d62108b46 (\"mmc: sdhci-omap: Add platform specific reset\ncallback\") skips data resets during tuning operation. Because of this,\na data error or data finish interrupt might still arrive after a command\nerror has been handled and the mrq ended. This ends up with a \"mmc0: Got\ndata interrupt 0x00000002 even though no data operation was in progress\"\nerror message.\n\nFix this by adding a platform specific callback for sdhci_irq. Mark the\nmrq as a failure but wait for a data interrupt instead of calling\nfinish_mrq().\n\nFixes: 5b0d62108b46 (\"mmc: sdhci-omap: Add platform specific reset\ncallback\")\nSigned-off-by: Faiz Abbas \u003cfaiz_abbas@ti.com\u003e\nAcked-by: Adrian Hunter \u003cadrian.hunter@intel.com\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Ulf Hansson \u003culf.hansson@linaro.org\u003e\n" }, { "commit": "d47703d43ecaa9189d70fb5d151a6883cc44afd3", "tree": "711250540d4ac3749c35de3f320264591c2ded12", "parents": [ "86dc6612abc15538d4dce79b9e1e5283c5659bd5", "bd13b2b874eceb4677cd26eebdc5f45cc52fa400" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Thu Apr 11 19:20:31 2019 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Thu Apr 11 19:20:39 2019 +1000" }, "message": "Merge branch \u0027drm-fixes-5.1\u0027 of git://people.freedesktop.org/~agd5f/linux into drm-fixes\n\nA few fixes for 5.1:\n- Cursor fixes\n- Add missing picasso pci id to KFD\n- XGMI fix\n- Shadow buffer handling fix for GPU reset\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\nFrom: Alex Deucher \u003calexdeucher@gmail.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20190410183031.3710-1-alexander.deucher@amd.com\n" }, { "commit": "86dc6612abc15538d4dce79b9e1e5283c5659bd5", "tree": "ddc17051dc60dc6f9aedacd84f6d43b3cb530311", "parents": [ "7a74198f587b3cfd5c4c5fe3cb8b46188c9ae02d", "9ee76098a1b8ae21cccac641b735ee4d3a77bccf" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Thu Apr 11 19:19:12 2019 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Thu Apr 11 19:19:53 2019 +1000" }, "message": "Merge branch \u0027mediatek-drm-fixes-5.1\u0027 of https://github.com/ckhu-mediatek/linux.git-tags into drm-fixes\n\nThis include stable MT2701 HDMI, framebuffer device and some fixes for\nmediatek drm driver.\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\nFrom: CK Hu \u003cck.hu@mediatek.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/1554860914.29842.4.camel@mtksdaap41\n" }, { "commit": "7a74198f587b3cfd5c4c5fe3cb8b46188c9ae02d", "tree": "c367e7e2cd5d9774edaed94f872c677fcc269a70", "parents": [ "9b39b013037fbfa8d4b999345d9e904d8a336fc2", "e154592a1d25fa1f50ac1bd8d132d0e1103442ba" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Thu Apr 11 19:17:31 2019 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Thu Apr 11 19:17:36 2019 +1000" }, "message": "Merge tag \u0027drm/tegra/for-5.1-rc5\u0027 of git://anongit.freedesktop.org/tegra/linux into drm-fixes\n\ndrm/tegra: Fixes for v5.1-rc5\n\nA single, one-line fix for a build error introduced in v5.1-rc1.\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n\nFrom: Thierry Reding \u003cthierry.reding@gmail.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20190411084106.7552-1-thierry.reding@gmail.com\n" }, { "commit": "e154592a1d25fa1f50ac1bd8d132d0e1103442ba", "tree": "ec5032a411edefc6a1eb137b71dd8983881e813b", "parents": [ "509869a2fec36ecb2b841180915995f41d5a0219" ], "author": { "name": "Stefan Agner", "email": "stefan@agner.ch", "time": "Thu Apr 11 00:47:46 2019 +0200" }, "committer": { "name": "Thierry Reding", "email": "treding@nvidia.com", "time": "Thu Apr 11 10:35:39 2019 +0200" }, "message": "gpu: host1x: Fix compile error when IOMMU API is not available\n\nIn case the IOMMU API is not available compiling host1x fails with\nthe following error:\n\n In file included from drivers/gpu/host1x/hw/host1x06.c:27:\n drivers/gpu/host1x/hw/channel_hw.c: In function ‘host1x_channel_set_streamid’:\n drivers/gpu/host1x/hw/channel_hw.c:118:30: error: implicit declaration of function\n ‘dev_iommu_fwspec_get’; did you mean ‘iommu_fwspec_free’? [-Werror\u003dimplicit-function-declaration]\n struct iommu_fwspec *spec \u003d dev_iommu_fwspec_get(channel-\u003edev-\u003eparent);\n ^~~~~~~~~~~~~~~~~~~~\n iommu_fwspec_free\n\nFixes: de5469c21ff9 (\"gpu: host1x: Program the channel stream ID\")\nSigned-off-by: Stefan Agner \u003cstefan@agner.ch\u003e\nSigned-off-by: Thierry Reding \u003ctreding@nvidia.com\u003e\n" }, { "commit": "cd7879f79f83aec4bb13f0f823f323911dc5397b", "tree": "84dfcd5181d7b3b4f082840de5032e718d23f1c8", "parents": [ "968a85b19d0a79dd8ed85f39e23eacd34b503e72" ], "author": { "name": "Xiong Zhang", "email": "xiong.y.zhang@intel.com", "time": "Wed Apr 10 12:16:33 2019 +0800" }, "committer": { "name": "Zhenyu Wang", "email": "zhenyuw@linux.intel.com", "time": "Thu Apr 11 11:09:53 2019 +0800" }, "message": "drm/i915/gvt: Roundup fb-\u003eheight into tile\u0027s height at calucation fb-\u003esize\n\nWhen fb is tiled and fb-\u003eheight isn\u0027t the multiple of tile\u0027s height,\nthe format fb-\u003esize \u003d fb-\u003estride * fb-\u003eheight, will get a smaller size\nthan the actual size. As the memory height of tiled fb should be multiple\nof tile\u0027s height.\n\nFixes: 7f1a93b1f1d1 (\"drm/i915/gvt: Correct the calculation of plane size\")\nReviewed-by: Zhenyu Wang \u003czhenyuw@linux.intel.com\u003e\nSigned-off-by: Xiong Zhang \u003cxiong.y.zhang@intel.com\u003e\nSigned-off-by: Zhenyu Wang \u003czhenyuw@linux.intel.com\u003e\n" }, { "commit": "7c2e07130090ae001a97a6b65597830d6815e93e", "tree": "49390d3305bda6d110ad8a71a871d36bc43192b3", "parents": [ "9f842abde84d4232d7a1951952dc148bd83f9ada" ], "author": { "name": "David Müller", "email": "dave.mueller@gmx.ch", "time": "Mon Apr 08 15:33:54 2019 +0200" }, "committer": { "name": "Stephen Boyd", "email": "sboyd@kernel.org", "time": "Wed Apr 10 15:54:12 2019 -0700" }, "message": "clk: x86: Add system specific quirk to mark clocks as critical\n\nSince commit 648e921888ad (\"clk: x86: Stop marking clocks as\nCLK_IS_CRITICAL\"), the pmc_plt_clocks of the Bay Trail SoC are\nunconditionally gated off. Unfortunately this will break systems where these\nclocks are used for external purposes beyond the kernel\u0027s knowledge. Fix it\nby implementing a system specific quirk to mark the necessary pmc_plt_clks as\ncritical.\n\nFixes: 648e921888ad (\"clk: x86: Stop marking clocks as CLK_IS_CRITICAL\")\nSigned-off-by: David Müller \u003cdave.mueller@gmx.ch\u003e\nSigned-off-by: Hans de Goede \u003chdegoede@redhat.com\u003e\nReviewed-by: Andy Shevchenko \u003candy.shevchenko@gmail.com\u003e\nSigned-off-by: Stephen Boyd \u003csboyd@kernel.org\u003e\n" }, { "commit": "a3761c3c91209b58b6f33bf69dd8bb8ec0c9d925", "tree": "a4308f3aae5a90ea3927becb776ff41acffe0163", "parents": [ "b2b3a70cd9984fe39ed5aaa9ce596476051ce5de" ], "author": { "name": "Jérôme Glisse", "email": "jglisse@redhat.com", "time": "Wed Apr 10 16:27:51 2019 -0400" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Wed Apr 10 16:14:40 2019 -0600" }, "message": "block: do not leak memory in bio_copy_user_iov()\n\nWhen bio_add_pc_page() fails in bio_copy_user_iov() we should free\nthe page we just allocated otherwise we are leaking it.\n\nCc: linux-block@vger.kernel.org\nCc: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nCc: stable@vger.kernel.org\nReviewed-by: Chaitanya Kulkarni \u003cchaitanya.kulkarni@wdc.com\u003e\nSigned-off-by: Jérôme Glisse \u003cjglisse@redhat.com\u003e\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "3943af9d01e94330d0cfac6fccdbc829aad50c92", "tree": "7cb9a83b339177e2ec64a7bc37edc3377d2fc86c", "parents": [ "9cde402a59770a0669d895399c13407f63d7d209" ], "author": { "name": "Sergey Miroshnichenko", "email": "s.miroshnichenko@yadro.com", "time": "Tue Mar 12 15:05:48 2019 +0300" }, "committer": { "name": "Bjorn Helgaas", "email": "bhelgaas@google.com", "time": "Wed Apr 10 16:06:43 2019 -0500" }, "message": "PCI: pciehp: Ignore Link State Changes after powering off a slot\n\nDuring a safe hot remove, the OS powers off the slot, which may cause a\nData Link Layer State Changed event. The slot has already been set to\nOFF_STATE, so that event results in re-enabling the device, making it\nimpossible to safely remove it.\n\nClear out the Presence Detect Changed and Data Link Layer State Changed\nevents when the disabled slot has settled down.\n\nIt is still possible to re-enable the device if it remains in the slot\nafter pressing the Attention Button by pressing it again.\n\nFixes the problem that Micah reported below: an NVMe drive power button may\nnot actually turn off the drive.\n\nLink: https://bugzilla.kernel.org/show_bug.cgi?id\u003d203237\nReported-by: Micah Parrish \u003cmicah.parrish@hpe.com\u003e\nTested-by: Micah Parrish \u003cmicah.parrish@hpe.com\u003e\nSigned-off-by: Sergey Miroshnichenko \u003cs.miroshnichenko@yadro.com\u003e\n[bhelgaas: changelog, add bugzilla URL]\nSigned-off-by: Bjorn Helgaas \u003cbhelgaas@google.com\u003e\nReviewed-by: Lukas Wunner \u003clukas@wunner.de\u003e\nCc: stable@vger.kernel.org\t# v4.19+\n" }, { "commit": "2a29e9f6b9b499f1fc5f4a48220dc3f4428499f9", "tree": "621a4c367696e9ddd877e037a867cf689207ff1f", "parents": [ "15ade5d2e7775667cf191cf2f94327a4889f8b9d" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Wed Apr 03 21:34:34 2019 +0200" }, "committer": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Wed Apr 10 21:42:04 2019 +0200" }, "message": "sparc64/pci_sun4v: fix ATU checks for large DMA masks\n\nNow that we allow drivers to always need to set larger than required\nDMA masks we need to be a little more careful in the sun4v PCI iommu\ndriver to chose when to select the ATU support - a larger DMA mask\ncan be set even when the platform does not support ATU, so we always\nhave to check if it is avaiable before using it. Add a little helper\nfor that and use it in all the places where we make ATU usage decisions\nbased on the DMA mask.\n\nFixes: 24132a419c68 (\"sparc64/pci_sun4v: allow large DMA masks\")\nReported-by: Meelis Roos \u003cmroos@linux.ee\u003e\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nTested-by: Meelis Roos \u003cmroos@linux.ee\u003e\nAcked-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "582549e3fbe137eb6ce9be591aca25c2222a36b4", "tree": "38edbb9c9e52d1a57a9a41bdea9473ec2424dda9", "parents": [ "ed79cc87302bf7fbc87f05d655b998f866b4fed8", "d737b25b1ae0540ba13cbd45ebb9b58a1d6d7f0d" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Apr 10 09:39:04 2019 -1000" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Apr 10 09:39:04 2019 -1000" }, "message": "Merge tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma\n\nPull rdma fixes from Jason Gunthorpe:\n \"Several driver bug fixes posted in the last several weeks\n\n - Several bug fixes for the hfi1 driver \u0027TID RDMA\u0027 functionality\n merged into 5.1. Since TID RDMA is on by default these all seem to\n be regressions.\n\n - Wrong software permission checks on memory in mlx5\n\n - Memory leak in vmw_pvrdma during driver remove\n\n - Several bug fixes for hns driver features merged into 5.1\"\n\n* tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:\n IB/hfi1: Do not flush send queue in the TID RDMA second leg\n RDMA/hns: Bugfix for SCC hem free\n RDMA/hns: Fix bug that caused srq creation to fail\n RDMA/vmw_pvrdma: Fix memory leak on pvrdma_pci_remove\n IB/mlx5: Reset access mask when looping inside page fault handler\n IB/hfi1: Fix the allocation of RSM table\n IB/hfi1: Eliminate opcode tests on mr deref\n IB/hfi1: Clear the IOWAIT pending bits when QP is put into error state\n IB/hfi1: Failed to drain send queue when QP is put into error state\n" }, { "commit": "b2b3a70cd9984fe39ed5aaa9ce596476051ce5de", "tree": "2b9f5049115efb26211ed7b921eda20e0905852e", "parents": [ "eb3afb75b57c28599af0dfa03a99579d410749e9" ], "author": { "name": "Hans Holmberg", "email": "hans.holmberg@cnexlabs.com", "time": "Wed Apr 10 19:56:43 2019 +0200" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Wed Apr 10 12:17:01 2019 -0600" }, "message": "lightnvm: pblk: fix crash in pblk_end_partial_read due to multipage bvecs\n\nThe introduction of multipage bio vectors broke pblk\u0027s partial read\nlogic due to it not being prepared for multipage bio vectors.\n\nUse bio vector iterators instead of direct bio vector indexing.\n\nFixes: 07173c3ec276 (\"block: enable multipage bvecs\")\nReported-by: Klaus Jensen \u003cklaus.jensen@cnexlabs.com\u003e\nSigned-off-by: Hans Holmberg \u003chans.holmberg@cnexlabs.com\u003e\nUpdated description.\nSigned-off-by: Matias Bjørling \u003cmb@lightnvm.io\u003e\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "d737b25b1ae0540ba13cbd45ebb9b58a1d6d7f0d", "tree": "8e81c561ee6a6793b3c2f8ec9e201c6180d67c05", "parents": [ "00fb67ec6b98114a887d9ef26fc7c3e566e7f665" ], "author": { "name": "Kaike Wan", "email": "kaike.wan@intel.com", "time": "Wed Apr 10 06:27:05 2019 -0700" }, "committer": { "name": "Jason Gunthorpe", "email": "jgg@mellanox.com", "time": "Wed Apr 10 15:09:30 2019 -0300" }, "message": "IB/hfi1: Do not flush send queue in the TID RDMA second leg\n\nWhen a QP is put into error state, the send queue will be flushed.\nThis mechanism is implemented in both the first and the second leg\nof the send engine. Since the second leg is only responsible for\ndata transactions in the KDETH space for the TID RDMA WRITE request,\nit should not perform the flushing of the send queue.\n\nThis patch removes the flushing function of the second leg, but\nstill keeps the bailing out of the QP if it is put into error state.\n\nFixes: 70dcb2e3dc6a (\"IB/hfi1: Add the TID second leg send packet builder\")\nReviewed-by: Mike Marciniszyn \u003cmike.marciniszyn@intel.com\u003e\nSigned-off-by: Kaike Wan \u003ckaike.wan@intel.com\u003e\nSigned-off-by: Dennis Dalessandro \u003cdennis.dalessandro@intel.com\u003e\nSigned-off-by: Jason Gunthorpe \u003cjgg@mellanox.com\u003e\n" }, { "commit": "ed79cc87302bf7fbc87f05d655b998f866b4fed8", "tree": "cec766af1bf668d2f34fdc7221e4ece20e609691", "parents": [ "e33c1b9923775d17ad246946fe67fcb9be288677", "cf94db21905333e610e479688add629397a4b384" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Apr 10 06:42:51 2019 -1000" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Apr 10 06:42:51 2019 -1000" }, "message": "Merge tag \u0027for_linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost\n\nPull virtio fixes from Michael Tsirkin:\n \"Several fixes, add more reviewers to the list\"\n\n* tag \u0027for_linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost:\n virtio: Honour \u0027may_reduce_num\u0027 in vring_create_virtqueue\n MAiNTAINERS: add Paolo, Stefan for virtio blk/scsi\n virtio_pci: fix a NULL pointer reference in vp_del_vqs\n" }, { "commit": "ac71317e6be01812cc0c54d8be6d3c1139c8380b", "tree": "dbc17e0d6f0381cda67b78868367dd62fa279207", "parents": [ "70802487bb9145a4f8b26f5a11d0e7f83c25100a" ], "author": { "name": "Marc Gonzalez", "email": "marc.w.gonzalez@free.fr", "time": "Wed Apr 10 16:23:38 2019 +0200" }, "committer": { "name": "Mark Brown", "email": "broonie@kernel.org", "time": "Wed Apr 10 17:07:59 2019 +0100" }, "message": "ASoC: wcd9335: Fix missing regmap requirement\n\nwcd9335.c: undefined reference to \u0027devm_regmap_add_irq_chip\u0027\n\nSigned-off-by: Marc Gonzalez \u003cmarc.w.gonzalez@free.fr\u003e\nSigned-off-by: Mark Brown \u003cbroonie@kernel.org\u003e\n" }, { "commit": "21635d7311734d2d1b177f8a95e2f9386174b76d", "tree": "075edf121fbe3a1b042d97e65b5aae63e78b2932", "parents": [ "4690985e00ac38334d4a68f99c56ac310ef0527b" ], "author": { "name": "Jani Nikula", "email": "jani.nikula@intel.com", "time": "Fri Apr 05 10:52:20 2019 +0300" }, "committer": { "name": "Rodrigo Vivi", "email": "rodrigo.vivi@intel.com", "time": "Wed Apr 10 09:06:39 2019 -0700" }, "message": "drm/i915/dp: revert back to max link rate and lane count on eDP\n\nCommit 7769db588384 (\"drm/i915/dp: optimize eDP 1.4+ link config fast\nand narrow\") started to optize the eDP 1.4+ link config, both per spec\nand as preparation for display stream compression support.\n\nSadly, we again face panels that flat out fail with parameters they\nclaim to support. Revert, and go back to the drawing board.\n\nv2: Actually revert to max params instead of just wide-and-slow.\n\nBugzilla: https://bugs.freedesktop.org/show_bug.cgi?id\u003d109959\nFixes: 7769db588384 (\"drm/i915/dp: optimize eDP 1.4+ link config fast and narrow\")\nCc: Ville Syrjälä \u003cville.syrjala@linux.intel.com\u003e\nCc: Manasi Navare \u003cmanasi.d.navare@intel.com\u003e\nCc: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\nCc: Matt Atwood \u003cmatthew.s.atwood@intel.com\u003e\nCc: \"Lee, Shawn C\" \u003cshawn.c.lee@intel.com\u003e\nCc: Dave Airlie \u003cairlied@gmail.com\u003e\nCc: intel-gfx@lists.freedesktop.org\nCc: \u003cstable@vger.kernel.org\u003e # v5.0+\nReviewed-by: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\nReviewed-by: Manasi Navare \u003cmanasi.d.navare@intel.com\u003e\nTested-by: Albert Astals Cid \u003caacid@kde.org\u003e # v5.0 backport\nTested-by: Emanuele Panigati \u003cilpanich@gmail.com\u003e # v5.0 backport\nTested-by: Matteo Iervasi \u003cmatteoiervasi@gmail.com\u003e # v5.0 backport\nSigned-off-by: Jani Nikula \u003cjani.nikula@intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/20190405075220.9815-1-jani.nikula@intel.com\n(cherry picked from commit f11cb1c19ad0563b3c1ea5eb16a6bac0e401f428)\nSigned-off-by: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\n" }, { "commit": "4690985e00ac38334d4a68f99c56ac310ef0527b", "tree": "5babe3cd336d8cd5b3fca2681aeaa0730c6a0666", "parents": [ "20eea462bf2fbff3a4be375cc8424a544235a432" ], "author": { "name": "Vandita Kulkarni", "email": "vandita.kulkarni@intel.com", "time": "Mon Mar 25 16:56:42 2019 +0530" }, "committer": { "name": "Rodrigo Vivi", "email": "rodrigo.vivi@intel.com", "time": "Wed Apr 10 09:06:35 2019 -0700" }, "message": "drm/i915/icl: Fix port disable sequence for mipi-dsi\n\nRe-enable clock gating of DDI clocks.\n\nv2: Fix the default ddi clk state for mipi-dsi (Imre)\n\nFixes: 1026bea00381 (\"drm/i915/icl: Ungate DSI clocks\")\nSigned-off-by: Vandita Kulkarni \u003cvandita.kulkarni@intel.com\u003e\nReviewed-by: Uma Shankar \u003cuma.shankar@intel.com\u003e\nSigned-off-by: Jani Nikula \u003cjani.nikula@intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/1553513202-13863-2-git-send-email-vandita.kulkarni@intel.com\n(cherry picked from commit 942d1cf48eae3fcd7e973cfb708d5c4860f0c713)\nSigned-off-by: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\n" }, { "commit": "20eea462bf2fbff3a4be375cc8424a544235a432", "tree": "eb81062e41a44a96e8bec44f13032c4a53c71c5d", "parents": [ "bef42cb2029c0ec8cd8c9e0545589d81913dec4e" ], "author": { "name": "Vandita Kulkarni", "email": "vandita.kulkarni@intel.com", "time": "Mon Mar 25 16:56:41 2019 +0530" }, "committer": { "name": "Rodrigo Vivi", "email": "rodrigo.vivi@intel.com", "time": "Wed Apr 10 09:06:29 2019 -0700" }, "message": "drm/i915/icl: Ungate ddi clocks before IO enable\n\nIO enable sequencing needs ddi clocks enabled.\nThese clocks will be gated at a later point in\nthe enable sequence.\n\nv2: Fix the commit header (Uma)\nv3: Remove the redundant read (Ville)\n\nFixes: 949fc52af19e (\"drm/i915/icl: add pll mapping for DSI\")\nSigned-off-by: Vandita Kulkarni \u003cvandita.kulkarni@intel.com\u003e\nReviewed-by: Uma Shankar \u003cuma.shankar@intel.com\u003e\nSigned-off-by: Jani Nikula \u003cjani.nikula@intel.com\u003e\nLink: https://patchwork.freedesktop.org/patch/msgid/1553513202-13863-1-git-send-email-vandita.kulkarni@intel.com\n(cherry picked from commit c5b81a325263a891d5811aabe938c87e03db4c37)\nSigned-off-by: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\n" }, { "commit": "eb3afb75b57c28599af0dfa03a99579d410749e9", "tree": "a56b6352e2f032e0ff91a73f58917bd7cbae7f26", "parents": [ "1b8f21b74c3c9c82fce5a751d7aefb7cc0b8d33d" ], "author": { "name": "Ming Lei", "email": "ming.lei@redhat.com", "time": "Tue Apr 09 06:31:22 2019 +0800" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Wed Apr 10 09:57:35 2019 -0600" }, "message": "nvme: cancel request synchronously\n\nnvme_cancel_request() is used in error handler, and it is always\nreliable to cancel request synchronously, and avoids possible race\nin which request may be completed after real hw queue is destroyed.\n\nOne issue is reported by our customer on NVMe RDMA, in which freed ib\nqueue pair may be used in nvme_rdma_complete_rq().\n\nCc: Sagi Grimberg \u003csagi@grimberg.me\u003e\nCc: Bart Van Assche \u003cbvanassche@acm.org\u003e\nCc: James Smart \u003cjames.smart@broadcom.com\u003e\nCc: linux-nvme@lists.infradead.org\nReviewed-by: Keith Busch \u003ckeith.busch@intel.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Ming Lei \u003cming.lei@redhat.com\u003e\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "1b8f21b74c3c9c82fce5a751d7aefb7cc0b8d33d", "tree": "b6e8eac9fe39c7d114587047c6addc7728009bbd", "parents": [ "1978f30a87732d4d9072a20abeded9fe17884f1b" ], "author": { "name": "Ming Lei", "email": "ming.lei@redhat.com", "time": "Tue Apr 09 06:31:21 2019 +0800" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Wed Apr 10 09:57:33 2019 -0600" }, "message": "blk-mq: introduce blk_mq_complete_request_sync()\n\nIn NVMe\u0027s error handler, follows the typical steps of tearing down\nhardware for recovering controller:\n\n1) stop blk_mq hw queues\n2) stop the real hw queues\n3) cancel in-flight requests via\n\tblk_mq_tagset_busy_iter(tags, cancel_request, ...)\ncancel_request():\n\tmark the request as abort\n\tblk_mq_complete_request(req);\n4) destroy real hw queues\n\nHowever, there may be race between #3 and #4, because blk_mq_complete_request()\nmay run q-\u003emq_ops-\u003ecomplete(rq) remotelly and asynchronously, and\n-\u003ecomplete(rq) may be run after #4.\n\nThis patch introduces blk_mq_complete_request_sync() for fixing the\nabove race.\n\nCc: Sagi Grimberg \u003csagi@grimberg.me\u003e\nCc: Bart Van Assche \u003cbvanassche@acm.org\u003e\nCc: James Smart \u003cjames.smart@broadcom.com\u003e\nCc: linux-nvme@lists.infradead.org\nReviewed-by: Keith Busch \u003ckeith.busch@intel.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Ming Lei \u003cming.lei@redhat.com\u003e\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "1978f30a87732d4d9072a20abeded9fe17884f1b", "tree": "df4d83ad5a1996234900cc4424f413d978eec139", "parents": [ "bf348f9b78d413e75bb079462751a1d86b6de36c" ], "author": { "name": "Dongli Zhang", "email": "dongli.zhang@oracle.com", "time": "Wed Mar 27 18:36:35 2019 +0800" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Wed Apr 10 08:18:25 2019 -0600" }, "message": "scsi: virtio_scsi: limit number of hw queues by nr_cpu_ids\n\nWhen tag_set-\u003enr_maps is 1, the block layer limits the number of hw queues\nby nr_cpu_ids. No matter how many hw queues are used by virtio-scsi, as it\nhas (tag_set-\u003enr_maps \u003d\u003d 1), it can use at most nr_cpu_ids hw queues.\n\nIn addition, specifically for pci scenario, when the \u0027num_queues\u0027 specified\nby qemu is more than maxcpus, virtio-scsi would not be able to allocate\nmore than maxcpus vectors in order to have a vector for each queue. As a\nresult, it falls back into MSI-X with one vector for config and one shared\nfor queues.\n\nConsidering above reasons, this patch limits the number of hw queues used\nby virtio-scsi by nr_cpu_ids.\n\nReviewed-by: Stefan Hajnoczi \u003cstefanha@redhat.com\u003e\nSigned-off-by: Dongli Zhang \u003cdongli.zhang@oracle.com\u003e\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "bf348f9b78d413e75bb079462751a1d86b6de36c", "tree": "ff62f039df5b908c1c02f8d4074e38ac3b8b9f2a", "parents": [ "eed47d19d9362bdd958e4ab56af480b9dbf6b2b6" ], "author": { "name": "Dongli Zhang", "email": "dongli.zhang@oracle.com", "time": "Wed Mar 27 18:36:34 2019 +0800" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Wed Apr 10 08:18:24 2019 -0600" }, "message": "virtio-blk: limit number of hw queues by nr_cpu_ids\n\nWhen tag_set-\u003enr_maps is 1, the block layer limits the number of hw queues\nby nr_cpu_ids. No matter how many hw queues are used by virtio-blk, as it\nhas (tag_set-\u003enr_maps \u003d\u003d 1), it can use at most nr_cpu_ids hw queues.\n\nIn addition, specifically for pci scenario, when the \u0027num-queues\u0027 specified\nby qemu is more than maxcpus, virtio-blk would not be able to allocate more\nthan maxcpus vectors in order to have a vector for each queue. As a result,\nit falls back into MSI-X with one vector for config and one shared for\nqueues.\n\nConsidering above reasons, this patch limits the number of hw queues used\nby virtio-blk by nr_cpu_ids.\n\nReviewed-by: Stefan Hajnoczi \u003cstefanha@redhat.com\u003e\nSigned-off-by: Dongli Zhang \u003cdongli.zhang@oracle.com\u003e\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "eed47d19d9362bdd958e4ab56af480b9dbf6b2b6", "tree": "ebb19cdcb30d8e16135544007babceb40da2d08b", "parents": [ "3ec482d15cb986bf08b923f9193eeddb3b9ca69f" ], "author": { "name": "Paolo Valente", "email": "paolo.valente@linaro.org", "time": "Wed Apr 10 10:38:33 2019 +0200" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Wed Apr 10 07:54:38 2019 -0600" }, "message": "block, bfq: fix use after free in bfq_bfqq_expire\n\nThe function bfq_bfqq_expire() invokes the function\n__bfq_bfqq_expire(), and the latter may free the in-service bfq-queue.\nIf this happens, then no other instruction of bfq_bfqq_expire() must\nbe executed, or a use-after-free will occur.\n\nBasing on the assumption that __bfq_bfqq_expire() invokes\nbfq_put_queue() on the in-service bfq-queue exactly once, the queue is\nassumed to be freed if its refcounter is equal to one right before\ninvoking __bfq_bfqq_expire().\n\nBut, since commit 9dee8b3b057e (\"block, bfq: fix queue removal from\nweights tree\") this assumption is false. __bfq_bfqq_expire() may also\ninvoke bfq_weights_tree_remove() and, since commit 9dee8b3b057e\n(\"block, bfq: fix queue removal from weights tree\"), also\nthe latter function may invoke bfq_put_queue(). So __bfq_bfqq_expire()\nmay invoke bfq_put_queue() twice, and this is the actual case where\nthe in-service queue may happen to be freed.\n\nTo address this issue, this commit moves the check on the refcounter\nof the queue right around the last bfq_put_queue() that may be invoked\non the queue.\n\nFixes: 9dee8b3b057e (\"block, bfq: fix queue removal from weights tree\")\nReported-by: Dmitrii Tcvetkov \u003cdemfloro@demfloro.ru\u003e\nReported-by: Douglas Anderson \u003cdianders@chromium.org\u003e\nTested-by: Dmitrii Tcvetkov \u003cdemfloro@demfloro.ru\u003e\nTested-by: Douglas Anderson \u003cdianders@chromium.org\u003e\nSigned-off-by: Paolo Valente \u003cpaolo.valente@linaro.org\u003e\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "d7a181da2dfa3190487c446042ba01e07d851c74", "tree": "c2c281018d0a0dd56e792e0fca49af8313d5a30b", "parents": [ "cae30527901d9590db0e12ace994c1d58bea87fd" ], "author": { "name": "Takashi Iwai", "email": "tiwai@suse.de", "time": "Wed Apr 10 12:49:55 2019 +0200" }, "committer": { "name": "Takashi Iwai", "email": "tiwai@suse.de", "time": "Wed Apr 10 15:23:36 2019 +0200" }, "message": "ALSA: hda: Fix racy display power access\n\nsnd_hdac_display_power() doesn\u0027t handle the concurrent calls carefully\nenough, and it may lead to the doubly get_power or put_power calls,\nwhen a runtime PM and an async work get called in racy way.\n\nThis patch addresses it by reusing the bus-\u003elock mutex that has been\nused for protecting the link state change in ext bus code, so that it\ncan protect against racy display state changes. The initialization of\nbus-\u003elock was moved from snd_hdac_ext_bus_init() to\nsnd_hdac_bus_init() as well accordingly.\n\nTestcase: igt/i915_pm_rpm/module-reload #glk-dsi\nReported-by: Chris Wilson \u003cchris@chris-wilson.co.uk\u003e\nReviewed-by: Chris Wilson \u003cchris@chris-wilson.co.uk\u003e\nCc: Imre Deak \u003cimre.deak@intel.com\u003e\nSigned-off-by: Takashi Iwai \u003ctiwai@suse.de\u003e\n" }, { "commit": "07d7e12091f4ab869cc6a4bb276399057e73b0b3", "tree": "dd505b68149ade6a64d4cce596329209b8cbbc24", "parents": [ "2a6a8e2d9004b5303fcb494588ba3a3b87a256c3" ], "author": { "name": "Andrei Vagin", "email": "avagin@gmail.com", "time": "Sun Apr 07 21:15:42 2019 -0700" }, "committer": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Wed Apr 10 15:23:26 2019 +0200" }, "message": "alarmtimer: Return correct remaining time\n\nTo calculate a remaining time, it\u0027s required to subtract the current time\nfrom the expiration time. In alarm_timer_remaining() the arguments of\nktime_sub are swapped.\n\nFixes: d653d8457c76 (\"alarmtimer: Implement remaining callback\")\nSigned-off-by: Andrei Vagin \u003cavagin@gmail.com\u003e\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nReviewed-by: Mukesh Ojha \u003cmojha@codeaurora.org\u003e\nCc: Stephen Boyd \u003csboyd@kernel.org\u003e\nCc: John Stultz \u003cjohn.stultz@linaro.org\u003e\nCc: stable@vger.kernel.org\nLink: https://lkml.kernel.org/r/20190408041542.26338-1-avagin@gmail.com\n\n" }, { "commit": "90c1cba2b3b3851c151229f61801919b2904d437", "tree": "57e97c7d41521ec536fa3d546e4e7e1639aa1520", "parents": [ "771acc7e4a6e5dba779cb1a7fd851a164bc81033" ], "author": { "name": "Bart Van Assche", "email": "bvanassche@acm.org", "time": "Wed Apr 03 16:35:52 2019 -0700" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Wed Apr 10 13:45:59 2019 +0200" }, "message": "locking/lockdep: Zap lock classes even with lock debugging disabled\n\nThe following commit:\n\n a0b0fd53e1e6 (\"locking/lockdep: Free lock classes that are no longer in use\")\n\nchanged the behavior of lockdep_free_key_range() from\nunconditionally zapping lock classes into only zapping lock classes if\ndebug_lock \u003d\u003d true. Not zapping lock classes if debug_lock \u003d\u003d false leaves\ndangling pointers in several lockdep datastructures, e.g. lock_class::name\nin the all_lock_classes list.\n\nThe shell command \"cat /proc/lockdep\" causes the kernel to iterate the\nall_lock_classes list. Hence the \"unable to handle kernel paging request\" cash\nthat Shenghui encountered by running cat /proc/lockdep.\n\nSince the new behavior can cause cat /proc/lockdep to crash, restore the\npre-v5.1 behavior.\n\nThis patch avoids that cat /proc/lockdep triggers the following crash\nwith debug_lock \u003d\u003d false:\n\n BUG: unable to handle kernel paging request at fffffbfff40ca448\n RIP: 0010:__asan_load1+0x28/0x50\n Call Trace:\n string+0xac/0x180\n vsnprintf+0x23e/0x820\n seq_vprintf+0x82/0xc0\n seq_printf+0x92/0xb0\n print_name+0x34/0xb0\n l_show+0x184/0x200\n seq_read+0x59e/0x6c0\n proc_reg_read+0x11f/0x170\n __vfs_read+0x4d/0x90\n vfs_read+0xc5/0x1f0\n ksys_read+0xab/0x130\n __x64_sys_read+0x43/0x50\n do_syscall_64+0x71/0x210\n entry_SYSCALL_64_after_hwframe+0x49/0xbe\n\nReported-by: shenghui \u003cshhuiw@foxmail.com\u003e\nSigned-off-by: Bart Van Assche \u003cbvanassche@acm.org\u003e\nSigned-off-by: Peter Zijlstra (Intel) \u003cpeterz@infradead.org\u003e\nCc: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nCc: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nCc: Waiman Long \u003clongman@redhat.com\u003e\nCc: Will Deacon \u003cwill.deacon@arm.com\u003e\nFixes: a0b0fd53e1e6 (\"locking/lockdep: Free lock classes that are no longer in use\") # v5.1-rc1.\nLink: https://lkml.kernel.org/r/20190403233552.124673-1-bvanassche@acm.org\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\n" }, { "commit": "70802487bb9145a4f8b26f5a11d0e7f83c25100a", "tree": "f43925307a15d57372c82db12db1dbb24fd6caf0", "parents": [ "e37c2deafe7058cf7989c4c47bbf1140cc867d89" ], "author": { "name": "Ranjani Sridharan", "email": "ranjani.sridharan@linux.intel.com", "time": "Mon Apr 08 12:30:25 2019 -0700" }, "committer": { "name": "Mark Brown", "email": "broonie@kernel.org", "time": "Wed Apr 10 12:28:00 2019 +0100" }, "message": "ASoC: pcm: fix error handling when try_module_get() fails.\n\nHandle error before returning when try_module_get() fails\nto prevent inconsistent mutex lock/unlock.\n\nFixes: 52034add7 (ASoC: pcm: update module refcount if\n\t\t module_get_upon_open is set)\nSigned-off-by: Ranjani Sridharan \u003cranjani.sridharan@linux.intel.com\u003e\nSigned-off-by: Mark Brown \u003cbroonie@kernel.org\u003e\n" }, { "commit": "e33c1b9923775d17ad246946fe67fcb9be288677", "tree": "11ff1477557e9608883f7d8986cb54f4fadfe239", "parents": [ "771acc7e4a6e5dba779cb1a7fd851a164bc81033" ], "author": { "name": "Kees Cook", "email": "keescook@chromium.org", "time": "Mon Apr 08 09:07:06 2019 -0700" }, "committer": { "name": "John Johansen", "email": "john.johansen@canonical.com", "time": "Wed Apr 10 04:24:48 2019 -0700" }, "message": "apparmor: Restore Y/N in /sys for apparmor\u0027s \"enabled\"\n\nBefore commit c5459b829b71 (\"LSM: Plumb visibility into optional \"enabled\"\nstate\"), /sys/module/apparmor/parameters/enabled would show \"Y\" or \"N\"\nsince it was using the \"bool\" handler. After being changed to \"int\",\nthis switched to \"1\" or \"0\", breaking the userspace AppArmor detection\nof dbus-broker. This restores the Y/N output while keeping the LSM\ninfrastructure happy.\n\nBefore:\n\t$ cat /sys/module/apparmor/parameters/enabled\n\t1\n\nAfter:\n\t$ cat /sys/module/apparmor/parameters/enabled\n\tY\n\nReported-by: David Rheinsberg \u003cdavid.rheinsberg@gmail.com\u003e\nReviewed-by: David Rheinsberg \u003cdavid.rheinsberg@gmail.com\u003e\nLink: https://lkml.kernel.org/r/CADyDSO6k8vYb1eryT4g6+EHrLCvb68GAbHVWuULkYjcZcYNhhw@mail.gmail.com\nFixes: c5459b829b71 (\"LSM: Plumb visibility into optional \"enabled\" state\")\nSigned-off-by: Kees Cook \u003ckeescook@chromium.org\u003e\nSigned-off-by: John Johansen \u003cjohn.johansen@canonical.com\u003e\n" }, { "commit": "e37c2deafe7058cf7989c4c47bbf1140cc867d89", "tree": "5ce535278ecf1bc90cbee48248f529d2c4485083", "parents": [ "d6ba3f815bc5f3c4249d15c8bc5fbb012651b4a4" ], "author": { "name": "Olivier Moysan", "email": "olivier.moysan@st.com", "time": "Wed Apr 10 10:08:36 2019 +0200" }, "committer": { "name": "Mark Brown", "email": "broonie@kernel.org", "time": "Wed Apr 10 12:16:15 2019 +0100" }, "message": "ASoC: stm32: sai: fix master clock management\n\nWhen master clock is used, master clock rate is set exclusively.\nParent clocks of master clock cannot be changed after a call to\nclk_set_rate_exclusive(). So the parent clock of SAI kernel clock\nmust be set before.\nEnsure also that exclusive rate operations are balanced\nin STM32 SAI driver.\n\nSigned-off-by: Olivier Moysan \u003colivier.moysan@st.com\u003e\nSigned-off-by: Mark Brown \u003cbroonie@kernel.org\u003e\n" }, { "commit": "d6ba3f815bc5f3c4249d15c8bc5fbb012651b4a4", "tree": "d4ceb4d3cf0108153b3725b91f4463e78fc8afac", "parents": [ "52034add758e268c39110f33d46e2a9492e82aef" ], "author": { "name": "Tzung-Bi Shih", "email": "tzungbi@google.com", "time": "Mon Apr 08 17:08:58 2019 +0800" }, "committer": { "name": "Mark Brown", "email": "broonie@kernel.org", "time": "Wed Apr 10 12:16:07 2019 +0100" }, "message": "ASoC: Intel: kbl: fix wrong number of channels\n\nFix wrong setting on number of channels. The context wants to set\nconstraint to 2 channels instead of 4.\n\nSigned-off-by: Tzung-Bi Shih \u003ctzungbi@google.com\u003e\nAcked-by: Pierre-Louis Bossart \u003cpierre-louis.bossart@linux.intel.com\u003e\nSigned-off-by: Mark Brown \u003cbroonie@kernel.org\u003e\n" }, { "commit": "3966c3feca3fd10b2935caa0b4a08c7dd59469e5", "tree": "2f19dfa77d22aad08853752cdd22ee509ef6b819", "parents": [ "6d3edaae16c6c7d238360f2841212c2b26774d5e" ], "author": { "name": "Lendacky, Thomas", "email": "Thomas.Lendacky@amd.com", "time": "Tue Apr 02 15:21:18 2019 +0000" }, "committer": { "name": "Ingo Molnar", "email": "mingo@kernel.org", "time": "Wed Apr 10 13:03:18 2019 +0200" }, "message": "x86/perf/amd: Remove need to check \"running\" bit in NMI handler\n\nSpurious interrupt support was added to perf in the following commit, almost\na decade ago:\n\n 63e6be6d98e1 (\"perf, x86: Catch spurious interrupts after disabling counters\")\n\nThe two previous patches (resolving the race condition when disabling a\nPMC and NMI latency mitigation) allow for the removal of this older\nspurious interrupt support.\n\nCurrently in x86_pmu_stop(), the bit for the PMC in the active_mask bitmap\nis cleared before disabling the PMC, which sets up a race condition. This\nrace condition was mitigated by introducing the running bitmap. That race\ncondition can be eliminated by first disabling the PMC, waiting for PMC\nreset on overflow and then clearing the bit for the PMC in the active_mask\nbitmap. The NMI handler will not re-enable a disabled counter.\n\nIf x86_pmu_stop() is called from the perf NMI handler, the NMI latency\nmitigation support will guard against any unhandled NMI messages.\n\nSigned-off-by: Tom Lendacky \u003cthomas.lendacky@amd.com\u003e\nSigned-off-by: Peter Zijlstra (Intel) \u003cpeterz@infradead.org\u003e\nCc: \u003cstable@vger.kernel.org\u003e # 4.14.x-\nCc: Alexander Shishkin \u003calexander.shishkin@linux.intel.com\u003e\nCc: Arnaldo Carvalho de Melo \u003cacme@kernel.org\u003e\nCc: Arnaldo Carvalho de Melo \u003cacme@redhat.com\u003e\nCc: Borislav Petkov \u003cbp@alien8.de\u003e\nCc: Jiri Olsa \u003cjolsa@redhat.com\u003e\nCc: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\nCc: Namhyung Kim \u003cnamhyung@kernel.org\u003e\nCc: Peter Zijlstra \u003cpeterz@infradead.org\u003e\nCc: Stephane Eranian \u003ceranian@google.com\u003e\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nCc: Vince Weaver \u003cvincent.weaver@maine.edu\u003e\nLink: https://lkml.kernel.org/r/Message-ID:\nSigned-off-by: Ingo Molnar \u003cmingo@kernel.org\u003e\n" }, { "commit": "cf7cf6977f531acd5dfe55250d0ee8cbbb6f1ae8", "tree": "57488b97ebf3c07b2ce5ca005b7296d02b251517", "parents": [ "7100e8704b61247649c50551b965e71d168df30b" ], "author": { "name": "Michael Ellerman", "email": "mpe@ellerman.id.au", "time": "Tue Apr 09 15:43:11 2019 +1000" }, "committer": { "name": "Michael Ellerman", "email": "mpe@ellerman.id.au", "time": "Wed Apr 10 14:45:57 2019 +1000" }, "message": "powerpc/mm: Define MAX_PHYSMEM_BITS for all 64-bit configs\n\nThe recent commit 8bc086899816 (\"powerpc/mm: Only define\nMAX_PHYSMEM_BITS in SPARSEMEM configurations\") removed our definition\nof MAX_PHYSMEM_BITS when SPARSEMEM is disabled.\n\nThis inadvertently broke some 64-bit FLATMEM using configs with eg:\n\n arch/powerpc/include/asm/book3s/64/mmu-hash.h:584:6: error: \"MAX_PHYSMEM_BITS\" is not defined, evaluates to 0\n #if (MAX_PHYSMEM_BITS \u003e MAX_EA_BITS_PER_CONTEXT)\n ^~~~~~~~~~~~~~~~\n\nFix it by making sure we define MAX_PHYSMEM_BITS for all 64-bit\nconfigs regardless of SPARSEMEM.\n\nFixes: 8bc086899816 (\"powerpc/mm: Only define MAX_PHYSMEM_BITS in SPARSEMEM configurations\")\nReported-by: Andreas Schwab \u003cschwab@linux-m68k.org\u003e\nReported-by: Hugh Dickins \u003chughd@google.com\u003e\nReviewed-by: Aneesh Kumar K.V \u003caneesh.kumar@linux.ibm.com\u003e\nSigned-off-by: Michael Ellerman \u003cmpe@ellerman.id.au\u003e\n" }, { "commit": "771acc7e4a6e5dba779cb1a7fd851a164bc81033", "tree": "a0574841f90b59fa08855d7eec95fd132063d645", "parents": [ "0ee7fb36f988539f52f83ce6048d696bd540066f" ], "author": { "name": "Brian Norris", "email": "briannorris@chromium.org", "time": "Tue Apr 09 11:49:17 2019 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Apr 09 17:38:24 2019 -1000" }, "message": "Bluetooth: btusb: request wake pin with NOAUTOEN\n\nBadly-designed systems might have (for example) active-high wake pins\nthat default to high (e.g., because of external pull ups) until they\nhave an active firmware which starts driving it low. This can cause an\ninterrupt storm in the time between request_irq() and disable_irq().\n\nWe don\u0027t support shared interrupts here, so let\u0027s just pre-configure the\ninterrupt to avoid auto-enabling it.\n\nFixes: fd913ef7ce61 (\"Bluetooth: btusb: Add out-of-band wakeup support\")\nFixes: 5364a0b4f4be (\"arm64: dts: rockchip: move QCA6174A wakeup pin into its USB node\")\nSigned-off-by: Brian Norris \u003cbriannorris@chromium.org\u003e\nReviewed-by: Matthias Kaehlcke \u003cmka@chromium.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "0ee7fb36f988539f52f83ce6048d696bd540066f", "tree": "111e8e16e1921d549c2d56881a6df90a8d527ec4", "parents": [ "972acfb49446b30a3533ceb5682bf8350c786bc8", "6e3572e83dc3563e3b7e742bcb225b42a60cdaeb" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Apr 09 16:27:18 2019 -1000" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Apr 09 16:27:18 2019 -1000" }, "message": "Merge tag \u0027mips_fixes_5.1_2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux\n\nPull MIPS fixes from Paul Burton:\n \"A few minor MIPS fixes:\n\n - Provide struct pt_regs * from get_irq_regs() to kgdb_nmicallback()\n when handling an IPI triggered by kgdb_roundup_cpus(), matching the\n behavior of other architectures \u0026 resolving kgdb issues for SMP\n systems.\n\n - Defer a pointer dereference until after a NULL check in the\n irq_shutdown callback for SGI IP27 HUB interrupts.\n\n - A defconfig update for the MSCC Ocelot to enable some necessary\n drivers\"\n\n* tag \u0027mips_fixes_5.1_2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux:\n MIPS: generic: Add switchdev, pinctrl and fit to ocelot_defconfig\n MIPS: SGI-IP27: Fix use of unchecked pointer in shutdown_bridge_irq\n MIPS: KGDB: fix kgdb support for SMP platforms.\n" }, { "commit": "972acfb49446b30a3533ceb5682bf8350c786bc8", "tree": "27f194c8b6e171bb800bfd611b8c7529ddff10b5", "parents": [ "869e3305f23dfeacdaa234717c92ccb237815d90", "6af1c849dfb1f1d326fbdd157c9bc882b921f450" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Apr 09 16:20:59 2019 -1000" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Apr 09 16:20:59 2019 -1000" }, "message": "Merge branch \u0027fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs\n\nPull misc fixes from Al Viro:\n \"A few regression fixes from this cycle\"\n\n* \u0027fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:\n aio: use kmem_cache_free() instead of kfree()\n iov_iter: Fix build error without CONFIG_CRYPTO\n aio: Fix an error code in __io_submit_one()\n" }, { "commit": "9ee76098a1b8ae21cccac641b735ee4d3a77bccf", "tree": "083e3312f604a323692227932ab48bb6f985404d", "parents": [ "8eeb3946feeb00486ac0909e2309da87db8988a5" ], "author": { "name": "Wangyan Wang", "email": "wangyan.wang@mediatek.com", "time": "Tue Apr 09 14:53:07 2019 +0800" }, "committer": { "name": "CK Hu", "email": "ck.hu@mediatek.com", "time": "Tue Apr 09 17:47:01 2019 +0800" }, "message": "drm/mediatek: no change parent rate in round_rate() for MT2701 hdmi phy\n\nThis is the third step to make MT2701 HDMI stable.\nWe should not change the rate of parent for hdmi phy when\ndoing round_rate for this clock. The parent clock of hdmi\nphy must be the same as it. We change it when doing set_rate\nonly.\n\nSigned-off-by: Wangyan Wang \u003cwangyan.wang@mediatek.com\u003e\nSigned-off-by: CK Hu \u003cck.hu@mediatek.com\u003e\n" }, { "commit": "8eeb3946feeb00486ac0909e2309da87db8988a5", "tree": "eea69ab79907588bec40e4ce9694ca65805723ab", "parents": [ "827abdd024207146822f66ba3ba74867135866b9" ], "author": { "name": "Wangyan Wang", "email": "wangyan.wang@mediatek.com", "time": "Tue Apr 09 14:53:05 2019 +0800" }, "committer": { "name": "CK Hu", "email": "ck.hu@mediatek.com", "time": "Tue Apr 09 17:47:01 2019 +0800" }, "message": "drm/mediatek: using new factor for tvdpll for MT2701 hdmi phy\n\nThis is the second step to make MT2701 HDMI stable.\nThe factor depends on the divider of DPI in MT2701, therefore,\nwe should fix this factor to the right and new one.\nTest: search ok\n\nSigned-off-by: Wangyan Wang \u003cwangyan.wang@mediatek.com\u003e\nSigned-off-by: CK Hu \u003cck.hu@mediatek.com\u003e\n" }, { "commit": "827abdd024207146822f66ba3ba74867135866b9", "tree": "40d7b501f77e4f6164b35bcaec52d13f11c644d0", "parents": [ "321b628e6f5a3af999f75eadd373adbcb8b4cb1f" ], "author": { "name": "Wangyan Wang", "email": "wangyan.wang@mediatek.com", "time": "Tue Apr 09 14:53:03 2019 +0800" }, "committer": { "name": "CK Hu", "email": "ck.hu@mediatek.com", "time": "Tue Apr 09 17:47:01 2019 +0800" }, "message": "drm/mediatek: remove flag CLK_SET_RATE_PARENT for MT2701 hdmi phy\n\nThis is the first step to make MT2701 hdmi stable.\nThe parent rate of hdmi phy had set by DPI driver.\nWe should not set or change the parent rate of MT2701 hdmi phy,\nas a result we should remove the flags of \"CLK_SET_RATE_PARENT\"\nfrom the clock of MT2701 hdmi phy.\n\nSigned-off-by: Wangyan Wang \u003cwangyan.wang@mediatek.com\u003e\nSigned-off-by: CK Hu \u003cck.hu@mediatek.com\u003e\n" }, { "commit": "321b628e6f5a3af999f75eadd373adbcb8b4cb1f", "tree": "f08e6986ea2112b2e98150e66d6d9465d288e315", "parents": [ "0c24613cda163dedfa229afc8eff6072e57fac8d" ], "author": { "name": "Wangyan Wang", "email": "wangyan.wang@mediatek.com", "time": "Tue Apr 09 14:53:06 2019 +0800" }, "committer": { "name": "CK Hu", "email": "ck.hu@mediatek.com", "time": "Tue Apr 09 17:46:54 2019 +0800" }, "message": "drm/mediatek: make implementation of recalc_rate() for MT2701 hdmi phy\n\nRecalculate the rate of this clock, by querying hardware to\nmake implementation of recalc_rate() to match the definition.\n\nSigned-off-by: Wangyan Wang \u003cwangyan.wang@mediatek.com\u003e\nSigned-off-by: CK Hu \u003cck.hu@mediatek.com\u003e\n" }, { "commit": "0c24613cda163dedfa229afc8eff6072e57fac8d", "tree": "ea80a1aa799e7e776af07793f33d3a2f443e9952", "parents": [ "2ae2c3316fb77dcf64275d011596b60104c45426" ], "author": { "name": "Wangyan Wang", "email": "wangyan.wang@mediatek.com", "time": "Tue Apr 09 14:53:04 2019 +0800" }, "committer": { "name": "CK Hu", "email": "ck.hu@mediatek.com", "time": "Tue Apr 09 17:15:40 2019 +0800" }, "message": "drm/mediatek: fix the rate and divder of hdmi phy for MT2701\n\nDue to a clerical error,there is one zero less for 12800000.\nFix it for 128000000\nFixes: 0fc721b2968e (\"drm/mediatek: add hdmi driver for MT2701 and MT7623\")\n\nSigned-off-by: Wangyan Wang \u003cwangyan.wang@mediatek.com\u003e\nSigned-off-by: CK Hu \u003cck.hu@mediatek.com\u003e\n" }, { "commit": "c5781ffbbd4f742a58263458145fe7f0ac01d9e0", "tree": "af293b3cea713c385afdf59961c18f8e0e2b3516", "parents": [ "15ade5d2e7775667cf191cf2f94327a4889f8b9d" ], "author": { "name": "Erik Schmauss", "email": "erik.schmauss@intel.com", "time": "Mon Apr 08 13:42:26 2019 -0700" }, "committer": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Tue Apr 09 10:05:11 2019 +0200" }, "message": "ACPICA: Namespace: remove address node from global list after method termination\n\nACPICA commit b233720031a480abd438f2e9c643080929d144c3\n\nASL operation_regions declare a range of addresses that it uses. In a\nperfect world, the range of addresses should be used exclusively by\nthe AML interpreter. The OS can use this information to decide which\ndrivers to load so that the AML interpreter and device drivers use\ndifferent regions of memory.\n\nDuring table load, the address information is added to a global\naddress range list. Each node in this list contains an address range\nas well as a namespace node of the operation_region. This list is\ndeleted at ACPI shutdown.\n\nUnfortunately, ASL operation_regions can be declared inside of control\nmethods. Although this is not recommended, modern firmware contains\nsuch code. New module level code changes unintentionally removed the\nfunctionality of adding and removing nodes to the global address\nrange list.\n\nA few months ago, support for adding addresses has been re-\nimplemented. However, the removal of the address range list was\nmissed and resulted in some systems to crash due to the address list\ncontaining bogus namespace nodes from operation_regions declared in\ncontrol methods. In order to fix the crash, this change removes\ndynamic operation_regions after control method termination.\n\nLink: https://github.com/acpica/acpica/commit/b2337200\nLink: https://bugzilla.kernel.org/show_bug.cgi?id\u003d202475\nFixes: 4abb951b73ff (\"ACPICA: AML interpreter: add region addresses in global list during initialization\")\nReported-by: Michael J Gruber \u003cmjg@fedoraproject.org\u003e\nSigned-off-by: Erik Schmauss \u003cerik.schmauss@intel.com\u003e\nSigned-off-by: Bob Moore \u003crobert.moore@intel.com\u003e\nCc: 4.20+ \u003cstable@vger.kernel.org\u003e # 4.20+\nSigned-off-by: Rafael J. Wysocki \u003crafael.j.wysocki@intel.com\u003e\n" } ], "next": "2ae2c3316fb77dcf64275d011596b60104c45426" }