kernel/linux.git/drivers/nvdimm, branch v4.4.171

libnvdimm: Hold reference on parent while scheduling async init

2018-11-21T08:27:34+00:00

commit b6eae0f61db27748606cc00dafcfd1e2c032f0a5 upstream. Unlike asynchronous initialization in the core we have not yet associated the device with the parent, and as such the device doesn't hold a reference to the parent. In order to resolve that we should be holding a reference on the parent until the asynchronous initialization has completed. Cc: Fixes: 4d88a97aa9e8 ("libnvdimm: ...base ... infrastructure") Signed-off-by: Alexander Duyck Signed-off-by: Dan Williams Signed-off-by: Greg Kroah-Hartman

linvdimm, pmem: Preserve read-only setting for pmem devices

2018-07-03T09:21:31+00:00

commit 254a4cd50b9fe2291a12b8902e08e56dcc4e9b10 upstream. The pmem driver does not honor a forced read-only setting for very long: $ blockdev --setro /dev/pmem0 $ blockdev --getro /dev/pmem0 1 followed by various commands like these: $ blockdev --rereadpt /dev/pmem0 or $ mkfs.ext4 /dev/pmem0 results in this in the kernel serial log: nd_pmem namespace0.0: region0 read-write, marking pmem0 read-write with the read-only setting lost: $ blockdev --getro /dev/pmem0 0 That's from bus.c nvdimm_revalidate_disk(), which always applies the setting from nd_region (which is initially based on the ACPI NFIT NVDIMM state flags not_armed bit). In contrast, commit 20bd1d026aac ("scsi: sd: Keep disk read-only when re-reading partition") fixed this issue for SCSI devices to preserve the previous setting if it was set to read-only. This patch modifies bus.c to preserve any previous read-only setting. It also eliminates the kernel serial log print except for cases where read-write is changed to read-only, so it doesn't print read-only to read-only non-changes. Cc: Fixes: 581388209405 ("libnvdimm, nfit: handle unarmed dimms, mark namespaces read-only") Signed-off-by: Robert Elliott Signed-off-by: Dan Williams Signed-off-by: Greg Kroah-Hartman

libnvdimm, namespace: make 'resource' attribute only readable by root

2017-11-30T08:37:23+00:00

commit c1fb3542074fd0c4d901d778bd52455111e4eb6f upstream. For the same reason that /proc/iomem returns 0's for non-root readers and acpi tables are root-only, make the 'resource' attribute for namespace devices only readable by root. Otherwise we disclose physical address information. Fixes: bf9bccc14c05 ("libnvdimm: pmem label sets and namespace instantiation") Reported-by: Dave Hansen Signed-off-by: Dan Williams Signed-off-by: Greg Kroah-Hartman

libnvdimm, namespace: fix label initialization to use valid seq numbers

2017-11-30T08:37:23+00:00

commit b18d4b8a25af6fe83d7692191d6ff962ea611c4f upstream. The set of valid sequence numbers is {1,2,3}. The specification indicates that an implementation should consider 0 a sign of a critical error: UEFI 2.7: 13.19 NVDIMM Label Protocol Software never writes the sequence number 00, so a correctly check-summed Index Block with this sequence number probably indicates a critical error. When software discovers this case it treats it as an invalid Index Block indication. While the expectation is that the invalid block is just thrown away, the Robustness Principle says we should fix this to make both sequence numbers valid. Fixes: f524bf271a5c ("libnvdimm: write pmem label set") Reported-by: Juston Li Signed-off-by: Dan Williams Signed-off-by: Greg Kroah-Hartman

libnvdimm, btt: fix btt_rw_page not returning errors

2017-08-07T02:19:42+00:00

commit c13c43d54f2c6a3be1c675766778ac1ad8dfbfcc upstream. btt_rw_page was not propagating errors frm btt_do_bvec, resulting in any IO errors via the rw_page path going unnoticed. the pmem driver recently fixed this in e10624f pmem: fail io-requests to known bad blocks but same problem in BTT went neglected. Fixes: 5212e11fde4d ("nd_btt: atomic sector updates") Cc: Toshi Kani Cc: Dan Williams Cc: Jeff Moyer Signed-off-by: Vishal Verma Signed-off-by: Dan Williams Signed-off-by: Greg Kroah-Hartman

libnvdimm: fix reconfig_mutex, mmap_sem, and jbd2_handle lockdep splat

2017-04-21T07:30:06+00:00

commit 0beb2012a1722633515c8aaa263c73449636c893 upstream. Holding the reconfig_mutex over a potential userspace fault sets up a lockdep dependency chain between filesystem-DAX and the libnvdimm ioctl path. Move the user access outside of the lock. [ INFO: possible circular locking dependency detected ] 4.11.0-rc3+ #13 Tainted: G W O ------------------------------------------------------- fallocate/16656 is trying to acquire lock: (&nvdimm_bus->reconfig_mutex){+.+.+.}, at: [] nvdimm_bus_lock+0x21/0x30 [libnvdimm] but task is already holding lock: (jbd2_handle){++++..}, at: [] start_this_handle+0x104/0x460 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #2 (jbd2_handle){++++..}: lock_acquire+0xbd/0x200 start_this_handle+0x16a/0x460 jbd2__journal_start+0xe9/0x2d0 __ext4_journal_start_sb+0x89/0x1c0 ext4_dirty_inode+0x32/0x70 __mark_inode_dirty+0x235/0x670 generic_update_time+0x87/0xd0 touch_atime+0xa9/0xd0 ext4_file_mmap+0x90/0xb0 mmap_region+0x370/0x5b0 do_mmap+0x415/0x4f0 vm_mmap_pgoff+0xd7/0x120 SyS_mmap_pgoff+0x1c5/0x290 SyS_mmap+0x22/0x30 entry_SYSCALL_64_fastpath+0x1f/0xc2 -> #1 (&mm->mmap_sem){++++++}: lock_acquire+0xbd/0x200 __might_fault+0x70/0xa0 __nd_ioctl+0x683/0x720 [libnvdimm] nvdimm_ioctl+0x8b/0xe0 [libnvdimm] do_vfs_ioctl+0xa8/0x740 SyS_ioctl+0x79/0x90 do_syscall_64+0x6c/0x200 return_from_SYSCALL_64+0x0/0x7a -> #0 (&nvdimm_bus->reconfig_mutex){+.+.+.}: __lock_acquire+0x16b6/0x1730 lock_acquire+0xbd/0x200 __mutex_lock+0x88/0x9b0 mutex_lock_nested+0x1b/0x20 nvdimm_bus_lock+0x21/0x30 [libnvdimm] nvdimm_forget_poison+0x25/0x50 [libnvdimm] nvdimm_clear_poison+0x106/0x140 [libnvdimm] pmem_do_bvec+0x1c2/0x2b0 [nd_pmem] pmem_make_request+0xf9/0x270 [nd_pmem] generic_make_request+0x118/0x3b0 submit_bio+0x75/0x150 Fixes: 62232e45f4a2 ("libnvdimm: control (ioctl) messages for nvdimm_bus and nvdimm devices") Cc: Dave Jiang Reported-by: Vishal Verma Signed-off-by: Dan Williams Signed-off-by: Greg Kroah-Hartman

nfit, libnvdimm: fix interleave set cookie calculation

2017-03-18T11:09:58+00:00

commit 86ef58a4e35e8fa66afb5898cf6dec6a3bb29f67 upstream. The interleave-set cookie is a sum that sanity checks the composition of an interleave set has not changed from when the namespace was initially created. The checksum is calculated by sorting the DIMMs by their location in the interleave-set. The comparison for the sort must be 64-bit wide, not byte-by-byte as performed by memcmp() in the broken case. Fix the implementation to accept correct cookie values in addition to the Linux "memcmp" order cookies, but only allow correct cookies to be generated going forward. It does mean that namespaces created by third-party-tooling, or created by newer kernels with this fix, will not validate on older kernels. However, there are a couple mitigating conditions: 1/ platforms with namespace-label capable NVDIMMs are not widely available. 2/ interleave-sets with a single-dimm are by definition not affected (nothing to sort). This covers the QEMU-KVM NVDIMM emulation case. The cookie stored in the namespace label will be fixed by any write the namespace label, the most straightforward way to achieve this is to write to the "alt_name" attribute of a namespace in sysfs. Fixes: eaf961536e16 ("libnvdimm, nfit: add interleave-set state-tracking infrastructure") Reported-by: Nicholas Moulin Tested-by: Nicholas Moulin Signed-off-by: Dan Williams Signed-off-by: Greg Kroah-Hartman

libnvdimm, pfn: fix uuid validation

2016-04-20T06:41:54+00:00

commit e5670563f588ed1c0603819350c0f02cec23f5c5 upstream. If we detect a namespace has a stale info block in the init path, we should overwrite with the latest configuration. In fact, we already return -ENODEV when the parent uuid is invalid, the same should be done for the 'self' uuid. Otherwise we can get into a condition where userspace is unable to reconfigure the pfn-device without directly / manually invalidating the info block. Reported-by: Jeff Moyer Signed-off-by: Dan Williams Signed-off-by: Greg Kroah-Hartman

libnvdimm: fix smart data retrieval

2016-04-20T06:41:54+00:00

commit 211291126698c8f047617565b2e2e7f822f86354 upstream. It appears that smart data retrieval has been broken the since the initial implementation. Fix the payload size to be 128-bytes per the specification. Signed-off-by: Dan Williams Signed-off-by: Greg Kroah-Hartman

libnvdimm: Fix security issue with DSM IOCTL.

2016-04-12T16:08:39+00:00

commit 07accfa9d1a8bac8262f6d24a94a54d2d1f35149 upstream. Code attempts to prevent certain IOCTL DSM from being called when device is opened read only. This security feature can be trivially overcome by changing the size portion of the ioctl_command which isn't used. Check only the _IOC_NR (i.e. the command). Signed-off-by: Jerry Hoemann Signed-off-by: Dan Williams Signed-off-by: Greg Kroah-Hartman