Pull device-dax updates from Dan Williams:
"New device-dax infrastructure to allow persistent memory and other
"reserved" / performance differentiated memories, to be assigned to
the core-mm as "System RAM".
Some users want to use persistent memory as additional volatile
memory. They are willing to cope with potential performance
differences, for example between DRAM and 3D Xpoint, and want to use
typical Linux memory management apis rather than a userspace memory
allocator layered over an mmap() of a dax file. The administration
model is to decide how much Persistent Memory (pmem) to use as System
RAM, create a device-dax-mode namespace of that size, and then assign
it to the core-mm. The rationale for device-dax is that it is a
generic memory-mapping driver that can be layered over any "special
purpose" memory, not just pmem. On subsequent boots udev rules can be
used to restore the memory assignment.
One implication of using pmem as RAM is that mlock() no longer keeps
data off persistent media. For this reason it is recommended to enable
NVDIMM Security (previously merged for 5.0) to encrypt pmem contents
at rest. We considered making this recommendation an actively enforced
requirement, but in the end decided to leave it as a distribution /
administrator policy to allow for emulation and test environments that
lack security capable NVDIMMs.
Summary:
- Replace the /sys/class/dax device model with /sys/bus/dax, and
include a compat driver so distributions can opt-in to the new ABI.
- Allow for an alternative driver for the device-dax address-range
- Introduce the 'kmem' driver to hotplug / assign a device-dax
address-range to the core-mm.
- Arrange for the device-dax target-node to be onlined so that the
newly added memory range can be uniquely referenced by numa apis"
NOTE! I'm not entirely happy with the whole "PMEM as RAM" model because
we currently have special - and very annoying rules in the kernel about
accessing PMEM only with the "MC safe" accessors, because machine checks
inside the regular repeat string copy functions can be fatal in some
(not described) circumstances.
And apparently the PMEM modules can cause that a lot more than regular
RAM. The argument is that this happens because PMEM doesn't necessarily
get scrubbed at boot like RAM does, but that is planned to be added for
the user space tooling.
Quoting Dan from another email:
"The exposure can be reduced in the volatile-RAM case by scanning for
and clearing errors before it is onlined as RAM. The userspace tooling
for that can be in place before v5.1-final. There's also runtime
notifications of errors via acpi_nfit_uc_error_notify() from
background scrubbers on the DIMM devices. With that mechanism the
kernel could proactively clear newly discovered poison in the volatile
case, but that would be additional development more suitable for v5.2.
I understand the concern, and the need to highlight this issue by
tapping the brakes on feature development, but I don't see PMEM as RAM
making the situation worse when the exposure is also there via DAX in
the PMEM case. Volatile-RAM is arguably a safer use case since it's
possible to repair pages where the persistent case needs active
application coordination"
* tag 'devdax-for-5.1' of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm:
device-dax: "Hotplug" persistent memory for use like normal RAM
mm/resource: Let walk_system_ram_range() search child resources
mm/memory-hotplug: Allow memory resources to be children
mm/resource: Move HMM pr_debug() deeper into resource code
mm/resource: Return real error codes from walk failures
device-dax: Add a 'modalias' attribute to DAX 'bus' devices
device-dax: Add a 'target_node' attribute
device-dax: Auto-bind device after successful new_id
acpi/nfit, device-dax: Identify differentiated memory with a unique numa-node
device-dax: Add /sys/class/dax backwards compatibility
device-dax: Add support for a dax override driver
device-dax: Move resource pinning+mapping into the common driver
device-dax: Introduce bus + driver model
device-dax: Start defining a dax bus model
device-dax: Remove multi-resource infrastructure
device-dax: Kill dax_region base
device-dax: Kill dax_region ida
I thought Josef Bacik's patch to drop the mmap_sem was buggy, because
when looking at the error cases, there was one case where we returned
VM_FAULT_RETRY without actually dropping the mmap_sem.
Josef had to explain to me (using small words) that yes, that's actually
what we're supposed to do, and his patch was correct. Which not only
convinced me he knew what he was doing and I should stop arguing with
him, but also that I should add a comment to the case I was confused
about.
Patiently-pointed-out-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Currently we only drop the mmap_sem if there is contention on the page
lock. The idea is that we issue readahead and then go to lock the page
while it is under IO and we want to not hold the mmap_sem during the IO.
The problem with this is the assumption that the readahead does anything.
In the case that the box is under extreme memory or IO pressure we may end
up not reading anything at all for readahead, which means we will end up
reading in the page under the mmap_sem.
Even if the readahead does something, it could get throttled because of io
pressure on the system and the process is in a lower priority cgroup.
Holding the mmap_sem while doing IO is problematic because it can cause
system-wide priority inversions. Consider some large company that does a
lot of web traffic. This large company has load balancing logic in it's
core web server, cause some engineer thought this was a brilliant plan.
This load balancing logic gets statistics from /proc about the system,
which trip over processes mmap_sem for various reasons. Now the web
server application is in a protected cgroup, but these other processes may
not be, and if they are being throttled while their mmap_sem is held we'll
stall, and cause this nice death spiral.
Instead rework filemap fault path to drop the mmap sem at any point that
we may do IO or block for an extended period of time. This includes while
issuing readahead, locking the page, or needing to call ->readpage because
readahead did not occur. Then once we have a fully uptodate page we can
return with VM_FAULT_RETRY and come back again to find our nicely in-cache
page that was gotten outside of the mmap_sem.
This patch also adds a new helper for locking the page with the mmap_sem
dropped. This doesn't make sense currently as generally speaking if the
page is already locked it'll have been read in (unless there was an error)
before it was unlocked. However a forthcoming patchset will change this
with the ability to abort read-ahead bio's if necessary, making it more
likely that we could contend for a page lock and still have a not uptodate
page. This allows us to deal with this case by grabbing the lock and
issuing the IO without the mmap_sem held, and then returning
VM_FAULT_RETRY to come back around.
[josef@toxicpanda.com: v6]
Link: http://lkml.kernel.org/r/20181212152757.10017-1-josef@toxicpanda.com
[kirill@shutemov.name: fix race in filemap_fault()]
Link: http://lkml.kernel.org/r/20181228235106.okk3oastsnpxusxs@kshutemo-mobl1
[akpm@linux-foundation.org: coding style fixes]
Link: http://lkml.kernel.org/r/20181211173801.29535-4-josef@toxicpanda.com
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Tested-by: syzbot+b437b5a429d680cf2217@syzkaller.appspotmail.com
Cc: Dave Chinner <david@fromorbit.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: "Kirill A. Shutemov" <kirill@shutemov.name>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Patch series "drop the mmap_sem when doing IO in the fault path", v6.
Now that we have proper isolation in place with cgroups2 we have started
going through and fixing the various priority inversions. Most are all
gone now, but this one is sort of weird since it's not necessarily a
priority inversion that happens within the kernel, but rather because of
something userspace does.
We have giant applications that we want to protect, and parts of these
giant applications do things like watch the system state to determine how
healthy the box is for load balancing and such. This involves running
'ps' or other such utilities. These utilities will often walk
/proc/<pid>/whatever, and these files can sometimes need to
down_read(&task->mmap_sem). Not usually a big deal, but we noticed when
we are stress testing that sometimes our protected application has latency
spikes trying to get the mmap_sem for tasks that are in lower priority
cgroups.
This is because any down_write() on a semaphore essentially turns it into
a mutex, so even if we currently have it held for reading, any new readers
will not be allowed on to keep from starving the writer. This is fine,
except a lower priority task could be stuck doing IO because it has been
throttled to the point that its IO is taking much longer than normal. But
because a higher priority group depends on this completing it is now stuck
behind lower priority work.
In order to avoid this particular priority inversion we want to use the
existing retry mechanism to stop from holding the mmap_sem at all if we
are going to do IO. This already exists in the read case sort of, but
needed to be extended for more than just grabbing the page lock. With
io.latency we throttle at submit_bio() time, so the readahead stuff can
block and even page_cache_read can block, so all these paths need to have
the mmap_sem dropped.
The other big thing is ->page_mkwrite. btrfs is particularly shitty here
because we have to reserve space for the dirty page, which can be a very
expensive operation. We use the same retry method as the read path, and
simply cache the page and verify the page is still setup properly the next
pass through ->page_mkwrite().
I've tested these patches with xfstests and there are no regressions.
This patch (of 3):
If we do not have a page at filemap_fault time we'll do this weird forced
page_cache_read thing to populate the page, and then drop it again and
loop around and find it. This makes for 2 ways we can read a page in
filemap_fault, and it's not really needed. Instead add a FGP_FOR_MMAP
flag so that pagecache_get_page() will return a unlocked page that's in
pagecache. Then use the normal page locking and readpage logic already in
filemap_fault. This simplifies the no page in page cache case
significantly.
[akpm@linux-foundation.org: fix comment text]
[josef@toxicpanda.com: don't unlock null page in FGP_FOR_MMAP case]
Link: http://lkml.kernel.org/r/20190312201742.22935-1-josef@toxicpanda.com
Link: http://lkml.kernel.org/r/20181211173801.29535-2-josef@toxicpanda.com
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Tejun Heo <tj@kernel.org>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: "Kirill A. Shutemov" <kirill@shutemov.name>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Merge misc updates from Andrew Morton:
- a few misc things
- the rest of MM
- remove flex_arrays, replace with new simple radix-tree implementation
* emailed patches from Andrew Morton <akpm@linux-foundation.org>: (38 commits)
Drop flex_arrays
sctp: convert to genradix
proc: commit to genradix
generic radix trees
selinux: convert to kvmalloc
md: convert to kvmalloc
openvswitch: convert to kvmalloc
of: fix kmemleak crash caused by imbalance in early memory reservation
mm: memblock: update comments and kernel-doc
memblock: split checks whether a region should be skipped to a helper function
memblock: remove memblock_{set,clear}_region_flags
memblock: drop memblock_alloc_*_nopanic() variants
memblock: memblock_alloc_try_nid: don't panic
treewide: add checks for the return value of memblock_alloc*()
swiotlb: add checks for the return value of memblock_alloc*()
init/main: add checks for the return value of memblock_alloc*()
mm/percpu: add checks for the return value of memblock_alloc*()
sparc: add checks for the return value of memblock_alloc*()
ia64: add checks for the return value of memblock_alloc*()
arch: don't memset(0) memory returned by memblock_alloc()
...
Currently, memblock has several internal functions with overlapping
functionality. They all call memblock_find_in_range_node() to find free
memory and then reserve the allocated range and mark it with kmemleak.
However, there is difference in the allocation constraints and in
fallback strategies.
The allocations returning physical address first attempt to find free
memory on the specified node within mirrored memory regions, then retry
on the same node without the requirement for memory mirroring and
finally fall back to all available memory.
The allocations returning virtual address start with clamping the
allowed range to memblock.current_limit, attempt to allocate from the
specified node from regions with mirroring and with user defined minimal
address. If such allocation fails, next attempt is done with node
restriction lifted. Next, the allocation is retried with minimal
address reset to zero and at last without the requirement for mirrored
regions.
Let's consolidate various fallbacks handling and make them more
consistent for physical and virtual variants. Most of the fallback
handling is moved to memblock_alloc_range_nid() and it now handles node
and mirror fallbacks.
The memblock_alloc_internal() uses memblock_alloc_range_nid() to get a
physical address of the allocated range and converts it to virtual
address.
The fallback for allocation below the specified minimal address remains
in memblock_alloc_internal() because memblock_alloc_range_nid() is used
by CMA with exact requirement for lower bounds.
The memblock_phys_alloc_nid() function is completely dropped as it is not
used anywhere outside memblock and its only usage can be replaced by a
call to memblock_alloc_range_nid().
[rppt@linux.ibm.com: fix parameter order in memblock_phys_alloc_try_nid()]
Link: http://lkml.kernel.org/r/20190203113915.GC8620@rapoport-lnx
Link: http://lkml.kernel.org/r/1548057848-15136-11-git-send-email-rppt@linux.ibm.com
Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Tested-by: Michael Ellerman <mpe@ellerman.id.au>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christophe Leroy <christophe.leroy@c-s.fr>
Cc: Christoph Hellwig <hch@lst.de>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Dennis Zhou <dennis@kernel.org>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Greentime Hu <green.hu@gmail.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Guan Xuetao <gxt@pku.edu.cn>
Cc: Guo Ren <guoren@kernel.org>
Cc: Guo Ren <ren_guo@c-sky.com> [c-sky]
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Juergen Gross <jgross@suse.com> [Xen]
Cc: Mark Salter <msalter@redhat.com>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Paul Burton <paul.burton@mips.com>
Cc: Petr Mladek <pmladek@suse.com>
Cc: Richard Weinberger <richard@nod.at>
Cc: Rich Felker <dalias@libc.org>
Cc: Rob Herring <robh+dt@kernel.org>
Cc: Rob Herring <robh@kernel.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Vineet Gupta <vgupta@synopsys.com>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull xen updates from Juergen Gross:
"xen fixes and features:
- remove fallback code for very old Xen hypervisors
- three patches for fixing Xen dom0 boot regressions
- an old patch for Xen PCI passthrough which was never applied for
unknown reasons
- some more minor fixes and cleanup patches"
* tag 'for-linus-5.1a-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
xen: fix dom0 boot on huge systems
xen, cpu_hotplug: Prevent an out of bounds access
xen: remove pre-xen3 fallback handlers
xen/ACPI: Switch to bitmap_zalloc()
x86/xen: dont add memory above max allowed allocation
x86: respect memory size limiting via mem= parameter
xen/gntdev: Check and release imported dma-bufs on close
xen/gntdev: Do not destroy context while dma-bufs are in use
xen/pciback: Don't disable PCI_COMMAND on PCI device reset.
xen-scsiback: mark expected switch fall-through
xen: mark expected switch fall-through
Pull rdma updates from Jason Gunthorpe:
"This has been a slightly more active cycle than normal with ongoing
core changes and quite a lot of collected driver updates.
- Various driver fixes for bnxt_re, cxgb4, hns, mlx5, pvrdma, rxe
- A new data transfer mode for HFI1 giving higher performance
- Significant functional and bug fix update to the mlx5
On-Demand-Paging MR feature
- A chip hang reset recovery system for hns
- Change mm->pinned_vm to an atomic64
- Update bnxt_re to support a new 57500 chip
- A sane netlink 'rdma link add' method for creating rxe devices and
fixing the various unregistration race conditions in rxe's
unregister flow
- Allow lookup up objects by an ID over netlink
- Various reworking of the core to driver interface:
- drivers should not assume umem SGLs are in PAGE_SIZE chunks
- ucontext is accessed via udata not other means
- start to make the core code responsible for object memory
allocation
- drivers should convert struct device to struct ib_device via a
helper
- drivers have more tools to avoid use after unregister problems"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma: (280 commits)
net/mlx5: ODP support for XRC transport is not enabled by default in FW
IB/hfi1: Close race condition on user context disable and close
RDMA/umem: Revert broken 'off by one' fix
RDMA/umem: minor bug fix in error handling path
RDMA/hns: Use GFP_ATOMIC in hns_roce_v2_modify_qp
cxgb4: kfree mhp after the debug print
IB/rdmavt: Fix concurrency panics in QP post_send and modify to error
IB/rdmavt: Fix loopback send with invalidate ordering
IB/iser: Fix dma_nents type definition
IB/mlx5: Set correct write permissions for implicit ODP MR
bnxt_re: Clean cq for kernel consumers only
RDMA/uverbs: Don't do double free of allocated PD
RDMA: Handle ucontext allocations by IB/core
RDMA/core: Fix a WARN() message
bnxt_re: fix the regression due to changes in alloc_pbl
IB/mlx4: Increase the timeout for CM cache
IB/core: Abort page fault handler silently during owning process exit
IB/mlx5: Validate correct PD before prefetch MR
IB/mlx5: Protect against prefetch of invalid MR
RDMA/uverbs: Store PR pointer before it is overwritten
...
Pull x86 mm cleanup from Ingo Molnar:
"A single GUP cleanup"
* 'x86-mm-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
mm/gup: Remove the 'write' parameter from gup_fast_permitted()
