The sendctx circular queue now guarantees that xprtrdma cannot
overflow the Send Queue, so remove the remaining bits of the
original Send WQE counting mechanism.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
When an RPC Call includes a file data payload, that payload can come
from pages in the page cache, or a user buffer (for direct I/O).
If the payload can fit inline, xprtrdma includes it in the Send
using a scatter-gather technique. xprtrdma mustn't allow the RPC
consumer to re-use the memory where that payload resides before the
Send completes. Otherwise, the new contents of that memory would be
exposed by an HCA retransmit of the Send operation.
So, block RPC completion on Send completion, but only in the case
where a separate file data payload is part of the Send. This
prevents the reuse of that memory while it is still part of a Send
operation without an undue cost to other cases.
Waiting is avoided in the common case because typically the Send
will have completed long before the RPC Reply arrives.
These days, an RPC timeout will trigger a disconnect, which tears
down the QP. The disconnect flushes all waiting Sends. This bounds
the amount of time the reply handler has to wait for a Send
completion.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Invoke a common routine for releasing hardware resources (for
example, invalidating MRs). This needs to be done whether an
RPC Reply has arrived or the RPC was terminated early.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
We have one boolean flag in rpcrdma_req today. I'd like to add more
flags, so convert that boolean to a bit flag.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Problem statement:
Recently Sagi Grimberg <sagi@grimberg.me> observed that kernel RDMA-
enabled storage initiators don't handle delayed Send completion
correctly. If Send completion is delayed beyond the end of a ULP
transaction, the ULP may release resources that are still being used
by the HCA to complete a long-running Send operation.
This is a common design trait amongst our initiators. Most Send
operations are faster than the ULP transaction they are part of.
Waiting for a completion for these is typically unnecessary.
Infrequently, a network partition or some other problem crops up
where an ordering problem can occur. In NFS parlance, the RPC Reply
arrives and completes the RPC, but the HCA is still retrying the
Send WR that conveyed the RPC Call. In this case, the HCA can try
to use memory that has been invalidated or DMA unmapped, and the
connection is lost. If that memory has been re-used for something
else (possibly not related to NFS), and the Send retransmission
exposes that data on the wire.
Thus we cannot assume that it is safe to release Send-related
resources just because a ULP reply has arrived.
After some analysis, we have determined that the completion
housekeeping will not be difficult for xprtrdma:
- Inline Send buffers are registered via the local DMA key, and
are already left DMA mapped for the lifetime of a transport
connection, thus no additional handling is necessary for those
- Gathered Sends involving page cache pages _will_ need to
DMA unmap those pages after the Send completes. But like
inline send buffers, they are registered via the local DMA key,
and thus will not need to be invalidated
In addition, RPC completion will need to wait for Send completion
in the latter case. However, nearly always, the Send that conveys
the RPC Call will have completed long before the RPC Reply
arrives, and thus no additional latency will be accrued.
Design notes:
In this patch, the rpcrdma_sendctx object is introduced, and a
lock-free circular queue is added to manage a set of them per
transport.
The RPC client's send path already prevents sending more than one
RPC Call at the same time. This allows us to treat the consumer
side of the queue (rpcrdma_sendctx_get_locked) as if there is a
single consumer thread.
The producer side of the queue (rpcrdma_sendctx_put_locked) is
invoked only from the Send completion handler, which is a single
thread of execution (soft IRQ).
The only care that needs to be taken is with the tail index, which
is shared between the producer and consumer. Only the producer
updates the tail index. The consumer compares the head with the
tail to ensure that the a sendctx that is in use is never handed
out again (or, expressed more conventionally, the queue is empty).
When the sendctx queue empties completely, there are enough Sends
outstanding that posting more Send operations can result in a Send
Queue overflow. In this case, the ULP is told to wait and try again.
This introduces strong Send Queue accounting to xprtrdma.
As a final touch, Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
suggested a mechanism that does not require signaling every Send.
We signal once every N Sends, and perform SGE unmapping of N Send
operations during that one completion.
Reported-by: Sagi Grimberg <sagi@grimberg.me>
Suggested-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Commit 655fec6987 ("xprtrdma: Use gathered Send for large inline
messages") assumed that, since the zeroeth element of the Send SGE
array always pointed to req->rl_rdmabuf, it needed to be initialized
just once. This was a valid assumption because the Send SGE array
and rl_rdmabuf both live in the same rpcrdma_req.
In a subsequent patch, the Send SGE array will be separated from the
rpcrdma_req, so the zeroeth element of the SGE array needs to be
initialized every time.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Clean up: Make rpcrdma_prepare_send_sges() return a negative errno
instead of a bool. Soon callers will want distinct treatments of
different types of failures.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
When this function fails, it needs to undo the DMA mappings it's
done so far. Otherwise these are leaked.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Clean up. rpcrdma_prepare_hdr_sge() sets num_sge to one, then
rpcrdma_prepare_msg_sges() sets num_sge again to the count of SGEs
it added, plus one for the header SGE just mapped in
rpcrdma_prepare_hdr_sge(). This is confusing, and nails in an
assumption about when these functions are called.
Instead, maintain a running count that both functions can update
with just the number of SGEs they have added to the SGE array.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
We need to decode and save the incoming rdma_credits field _after_
we know that the direction of the message is "forward direction
Reply". Otherwise, the credits value in reverse direction Calls is
also used to update the forward direction credits.
It is safe to decode the rdma_credits field in rpcrdma_reply_handler
now that rpcrdma_reply_handler is single-threaded. Receives complete
in the same order as they were sent on the NFS server.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
I noticed that the soft IRQ thread looked pretty busy under heavy
I/O workloads. perf suggested one area that was expensive was the
queue_work() call in rpcrdma_wc_receive. That gave me some ideas.
Instead of scheduling a separate worker to process RPC Replies,
promote the Receive completion handler to IB_POLL_WORKQUEUE, and
invoke rpcrdma_reply_handler directly.
Note that the poll workqueue is single-threaded. In order to keep
memory invalidation from serializing all RPC Replies, handle any
necessary invalidation tasks in a separate multi-threaded workqueue.
This provides a two-tier scheme, similar to OS I/O interrupt
handlers: A fast interrupt handler that schedules the slow handler
and re-enables the interrupt, and a slower handler that is invoked
for any needed heavy lifting.
Benefits include:
- One less context switch for RPCs that don't register memory
- Receive completion handling is moved out of soft IRQ context to
make room for other users of soft IRQ
- The same CPU core now DMA syncs and XDR decodes the Receive buffer
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Clean up: I'd like to be able to invoke the tail of
rpcrdma_reply_handler in two different places. Split the tail out
into its own helper function.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Clean up: Make it easier to pass the decoded XID, vers, credits, and
proc fields around by moving these variables into struct rpcrdma_rep.
Note: the credits field will be handled in a subsequent patch.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
A reply with an unrecognized value in the version field means the
transport header is potentially garbled and therefore all the fields
are untrustworthy.
Fixes: 59aa1f9a3c ("xprtrdma: Properly handle RDMA_ERROR ... ")
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Clean up: There are no remaining callers of this method.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
The "safe" version of ro_unmap is used here to avoid waiting
unnecessarily. However:
- It is safe to wait. After all, we have to wait anyway when using
FMR to register memory.
- This case is rare: it occurs only after a reconnect.
By switching this call site to ro_unmap_sync, the final use of
ro_unmap_safe is removed.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
In current kernels, waiting in xprt_release appears to be safe to
do. I had erroneously believed that for ASYNC RPCs, waiting of any
kind in xprt_release->xprt_rdma_free would result in deadlock. I've
done injection testing and consulted with Trond to confirm that
waiting in the RPC release path is safe.
For the very few times where RPC resources haven't yet been released
earlier by the reply handler, it is safe to wait synchronously in
xprt_rdma_free for invalidation rather than defering it to MR
recovery.
Note: When the QP is error state, posting a LocalInvalidate should
flush and mark the MR as bad. There is no way the remote HCA can
access that MR via a QP in error state, so it is effectively already
inaccessible and thus safe for the Upper Layer to access. The next
time the MR is used it should be recognized and cleaned up properly
by frwr_op_map.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Commit f5a73672d1 ("NFS: allow close-to-open cache semantics to
apply to root of NFS filesystem") added a call to
__nfs_revalidate_inode() to nfs_opendir to as the lookup
process wouldn't reliable do this.
Subsequent commit a3fbbde70a ("VFS: we need to set LOOKUP_JUMPED
on mountpoint crossing") make this unnecessary. So remove the
unnecessary code.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
For correct close-to-open semantics, NFS must validate
the change attribute of a directory (or file) on open.
Since commit ecf3d1f1aa ("vfs: kill FS_REVAL_DOT by adding a
d_weak_revalidate dentry op"), open() of "." or a path ending ".." is
not revalidated reliably (except when that direct is a mount point).
Prior to that commit, "." was revalidated using nfs_lookup_revalidate()
which checks the LOOKUP_OPEN flag and forces revalidation if the flag is
set.
Since that commit, nfs_weak_revalidate() is used for NFSv3 (which
ignores the flags) and nothing is used for NFSv4.
This is fixed by using nfs_lookup_verify_inode() in
nfs_weak_revalidate(). This does the revalidation exactly when needed.
Also, add a definition of .d_weak_revalidate for NFSv4.
The incorrect behavior is easily demonstrated by running "echo *" in
some non-mountpoint NFS directory while watching network traffic.
Without this patch, "echo *" sometimes doesn't produce any traffic.
With the patch it always does.
Fixes: ecf3d1f1aa ("vfs: kill FS_REVAL_DOT by adding a d_weak_revalidate dentry op")
cc: stable@vger.kernel.org (3.9+)
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
The NFS_ACCESS_* flags aren't a 1:1 mapping to the MAY_* flags, so
checking for MAY_WHATEVER might have surprising results in
nfs*_proc_access(). Let's simplify this check when determining which
bits to ask for, and do it in a generic place instead of copying code
for each NFS version.
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Passing the NFS v4 flags into the v3 code seems weird to me, even if
they are defined to the same values. This patch adds in generic flags
to help me feel better
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Here are 4 patches to resolve some char/misc driver issues found these
past weeks.
One of them is a mei bugfix and another is a new mei device id. There
is also a hyper-v fix for a reported issue, and a binder issue fix for a
problem reported by a few people.
All of these have been in my tree for a while, I don't know if
linux-next is really testing much this month. But 0-day is happy with
them :)
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-----BEGIN PGP SIGNATURE-----
iG0EABECAC0WIQT0tgzFv3jCIUoxPcsxR9QN2y37KQUCWeNCTw8cZ3JlZ0Brcm9h
aC5jb20ACgkQMUfUDdst+ykNOQCgzskhoji+7Fsl+xc3SX/3ntJwoecAnAz5xQIA
grpH4IOHq/0aPbQj8g3Y
=F3xc
-----END PGP SIGNATURE-----
Merge tag 'char-misc-4.14-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc
Pull char/misc driver fixes from Greg KH:
"Here are 4 patches to resolve some char/misc driver issues found these
past weeks.
One of them is a mei bugfix and another is a new mei device id. There
is also a hyper-v fix for a reported issue, and a binder issue fix for
a problem reported by a few people.
All of these have been in my tree for a while, I don't know if
linux-next is really testing much this month. But 0-day is happy with
them :)"
* tag 'char-misc-4.14-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc:
binder: fix use-after-free in binder_transaction()
Drivers: hv: vmbus: Fix bugs in rescind handling
mei: me: add gemini lake devices id
mei: always use domain runtime pm callbacks.
Here are a handful of USB driver fixes for 4.14-rc5.
There is the "usual" usb-serial fixes and device ids, USB gadget fixes,
and some more fixes found by the fuzz testing that is happening on the
USB layer right now.
All of these have been in my tree this week with no reported issues.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-----BEGIN PGP SIGNATURE-----
iG0EABECAC0WIQT0tgzFv3jCIUoxPcsxR9QN2y37KQUCWeNDHw8cZ3JlZ0Brcm9h
aC5jb20ACgkQMUfUDdst+yngbQCbBhbyMDZskwrjemqopL4ML1pGrAkAoKwcX2VK
lq1pOJB2GCMeWuUrFLzr
=MIeM
-----END PGP SIGNATURE-----
Merge tag 'usb-4.14-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb
Pull USB fixes from Greg KH:
"Here are a handful of USB driver fixes for 4.14-rc5.
There is the "usual" usb-serial fixes and device ids, USB gadget
fixes, and some more fixes found by the fuzz testing that is happening
on the USB layer right now.
All of these have been in my tree this week with no reported issues"
* tag 'usb-4.14-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb:
usb: usbtest: fix NULL pointer dereference
usb: gadget: configfs: Fix memory leak of interface directory data
usb: gadget: composite: Fix use-after-free in usb_composite_overwrite_options
usb: misc: usbtest: Fix overflow in usbtest_do_ioctl()
usb: renesas_usbhs: Fix DMAC sequence for receiving zero-length packet
USB: dummy-hcd: Fix deadlock caused by disconnect detection
usb: phy: tegra: Fix phy suspend for UDC
USB: serial: console: fix use-after-free after failed setup
USB: serial: console: fix use-after-free on disconnect
USB: serial: qcserial: add Dell DW5818, DW5819
USB: serial: cp210x: add support for ELV TFD500
USB: serial: cp210x: fix partnum regression
USB: serial: option: add support for TP-Link LTE module
USB: serial: ftdi_sio: add id for Cypress WICED dev board
Pull x86 fixes from Ingo Molnar:
"A landry list of fixes:
- fix reboot breakage on some PCID-enabled system
- fix crashes/hangs on some PCID-enabled systems
- fix microcode loading on certain older CPUs
- various unwinder fixes
- extend an APIC quirk to more hardware systems and disable APIC
related warning on virtualized systems
- various Hyper-V fixes
- a macro definition robustness fix
- remove jprobes IRQ disabling
- various mem-encryption fixes"
* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/microcode: Do the family check first
x86/mm: Flush more aggressively in lazy TLB mode
x86/apic: Update TSC_DEADLINE quirk with additional SKX stepping
x86/apic: Silence "FW_BUG TSC_DEADLINE disabled due to Errata" on hypervisors
x86/mm: Disable various instrumentations of mm/mem_encrypt.c and mm/tlb.c
x86/hyperv: Fix hypercalls with extended CPU ranges for TLB flushing
x86/hyperv: Don't use percpu areas for pcpu_flush/pcpu_flush_ex structures
x86/hyperv: Clear vCPU banks between calls to avoid flushing unneeded vCPUs
x86/unwind: Disable unwinder warnings on 32-bit
x86/unwind: Align stack pointer in unwinder dump
x86/unwind: Use MSB for frame pointer encoding on 32-bit
x86/unwind: Fix dereference of untrusted pointer
x86/alternatives: Fix alt_max_short macro to really be a max()
x86/mm/64: Fix reboot interaction with CR4.PCIDE
kprobes/x86: Remove IRQ disabling from jprobe handlers
kprobes/x86: Set up frame pointer in kprobe trampoline
Pull RAS fixes from Ingo Molnar:
"A boot parameter fix, plus a header export fix"
* 'ras-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/mce: Hide mca_cfg
RAS/CEC: Use the right length for "cec_disable"
Pull perf fixes from Ingo Molnar:
"Some tooling fixes plus three kernel fixes: a memory leak fix, a
statistics fix and a crash fix"
* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
perf/x86/intel/uncore: Fix memory leaks on allocation failures
perf/core: Fix cgroup time when scheduling descendants
perf/core: Avoid freeing static PMU contexts when PMU is unregistered
tools include uapi bpf.h: Sync kernel ABI header with tooling header
perf pmu: Unbreak perf record for arm/arm64 with events with explicit PMU
perf script: Add missing separator for "-F ip,brstack" (and brstackoff)
perf callchain: Compare dsos (as well) for CCKEY_FUNCTION
Pull locking fixes from Ingo Molnar:
"Two lockdep fixes for bugs introduced by the cross-release dependency
tracking feature - plus a commit that disables it because performance
regressed in an absymal fashion on some systems"
* 'locking-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
locking/lockdep: Disable cross-release features for now
locking/selftest: Avoid false BUG report
locking/lockdep: Fix stacktrace mess
Pull irq fixes from Ingo Molnar:
"A CPU hotplug related fix, plus two related sanity checks"
* 'irq-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
genirq/cpuhotplug: Enforce affinity setting on startup of managed irqs
genirq/cpuhotplug: Add sanity check for effective affinity mask
genirq: Warn when effective affinity is not updated
Pull objtool fix from Ingo Molnar:
"A single objtool fix: avoid silently broken ORC debuginfo builds and
error out instead"
* 'core-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
objtool: Upgrade libelf-devel warning to error for CONFIG_ORC_UNWINDER
On CPUs like AMD's Geode, for example, we shouldn't even try to load
microcode because they do not support the modern microcode loading
interface.
However, we do the family check *after* the other checks whether the
loader has been disabled on the command line or whether we're running in
a guest.
So move the family checks first in order to exit early if we're being
loaded on an unsupported family.
Reported-and-tested-by: Sven Glodowski <glodi1@arcor.de>
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: <stable@vger.kernel.org> # 4.11..
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://bugzilla.suse.com/show_bug.cgi?id=1061396
Link: http://lkml.kernel.org/r/20171012112316.977-1-bp@alien8.de
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Johan Hovold reported a big lockdep slowdown on his system, caused by lockdep:
> I had noticed that the BeagleBone Black boot time appeared to have
> increased significantly with 4.14 and yesterday I finally had time to
> investigate it.
>
> Boot time (from "Linux version" to login prompt) had in fact doubled
> since 4.13 where it took 17 seconds (with my current config) compared to
> the 35 seconds I now see with 4.14-rc4.
>
> I quick bisect pointed to lockdep and specifically the following commit:
>
> 28a903f63e ("locking/lockdep: Handle non(or multi)-acquisition of a crosslock")
Because the final v4.14 release is close, disable the cross-release lockdep
features for now.
Bisected-by: Johan Hovold <johan@kernel.org>
Debugged-by: Johan Hovold <johan@kernel.org>
Reported-by: Johan Hovold <johan@kernel.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Byungchul Park <byungchul.park@lge.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Tony Lindgren <tony@atomide.com>
Cc: kernel-team@lge.com
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-mm@kvack.org
Cc: linux-omap@vger.kernel.org
Link: http://lkml.kernel.org/r/20171014072659.f2yr6mhm5ha3eou7@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Pull MIPS fixes from Ralf Baechle:
"More MIPS fixes for 4.14:
- Loongson 1: Set the default number of RX and TX queues to
accomodate for recent changes of stmmac driver.
- BPF: Fix uninitialised target compiler error.
- Fix cmpxchg on 32 bit signed ints for 64 bit kernels with
!kernel_uses_llsc
- Fix generic-board-config.sh for builds using O=
- Remove pr_err() calls from fpu_emu() for a case which is not a
kernel error"
* '4.14-fixes' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus:
MIPS: math-emu: Remove pr_err() calls from fpu_emu()
MIPS: Fix generic-board-config.sh for builds using O=
MIPS: Fix cmpxchg on 32b signed ints for 64b kernel with !kernel_uses_llsc
MIPS: loongson1: set default number of rx and tx queues for stmmac
MIPS: bpf: Fix uninitialised target compiler error
Since commit:
94b1b03b51 ("x86/mm: Rework lazy TLB mode and TLB freshness tracking")
x86's lazy TLB mode has been all the way lazy: when running a kernel thread
(including the idle thread), the kernel keeps using the last user mm's
page tables without attempting to maintain user TLB coherence at all.
From a pure semantic perspective, this is fine -- kernel threads won't
attempt to access user pages, so having stale TLB entries doesn't matter.
Unfortunately, I forgot about a subtlety. By skipping TLB flushes,
we also allow any paging-structure caches that may exist on the CPU
to become incoherent. This means that we can have a
paging-structure cache entry that references a freed page table, and
the CPU is within its rights to do a speculative page walk starting
at the freed page table.
I can imagine this causing two different problems:
- A speculative page walk starting from a bogus page table could read
IO addresses. I haven't seen any reports of this causing problems.
- A speculative page walk that involves a bogus page table can install
garbage in the TLB. Such garbage would always be at a user VA, but
some AMD CPUs have logic that triggers a machine check when it notices
these bogus entries. I've seen a couple reports of this.
Boris further explains the failure mode:
> It is actually more of an optimization which assumes that paging-structure
> entries are in WB DRAM:
>
> "TlbCacheDis: cacheable memory disable. Read-write. 0=Enables
> performance optimization that assumes PML4, PDP, PDE, and PTE entries
> are in cacheable WB-DRAM; memory type checks may be bypassed, and
> addresses outside of WB-DRAM may result in undefined behavior or NB
> protocol errors. 1=Disables performance optimization and allows PML4,
> PDP, PDE and PTE entries to be in any memory type. Operating systems
> that maintain page tables in memory types other than WB- DRAM must set
> TlbCacheDis to insure proper operation."
>
> The MCE generated is an NB protocol error to signal that
>
> "Link: A specific coherent-only packet from a CPU was issued to an
> IO link. This may be caused by software which addresses page table
> structures in a memory type other than cacheable WB-DRAM without
> properly configuring MSRC001_0015[TlbCacheDis]. This may occur, for
> example, when page table structure addresses are above top of memory. In
> such cases, the NB will generate an MCE if it sees a mismatch between
> the memory operation generated by the core and the link type."
>
> I'm assuming coherent-only packets don't go out on IO links, thus the
> error.
To fix this, reinstate TLB coherence in lazy mode. With this patch
applied, we do it in one of two ways:
- If we have PCID, we simply switch back to init_mm's page tables
when we enter a kernel thread -- this seems to be quite cheap
except for the cost of serializing the CPU.
- If we don't have PCID, then we set a flag and switch to init_mm
the first time we would otherwise need to flush the TLB.
The /sys/kernel/debug/x86/tlb_use_lazy_mode debug switch can be changed
to override the default mode for benchmarking.
In theory, we could optimize this better by only flushing the TLB in
lazy CPUs when a page table is freed. Doing that would require
auditing the mm code to make sure that all page table freeing goes
through tlb_remove_page() as well as reworking some data structures
to implement the improved flush logic.
Reported-by: Markus Trippelsdorf <markus@trippelsdorf.de>
Reported-by: Adam Borowski <kilobyte@angband.pl>
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Eric Biggers <ebiggers@google.com>
Cc: Johannes Hirte <johannes.hirte@datenkhaos.de>
Cc: Kees Cook <keescook@chromium.org>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Nadav Amit <nadav.amit@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Roman Kagan <rkagan@virtuozzo.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Fixes: 94b1b03b51 ("x86/mm: Rework lazy TLB mode and TLB freshness tracking")
Link: http://lkml.kernel.org/r/20171009170231.fkpraqokz6e4zeco@pd.tnic
Signed-off-by: Ingo Molnar <mingo@kernel.org>
-----BEGIN PGP SIGNATURE-----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=yrkB
-----END PGP SIGNATURE-----
Merge tag 'drm-fixes-for-v4.14-rc5' of git://people.freedesktop.org/~airlied/linux
Pull drm fixes from Dave Airlie:
"Couple of the arm people seem to wake up so this has imx and msm
fixes, along with a bunch of i915 stable bounds fixes and an amdgpu
regression fix.
All seems pretty okay for now"
* tag 'drm-fixes-for-v4.14-rc5' of git://people.freedesktop.org/~airlied/linux:
drm/msm: fix _NO_IMPLICIT fencing case
drm/msm: fix error path cleanup
drm/msm/mdp5: Remove extra pm_runtime_put call in mdp5_crtc_cursor_set()
drm/msm/dsi: Use correct pm_runtime_put variant during host_init
drm/msm: fix return value check in _msm_gem_kernel_new()
drm/msm: use proper memory barriers for updating tail/head
drm/msm/mdp5: add missing max size for 8x74 v1
drm/amdgpu: fix placement flags in amdgpu_ttm_bind
drm/i915/bios: parse DDI ports also for CHV for HDMI DDC pin and DP AUX channel
gpu: ipu-v3: pre: implement workaround for ERR009624
gpu: ipu-v3: prg: wait for double buffers to be filled on channel startup
gpu: ipu-v3: Allow channel burst locking on i.MX6 only
drm/i915: Read timings from the correct transcoder in intel_crtc_mode_get()
drm/i915: Order two completing nop_submit_request
drm/i915: Silence compiler warning for hsw_power_well_enable()
drm/i915: Use crtc_state_is_legacy_gamma in intel_color_check
drm/i915/edp: Increase the T12 delay quirk to 1300ms
drm/i915/edp: Get the Panel Power Off timestamp after panel is off
sync_file: Return consistent status in SYNC_IOC_FILE_INFO
drm/atomic: Unref duplicated drm_atomic_state in drm_atomic_helper_resume()
drm/i915 fixes for 4.14-rc5:
Three fixes for stable:
- Use crtc_state_is_legacy_gamma in intel_color_check (Maarten)
- Read timings from the correct transcoder (Ville).
- Fix HDMI on BSW (Jani).
Other fixes:
- eDP fixes (Manasi)
- Silence compiler warnings (Chris)
- Order two completing nop_submit_request (Chris)
* tag 'drm-intel-fixes-2017-10-11' of git://anongit.freedesktop.org/drm/drm-intel:
drm/i915/bios: parse DDI ports also for CHV for HDMI DDC pin and DP AUX channel
drm/i915: Read timings from the correct transcoder in intel_crtc_mode_get()
drm/i915: Order two completing nop_submit_request
drm/i915: Silence compiler warning for hsw_power_well_enable()
drm/i915: Use crtc_state_is_legacy_gamma in intel_color_check
drm/i915/edp: Increase the T12 delay quirk to 1300ms
drm/i915/edp: Get the Panel Power Off timestamp after panel is off
bunch of msm fixes
* 'msm-fixes-4.14-rc4' of git://people.freedesktop.org/~robclark/linux:
drm/msm: fix _NO_IMPLICIT fencing case
drm/msm: fix error path cleanup
drm/msm/mdp5: Remove extra pm_runtime_put call in mdp5_crtc_cursor_set()
drm/msm/dsi: Use correct pm_runtime_put variant during host_init
drm/msm: fix return value check in _msm_gem_kernel_new()
drm/msm: use proper memory barriers for updating tail/head
drm/msm/mdp5: add missing max size for 8x74 v1
Merge misc fixes from Andrew Morton:
"18 fixes"
* emailed patches from Andrew Morton <akpm@linux-foundation.org>:
mm, swap: use page-cluster as max window of VMA based swap readahead
mm: page_vma_mapped: ensure pmd is loaded with READ_ONCE outside of lock
kmemleak: clear stale pointers from task stacks
fs/binfmt_misc.c: node could be NULL when evicting inode
fs/mpage.c: fix mpage_writepage() for pages with buffers
linux/kernel.h: add/correct kernel-doc notation
tty: fall back to N_NULL if switching to N_TTY fails during hangup
Revert "vmalloc: back off when the current task is killed"
mm/cma.c: take __GFP_NOWARN into account in cma_alloc()
scripts/kallsyms.c: ignore symbol type 'n'
userfaultfd: selftest: exercise -EEXIST only in background transfer
mm: only display online cpus of the numa node
mm: remove unnecessary WARN_ONCE in page_vma_mapped_walk().
mm/mempolicy: fix NUMA_INTERLEAVE_HIT counter
include/linux/of.h: provide of_n_{addr,size}_cells wrappers for !CONFIG_OF
mm/madvise.c: add description for MADV_WIPEONFORK and MADV_KEEPONFORK
lib/Kconfig.debug: kernel hacking menu: runtime testing: keep tests together
mm/migrate: fix indexing bug (off by one) and avoid out of bound access
When the VMA based swap readahead was introduced, a new knob
/sys/kernel/mm/swap/vma_ra_max_order
was added as the max window of VMA swap readahead. This is to make it
possible to use different max window for VMA based readahead and
original physical readahead. But Minchan Kim pointed out that this will
cause a regression because setting page-cluster sysctl to zero cannot
disable swap readahead with the change.
To fix the regression, the page-cluster sysctl is used as the max window
of both the VMA based swap readahead and original physical swap
readahead. If more fine grained control is needed in the future, more
knobs can be added as the subordinate knobs of the page-cluster sysctl.
The vma_ra_max_order knob is deleted. Because the knob was introduced
in v4.14-rc1, and this patch is targeting being merged before v4.14
releasing, there should be no existing users of this newly added ABI.
Link: http://lkml.kernel.org/r/20171011070847.16003-1-ying.huang@intel.com
Fixes: ec560175c0 ("mm, swap: VMA based swap readahead")
Signed-off-by: "Huang, Ying" <ying.huang@intel.com>
Reported-by: Minchan Kim <minchan@kernel.org>
Acked-by: Minchan Kim <minchan@kernel.org>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Shaohua Li <shli@kernel.org>
Cc: Hugh Dickins <hughd@google.com>
Cc: Fengguang Wu <fengguang.wu@intel.com>
Cc: Tim Chen <tim.c.chen@intel.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Loading the pmd without holding the pmd_lock exposes us to races with
concurrent updaters of the page tables but, worse still, it also allows
the compiler to cache the pmd value in a register and reuse it later on,
even if we've performed a READ_ONCE in between and seen a more recent
value.
In the case of page_vma_mapped_walk, this leads to the following crash
when the pmd loaded for the initial pmd_trans_huge check is all zeroes
and a subsequent valid table entry is loaded by check_pmd. We then
proceed into map_pte, but the compiler re-uses the zero entry inside
pte_offset_map, resulting in a junk pointer being installed in
pvmw->pte:
PC is at check_pte+0x20/0x170
LR is at page_vma_mapped_walk+0x2e0/0x540
[...]
Process doio (pid: 2463, stack limit = 0xffff00000f2e8000)
Call trace:
check_pte+0x20/0x170
page_vma_mapped_walk+0x2e0/0x540
page_mkclean_one+0xac/0x278
rmap_walk_file+0xf0/0x238
rmap_walk+0x64/0xa0
page_mkclean+0x90/0xa8
clear_page_dirty_for_io+0x84/0x2a8
mpage_submit_page+0x34/0x98
mpage_process_page_bufs+0x164/0x170
mpage_prepare_extent_to_map+0x134/0x2b8
ext4_writepages+0x484/0xe30
do_writepages+0x44/0xe8
__filemap_fdatawrite_range+0xbc/0x110
file_write_and_wait_range+0x48/0xd8
ext4_sync_file+0x80/0x4b8
vfs_fsync_range+0x64/0xc0
SyS_msync+0x194/0x1e8
This patch fixes the problem by ensuring that READ_ONCE is used before
the initial checks on the pmd, and this value is subsequently used when
checking whether or not the pmd is present. pmd_check is removed and
the pmd_present check is inlined directly.
Link: http://lkml.kernel.org/r/1507222630-5839-1-git-send-email-will.deacon@arm.com
Fixes: f27176cfc3 ("mm: convert page_mkclean_one() to use page_vma_mapped_walk()")
Signed-off-by: Will Deacon <will.deacon@arm.com>
Tested-by: Yury Norov <ynorov@caviumnetworks.com>
Tested-by: Richard Ruigrok <rruigrok@codeaurora.org>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
inode->i_private is assigned by a Node pointer only after registering a
new binary format, so it could be NULL if inode was created by
bm_fill_super() (or iput() was called by the error path in
bm_register_write()), and this could result in NULL pointer dereference
when evicting such an inode. e.g. mount binfmt_misc filesystem then
umount it immediately:
mount -t binfmt_misc binfmt_misc /proc/sys/fs/binfmt_misc
umount /proc/sys/fs/binfmt_misc
will result in
BUG: unable to handle kernel NULL pointer dereference at 0000000000000013
IP: bm_evict_inode+0x16/0x40 [binfmt_misc]
...
Call Trace:
evict+0xd3/0x1a0
iput+0x17d/0x1d0
dentry_unlink_inode+0xb9/0xf0
__dentry_kill+0xc7/0x170
shrink_dentry_list+0x122/0x280
shrink_dcache_parent+0x39/0x90
do_one_tree+0x12/0x40
shrink_dcache_for_umount+0x2d/0x90
generic_shutdown_super+0x1f/0x120
kill_litter_super+0x29/0x40
deactivate_locked_super+0x43/0x70
deactivate_super+0x45/0x60
cleanup_mnt+0x3f/0x70
__cleanup_mnt+0x12/0x20
task_work_run+0x86/0xa0
exit_to_usermode_loop+0x6d/0x99
syscall_return_slowpath+0xba/0xf0
entry_SYSCALL_64_fastpath+0xa3/0xa
Fix it by making sure Node (e) is not NULL.
Link: http://lkml.kernel.org/r/20171010100642.31786-1-eguan@redhat.com
Fixes: 83f918274e ("exec: binfmt_misc: shift filp_close(interp_file) from kill_node() to bm_evict_inode()")
Signed-off-by: Eryu Guan <eguan@redhat.com>
Acked-by: Oleg Nesterov <oleg@redhat.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When using FAT on a block device which supports rw_page, we can hit
BUG_ON(!PageLocked(page)) in try_to_free_buffers(). This is because we
call clean_buffers() after unlocking the page we've written. Introduce
a new clean_page_buffers() which cleans all buffers associated with a
page and call it from within bdev_write_page().
[akpm@linux-foundation.org: s/PAGE_SIZE/~0U/ per Linus and Matthew]
Link: http://lkml.kernel.org/r/20171006211541.GA7409@bombadil.infradead.org
Signed-off-by: Matthew Wilcox <mawilcox@microsoft.com>
Reported-by: Toshi Kani <toshi.kani@hpe.com>
Reported-by: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
Tested-by: Toshi Kani <toshi.kani@hpe.com>
Acked-by: Johannes Thumshirn <jthumshirn@suse.de>
Cc: Ross Zwisler <ross.zwisler@linux.intel.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Add kernel-doc notation for some macros. Correct kernel-doc comments &
typos for a few macros.
Link: http://lkml.kernel.org/r/76fa1403-1511-be4c-e9c4-456b43edfad3@infradead.org
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
We have seen NULL-pointer dereference crashes in tty->disc_data when the
N_TTY fallback driver failed to open during hangup. The immediate cause
of this open to fail has been addressed in the preceding patch to
vmalloc(), but this code could be more robust.
As Alan pointed out in commit 8a8dabf2dd ("tty: handle the case where
we cannot restore a line discipline"), the N_TTY driver, historically
the safe fallback that could never fail, can indeed fail, but the
surrounding code is not prepared to handle this. To avoid crashes he
added a new N_NULL driver to take N_TTY's place as the last resort.
Hook that fallback up to the hangup path. Update tty_ldisc_reinit() to
reflect the reality that n_tty_open can indeed fail.
Link: http://lkml.kernel.org/r/20171004185959.GC2136@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Alan Cox <alan@llwyncelyn.cymru>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Michal Hocko <mhocko@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This reverts commits 5d17a73a2e ("vmalloc: back off when the current
task is killed") and 171012f561 ("mm: don't warn when vmalloc() fails
due to a fatal signal").
Commit 5d17a73a2e ("vmalloc: back off when the current task is
killed") made all vmalloc allocations from a signal-killed task fail.
We have seen crashes in the tty driver from this, where a killed task
exiting tries to switch back to N_TTY, fails n_tty_open because of the
vmalloc failing, and later crashes when dereferencing tty->disc_data.
Arguably, relying on a vmalloc() call to succeed in order to properly
exit a task is not the most robust way of doing things. There will be a
follow-up patch to the tty code to fall back to the N_NULL ldisc.
But the justification to make that vmalloc() call fail like this isn't
convincing, either. The patch mentions an OOM victim exhausting the
memory reserves and thus deadlocking the machine. But the OOM killer is
only one, improbable source of fatal signals. It doesn't make sense to
fail allocations preemptively with plenty of memory in most cases.
The patch doesn't mention real-life instances where vmalloc sites would
exhaust memory, which makes it sound more like a theoretical issue to
begin with. But just in case, the OOM access to memory reserves has
been restricted on the allocator side in cd04ae1e2d ("mm, oom: do not
rely on TIF_MEMDIE for memory reserves access"), which should take care
of any theoretical concerns on that front.
Revert this patch, and the follow-up that suppresses the allocation
warnings when we fail the allocations due to a signal.
Link: http://lkml.kernel.org/r/20171004185906.GB2136@cmpxchg.org
Fixes: 171012f561 ("mm: don't warn when vmalloc() fails due to a fatal signal")
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: Alan Cox <alan@llwyncelyn.cymru>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
cma_alloc() unconditionally prints an INFO message when the CMA
allocation fails. Make this message conditional on the non-presence of
__GFP_NOWARN in gfp_mask.
This patch aims at removing INFO messages that are displayed when the
VC4 driver tries to allocate buffer objects. From the driver
perspective an allocation failure is acceptable, and the driver can
possibly do something to make following allocation succeed (like
flushing the VC4 internal cache).
Link: http://lkml.kernel.org/r/20171004125447.15195-1-boris.brezillon@free-electrons.com
Signed-off-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Acked-by: Laura Abbott <labbott@redhat.com>
Cc: Jaewon Kim <jaewon31.kim@samsung.com>
Cc: David Airlie <airlied@linux.ie>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Eric Anholt <eric@anholt.net>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
gcc on aarch64 may emit synbols of type 'n' if the kernel is built with
'-frecord-gcc-switches'. In most cases, those symbols are reported with
nm as
000000000000000e n $d
and with objdump as
0000000000000000 l d .GCC.command.line 0000000000000000 .GCC.command.line
000000000000000e l .GCC.command.line 0000000000000000 $d
Those symbols are detected in is_arm_mapping_symbol() and ignored.
However, if "--prefix-symbols=<prefix>" is configured as well, the
situation is different. For example, in efi/libstub, arm64 images are
built with
'--prefix-alloc-sections=.init --prefix-symbols=__efistub_'.
In combination with '-frecord-gcc-switches', the symbols are now reported
by nm as:
000000000000000e n __efistub_$d
and by objdump as:
0000000000000000 l d .GCC.command.line 0000000000000000 .GCC.command.line
000000000000000e l .GCC.command.line 0000000000000000 __efistub_$d
Those symbols are no longer ignored and included in the base address
calculation. This results in a base address of 000000000000000e, which
in turn causes kallsyms to abort with
kallsyms failure:
relative symbol value 0xffffff900800a000 out of range in relative mode
The problem is seen in little endian arm64 builds with CONFIG_EFI
enabled and with '-frecord-gcc-switches' set in KCFLAGS.
Explicitly ignore symbols of type 'n' since those are clearly debug
symbols.
Link: http://lkml.kernel.org/r/1507136063-3139-1-git-send-email-linux@roeck-us.net
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
