In a cluster failover scenario, it is desirable for the client to
attempt to reconnect quickly, as an alternate NFS server is already
waiting to take over for the down server. The client can't see that
a server IP address has moved to a new server until the existing
connection is gone.
For fabrics and devices where it is meaningful, set a definite upper
bound on the amount of time before it is determined that a
connection is no longer valid. This allows the RPC client to detect
connection loss in a timely matter, then perform a fresh resolution
of the server GUID in case it has changed (cluster failover).
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Tested-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
There needs to be a safe method of releasing registered memory
resources when an RPC terminates. Safe can mean a number of things:
+ Doesn't have to sleep
+ Doesn't rely on having a QP in RTS
ro_unmap_safe will be that safe method. It can be used in cases
where synchronous memory invalidation can deadlock, or needs to have
an active QP.
The important case is fencing an RPC's memory regions after it is
signaled (^C) and before it exits. If this is not done, there is a
window where the server can write an RPC reply into memory that the
client has released and re-used for some other purpose.
Note that this is a full solution for FRWR, but FMR and physical
still have some gaps where a particularly bad server can wreak
some havoc on the client. These gaps are not made worse by this
patch and are expected to be exceptionally rare and timing-based.
They are noted in documenting comments.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Tested-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Separate the DMA unmap operation from freeing the MW. In a
subsequent patch they will not always be done at the same time,
and they are not related operations (except by order; freeing
the MW must be the last step during invalidation).
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Tested-by: Steve Wise <swise@opengridcomputing.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
In a subsequent patch, the fr_xprt and fr_worker fields will be
needed by another memory registration mode. Move them into the
generic rpcrdma_mw structure that wraps struct rpcrdma_frmr.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Tested-by: Steve Wise <swise@opengridcomputing.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
frwr_op_unmap_sync() is now invoked in a workqueue context, the same
as __frwr_queue_recovery(). There's no need to defer MR reset if
posting LOCAL_INV MRs fails.
This means that even when ib_post_send() fails (which should occur
very rarely) the invalidation and DMA unmapping steps are still done
in the correct order.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Tested-by: Steve Wise <swise@opengridcomputing.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Move the the I/O direction field from rpcrdma_mr_seg into the
rpcrdma_frmr.
This makes it possible to DMA-unmap the frwr long after an RPC has
exited and its rpcrdma_mr_seg array has been released and re-used.
This might occur if an RPC times out while waiting for a new
connection to be established.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Tested-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
rpcrdma_marshal_req() makes a simplifying assumption: that NFS
operations with large Call messages have small Reply messages, and
vice versa. Therefore with RPC-over-RDMA, only one chunk type is
ever needed for each Call/Reply pair, because one direction needs
chunks, the other direction will always fit inline.
In fact, this assumption is asserted in the code:
if (rtype != rpcrdma_noch && wtype != rpcrdma_noch) {
dprintk("RPC: %s: cannot marshal multiple chunk lists\n",
__func__);
return -EIO;
}
But RPCGSS_SEC breaks this assumption. Because krb5i and krb5p
perform data transformation on RPC messages before they are
transmitted, direct data placement techniques cannot be used, thus
RPC messages must be sent via a Long call in both directions.
All such calls are sent with a Position Zero Read chunk, and all
such replies are handled with a Reply chunk. Thus the client must
provide every Call/Reply pair with both a Read list and a Reply
chunk.
Without any special security in effect, NFSv4 WRITEs may now also
use the Read list and provide a Reply chunk. The marshal_req
logic was preventing that, meaning an NFSv4 WRITE with a large
payload that included a GETATTR result larger than the inline
threshold would fail.
The code that encodes each chunk list is now completely contained in
its own function. There is some code duplication, but the trade-off
is that the overall logic should be more clear.
Note that all three chunk lists now share the rl_segments array.
Some additional per-req accounting is necessary to track this
usage. For the same reasons that the above simplifying assumption
has held true for so long, I don't expect more array elements are
needed at this time.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Tested-by: Steve Wise <swise@opengridcomputing.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Avoid the latency and interrupt overhead of registering a Write
chunk when handling NFS READ requests of a few hundred bytes or
less.
This change does not interoperate with Linux NFS/RDMA servers
that do not have commit 9d11b51ce7 ('svcrdma: Fix send_reply()
scatter/gather set-up'). Commit 9d11b51ce7 was introduced in v4.3,
and is included in 4.2.y, 4.1.y, and 3.18.y.
Oracle bug 22925946 has been filed to request that the above fix
be included in the Oracle Linux UEK4 NFS/RDMA server.
Red Hat bugzillas 1327280 and 1327554 have been filed to request
that RHEL NFS/RDMA server backports include the above fix.
Workaround: Replace the "proto=rdma,port=20049" mount options
with "proto=tcp" until commit 9d11b51ce7 is applied to your
NFS server.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Tested-by: Steve Wise <swise@opengridcomputing.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
When deciding whether to send a Call inline, rpcrdma_marshal_req
doesn't take into account header bytes consumed by chunk lists.
This results in Call messages on the wire that are sometimes larger
than the inline threshold.
Likewise, when a Write list or Reply chunk is in play, the server's
reply has to emit an RDMA Send that includes a larger-than-minimal
RPC-over-RDMA header.
The actual size of a Call message cannot be estimated until after
the chunk lists have been registered. Thus the size of each
RPC-over-RDMA header can be estimated only after chunks are
registered; but the decision to register chunks is based on the size
of that header. Chicken, meet egg.
The best a client can do is estimate header size based on the
largest header that might occur, and then ensure that inline content
is always smaller than that.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Tested-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Send buffer space is shared between the RPC-over-RDMA header and
an RPC message. A large RPC-over-RDMA header means less space is
available for the associated RPC message, which then has to be
moved via an RDMA Read or Write.
As more segments are added to the chunk lists, the header increases
in size. Typical modern hardware needs only a few segments to
convey the maximum payload size, but some devices and registration
modes may need a lot of segments to convey data payload. Sometimes
so many are needed that the remaining space in the Send buffer is
not enough for the RPC message. Sending such a message usually
fails.
To ensure a transport can always make forward progress, cap the
number of RDMA segments that are allowed in chunk lists. This
prevents less-capable devices and memory registrations from
consuming a large portion of the Send buffer by reducing the
maximum data payload that can be conveyed with such devices.
For now I choose an arbitrary maximum of 8 RDMA segments. This
allows a maximum size RPC-over-RDMA header to fit nicely in the
current 1024 byte inline threshold with over 700 bytes remaining
for an inline RPC message.
The current maximum data payload of NFS READ or WRITE requests is
one megabyte. To convey that payload on a client with 4KB pages,
each chunk segment would need to handle 32 or more data pages. This
is well within the capabilities of FMR. For physical registration,
the maximum payload size on platforms with 4KB pages is reduced to
32KB.
For FRWR, a device's maximum page list depth would need to be at
least 34 to support the maximum 1MB payload. A device with a smaller
maximum page list depth means the maximum data payload is reduced
when using that device.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Tested-by: Steve Wise <swise@opengridcomputing.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Currently the sysctls that allow setting the inline threshold allow
any value to be set.
Small values only make the transport run slower. The default 1KB
setting is as low as is reasonable. And the logic that decides how
to divide a Send buffer between RPC-over-RDMA header and RPC message
assumes (but does not check) that the lower bound is not crazy (say,
57 bytes).
Send and receive buffers share a page with some control information.
Values larger than about 3KB can't be supported, currently.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Tested-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
RPC-over-RDMA transports have a limit on how large a backward
direction (backchannel) RPC message can be. Ensure that the NFSv4.x
CREATE_SESSION operation advertises this limit to servers.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Tested-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Commit 176e21ee2e ("SUNRPC: Support for RPC over AF_LOCAL
transports") added a 5-character netid, but did not bump
RPCBIND_MAXNETIDLEN from 4 to 5.
Fixes: 176e21ee2e ("SUNRPC: Support for RPC over AF_LOCAL ...")
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
RFC 5666: The "rdma" netid is to be used when IPv4 addressing
is employed by the underlying transport, and "rdma6" for IPv6
addressing.
Add mount -o proto=rdma6 option to support NFS/RDMA IPv6 addressing.
Changes from v2:
- Integrated comments from Chuck Level, Anna Schumaker, Trodt Myklebust
- Add a little more to the patch description to describe NFS/RDMA
IPv6 suggested by Chuck Level and Anna Schumaker
- Removed duplicated rdma6 define
- Remove Opt_xprt_rdma mountfamily since it doesn't support
Signed-off-by: Shirley Ma <shirley.ma@oracle.com>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
OPEN_CREATE with EXCLUSIVE4_1 sends initial file permission.
Ignoring fact, that server have indicated that file mod is set, client
will send yet another SETATTR request, but, as mode is already set,
new SETATTR will be empty. This is not a problem, nevertheless
an extra roundtrip and slow open on high latency networks.
This change is aims to skip extra setattr after open if there are
no attributes to be set.
Signed-off-by: Tigran Mkrtchyan <tigran.mkrtchyan@desy.de>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
This adds the copy_range file_ops function pointer used by the
sys_copy_range() function call. This patch only implements sync copies,
so if an async copy happens we decode the stateid and ignore it.
Signed-off-by: Anna Schumaker <bjschuma@netapp.com>
Copy will use this to set up a commit request for a generic range. I
don't want to allocate a new pagecache entry for the file, so I needed
to change parts of the commit path to handle requests with a null
wb_page.
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
* pci/arm64:
PCI, of: Move PCI I/O space management to PCI core code
PCI: generic, thunder: Use generic ECAM API
PCI: Provide common functions for ECAM mapping
* pci/host-hv:
PCI: hv: Add explicit barriers to config space access
* pci/hotplug:
PCI: Use cached copy of PCI_EXP_SLTCAP_HPC bit
* pci/resource:
PCI: Disable all BAR sizing for devices with non-compliant BARs
x86/PCI: Mark Broadwell-EP Home Agent 1 as having non-compliant BARs
PCI: Identify Enhanced Allocation (EA) BAR Equivalent resources in sysfs
Wrong return code was being returned on SMB3 rmdir of
non-empty directory.
For SMB3 (unlike for cifs), we attempt to delete a directory by
set of delete on close flag on the open. Windows clients set
this flag via a set info (SET_FILE_DISPOSITION to set this flag)
which properly checks if the directory is empty.
With this patch on smb3 mounts we correctly return
"DIRECTORY NOT EMPTY"
on attempts to remove a non-empty directory.
Signed-off-by: Steve French <steve.french@primarydata.com>
CC: Stable <stable@vger.kernel.org>
Acked-by: Sachin Prabhu <sprabhu@redhat.com>
See [MS-NLMP] 3.2.5.1.2 Server Receives an AUTHENTICATE_MESSAGE from the Client:
...
Set NullSession to FALSE
If (AUTHENTICATE_MESSAGE.UserNameLen == 0 AND
AUTHENTICATE_MESSAGE.NtChallengeResponse.Length == 0 AND
(AUTHENTICATE_MESSAGE.LmChallengeResponse == Z(1)
OR
AUTHENTICATE_MESSAGE.LmChallengeResponse.Length == 0))
-- Special case: client requested anonymous authentication
Set NullSession to TRUE
...
Only server which map unknown users to guest will allow
access using a non-null NTChallengeResponse.
For Samba it's the "map to guest = bad user" option.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=11913
CC: Stable <stable@vger.kernel.org>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Steve French <smfrench@gmail.com>
We currently do not check if any delimiter exists before the prefix
path in cifs_compose_mount_options(). Consequently when building the
devname using cifs_build_devname() we can end up with multiple
delimiters separating the UNC and the prefix path.
An issue was reported by the customer mounting a folder within a DFS
share from a Netapp server which uses McAfee antivirus. We have
narrowed down the cause to the use of double backslashes in the file
name used to open the file. This was determined to be caused because of
additional delimiters as a result of the bug.
In addition to changes in cifs_build_devname(), we also fix
cifs_parse_devname() to ignore any preceding delimiter for the prefix
path.
The problem was originally reported on RHEL 6 in RHEL bz 1252721. This
is the upstream version of the fix. The fix was confirmed by looking at
the packet capture of a DFS mount.
Signed-off-by: Sachin Prabhu <sprabhu@redhat.com>
Signed-off-by: Steve French <smfrench@gmail.com>
CIFS may be used as lower layer of overlayfs and accessing f_path.dentry can
lead to a crash.
Fix by replacing direct access of file->f_path.dentry with the
file_dentry() accessor, which will always return a native object.
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
Acked-by: Shirish Pargaonkar <shirishpargaonkar@gmail.com>
Signed-off-by: Steve French <smfrench@gmail.com>
The current logic around the 'use_dac' module parameter prevents the
r81969 driver from being loadable on 64-bit systems without any RAM
below 4 GB when the parameter is left at its default value.
So introduce a new default value -1 which indicates that 64-bit DMA
should be enabled on sufficiently recent PCIe chips, i.e., versions
RTL_GIGA_MAC_VER_18 or later. Explicit param values of 0 or 1 retain
the existing behavior of unconditionally enabling/disabling 64-bit DMA
on 64-bit architectures (i.e., regardless of the type and version of the
chip)
Since PCIe chips do not need to CPlusCmd Dual Address Cycle to be set,
make that conditional on the device type as well.
Cc: Realtek linux nic maintainers <nic_swsd@realtek.com>
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
If you compile without OF_MDIO support in an RGMII configuration, we fail
to configure the dp83867 phy today by writing garbage into its configuration
registers.
On the other hand if you do compile with OF_MDIO and the phy gets loaded via
device tree, you have to have the properties set in the device tree, otherwise
we fail to load the driver and don't even attach the generic phy driver to
the interface anymore.
To make things slightly more consistent, make the rgmii configuration properties
optional and allow a user to omit them in their device tree.
Signed-off-by: Alexander Graf <agraf@suse.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
When CONFIG_OF_MDIO is configured as module, the #define for it really
is CONFIG_OF_MDIO_MODULE, not CONFIG_OF_MDIO. So if we are compiling it
as module, the dp83867 doesn't see that OF_MDIO was selected and doesn't
read the dt rgmii parameters.
The fix is simple: Use IS_ENABLED(). It checks for both - module as well
as compiled in code.
Signed-off-by: Alexander Graf <agraf@suse.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Merge tag 'qcom-soc-for-4.7-2' into net-next
This merges the Qualcomm SOC tree with the net-next, solving the
merge conflict in the SMD API between the two.
In the current implementation of ARM64 eBPF JIT, R23 and R24 are used for
tmp registers, which are callee-saved registers. This leads to variable size
of JIT prologue and epilogue. The latest blinding constant change prefers to
constant size of prologue and epilogue. AAPCS reserves R9 ~ R15 for temp
registers which not need to be saved/restored during function call. So, replace
R23 and R24 to R10 and R11, and remove tmp_used flag to save 2 instructions for
some jited BPF program.
CC: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Zi Shen Lim <zlim.lnx@gmail.com>
Signed-off-by: Yang Shi <yang.shi@linaro.org>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
In testing with HiKey, we found that since
commit 3f30b158eb ("asix: On RX avoid creating bad Ethernet
frames"),
we're seeing lots of noise during network transfers:
[ 239.027993] asix 1-1.1:1.0 eth0: asix_rx_fixup() Data Header synchronisation was lost, remaining 988
[ 239.037310] asix 1-1.1:1.0 eth0: asix_rx_fixup() Bad Header Length 0x54ebb5ec, offset 4
[ 239.045519] asix 1-1.1:1.0 eth0: asix_rx_fixup() Bad Header Length 0xcdffe7a2, offset 4
[ 239.275044] asix 1-1.1:1.0 eth0: asix_rx_fixup() Data Header synchronisation was lost, remaining 988
[ 239.284355] asix 1-1.1:1.0 eth0: asix_rx_fixup() Bad Header Length 0x1d36f59d, offset 4
[ 239.292541] asix 1-1.1:1.0 eth0: asix_rx_fixup() Bad Header Length 0xaef3c1e9, offset 4
[ 239.518996] asix 1-1.1:1.0 eth0: asix_rx_fixup() Data Header synchronisation was lost, remaining 988
[ 239.528300] asix 1-1.1:1.0 eth0: asix_rx_fixup() Bad Header Length 0x2881912, offset 4
[ 239.536413] asix 1-1.1:1.0 eth0: asix_rx_fixup() Bad Header Length 0x5638f7e2, offset 4
And network throughput ends up being pretty bursty and slow with
a overall throughput of at best ~30kB/s (where as previously we
got 1.1MB/s with the slower USB1.1 "full speed" host).
We found the issue also was reproducible on a x86_64 system,
using a "high-speed" USB2.0 port but the throughput did not
measurably drop (possibly due to the scp transfer being cpu
bound on my slow test hardware).
After lots of debugging, I found the check added in the
problematic commit seems to be calculating the offset
incorrectly.
In the normal case, in the main loop of the function, we do:
(where offset is zero, or set to "offset += (copy_length + 1) &
0xfffe" in the previous loop)
rx->header = get_unaligned_le32(skb->data +
offset);
offset += sizeof(u32);
But the problematic patch calculates:
offset = ((rx->remaining + 1) & 0xfffe) + sizeof(u32);
rx->header = get_unaligned_le32(skb->data + offset);
Adding some debug logic to check those offset calculation used
to find rx->header, the one in problematic code is always too
large by sizeof(u32).
Thus, this patch removes the incorrect " + sizeof(u32)" addition
in the problematic calculation, and resolves the issue.
Cc: Dean Jenkins <Dean_Jenkins@mentor.com>
Cc: "David B. Robins" <linux@davidrobins.net>
Cc: Mark Craske <Mark_Craske@mentor.com>
Cc: Emil Goode <emilgoode@gmail.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: YongQin Liu <yongqin.liu@linaro.org>
Cc: Guodong Xu <guodong.xu@linaro.org>
Cc: Ivan Vecera <ivecera@redhat.com>
Cc: linux-usb@vger.kernel.org
Cc: netdev@vger.kernel.org
Cc: stable <stable@vger.kernel.org> #4.4+
Reported-by: Yongqin Liu <yongqin.liu@linaro.org>
Signed-off-by: John Stultz <john.stultz@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pull parallel filesystem directory handling update from Al Viro.
This is the main parallel directory work by Al that makes the vfs layer
able to do lookup and readdir in parallel within a single directory.
That's a big change, since this used to be all protected by the
directory inode mutex.
The inode mutex is replaced by an rwsem, and serialization of lookups of
a single name is done by a "in-progress" dentry marker.
The series begins with xattr cleanups, and then ends with switching
filesystems over to actually doing the readdir in parallel (switching to
the "iterate_shared()" that only takes the read lock).
A more detailed explanation of the process from Al Viro:
"The xattr work starts with some acl fixes, then switches ->getxattr to
passing inode and dentry separately. This is the point where the
things start to get tricky - that got merged into the very beginning
of the -rc3-based #work.lookups, to allow untangling the
security_d_instantiate() mess. The xattr work itself proceeds to
switch a lot of filesystems to generic_...xattr(); no complications
there.
After that initial xattr work, the series then does the following:
- untangle security_d_instantiate()
- convert a bunch of open-coded lookup_one_len_unlocked() to calls of
that thing; one such place (in overlayfs) actually yields a trivial
conflict with overlayfs fixes later in the cycle - overlayfs ended
up switching to a variant of lookup_one_len_unlocked() sans the
permission checks. I would've dropped that commit (it gets
overridden on merge from #ovl-fixes in #for-next; proper resolution
is to use the variant in mainline fs/overlayfs/super.c), but I
didn't want to rebase the damn thing - it was fairly late in the
cycle...
- some filesystems had managed to depend on lookup/lookup exclusion
for *fs-internal* data structures in a way that would break if we
relaxed the VFS exclusion. Fixing hadn't been hard, fortunately.
- core of that series - parallel lookup machinery, replacing
->i_mutex with rwsem, making lookup_slow() take it only shared. At
that point lookups happen in parallel; lookups on the same name
wait for the in-progress one to be done with that dentry.
Surprisingly little code, at that - almost all of it is in
fs/dcache.c, with fs/namei.c changes limited to lookup_slow() -
making it use the new primitive and actually switching to locking
shared.
- parallel readdir stuff - first of all, we provide the exclusion on
per-struct file basis, same as we do for read() vs lseek() for
regular files. That takes care of most of the needed exclusion in
readdir/readdir; however, these guys are trickier than lookups, so
I went for switching them one-by-one. To do that, a new method
'->iterate_shared()' is added and filesystems are switched to it
as they are either confirmed to be OK with shared lock on directory
or fixed to be OK with that. I hope to kill the original method
come next cycle (almost all in-tree filesystems are switched
already), but it's still not quite finished.
- several filesystems get switched to parallel readdir. The
interesting part here is dealing with dcache preseeding by readdir;
that needs minor adjustment to be safe with directory locked only
shared.
Most of the filesystems doing that got switched to in those
commits. Important exception: NFS. Turns out that NFS folks, with
their, er, insistence on VFS getting the fuck out of the way of the
Smart Filesystem Code That Knows How And What To Lock(tm) have
grown the locking of their own. They had their own homegrown
rwsem, with lookup/readdir/atomic_open being *writers* (sillyunlink
is the reader there). Of course, with VFS getting the fuck out of
the way, as requested, the actual smarts of the smart filesystem
code etc. had become exposed...
- do_last/lookup_open/atomic_open cleanups. As the result, open()
without O_CREAT locks the directory only shared. Including the
->atomic_open() case. Backmerge from #for-linus in the middle of
that - atomic_open() fix got brought in.
- then comes NFS switch to saner (VFS-based ;-) locking, killing the
homegrown "lookup and readdir are writers" kinda-sorta rwsem. All
exclusion for sillyunlink/lookup is done by the parallel lookups
mechanism. Exclusion between sillyunlink and rmdir is a real rwsem
now - rmdir being the writer.
Result: NFS lookups/readdirs/O_CREAT-less opens happen in parallel
now.
- the rest of the series consists of switching a lot of filesystems
to parallel readdir; in a lot of cases ->llseek() gets simplified
as well. One backmerge in there (again, #for-linus - rockridge
fix)"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (74 commits)
ext4: switch to ->iterate_shared()
hfs: switch to ->iterate_shared()
hfsplus: switch to ->iterate_shared()
hostfs: switch to ->iterate_shared()
hpfs: switch to ->iterate_shared()
hpfs: handle allocation failures in hpfs_add_pos()
gfs2: switch to ->iterate_shared()
f2fs: switch to ->iterate_shared()
afs: switch to ->iterate_shared()
befs: switch to ->iterate_shared()
befs: constify stuff a bit
isofs: switch to ->iterate_shared()
get_acorn_filename(): deobfuscate a bit
btrfs: switch to ->iterate_shared()
logfs: no need to lock directory in lseek
switch ecryptfs to ->iterate_shared
9p: switch to ->iterate_shared()
fat: switch to ->iterate_shared()
romfs, squashfs: switch to ->iterate_shared()
more trivial ->iterate_shared conversions
...
The problem is that fib_info->nh is [0] so the struct fib_info
allocation size depends on number of nexthops. If we just copy fib_info,
we do not copy the nexthops info and driver accesses memory which is not
ours.
Given the fact that fib4 does not defer operations and therefore it does
not need copy, just pass the pointer down to drivers as it was done
before.
Fixes: 850d0cbc91 ("switchdev: remove pointers from switchdev objects")
Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pull irq updates from Thomas Gleixner:
"This update delivers:
- Yet another interrupt chip diver (LPC32xx)
- Core functions to handle partitioned per-cpu interrupts
- Enhancements to the IPI core
- Proper handling of irq type configuration
- A large set of ARM GIC enhancements
- The usual pile of small fixes, cleanups and enhancements"
* 'irq-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (31 commits)
irqchip/bcm2836: Use a more generic memory barrier call
irqchip/bcm2836: Fix compiler warning on 64-bit build
irqchip/bcm2836: Drop smp_set_ops on arm64 builds
irqchip/gic: Add helper functions for GIC setup and teardown
irqchip/gic: Store GIC configuration parameters
irqchip/gic: Pass GIC pointer to save/restore functions
irqchip/gic: Return an error if GIC initialisation fails
irqchip/gic: Remove static irq_chip definition for eoimode1
irqchip/gic: Don't initialise chip if mapping IO space fails
irqchip/gic: WARN if setting the interrupt type for a PPI fails
irqchip/gic: Don't unnecessarily write the IRQ configuration
irqchip: Mask the non-type/sense bits when translating an IRQ
genirq: Ensure IRQ descriptor is valid when setting-up the IRQ
irqchip/gic-v3: Configure all interrupts as non-secure Group-1
irqchip/gic-v2m: Add workaround for Broadcom NS2 GICv2m erratum
irqchip/irq-alpine-msi: Don't use <asm-generic/msi.h>
irqchip/mbigen: Checking for IS_ERR() instead of NULL
irqchip/gic-v3: Remove inexistant register definition
irqchip/gicv3-its: Don't allow devices whose ID is outside range
irqchip: Add LPC32xx interrupt controller driver
...
Cut down on noise for mainstream users of the API and people doing build
testing by dropping the deprecated flag from regulator_can_change_voltage()
as it triggers even on the EXPORT_SYMBOL_GPL() which affects all builds
rather than just the remaining drivers with calls to it (for which fixes
are currently pending).
The function remains deprecated and is expected to be removed entirely
in v4.8.
Signed-off-by: Mark Brown <broonie@kernel.org>
Pull timer updates from Thomas Gleixner:
"A rather small set of patches from the timer departement:
- Some more y2038 work
- Yet another new clocksource driver
- The usual set of small fixes, cleanups and enhancements"
* 'timers-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
clocksource/drivers/tegra: Remove unused suspend/resume code
clockevents/driversi/mps2: add MPS2 Timer driver
dt-bindings: document the MPS2 timer bindings
clocksource/drivers/mtk_timer: Add __init attribute
clockevents/drivers/dw_apb_timer: Implement ->set_state_oneshot_stopped()
time: Introduce do_sys_settimeofday64()
security: Introduce security_settime64()
clocksource: Add missing include of of.h.
Pull tracing ring-buffer fixes from Steven Rostedt:
"Hao Qin reported an integer overflow possibility with signed and
unsigned numbers in the ring-buffer code.
https://bugzilla.kernel.org/show_bug.cgi?id=118001
At first I did not think this was too much of an issue, because the
overflow would be caught later when either too much data was allocated
or it would trigger RB_WARN_ON() which shuts down the ring buffer.
But looking closer into it, I found that the right settings could
bypass the checks and crash the kernel. Luckily, this is only
accessible by root.
The first fix is to convert all the variables into long, such that we
don't get into issues between 32 bit variables being assigned 64 bit
ones. This fixes the RB_WARN_ON() triggering.
The next fix is to get rid of a duplicate DIV_ROUND_UP() that when
called twice with the right value, can cause a kernel crash.
The first DIV_ROUND_UP() is to normalize the input and it is checked
against the minimum allowable value. But then DIV_ROUND_UP() is
called again, which can overflow due to the (a + b - 1)/b, logic. The
first called upped the value, the second can overflow (with the +b
part).
The second call to DIV_ROUND_UP() came in via a second change a while
ago and the code is cleaned up to remove it"
* tag 'trace-fixes-v4.6-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
ring-buffer: Prevent overflow of size in ring_buffer_resize()
ring-buffer: Use long for nr_pages to avoid overflow failures
We saw the following extra refcount release on veth device:
kernel: [7957821.463992] unregister_netdevice: waiting for mesos50284 to become free. Usage count = -1
Since we heavily use mirred action to redirect packets to veth, I think
this is caused by the following race condition:
CPU0:
tcf_mirred_release(): (in RCU callback)
struct net_device *dev = rcu_dereference_protected(m->tcfm_dev, 1);
CPU1:
mirred_device_event():
spin_lock_bh(&mirred_list_lock);
list_for_each_entry(m, &mirred_list, tcfm_list) {
if (rcu_access_pointer(m->tcfm_dev) == dev) {
dev_put(dev);
/* Note : no rcu grace period necessary, as
* net_device are already rcu protected.
*/
RCU_INIT_POINTER(m->tcfm_dev, NULL);
}
}
spin_unlock_bh(&mirred_list_lock);
CPU0:
tcf_mirred_release():
spin_lock_bh(&mirred_list_lock);
list_del(&m->tcfm_list);
spin_unlock_bh(&mirred_list_lock);
if (dev) // <======== Stil refers to the old m->tcfm_dev
dev_put(dev); // <======== dev_put() is called on it again
The action init code path is good because it is impossible to modify
an action that is being removed.
So, fix this by moving everything under the spinlock.
Fixes: 2ee22a90c7 ("net_sched: act_mirred: remove spinlock in fast path")
Fixes: 6bd00b8506 ("act_mirred: fix a race condition on mirred_list")
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The publication field of the old netlink API should contain the
publication key and not the publication reference.
Fixes: 44a8ae94fd (tipc: convert legacy nl name table dump to nl compat)
Signed-off-by: Richard Alpe <richard.alpe@ericsson.com>
Acked-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
