Add plumbing to allow DSA drivers to register parameters with devlink.
To keep with the abstraction, the DSA drivers pass the ds structure to
these helpers, and the DSA core then translates that to the devlink
structure associated to the device.
Signed-off-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
One place in the driver was left where the open-coded functionality
hasn't been replaced with helper rtl_hw_aspm_clkreq_enable yet.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
sk_page_frag() optimizes skb_frag allocations by using per-task
skb_frag cache when it knows it's the only user. The condition is
determined by seeing whether the socket allocation mask allows
blocking - if the allocation may block, it obviously owns the task's
context and ergo exclusively owns current->task_frag.
Unfortunately, this misses recursion through memory reclaim path.
Please take a look at the following backtrace.
[2] RIP: 0010:tcp_sendmsg_locked+0xccf/0xe10
...
tcp_sendmsg+0x27/0x40
sock_sendmsg+0x30/0x40
sock_xmit.isra.24+0xa1/0x170 [nbd]
nbd_send_cmd+0x1d2/0x690 [nbd]
nbd_queue_rq+0x1b5/0x3b0 [nbd]
__blk_mq_try_issue_directly+0x108/0x1b0
blk_mq_request_issue_directly+0xbd/0xe0
blk_mq_try_issue_list_directly+0x41/0xb0
blk_mq_sched_insert_requests+0xa2/0xe0
blk_mq_flush_plug_list+0x205/0x2a0
blk_flush_plug_list+0xc3/0xf0
[1] blk_finish_plug+0x21/0x2e
_xfs_buf_ioapply+0x313/0x460
__xfs_buf_submit+0x67/0x220
xfs_buf_read_map+0x113/0x1a0
xfs_trans_read_buf_map+0xbf/0x330
xfs_btree_read_buf_block.constprop.42+0x95/0xd0
xfs_btree_lookup_get_block+0x95/0x170
xfs_btree_lookup+0xcc/0x470
xfs_bmap_del_extent_real+0x254/0x9a0
__xfs_bunmapi+0x45c/0xab0
xfs_bunmapi+0x15/0x30
xfs_itruncate_extents_flags+0xca/0x250
xfs_free_eofblocks+0x181/0x1e0
xfs_fs_destroy_inode+0xa8/0x1b0
destroy_inode+0x38/0x70
dispose_list+0x35/0x50
prune_icache_sb+0x52/0x70
super_cache_scan+0x120/0x1a0
do_shrink_slab+0x120/0x290
shrink_slab+0x216/0x2b0
shrink_node+0x1b6/0x4a0
do_try_to_free_pages+0xc6/0x370
try_to_free_mem_cgroup_pages+0xe3/0x1e0
try_charge+0x29e/0x790
mem_cgroup_charge_skmem+0x6a/0x100
__sk_mem_raise_allocated+0x18e/0x390
__sk_mem_schedule+0x2a/0x40
[0] tcp_sendmsg_locked+0x8eb/0xe10
tcp_sendmsg+0x27/0x40
sock_sendmsg+0x30/0x40
___sys_sendmsg+0x26d/0x2b0
__sys_sendmsg+0x57/0xa0
do_syscall_64+0x42/0x100
entry_SYSCALL_64_after_hwframe+0x44/0xa9
In [0], tcp_send_msg_locked() was using current->page_frag when it
called sk_wmem_schedule(). It already calculated how many bytes can
be fit into current->page_frag. Due to memory pressure,
sk_wmem_schedule() called into memory reclaim path which called into
xfs and then IO issue path. Because the filesystem in question is
backed by nbd, the control goes back into the tcp layer - back into
tcp_sendmsg_locked().
nbd sets sk_allocation to (GFP_NOIO | __GFP_MEMALLOC) which makes
sense - it's in the process of freeing memory and wants to be able to,
e.g., drop clean pages to make forward progress. However, this
confused sk_page_frag() called from [2]. Because it only tests
whether the allocation allows blocking which it does, it now thinks
current->page_frag can be used again although it already was being
used in [0].
After [2] used current->page_frag, the offset would be increased by
the used amount. When the control returns to [0],
current->page_frag's offset is increased and the previously calculated
number of bytes now may overrun the end of allocated memory leading to
silent memory corruptions.
Fix it by adding gfpflags_normal_context() which tests sleepable &&
!reclaim and use it to determine whether to use current->task_frag.
v2: Eric didn't like gfp flags being tested twice. Introduce a new
helper gfpflags_normal_context() and combine the two tests.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Josef Bacik <josef@toxicpanda.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: stable@vger.kernel.org
Signed-off-by: David S. Miller <davem@davemloft.net>
Remove xfs_zero_file_space and reorganize xfs_file_fallocate so that a
single call to xfs_alloc_file_space covers all modes that preallocate
blocks.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
If we always have to write out of place preallocating blocks is
pointless. We already check for this in the normal falloc path, but
the check was missig in the legacy ALLOCSP path.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
ACPICA commit 3d70fd4894824ed1e685f2d059ca22ccd9ac6163
Version 20191018.
Link: 3d70fd48
Signed-off-by: Bob Moore <robert.moore@intel.com>
Signed-off-by: Erik Schmauss <erik.schmauss@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ACPICA commit 1b7228072f254a5b02625586ff7d561757b7fc2d
By removing leading whitespaces, the conversion computes the
correct number of elements in a given buffer or field encoding
that contains leading whitespaces.
Link: 1b722807
Signed-off-by: Erik Schmauss <erik.schmauss@intel.com>
Signed-off-by: Bob Moore <robert.moore@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ACPICA commit 367b363edc5fa1f93bbc14e4a1e05f34fef765a2
acpiexec allows a user to provide a file that indicates values to
initialize named objects during table load with the -fi option. This
can provide more accurate simulation by setting named objects to
values found during OS runtime.
Previously, this option only supported integer objects. This change
adds user initialization support for field units, strings, buffers,
and packages.
Link: 367b363e
Signed-off-by: Erik Schmauss <erik.schmauss@intel.com>
Signed-off-by: Bob Moore <robert.moore@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ACPICA commit d509afa88e9415f13a3283c38ce9ee034634ae24
Since field unit data output from the debugger are now surrounded by
braces '{', support has been added to acpi_db_get_next_token to recognize
strings beginning with this character as a ACPI_TYPE_FIELD_UNIT.
Link: d509afa8
Signed-off-by: Erik Schmauss <erik.schmauss@intel.com>
Signed-off-by: Bob Moore <robert.moore@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ACPICA commit 76ca57291d007d33087982a4b28cd1ee9bcd37a6
This helps differentiate the type of named objects between field
units and buffers. In other words, without this symbol, it would be
difficult to tell whether a particular named object is a buffer or a
field unit.
Link: 76ca5729
Signed-off-by: Erik Schmauss <erik.schmauss@intel.com>
Signed-off-by: Bob Moore <robert.moore@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
In acpiexec, this can be invoked by typing "fields" followed by a
number representing the address space ID of that field.
Signed-off-by: Erik Schmauss <erik.schmauss@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ACPICA commit fb18935fcf940c5854a055975c6b9ee31f0e1a5a
Link: fb18935f
Signed-off-by: Erik Schmauss <erik.schmauss@intel.com>
Signed-off-by: Bob Moore <robert.moore@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ACPICA commit d1716a829d19be23277d9157c575a03b9abb7457
For unloading an ACPI table, it is necessary to provide the index of
the table. The method intended for dynamically loading or hotplug
addition of tables, acpi_load_table(), should provide this information
via an optional pointer to the loaded table index.
This patch fixes the table unload function of acpi_configfs.
Reported-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Fixes: d06c47e3dd ("ACPI: configfs: Resolve objects on host-directed table loads")
Link: d1716a82
Signed-off-by: Nikolaus Voss <nikolaus.voss@loewensteinmedical.de>
Signed-off-by: Erik Schmauss <erik.schmauss@intel.com>
Signed-off-by: Bob Moore <robert.moore@intel.com>
Tested-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ACPICA commit c69369cd9cf0134e1aac516e97d612947daa8dc2
Unload a table via the table_index.
Link: c69369cd
Signed-off-by: Bob Moore <robert.moore@intel.com>
Signed-off-by: Erik Schmauss <erik.schmauss@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ACPICA commit 54b3aefb5de860306951c8c3339b1c37dcdf1b39
V8.0.1: Fixed all "dead assignment" warnings.
Link: 54b3aefb
Signed-off-by: Bob Moore <robert.moore@intel.com>
Signed-off-by: Erik Schmauss <erik.schmauss@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ACPICA commit 7bc16c650317001bc82d4bae227b888a49c51f5e
Avoid possible overflow from get_tick_count. Also, cast math
using ACPI_100NSEC_PER_MSEC to uint64.
Link: 7bc16c65
Signed-off-by: Bob Moore <robert.moore@intel.com>
Signed-off-by: Erik Schmauss <erik.schmauss@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
ACPICA commit 1f08279b3eb13f17004159c28c391a390cd68feb
Changes/fixes From Clang V5.0.1. Mostly "set but never read"
warnings.
Link: 1f08279b
Signed-off-by: Bob Moore <robert.moore@intel.com>
Signed-off-by: Erik Schmauss <erik.schmauss@intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Florian Fainelli says:
====================
net: dsa: b53: Add support for MDB
This patch series adds support for programming multicast database
entries on b53 and bcm_sf2. This is extracted from a previously
submitted series that added managed mode support, but these patches are
usable in isolation. The larger series still needs to be reworked.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Leverage the recently add b53_mdb_{add,del,prepare} functions since they
work as-is for bcm_sf2.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Vivien Didelot <vivien.didelot@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In preparation for supporting IGMP snooping with or without the use of
a bridge, add support within b53_common.c to program the ARL entries for
multicast operations. The key difference is that a multicast ARL entry
is comprised of a bitmask of enabled ports, instead of a port number.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Vivien Didelot <vivien.didelot@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
KCSAN reported a data-race in udp_set_dev_scratch() [1]
The issue here is that we must not write over skb fields
if skb is shared. A similar issue has been fixed in commit
89c22d8c3b ("net: Fix skb csum races when peeking")
While we are at it, use a helper only dealing with
udp_skb_scratch(skb)->csum_unnecessary, as this allows
udp_set_dev_scratch() to be called once and thus inlined.
[1]
BUG: KCSAN: data-race in udp_set_dev_scratch / udpv6_recvmsg
write to 0xffff888120278317 of 1 bytes by task 10411 on cpu 1:
udp_set_dev_scratch+0xea/0x200 net/ipv4/udp.c:1308
__first_packet_length+0x147/0x420 net/ipv4/udp.c:1556
first_packet_length+0x68/0x2a0 net/ipv4/udp.c:1579
udp_poll+0xea/0x110 net/ipv4/udp.c:2720
sock_poll+0xed/0x250 net/socket.c:1256
vfs_poll include/linux/poll.h:90 [inline]
do_select+0x7d0/0x1020 fs/select.c:534
core_sys_select+0x381/0x550 fs/select.c:677
do_pselect.constprop.0+0x11d/0x160 fs/select.c:759
__do_sys_pselect6 fs/select.c:784 [inline]
__se_sys_pselect6 fs/select.c:769 [inline]
__x64_sys_pselect6+0x12e/0x170 fs/select.c:769
do_syscall_64+0xcc/0x370 arch/x86/entry/common.c:290
entry_SYSCALL_64_after_hwframe+0x44/0xa9
read to 0xffff888120278317 of 1 bytes by task 10413 on cpu 0:
udp_skb_csum_unnecessary include/net/udp.h:358 [inline]
udpv6_recvmsg+0x43e/0xe90 net/ipv6/udp.c:310
inet6_recvmsg+0xbb/0x240 net/ipv6/af_inet6.c:592
sock_recvmsg_nosec+0x5c/0x70 net/socket.c:871
___sys_recvmsg+0x1a0/0x3e0 net/socket.c:2480
do_recvmmsg+0x19a/0x5c0 net/socket.c:2601
__sys_recvmmsg+0x1ef/0x200 net/socket.c:2680
__do_sys_recvmmsg net/socket.c:2703 [inline]
__se_sys_recvmmsg net/socket.c:2696 [inline]
__x64_sys_recvmmsg+0x89/0xb0 net/socket.c:2696
do_syscall_64+0xcc/0x370 arch/x86/entry/common.c:290
entry_SYSCALL_64_after_hwframe+0x44/0xa9
Reported by Kernel Concurrency Sanitizer on:
CPU: 0 PID: 10413 Comm: syz-executor.0 Not tainted 5.4.0-rc3+ #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Fixes: 2276f58ac5 ("udp: use a separate rx queue for packet reception")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Cc: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Matteo Croce says:
====================
mvpp2 improvements in rx path
Refactor some code in the RX path to allow prefetching some data from the
packet header. The first patch is only a refactor, the second one
reduces the data synced, while the third one adds the prefetch.
The packet rate improvement with the second patch is very small (1606 => 1620 kpps),
while the prefetch bumps it up by 14%: 1620 => 1853 kpps.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
When receiving traffic, eth_type_trans() is high up on the perf top list,
because it's the first function which access the packet data.
Move the DMA unmap a bit higher, and put a prefetch just after it, so we
have more time to load the data into the cache.
The packet rate increase is about 14% with a tc drop test: 1620 => 1853 kpps
Signed-off-by: Matteo Croce <mcroce@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In the RX path we always sync against the maximum frame size for that pool.
Do the DMA sync and the unmap separately, so we can only sync by the
size of the received frame.
Signed-off-by: Matteo Croce <mcroce@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix misspellings of "endpoints", "configuration", and "device's".
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix misspellings of "disconnect", "disconnecting", "connections", and
"disconnected".
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Acked-by: Kalle Valo <kvalo@codeaurora.org>
Acked-by: Simon Horman <horms@verge.net.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix various misspellings of "configuration" and "configure".
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Acked-by: Kalle Valo <kvalo@codeaurora.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently we insert a arbitration point every 128MiB during a blitter
copy. At 8GiB/s, this is around 30ms. This is a little on the large side
if we need to inject a high priority work, so reduced it down to 8MiB or
roughly 1ms.
v2: Don't forget both fill/copy.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20191028203012.14566-1-chris@chris-wilson.co.uk
This patch corrects the SPDX License Identifier style in
header files related to DPAA2 Ethernet driver supporting
Freescale SoCs with DPAA2. For C header files
Documentation/process/license-rules.rst mandates C-like comments
(opposed to C source files where C++ style should be used)
Changes made by using a script provided by Joe Perches here:
https://lkml.org/lkml/2019/2/7/46.
Suggested-by: Joe Perches <joe@perches.com>
Signed-off-by: Nishad Kamdar <nishadkamdar@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While processing CSB there is no need to look at GuC submission
settings, just check if engine is configured for execlists mode.
While today GuC submission is disabled it's settings are still
based on modparam values that might not correctly reflect actual
submission status in case of any fallback. Until that is fully
fixed, use alternate method to confirm that engine really runs in
execlists mode by comparing set_default_submission vfunc.
v2: add other immediate use of new helper
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Janusz Krzysztofik <janusz.krzysztofik@linux.intel.com>
Reviewed-by: Janusz Krzysztofik <janusz.krzysztofik@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20191028164520.31772-1-michal.wajdeczko@intel.com
Since commit 0394a63acf ("net: dsa: enable and disable all ports")
the dsa core disables all unused ports of a switch. In this case
disabling ports with numbers higher than QCA8K_NUM_PORTS causes that
some switch registers are overwritten with incorrect content.
To fix this, initialize the dsa_switch->num_ports with correct number
of ports.
Fixes: 7e99e34701 ("net: dsa: remove dsa_switch_alloc helper")
Signed-off-by: Michal Vokáč <michal.vokac@ysoft.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently ds->dev is dereferenced on the assignments of pdata and
np before ds->dev is null checked, hence there is a potential null
pointer dereference on ds->dev. Fix this by assigning pdata and
np after the ds->dev null pointer sanity check.
Addresses-Coverity: ("Dereference before null check")
Fixes: 7e99e34701 ("net: dsa: remove dsa_switch_alloc helper")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Reported-by: kbuild test robot <lkp@intel.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Vivien Didelot <vivien.didelot@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet says:
====================
net: avoid KCSAN splats
Often times we use skb_queue_empty() without holding a lock,
meaning that other cpus (or interrupt) can change the queue
under us. This is fine, but we need to properly annotate
the lockless intent to make sure the compiler wont over
optimize things.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Busy polling usually runs without locks.
Let's use skb_queue_empty_lockless() instead of skb_queue_empty()
Also uses READ_ONCE() in __skb_try_recv_datagram() to address
a similar potential problem.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Many poll() handlers are lockless. Using skb_queue_empty_lockless()
instead of skb_queue_empty() is more appropriate.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Some paths call skb_queue_empty() without holding
the queue lock. We must use a barrier in order
to not let the compiler do strange things, and avoid
KCSAN splats.
Adding a barrier in skb_queue_empty() might be overkill,
I prefer adding a new helper to clearly identify
points where the callers might be lockless. This might
help us finding real bugs.
The corresponding WRITE_ONCE() should add zero cost
for current compilers.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jason Gunthorpe says:
====================
In order to hoist the interval tree code out of the drivers and into the
mmu_notifiers it is necessary for the drivers to not use the interval tree
for other things.
This series replaces the interval tree with an xarray and along the way
re-aligns all the locking to use a sensible SRCU model where the 'update'
step is done by modifying an xarray.
The result is overall much simpler and with less locking in the critical
path. Many functions were reworked for clarity and small details like
using 'imr' to refer to the implicit MR make the entire code flow here
more readable.
This also squashes at least two race bugs on its own, and quite possibily
more that haven't been identified.
====================
Merge conflicts with the odp statistics patch resolved.
* branch 'odp_rework':
RDMA/odp: Remove broken debugging call to invalidate_range
RDMA/mlx5: Do not race with mlx5_ib_invalidate_range during create and destroy
RDMA/mlx5: Do not store implicit children in the odp_mkeys xarray
RDMA/mlx5: Rework implicit ODP destroy
RDMA/mlx5: Avoid double lookups on the pagefault path
RDMA/mlx5: Reduce locking in implicit_mr_get_data()
RDMA/mlx5: Use an xarray for the children of an implicit ODP
RDMA/mlx5: Split implicit handling from pagefault_mr
RDMA/mlx5: Set the HW IOVA of the child MRs to their place in the tree
RDMA/mlx5: Lift implicit_mr_alloc() into the two routines that call it
RDMA/mlx5: Rework implicit_mr_get_data
RDMA/mlx5: Delete struct mlx5_priv->mkey_table
RDMA/mlx5: Use a dedicated mkey xarray for ODP
RDMA/mlx5: Split sig_err MR data into its own xarray
RDMA/mlx5: Use SRCU properly in ODP prefetch
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
invalidate_range() also obtains the umem_mutex which is being held at this
point, so if this path were was ever called it would deadlock. Thus
conclude the debugging never triggers and rework it into a simple WARN_ON
and leave things as they are.
While here add a note to explain how we could possibly get inconsistent
page pointers.
Link: https://lore.kernel.org/r/20191009160934.3143-16-jgg@ziepe.ca
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
For creation, as soon as the umem_odp is created the notifier can be
called, however the underlying MR may not have been setup yet. This would
cause problems if mlx5_ib_invalidate_range() runs. There is some
confusing/ulocked/racy code that might by trying to solve this, but
without locks it isn't going to work right.
Instead trivially solve the problem by short-circuiting the invalidation
if there are not yet any DMA mapped pages. By definition there is nothing
to invalidate in this case.
The create code will have the umem fully setup before anything is DMA
mapped, and npages is fully locked by the umem_mutex.
For destroy, invalidate the entire MR at the HW to stop DMA then DMA unmap
the pages before destroying the MR. This drives npages to zero and
prevents similar racing with invalidate while the MR is undergoing
destruction.
Arguably it would be better if the umem was created after the MR and
destroyed before, but that would require a big rework of the MR code.
Fixes: 6aec21f6a8 ("IB/mlx5: Page faults handling infrastructure")
Link: https://lore.kernel.org/r/20191009160934.3143-15-jgg@ziepe.ca
Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
These mkeys are entirely internal and are never used by the HW for
page fault. They should also never be used by userspace for prefetch.
Simplify & optimize things by not including them in the xarray.
Since the prefetch path can now never see a child mkey there is no need
for the second synchronize_srcu() during imr destroy.
Link: https://lore.kernel.org/r/20191009160934.3143-14-jgg@ziepe.ca
Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Use SRCU in a sensible way by removing all MRs in the implicit tree from
the two xarrays (the update operation), then a synchronize, followed by a
normal single threaded teardown.
This is only a little unusual from the normal pattern as there can still
be some work pending in the unbound wq that may also require a workqueue
flush. This is tracked with a single atomic, consolidating the redundant
existing atomics and wait queue.
For understand-ability the entire ODP implicit create/destroy flow now
largely exists in a single pair of functions within odp.c, with a few
support functions for tearing down an unused child.
Link: https://lore.kernel.org/r/20191009160934.3143-13-jgg@ziepe.ca
Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Now that the locking is simplified combine pagefault_implicit_mr() with
implicit_mr_get_data() so that we sweep over the idx range only once,
and do the single xlt update at the end, after the child umems are
setup.
This avoids double iteration/xa_loads plus the sketchy failure path if the
xa_load() fails.
Link: https://lore.kernel.org/r/20191009160934.3143-12-jgg@ziepe.ca
Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Now that the child MRs are stored in an xarray we can rely on the SRCU
lock to protect the xa_load and use xa_cmpxchg on the slow allocation path
to resolve races with concurrent page fault.
This reduces the scope of the critical section of umem_mutex for implicit
MRs to only cover mlx5_ib_update_xlt, and avoids taking a lock at all if
the child MR is already in the xarray. This makes it consistent with the
normal ODP MR critical section for umem_lock, and the locking approach
used for destroying an unusued implicit child MR.
The MLX5_IB_UPD_XLT_ATOMIC is no longer needed in implicit_get_child_mr()
since it is no longer called with any locks.
Link: https://lore.kernel.org/r/20191009160934.3143-11-jgg@ziepe.ca
Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Currently the child leaves are stored in the shared interval tree and
every lookup for a child must be done under the interval tree rwsem.
This is further complicated by dropping the rwsem during iteration (ie the
odp_lookup(), odp_next() pattern), which requires a very tricky an
difficult to understand locking scheme with SRCU.
Instead reserve the interval tree for the exclusive use of the mmu
notifier related code in umem_odp.c and give each implicit MR a xarray
containing all the child MRs.
Since the size of each child is 1GB of VA, a 1 level xarray will index 64G
of VA, and a 2 level will index 2TB, making xarray a much better
data structure choice than an interval tree.
The locking properties of xarray will be used in the next patches to
rework the implicit ODP locking scheme into something simpler.
At this point, the xarray is locked by the implicit MR's umem_mutex, and
read can also be locked by the odp_srcu.
Link: https://lore.kernel.org/r/20191009160934.3143-10-jgg@ziepe.ca
Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
The single routine has a very confusing scheme to advance to the next
child MR when working on an implicit parent. This scheme can only be used
when working with an implicit parent and must not be triggered when
working on a normal MR.
Re-arrange things by directly putting all the single-MR stuff into one
function and calling it in a loop for the implicit case. Simplify some of
the error handling in the new pagefault_real_mr() to remove unneeded gotos.
Link: https://lore.kernel.org/r/20191009160934.3143-9-jgg@ziepe.ca
Reviewed-by: Artemy Kovalyov <artemyko@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
