Age | Commit message (Collapse) | Author | Files | Lines |
|
Add debugfs to hns RoCE. This patch only adds an empty directory
"hns_roce" to debugs root directory.
Signed-off-by: Junxian Huang <huangjunxian6@hisilicon.com>
Link: https://lore.kernel.org/r/20231114123449.1106162-3-huangjunxian6@hisilicon.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
EOPNOTSUPP is more situable than EINVAL for allocating XRCD while XRC
is not supported and unsupported resizing SRQ.
Fixes: 32548870d438 ("RDMA/hns: Add support for XRC on HIP09")
Fixes: 221109e64316 ("RDMA/hns: Add interception for resizing SRQs")
Signed-off-by: Junxian Huang <huangjunxian6@hisilicon.com>
Link: https://lore.kernel.org/r/20231114123449.1106162-2-huangjunxian6@hisilicon.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Currently, there is no wait for the QP suspend to complete on a modify
to SQD state. Add a wait, after the modify to SQD state, for the Suspend
Complete AE. While we are at it, update the suspend timeout value in
irdma_prep_tc_change to use IRDMA_EVENT_TIMEOUT_MS too.
Fixes: b48c24c2d710 ("RDMA/irdma: Implement device supported verb APIs")
Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Link: https://lore.kernel.org/r/20231114170246.238-3-shiraz.saleem@intel.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Remove the modify to SQD before going to ERROR state. It is not needed.
Fixes: b48c24c2d710 ("RDMA/irdma: Implement device supported verb APIs")
Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Link: https://lore.kernel.org/r/20231114170246.238-2-shiraz.saleem@intel.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
The patch 1da177e4c3f4: "Linux-2.6.12-rc2" from Apr 16, 2005
(linux-next), leads to the following Smatch static checker warning:
drivers/infiniband/hw/mthca/mthca_cmd.c:644 mthca_SYS_EN()
error: uninitialized symbol 'out'.
drivers/infiniband/hw/mthca/mthca_cmd.c
636 int mthca_SYS_EN(struct mthca_dev *dev)
637 {
638 u64 out;
639 int ret;
640
641 ret = mthca_cmd_imm(dev, 0, &out, 0, 0, CMD_SYS_EN, CMD_TIME_CLASS_D);
We pass out here and it gets used without being initialized.
err = mthca_cmd_post(dev, in_param,
out_param ? *out_param : 0,
^^^^^^^^^^
in_modifier, op_modifier,
op, context->token, 1);
It's the same in mthca_cmd_wait() and mthca_cmd_poll().
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Reported-by: Dan Carpenter <dan.carpenter@linaro.org>
Closes: https://lore.kernel.org/all/533bc3df-8078-4397-b93d-d1f6cec9b636@moroto.mountain
Link: https://lore.kernel.org/r/c559cb7113158c02d75401ac162652072ef1b5f0.1699867650.git.leon@kernel.org
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
|
|
Simplify irdma_ieq_check_mpacrc() by using crypto_shash_digest() instead
of an init+update+final sequence. This should also improve performance.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Link: https://lore.kernel.org/r/20231029045756.153943-1-ebiggers@kernel.org
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Rounding up the queue depth to power of two is not a hardware requirement.
In order to optimize the per connection memory usage, removing drivers
implementation which round up to the queue depths to the power of 2.
Implements a mask to maintain backward compatibility with older
library.
Signed-off-by: Chandramohan Akula <chandramohan.akula@broadcom.com>
Signed-off-by: Selvin Xavier <selvin.xavier@broadcom.com>
Link: https://lore.kernel.org/r/1698069803-1787-3-git-send-email-selvin.xavier@broadcom.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
The queue index wrap around logic is based on power of 2 size depth.
All queues are created with power of 2 depth. This increases the
memory usage by the driver. This change is required for the next
patches that avoids the power of 2 depth requirement for each of
the queues.
Update the function that increments producer index and consumer
index during wrap around. Also, changes the index handling across
multiple functions.
Signed-off-by: Chandramohan Akula <chandramohan.akula@broadcom.com>
Signed-off-by: Selvin Xavier <selvin.xavier@broadcom.com>
Link: https://lore.kernel.org/r/1698069803-1787-2-git-send-email-selvin.xavier@broadcom.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
algorithm
Add a default congest control algorithm so that driver won't return
an error when the configured algorithm is invalid.
Fixes: f91696f2f053 ("RDMA/hns: Support congestion control type selection according to the FW")
Signed-off-by: Junxian Huang <huangjunxian6@hisilicon.com>
Link: https://lore.kernel.org/r/20231028093242.670325-1-huangjunxian6@hisilicon.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Currently, memdup_user() is utilized at two positions to copy userspace
arrays. This is done without overflow checks.
Use the new wrapper memdup_array_user() to copy the arrays more safely.
Suggested-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Philipp Stanner <pstanner@redhat.com>
Link: https://lore.kernel.org/r/20231102191308.52046-2-pstanner@redhat.com
Acked-by: Dennis Dalessandro <dennis.dalessandro@cornelisnetworks.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Pull rdma updates from Jason Gunthorpe:
"Nothing exciting this cycle, most of the diffstat is changing SPDX
'or' to 'OR'.
Summary:
- Bugfixes for hns, mlx5, and hfi1
- Hardening patches for size_*, counted_by, strscpy
- rts fixes from static analysis
- Dump SRQ objects in rdma netlink, with hns support
- Fix a performance regression in mlx5 MR deregistration
- New XDR (200Gb/lane) link speed
- SRQ record doorbell latency optimization for hns
- IPSEC support for mlx5 multi-port mode
- ibv_rereg_mr() support for irdma
- Affiliated event support for bnxt_re
- Opt out for the spec compliant qkey security enforcement as we
discovered SW that breaks under enforcement
- Comment and trivial updates"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma: (50 commits)
IB/mlx5: Fix init stage error handling to avoid double free of same QP and UAF
RDMA/mlx5: Fix mkey cache WQ flush
RDMA/hfi1: Workaround truncation compilation error
IB/hfi1: Fix potential deadlock on &irq_src_lock and &dd->uctxt_lock
RDMA/core: Remove NULL check before dev_{put, hold}
RDMA/hfi1: Remove redundant assignment to pointer ppd
RDMA/mlx5: Change the key being sent for MPV device affiliation
RDMA/bnxt_re: Fix clang -Wimplicit-fallthrough in bnxt_re_handle_cq_async_error()
RDMA/hns: Fix init failure of RoCE VF and HIP08
RDMA/hns: Fix unnecessary port_num transition in HW stats allocation
RDMA/hns: The UD mode can only be configured with DCQCN
RDMA/hns: Add check for SL
RDMA/hns: Fix signed-unsigned mixed comparisons
RDMA/hns: Fix uninitialized ucmd in hns_roce_create_qp_common()
RDMA/hns: Fix printing level of asynchronous events
RDMA/core: Add support to set privileged QKEY parameter
RDMA/bnxt_re: Do not report SRQ error in srq notification
RDMA/bnxt_re: Report async events and errors
RDMA/bnxt_re: Update HW interface headers
IB/mlx5: Fix rdma counter binding for RAW QP
...
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next
Pull networking updates from Jakub Kicinski:
"Core & protocols:
- Support usec resolution of TCP timestamps, enabled selectively by a
route attribute.
- Defer regular TCP ACK while processing socket backlog, try to send
a cumulative ACK at the end. Increase single TCP flow performance
on a 200Gbit NIC by 20% (100Gbit -> 120Gbit).
- The Fair Queuing (FQ) packet scheduler:
- add built-in 3 band prio / WRR scheduling
- support bypass if the qdisc is mostly idle (5% speed up for TCP RR)
- improve inactive flow reporting
- optimize the layout of structures for better cache locality
- Support TCP Authentication Option (RFC 5925, TCP-AO), a more modern
replacement for the old MD5 option.
- Add more retransmission timeout (RTO) related statistics to
TCP_INFO.
- Support sending fragmented skbs over vsock sockets.
- Make sure we send SIGPIPE for vsock sockets if socket was
shutdown().
- Add sysctl for ignoring lower limit on lifetime in Router
Advertisement PIO, based on an in-progress IETF draft.
- Add sysctl to control activation of TCP ping-pong mode.
- Add sysctl to make connection timeout in MPTCP configurable.
- Support rcvlowat and notsent_lowat on MPTCP sockets, to help apps
limit the number of wakeups.
- Support netlink GET for MDB (multicast forwarding), allowing user
space to request a single MDB entry instead of dumping the entire
table.
- Support selective FDB flushing in the VXLAN tunnel driver.
- Allow limiting learned FDB entries in bridges, prevent OOM attacks.
- Allow controlling via configfs netconsole targets which were
created via the kernel cmdline at boot, rather than via configfs at
runtime.
- Support multiple PTP timestamp event queue readers with different
filters.
- MCTP over I3C.
BPF:
- Add new veth-like netdevice where BPF program defines the logic of
the xmit routine. It can operate in L3 and L2 mode.
- Support exceptions - allow asserting conditions which should never
be true but are hard for the verifier to infer. With some extra
flexibility around handling of the exit / failure:
https://lwn.net/Articles/938435/
- Add support for local per-cpu kptr, allow allocating and storing
per-cpu objects in maps. Access to those objects operates on the
value for the current CPU.
This allows to deprecate local one-off implementations of per-CPU
storage like BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE maps.
- Extend cgroup BPF sockaddr hooks for UNIX sockets. The use case is
for systemd to re-implement the LogNamespace feature which allows
running multiple instances of systemd-journald to process the logs
of different services.
- Enable open-coded task_vma iteration, after maple tree conversion
made it hard to directly walk VMAs in tracing programs.
- Add open-coded task, css_task and css iterator support. One of the
use cases is customizable OOM victim selection via BPF.
- Allow source address selection with bpf_*_fib_lookup().
- Add ability to pin BPF timer to the current CPU.
- Prevent creation of infinite loops by combining tail calls and
fentry/fexit programs.
- Add missed stats for kprobes to retrieve the number of missed
kprobe executions and subsequent executions of BPF programs.
- Inherit system settings for CPU security mitigations.
- Add BPF v4 CPU instruction support for arm32 and s390x.
Changes to common code:
- overflow: add DEFINE_FLEX() for on-stack definition of structs with
flexible array members.
- Process doc update with more guidance for reviewers.
Driver API:
- Simplify locking in WiFi (cfg80211 and mac80211 layers), use wiphy
mutex in most places and remove a lot of smaller locks.
- Create a common DPLL configuration API. Allow configuring and
querying state of PLL circuits used for clock syntonization, in
network time distribution.
- Unify fragmented and full page allocation APIs in page pool code.
Let drivers be ignorant of PAGE_SIZE.
- Rework PHY state machine to avoid races with calls to phy_stop().
- Notify DSA drivers of MAC address changes on user ports, improve
correctness of offloads which depend on matching port MAC
addresses.
- Allow antenna control on injected WiFi frames.
- Reduce the number of variants of napi_schedule().
- Simplify error handling when composing devlink health messages.
Misc:
- A lot of KCSAN data race "fixes", from Eric.
- A lot of __counted_by() annotations, from Kees.
- A lot of strncpy -> strscpy and printf format fixes.
- Replace master/slave terminology with conduit/user in DSA drivers.
- Handful of KUnit tests for netdev and WiFi core.
Removed:
- AppleTalk COPS.
- AppleTalk ipddp.
- TI AR7 CPMAC Ethernet driver.
Drivers:
- Ethernet high-speed NICs:
- Intel (100G, ice, idpf):
- add a driver for the Intel E2000 IPUs
- make CRC/FCS stripping configurable
- cross-timestamping for E823 devices
- basic support for E830 devices
- use aux-bus for managing client drivers
- i40e: report firmware versions via devlink
- nVidia/Mellanox:
- support 4-port NICs
- increase max number of channels to 256
- optimize / parallelize SF creation flow
- Broadcom (bnxt):
- enhance NIC temperature reporting
- support PAM4 speeds and lane configuration
- Marvell OcteonTX2:
- PTP pulse-per-second output support
- enable hardware timestamping for VFs
- Solarflare/AMD:
- conntrack NAT offload and offload for tunnels
- Wangxun (ngbe/txgbe):
- expose HW statistics
- Pensando/AMD:
- support PCI level reset
- narrow down the condition under which skbs are linearized
- Netronome/Corigine (nfp):
- support CHACHA20-POLY1305 crypto in IPsec offload
- Ethernet NICs embedded, slower, virtual:
- Synopsys (stmmac):
- add Loongson-1 SoC support
- enable use of HW queues with no offload capabilities
- enable PPS input support on all 5 channels
- increase TX coalesce timer to 5ms
- RealTek USB (r8152): improve efficiency of Rx by using GRO frags
- xen: support SW packet timestamping
- add drivers for implementations based on TI's PRUSS (AM64x EVM)
- nVidia/Mellanox Ethernet datacenter switches:
- avoid poor HW resource use on Spectrum-4 by better block
selection for IPv6 multicast forwarding and ordering of blocks
in ACL region
- Ethernet embedded switches:
- Microchip:
- support configuring the drive strength for EMI compliance
- ksz9477: partial ACL support
- ksz9477: HSR offload
- ksz9477: Wake on LAN
- Realtek:
- rtl8366rb: respect device tree config of the CPU port
- Ethernet PHYs:
- support Broadcom BCM5221 PHYs
- TI dp83867: support hardware LED blinking
- CAN:
- add support for Linux-PHY based CAN transceivers
- at91_can: clean up and use rx-offload helpers
- WiFi:
- MediaTek (mt76):
- new sub-driver for mt7925 USB/PCIe devices
- HW wireless <> Ethernet bridging in MT7988 chips
- mt7603/mt7628 stability improvements
- Qualcomm (ath12k):
- WCN7850:
- enable 320 MHz channels in 6 GHz band
- hardware rfkill support
- enable IEEE80211_HW_SINGLE_SCAN_ON_ALL_BANDS to
make scan faster
- read board data variant name from SMBIOS
- QCN9274: mesh support
- RealTek (rtw89):
- TDMA-based multi-channel concurrency (MCC)
- Silicon Labs (wfx):
- Remain-On-Channel (ROC) support
- Bluetooth:
- ISO: many improvements for broadcast support
- mark BCM4378/BCM4387 as BROKEN_LE_CODED
- add support for QCA2066
- btmtksdio: enable Bluetooth wakeup from suspend"
* tag 'net-next-6.7' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (1816 commits)
net: pcs: xpcs: Add 2500BASE-X case in get state for XPCS drivers
net: bpf: Use sockopt_lock_sock() in ip_sock_set_tos()
net: mana: Use xdp_set_features_flag instead of direct assignment
vxlan: Cleanup IFLA_VXLAN_PORT_RANGE entry in vxlan_get_size()
iavf: delete the iavf client interface
iavf: add a common function for undoing the interrupt scheme
iavf: use unregister_netdev
iavf: rely on netdev's own registered state
iavf: fix the waiting time for initial reset
iavf: in iavf_down, don't queue watchdog_task if comms failed
iavf: simplify mutex_trylock+sleep loops
iavf: fix comments about old bit locks
doc/netlink: Update schema to support cmd-cnt-name and cmd-max-name
tools: ynl: introduce option to process unknown attributes or types
ipvlan: properly track tx_errors
netdevsim: Block until all devices are released
nfp: using napi_build_skb() to replace build_skb()
net: dsa: microchip: ksz9477: Fix spelling mistake "Enery" -> "Energy"
net: dsa: microchip: Ensure Stable PME Pin State for Wake-on-LAN
net: dsa: microchip: Refactor switch shutdown routine for WoL preparation
...
|
|
In the unlikely event that workqueue allocation fails and returns NULL in
mlx5_mkey_cache_init(), delete the call to
mlx5r_umr_resource_cleanup() (which frees the QP) in
mlx5_ib_stage_post_ib_reg_umr_init(). This will avoid attempted double
free of the same QP when __mlx5_ib_add() does its cleanup.
Resolves a splat:
Syzkaller reported a UAF in ib_destroy_qp_user
workqueue: Failed to create a rescuer kthread for wq "mkey_cache": -EINTR
infiniband mlx5_0: mlx5_mkey_cache_init:981:(pid 1642):
failed to create work queue
infiniband mlx5_0: mlx5_ib_stage_post_ib_reg_umr_init:4075:(pid 1642):
mr cache init failed -12
==================================================================
BUG: KASAN: slab-use-after-free in ib_destroy_qp_user (drivers/infiniband/core/verbs.c:2073)
Read of size 8 at addr ffff88810da310a8 by task repro_upstream/1642
Call Trace:
<TASK>
kasan_report (mm/kasan/report.c:590)
ib_destroy_qp_user (drivers/infiniband/core/verbs.c:2073)
mlx5r_umr_resource_cleanup (drivers/infiniband/hw/mlx5/umr.c:198)
__mlx5_ib_add (drivers/infiniband/hw/mlx5/main.c:4178)
mlx5r_probe (drivers/infiniband/hw/mlx5/main.c:4402)
...
</TASK>
Allocated by task 1642:
__kmalloc (./include/linux/kasan.h:198 mm/slab_common.c:1026
mm/slab_common.c:1039)
create_qp (./include/linux/slab.h:603 ./include/linux/slab.h:720
./include/rdma/ib_verbs.h:2795 drivers/infiniband/core/verbs.c:1209)
ib_create_qp_kernel (drivers/infiniband/core/verbs.c:1347)
mlx5r_umr_resource_init (drivers/infiniband/hw/mlx5/umr.c:164)
mlx5_ib_stage_post_ib_reg_umr_init (drivers/infiniband/hw/mlx5/main.c:4070)
__mlx5_ib_add (drivers/infiniband/hw/mlx5/main.c:4168)
mlx5r_probe (drivers/infiniband/hw/mlx5/main.c:4402)
...
Freed by task 1642:
__kmem_cache_free (mm/slub.c:1826 mm/slub.c:3809 mm/slub.c:3822)
ib_destroy_qp_user (drivers/infiniband/core/verbs.c:2112)
mlx5r_umr_resource_cleanup (drivers/infiniband/hw/mlx5/umr.c:198)
mlx5_ib_stage_post_ib_reg_umr_init (drivers/infiniband/hw/mlx5/main.c:4076
drivers/infiniband/hw/mlx5/main.c:4065)
__mlx5_ib_add (drivers/infiniband/hw/mlx5/main.c:4168)
mlx5r_probe (drivers/infiniband/hw/mlx5/main.c:4402)
...
Fixes: 04876c12c19e ("RDMA/mlx5: Move init and cleanup of UMR to umr.c")
Link: https://lore.kernel.org/r/1698170518-4006-1-git-send-email-george.kennedy@oracle.com
Suggested-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: George Kennedy <george.kennedy@oracle.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
|
|
The cited patch tries to ensure no pending works on the mkey cache
workqueue by disabling adding new works and call flush_workqueue().
But this workqueue also has delayed works which might still be pending
the delay time to be queued.
Add cancel_delayed_work() for the delayed works which waits to be queued
and then the flush_workqueue() will flush all works which are already
queued and running.
Fixes: 374012b00457 ("RDMA/mlx5: Fix mkey cache possible deadlock on cleanup")
Link: https://lore.kernel.org/r/b8722f14e7ed81452f791764a26d2ed4cfa11478.1698256179.git.leon@kernel.org
Signed-off-by: Moshe Shemesh <moshe@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
|
|
Resolve conflict by taking the spin_lock hunk from for-next:
https://lore.kernel.org/r/20230928113851.5197a1ec@canb.auug.org.au
Required for the next patch.
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
|
|
gitolite.kernel.org:pub/scm/linux/kernel/git/vfs/vfs
Pull vfs inode time accessor updates from Christian Brauner:
"This finishes the conversion of all inode time fields to accessor
functions as discussed on list. Changing timestamps manually as we
used to do before is error prone. Using accessors function makes this
robust.
It does not contain the switch of the time fields to discrete 64 bit
integers to replace struct timespec and free up space in struct inode.
But after this, the switch can be trivially made and the patch should
only affect the vfs if we decide to do it"
* tag 'vfs-6.7.ctime' of gitolite.kernel.org:pub/scm/linux/kernel/git/vfs/vfs: (86 commits)
fs: rename inode i_atime and i_mtime fields
security: convert to new timestamp accessors
selinux: convert to new timestamp accessors
apparmor: convert to new timestamp accessors
sunrpc: convert to new timestamp accessors
mm: convert to new timestamp accessors
bpf: convert to new timestamp accessors
ipc: convert to new timestamp accessors
linux: convert to new timestamp accessors
zonefs: convert to new timestamp accessors
xfs: convert to new timestamp accessors
vboxsf: convert to new timestamp accessors
ufs: convert to new timestamp accessors
udf: convert to new timestamp accessors
ubifs: convert to new timestamp accessors
tracefs: convert to new timestamp accessors
sysv: convert to new timestamp accessors
squashfs: convert to new timestamp accessors
server: convert to new timestamp accessors
client: convert to new timestamp accessors
...
|
|
gitolite.kernel.org:pub/scm/linux/kernel/git/vfs/vfs
Pull iov_iter updates from Christian Brauner:
"This contain's David's iov_iter cleanup work to convert the iov_iter
iteration macros to inline functions:
- Remove last_offset from iov_iter as it was only used by ITER_PIPE
- Add a __user tag on copy_mc_to_user()'s dst argument on x86 to
match that on powerpc and get rid of a sparse warning
- Convert iter->user_backed to user_backed_iter() in the sound PCM
driver
- Convert iter->user_backed to user_backed_iter() in a couple of
infiniband drivers
- Renumber the type enum so that the ITER_* constants match the order
in iterate_and_advance*()
- Since the preceding patch puts UBUF and IOVEC at 0 and 1, change
user_backed_iter() to just use the type value and get rid of the
extra flag
- Convert the iov_iter iteration macros to always-inline functions to
make the code easier to follow. It uses function pointers, but they
get optimised away
- Move the check for ->copy_mc to _copy_from_iter() and
copy_page_from_iter_atomic() rather than in memcpy_from_iter_mc()
where it gets repeated for every segment. Instead, we check once
and invoke a side function that can use iterate_bvec() rather than
iterate_and_advance() and supply a different step function
- Move the copy-and-csum code to net/ where it can be in proximity
with the code that uses it
- Fold memcpy_and_csum() in to its two users
- Move csum_and_copy_from_iter_full() out of line and merge in
csum_and_copy_from_iter() since the former is the only caller of
the latter
- Move hash_and_copy_to_iter() to net/ where it can be with its only
caller"
* tag 'vfs-6.7.iov_iter' of gitolite.kernel.org:pub/scm/linux/kernel/git/vfs/vfs:
iov_iter, net: Move hash_and_copy_to_iter() to net/
iov_iter, net: Merge csum_and_copy_from_iter{,_full}() together
iov_iter, net: Fold in csum_and_memcpy()
iov_iter, net: Move csum_and_copy_to/from_iter() to net/
iov_iter: Don't deal with iter->copy_mc in memcpy_from_iter_mc()
iov_iter: Convert iterate*() to inline funcs
iov_iter: Derive user-backedness from the iterator type
iov_iter: Renumber ITER_* constants
infiniband: Use user_backed_iter() to see if iterator is UBUF/IOVEC
sound: Fix snd_pcm_readv()/writev() to use iov access functions
iov_iter, x86: Be consistent about the __user tag on copy_mc_to_user()
iov_iter: Remove last_offset from iov_iter as it was for ITER_PIPE
|
|
Increase name array to be large enough to overcome the following
compilation error.
drivers/infiniband/hw/hfi1/efivar.c: In function ‘read_hfi1_efi_var’:
drivers/infiniband/hw/hfi1/efivar.c:124:44: error: ‘snprintf’ output may be truncated before the last format character [-Werror=format-truncation=]
124 | snprintf(name, sizeof(name), "%s-%s", prefix_name, kind);
| ^
drivers/infiniband/hw/hfi1/efivar.c:124:9: note: ‘snprintf’ output 2 or more bytes (assuming 65) into a destination of size 64
124 | snprintf(name, sizeof(name), "%s-%s", prefix_name, kind);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/infiniband/hw/hfi1/efivar.c:133:52: error: ‘snprintf’ output may be truncated before the last format character [-Werror=format-truncation=]
133 | snprintf(name, sizeof(name), "%s-%s", prefix_name, kind);
| ^
drivers/infiniband/hw/hfi1/efivar.c:133:17: note: ‘snprintf’ output 2 or more bytes (assuming 65) into a destination of size 64
133 | snprintf(name, sizeof(name), "%s-%s", prefix_name, kind);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors
make[6]: *** [scripts/Makefile.build:243: drivers/infiniband/hw/hfi1/efivar.o] Error 1
Fixes: c03c08d50b3d ("IB/hfi1: Check upper-case EFI variables")
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Link: https://lore.kernel.org/r/238fa39a8fd60e87a5ad7e1ca6584fcdf32e9519.1698159993.git.leonro@nvidia.com
Acked-by: Dennis Dalessandro <dennis.dalessandro@cornelisnetworks.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
handle_receive_interrupt_napi_sp() running inside interrupt handler
could introduce inverse lock ordering between &dd->irq_src_lock
and &dd->uctxt_lock, if read_mod_write() is preempted by the isr.
[CPU0] | [CPU1]
hfi1_ipoib_dev_open() |
--> hfi1_netdev_enable_queues() |
--> enable_queues(rx) |
--> hfi1_rcvctrl() |
--> set_intr_bits() |
--> read_mod_write() |
--> spin_lock(&dd->irq_src_lock) |
| hfi1_poll()
| --> poll_next()
| --> spin_lock_irq(&dd->uctxt_lock)
|
| --> hfi1_rcvctrl()
| --> set_intr_bits()
| --> read_mod_write()
| --> spin_lock(&dd->irq_src_lock)
<interrupt> |
--> handle_receive_interrupt_napi_sp() |
--> set_all_fastpath() |
--> hfi1_rcd_get_by_index() |
--> spin_lock_irqsave(&dd->uctxt_lock) |
This flaw was found by an experimental static analysis tool I am
developing for irq-related deadlock.
To prevent the potential deadlock, the patch use spin_lock_irqsave()
on &dd->irq_src_lock inside read_mod_write() to prevent the possible
deadlock scenario.
Signed-off-by: Chengfeng Ye <dg573847474@gmail.com>
Link: https://lore.kernel.org/r/20230926101116.2797-1-dg573847474@gmail.com
Acked-by: Dennis Dalessandro <dennis.dalessandro@cornelisnetworks.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
The call netdev_{put, hold} of dev_{put, hold} will check NULL,
so there is no need to check before using dev_{put, hold},
remove it to silence the warning:
./drivers/infiniband/core/nldev.c:375:2-9: WARNING: NULL check before dev_{put, hold} functions is not needed.
Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Closes: https://bugzilla.openanolis.cn/show_bug.cgi?id=7047
Signed-off-by: Yang Li <yang.lee@linux.alibaba.com>
Link: https://lore.kernel.org/r/20231024003815.89742-1-yang.lee@linux.alibaba.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Pointer ppd is being assigned a value in a for-loop however it
is never read. The assignment is redundant and can be removed.
Cleans up clang scan build warning:
drivers/infiniband/hw/hfi1/init.c:1030:3: warning: Value stored
to 'ppd' is never read [deadcode.DeadStores]
Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Link: https://lore.kernel.org/r/20231023141733.667807-1-colin.i.king@gmail.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Change the key that we send from IB driver to EN driver regarding the
MPV device affiliation, since at that stage the IB device is not yet
initialized, so its index would be zero for different IB devices and
cause wrong associations between unrelated master and slave devices.
Instead use a unique value from inside the core device which is already
initialized at this stage.
Fixes: 0d293714ac32 ("RDMA/mlx5: Send events from IB driver about device affiliation state")
Signed-off-by: Patrisious Haddad <phaddad@nvidia.com>
Link: https://lore.kernel.org/r/ac7e66357d963fc68d7a419515180212c96d137d.1697705185.git.leon@kernel.org
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
bnxt_re_handle_cq_async_error()
Clang warns (or errors with CONFIG_WERROR=y):
drivers/infiniband/hw/bnxt_re/main.c:1114:2: error: unannotated fall-through between switch labels [-Werror,-Wimplicit-fallthrough]
1114 | default:
| ^
drivers/infiniband/hw/bnxt_re/main.c:1114:2: note: insert 'break;' to avoid fall-through
1114 | default:
| ^
| break;
1 error generated.
Clang is a little more pedantic than GCC, which does not warn when
falling through to a case that is just break or return. Clang's version
is more in line with the kernel's own stance in deprecated.rst, which
states that all switch/case blocks must end in either break,
fallthrough, continue, goto, or return. Add the missing break to silence
the warning.
Closes: https://github.com/ClangBuiltLinux/linux/issues/1950
Fixes: b02fd3f79ec3 ("RDMA/bnxt_re: Report async events and errors")
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Link: https://lore.kernel.org/r/20231020-bnxt_re-implicit-fallthrough-v1-1-b5c19534a6c9@kernel.org
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
During device init, a struct for HW stats will be allocated. As HW
stats are not supported for VF and HIP08, currently
hns_roce_alloc_hw_port_stats() returns NULL in this case. However,
ib-core considers the returned NULL pointer as memory allocation
failure and returns ENOMEM, eventually leading to the failure of VF
and HIP08 init.
In the case where the driver does not support the .alloc_hw_port_stats()
ops, ib-core will return EOPNOTSUPP and ignore this error code in the
upper layer function. So for VF and HIP08, just don't set the HW stats
ops to ib-core.
Fixes: 5a87279591a1 ("RDMA/hns: Support hns HW stats")
Signed-off-by: Junxian Huang <huangjunxian6@hisilicon.com>
Link: https://lore.kernel.org/r/20231017125239.164455-8-huangjunxian6@hisilicon.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
The num_ports capability of devices should be compared with the
number of port(i.e. the input param "port_num") but not the port
index(i.e. port_num - 1).
Fixes: 5a87279591a1 ("RDMA/hns: Support hns HW stats")
Signed-off-by: Junxian Huang <huangjunxian6@hisilicon.com>
Link: https://lore.kernel.org/r/20231017125239.164455-7-huangjunxian6@hisilicon.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Due to hardware limitations, only DCQCN is supported for UD. Therefore, the
default algorithm for UD is set to DCQCN.
Fixes: f91696f2f053 ("RDMA/hns: Support congestion control type selection according to the FW")
Signed-off-by: Luoyouming <luoyouming@huawei.com>
Signed-off-by: Junxian Huang <huangjunxian6@hisilicon.com>
Link: https://lore.kernel.org/r/20231017125239.164455-6-huangjunxian6@hisilicon.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
SL set by users may exceed the capability of devices. So add check
for this situation.
Fixes: fba429fcf9a5 ("RDMA/hns: Fix missing fields in address vector")
Fixes: 70f92521584f ("RDMA/hns: Use the reserved loopback QPs to free MR before destroying MPT")
Fixes: f0cb411aad23 ("RDMA/hns: Use new interface to modify QP context")
Signed-off-by: Luoyouming <luoyouming@huawei.com>
Signed-off-by: Junxian Huang <huangjunxian6@hisilicon.com>
Link: https://lore.kernel.org/r/20231017125239.164455-5-huangjunxian6@hisilicon.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
The ib_mtu_enum_to_int() and uverbs_attr_get_len() may returns a negative
value. In this case, mixed comparisons of signed and unsigned types will
throw wrong results.
This patch adds judgement for this situation.
Fixes: 30b707886aeb ("RDMA/hns: Support inline data in extented sge space for RC")
Signed-off-by: Chengchang Tang <tangchengchang@huawei.com>
Signed-off-by: Junxian Huang <huangjunxian6@hisilicon.com>
Link: https://lore.kernel.org/r/20231017125239.164455-4-huangjunxian6@hisilicon.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
ucmd in hns_roce_create_qp_common() are not initialized. But it works fine
until new member sdb_addr is added to struct hns_roce_ib_create_qp.
If the user-mode driver uses an old version ABI, then the value of the new
member will be undefined after ib_copy_from_udata().
This patch fixes it by initialize this variable to 0. And the default value
of the new member sdb_addr will be 0 which is invalid.
Fixes: 0425e3e6e0c7 ("RDMA/hns: Support flush cqe for hip08 in kernel space")
Signed-off-by: Chengchang Tang <tangchengchang@huawei.com>
Signed-off-by: Junxian Huang <huangjunxian6@hisilicon.com>
Link: https://lore.kernel.org/r/20231017125239.164455-3-huangjunxian6@hisilicon.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
The current driver will print all asynchronous events. Some of the
print levels are set improperly, e.g. SRQ limit reach and SRQ last
wqe reach, which may also occur during normal operation of the software.
Currently, the information of these event is printed as a warning,
which causes a large amount of printing even during normal use of the
application. As a result, the service performance deteriorates.
This patch fixes the printing storms by modifying the print level.
Fixes: b00a92c8f2ca ("RDMA/hns: Move all prints out of irq handle")
Signed-off-by: Chengchang Tang <tangchengchang@huawei.com>
Signed-off-by: Junxian Huang <huangjunxian6@hisilicon.com>
Link: https://lore.kernel.org/r/20231017125239.164455-2-huangjunxian6@hisilicon.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Convert to using the new inode timestamp accessor functions.
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Link: https://lore.kernel.org/r/20231004185347.80880-5-jlayton@kernel.org
Signed-off-by: Christian Brauner <brauner@kernel.org>
|
|
In the SRQ notification handler, do not report the SRQ_ERROR
in the default event case, as there was no error.
Signed-off-by: Chandramohan Akula <chandramohan.akula@broadcom.com>
Signed-off-by: Selvin Xavier <selvin.xavier@broadcom.com>
Link: https://lore.kernel.org/r/1697049097-31992-4-git-send-email-selvin.xavier@broadcom.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Report QP, SRQ and CQ async events and errors.
Signed-off-by: Chandramohan Akula <chandramohan.akula@broadcom.com>
Signed-off-by: Selvin Xavier <selvin.xavier@broadcom.com>
Link: https://lore.kernel.org/r/1697049097-31992-3-git-send-email-selvin.xavier@broadcom.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Updating the HW structures for the affiliated event and error
reporting. Newly added interface structures will be used in the
followup patch.
Signed-off-by: Chandramohan Akula <chandramohan.akula@broadcom.com>
Signed-off-by: Selvin Xavier <selvin.xavier@broadcom.com>
Link: https://lore.kernel.org/r/1697049097-31992-2-git-send-email-selvin.xavier@broadcom.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Previously when we had a RAW QP, we bound a counter to it when it moved
to INIT state, using the counter context inside RQC.
But when we try to modify that counter later in RTS state we used
modify QP which tries to change the counter inside QPC instead of RQC.
Now we correctly modify the counter set_id inside of RQC instead of QPC
for the RAW QP.
Fixes: d14133dd4161 ("IB/mlx5: Support set qp counter")
Signed-off-by: Patrisious Haddad <phaddad@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Link: https://lore.kernel.org/r/2e5ab6713784a8fe997d19c508187a0dfecf2dfc.1696847964.git.leon@kernel.org
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
https://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux
Leon Romanovsky says:
====================
This PR is collected from
https://lore.kernel.org/all/cover.1695296682.git.leon@kernel.org
This series from Patrisious extends mlx5 to support IPsec packet offload
in multiport devices (MPV, see [1] for more details).
These devices have single flow steering logic and two netdev interfaces,
which require extra logic to manage IPsec configurations as they performed
on netdevs.
[1] https://lore.kernel.org/linux-rdma/20180104152544.28919-1-leon@kernel.org/
* 'mlx5-next' of https://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux:
net/mlx5: Handle IPsec steering upon master unbind/bind
net/mlx5: Configure IPsec steering for ingress RoCEv2 MPV traffic
net/mlx5: Configure IPsec steering for egress RoCEv2 MPV traffic
net/mlx5: Add create alias flow table function to ipsec roce
net/mlx5: Implement alias object allow and create functions
net/mlx5: Add alias flow table bits
net/mlx5: Store devcom pointer inside IPsec RoCE
net/mlx5: Register mlx5e priv to devcom in MPV mode
RDMA/mlx5: Send events from IB driver about device affiliation state
net/mlx5: Introduce ifc bits for migration in a chunk mode
====================
Link: https://lore.kernel.org/r/20231002083832.19746-1-leon@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Add support for reregister MR verb API by doing a de-register
followed by a register MR with the new attributes. Reuse resources
like iwmr handle and HW stag where possible.
Signed-off-by: Sindhu Devale <sindhu.devale@intel.com>
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Link: https://lore.kernel.org/r/20231004151306.228-1-shiraz.saleem@intel.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
This series from Patrisious extends mlx5 to support IPsec packet offload
in multiport devices (MPV, see [1] for more details).
These devices have single flow steering logic and two netdev interfaces,
which require extra logic to manage IPsec configurations as they performed
on netdevs.
Thanks
[1] https://lore.kernel.org/linux-rdma/20180104152544.28919-1-leon@kernel.org/
Link: https://lore.kernel.org/all/20231002083832.19746-1-leon@kernel.org
Signed-of-by: Leon Romanovsky <leon@kernel.org>
* mlx5-next: (576 commits)
net/mlx5: Handle IPsec steering upon master unbind/bind
net/mlx5: Configure IPsec steering for ingress RoCEv2 MPV traffic
net/mlx5: Configure IPsec steering for egress RoCEv2 MPV traffic
net/mlx5: Add create alias flow table function to ipsec roce
net/mlx5: Implement alias object allow and create functions
net/mlx5: Add alias flow table bits
net/mlx5: Store devcom pointer inside IPsec RoCE
net/mlx5: Register mlx5e priv to devcom in MPV mode
RDMA/mlx5: Send events from IB driver about device affiliation state
net/mlx5: Introduce ifc bits for migration in a chunk mode
Linux 6.6-rc3
...
|
|
Prepare for the coming implementation by GCC and Clang of the __counted_by
attribute. Flexible array members annotated with __counted_by can have
their accesses bounds-checked at run-time checking via CONFIG_UBSAN_BOUNDS
(for array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family
functions).
As found with Coccinelle[1], add __counted_by for struct tid_rb_node.
[1] https://github.com/kees/kernel-tools/blob/trunk/coccinelle/examples/counted_by.cocci
Cc: Dennis Dalessandro <dennis.dalessandro@cornelisnetworks.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Leon Romanovsky <leon@kernel.org>
Cc: linux-rdma@vger.kernel.org
Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20230929180431.3005464-7-keescook@chromium.org
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Prepare for the coming implementation by GCC and Clang of the __counted_by
attribute. Flexible array members annotated with __counted_by can have
their accesses bounds-checked at run-time checking via CONFIG_UBSAN_BOUNDS
(for array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family
functions).
As found with Coccinelle[1], add __counted_by for struct mthca_icm_table.
[1] https://github.com/kees/kernel-tools/blob/trunk/coccinelle/examples/counted_by.cocci
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Leon Romanovsky <leon@kernel.org>
Cc: linux-rdma@vger.kernel.org
Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20230929180431.3005464-6-keescook@chromium.org
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Prepare for the coming implementation by GCC and Clang of the __counted_by
attribute. Flexible array members annotated with __counted_by can have
their accesses bounds-checked at run-time checking via CONFIG_UBSAN_BOUNDS
(for array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family
functions).
As found with Coccinelle[1], add __counted_by for struct usnic_uiom_chunk.
[1] https://github.com/kees/kernel-tools/blob/trunk/coccinelle/examples/counted_by.cocci
Cc: Christian Benvenuti <benve@cisco.com>
Cc: Nelson Escobar <neescoba@cisco.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Leon Romanovsky <leon@kernel.org>
Cc: linux-rdma@vger.kernel.org
Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20230929180431.3005464-3-keescook@chromium.org
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
During execution of mlx5_mkey_cache_cleanup(), there is a guarantee
that MR are not registered and/or destroyed. It means that we don't
need newly introduced cache disable flag.
Fixes: 374012b00457 ("RDMA/mlx5: Fix mkey cache possible deadlock on cleanup")
Link: https://lore.kernel.org/r/c7e9c9f98c8ae4a7413d97d9349b29f5b0a23dbe.1695921626.git.leon@kernel.org
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
|
|
Compared with normal doorbell, using record doorbell can shorten the
process of ringing the doorbell and reduce the latency.
Add a flag HNS_ROCE_CAP_FLAG_SRQ_RECORD_DB to allow FW to
enable/disable SRQ record doorbell.
If the flag above is set, allocate the dma buffer for SRQ record
doorbell and write the buffer address into SRQC during SRQ creation.
For userspace SRQ, add a flag HNS_ROCE_RSP_SRQ_CAP_RECORD_DB to notify
userspace whether the SRQ record doorbell is enabled.
Signed-off-by: Yangyang Li <liyangyang20@huawei.com>
Signed-off-by: Junxian Huang <huangjunxian6@hisilicon.com>
Link: https://lore.kernel.org/r/20230926130026.583088-1-huangjunxian6@hisilicon.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Send blocking events from IB driver whenever the device is done being
affiliated or if it is removed from an affiliation.
This is useful since now the EN driver can register to those event and
know when a device is affiliated or not.
Signed-off-by: Patrisious Haddad <phaddad@nvidia.com>
Reviewed-by: Mark Bloch <mbloch@nvidia.com>
Link: https://lore.kernel.org/r/a7491c3e483cfd8d962f5f75b9a25f253043384a.1695296682.git.leon@kernel.org
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Adjust mlx5 function which maps the speed rate from IB spec values
to internal driver values to be able to handle speeds up to 800Gb.
Signed-off-by: Patrisious Haddad <phaddad@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Link: https://lore.kernel.org/r/301c803d8486b0df8aefad3fb3cc10dc58671985.1695204156.git.leon@kernel.org
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Rename 400G_8X speed to comply to naming convention.
Signed-off-by: Patrisious Haddad <phaddad@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Link: https://lore.kernel.org/r/ac98447cac8379a43fbdb36d56e5fb2b741a97ff.1695204156.git.leon@kernel.org
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Add a check for 800G_8X speed when querying PTYS and report it back
correctly when needed.
Signed-off-by: Patrisious Haddad <phaddad@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Link: https://lore.kernel.org/r/26fd0b6e1fac071c3eb779657bb3d8ba47f47c4f.1695204156.git.leon@kernel.org
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Under MAD query port, Report NDR speed when NDR is supported in the port
capability mask.
Signed-off-by: Or Har-Toov <ohartoov@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Link: https://lore.kernel.org/r/d30bdec2a66a8a2edd1d84ee61453c58cf346b43.1695204156.git.leon@kernel.org
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Currently, mkeys are managed via xarray. This implementation leads to
a degradation in cases many MRs are unregistered in parallel, due to xarray
internal implementation, for example: deregistration 1M MRs via 64 threads
is taking ~15% more time[1].
Hence, implement mkeys management via LIFO queue, which solved the
degradation.
[1]
2.8us in kernel v5.19 compare to 3.2us in kernel v6.4
Signed-off-by: Shay Drory <shayd@nvidia.com>
Link: https://lore.kernel.org/r/fde3d4cfab0f32f0ccb231cd113298256e1502c5.1695283384.git.leon@kernel.org
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
Fix the deadlock by refactoring the MR cache cleanup flow to flush the
workqueue without holding the rb_lock.
This adds a race between cache cleanup and creation of new entries which
we solve by denied creation of new entries after cache cleanup started.
Lockdep:
WARNING: possible circular locking dependency detected
[ 2785.326074 ] 6.2.0-rc6_for_upstream_debug_2023_01_31_14_02 #1 Not tainted
[ 2785.339778 ] ------------------------------------------------------
[ 2785.340848 ] devlink/53872 is trying to acquire lock:
[ 2785.341701 ] ffff888124f8c0c8 ((work_completion)(&(&ent->dwork)->work)){+.+.}-{0:0}, at: __flush_work+0xc8/0x900
[ 2785.343403 ]
[ 2785.343403 ] but task is already holding lock:
[ 2785.344464 ] ffff88817e8f1260 (&dev->cache.rb_lock){+.+.}-{3:3}, at: mlx5_mkey_cache_cleanup+0x77/0x250 [mlx5_ib]
[ 2785.346273 ]
[ 2785.346273 ] which lock already depends on the new lock.
[ 2785.346273 ]
[ 2785.347720 ]
[ 2785.347720 ] the existing dependency chain (in reverse order) is:
[ 2785.349003 ]
[ 2785.349003 ] -> #1 (&dev->cache.rb_lock){+.+.}-{3:3}:
[ 2785.350160 ] __mutex_lock+0x14c/0x15c0
[ 2785.350962 ] delayed_cache_work_func+0x2d1/0x610 [mlx5_ib]
[ 2785.352044 ] process_one_work+0x7c2/0x1310
[ 2785.352879 ] worker_thread+0x59d/0xec0
[ 2785.353636 ] kthread+0x28f/0x330
[ 2785.354370 ] ret_from_fork+0x1f/0x30
[ 2785.355135 ]
[ 2785.355135 ] -> #0 ((work_completion)(&(&ent->dwork)->work)){+.+.}-{0:0}:
[ 2785.356515 ] __lock_acquire+0x2d8a/0x5fe0
[ 2785.357349 ] lock_acquire+0x1c1/0x540
[ 2785.358121 ] __flush_work+0xe8/0x900
[ 2785.358852 ] __cancel_work_timer+0x2c7/0x3f0
[ 2785.359711 ] mlx5_mkey_cache_cleanup+0xfb/0x250 [mlx5_ib]
[ 2785.360781 ] mlx5_ib_stage_pre_ib_reg_umr_cleanup+0x16/0x30 [mlx5_ib]
[ 2785.361969 ] __mlx5_ib_remove+0x68/0x120 [mlx5_ib]
[ 2785.362960 ] mlx5r_remove+0x63/0x80 [mlx5_ib]
[ 2785.363870 ] auxiliary_bus_remove+0x52/0x70
[ 2785.364715 ] device_release_driver_internal+0x3c1/0x600
[ 2785.365695 ] bus_remove_device+0x2a5/0x560
[ 2785.366525 ] device_del+0x492/0xb80
[ 2785.367276 ] mlx5_detach_device+0x1a9/0x360 [mlx5_core]
[ 2785.368615 ] mlx5_unload_one_devl_locked+0x5a/0x110 [mlx5_core]
[ 2785.369934 ] mlx5_devlink_reload_down+0x292/0x580 [mlx5_core]
[ 2785.371292 ] devlink_reload+0x439/0x590
[ 2785.372075 ] devlink_nl_cmd_reload+0xaef/0xff0
[ 2785.372973 ] genl_family_rcv_msg_doit.isra.0+0x1bd/0x290
[ 2785.374011 ] genl_rcv_msg+0x3ca/0x6c0
[ 2785.374798 ] netlink_rcv_skb+0x12c/0x360
[ 2785.375612 ] genl_rcv+0x24/0x40
[ 2785.376295 ] netlink_unicast+0x438/0x710
[ 2785.377121 ] netlink_sendmsg+0x7a1/0xca0
[ 2785.377926 ] sock_sendmsg+0xc5/0x190
[ 2785.378668 ] __sys_sendto+0x1bc/0x290
[ 2785.379440 ] __x64_sys_sendto+0xdc/0x1b0
[ 2785.380255 ] do_syscall_64+0x3d/0x90
[ 2785.381031 ] entry_SYSCALL_64_after_hwframe+0x46/0xb0
[ 2785.381967 ]
[ 2785.381967 ] other info that might help us debug this:
[ 2785.381967 ]
[ 2785.383448 ] Possible unsafe locking scenario:
[ 2785.383448 ]
[ 2785.384544 ] CPU0 CPU1
[ 2785.385383 ] ---- ----
[ 2785.386193 ] lock(&dev->cache.rb_lock);
[ 2785.386940 ] lock((work_completion)(&(&ent->dwork)->work));
[ 2785.388327 ] lock(&dev->cache.rb_lock);
[ 2785.389425 ] lock((work_completion)(&(&ent->dwork)->work));
[ 2785.390414 ]
[ 2785.390414 ] *** DEADLOCK ***
[ 2785.390414 ]
[ 2785.391579 ] 6 locks held by devlink/53872:
[ 2785.392341 ] #0: ffffffff84c17a50 (cb_lock){++++}-{3:3}, at: genl_rcv+0x15/0x40
[ 2785.393630 ] #1: ffff888142280218 (&devlink->lock_key){+.+.}-{3:3}, at: devlink_get_from_attrs_lock+0x12d/0x2d0
[ 2785.395324 ] #2: ffff8881422d3c38 (&dev->lock_key){+.+.}-{3:3}, at: mlx5_unload_one_devl_locked+0x4a/0x110 [mlx5_core]
[ 2785.397322 ] #3: ffffffffa0e59068 (mlx5_intf_mutex){+.+.}-{3:3}, at: mlx5_detach_device+0x60/0x360 [mlx5_core]
[ 2785.399231 ] #4: ffff88810e3cb0e8 (&dev->mutex){....}-{3:3}, at: device_release_driver_internal+0x8d/0x600
[ 2785.400864 ] #5: ffff88817e8f1260 (&dev->cache.rb_lock){+.+.}-{3:3}, at: mlx5_mkey_cache_cleanup+0x77/0x250 [mlx5_ib]
Fixes: b95845178328 ("RDMA/mlx5: Change the cache structure to an RB-tree")
Signed-off-by: Shay Drory <shayd@nvidia.com>
Signed-off-by: Michael Guralnik <michaelgur@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
|