Now that req.task and comp.task are merged it is no longer necessary to
call save_state() before calling rxe_xmit_pkt() and rollback_state() if
rxe_xmit_pkt() fails. This was done originally to prevent races between
rxe_completer() and rxe_requester() which now cannot happen.
Link: https://lore.kernel.org/r/20240329145513.35381-8-rpearsonhpe@gmail.com
Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Currently the rxe driver has three work queue tasks per qp. These are the
req.task, comp.task and resp.task which call rxe_requester(),
rxe_completer() and rxe_responder() respectively directly or on work
queues. Each of these subroutines checks to see if there is work to be
performed on the send queue or on the response packet queue or the request
packet queue and will run until there is no work remaining or yield the
cpu and reschedule itself until there is no work remaining.
This commit combines the req.task and comp.task into a single send.task
and renames the resp.task to the recv.task. The combined send.task calls
rxe_requester() and rxe_completer() serially and continues until all work
on both the send queue and the response packet queue are done.
In various benchmarks the performance is either improved or left the
same. At high scale there is a significant reduction in the load on the
cpu.
This is the first step in combining these two tasks. Once they are
serialized cross rescheduling of req.task and comp.task can be more
efficiently handled by just letting the send.task continue to run. This
will be done in the next several patches.
Link: https://lore.kernel.org/r/20240329145513.35381-7-rpearsonhpe@gmail.com
Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
In rxe_post_send_kernel() if the qp is in the error state after posting
the work requests the rxe_completer() task is scheduled.
But, the only way to move the qp into the error state is to call
rxe_qp_error() which also schedules the rxe_completer() task to drain the
queues. Calling it a second time has no effect. This commit removes the
redundant call.
Link: https://lore.kernel.org/r/20240329145513.35381-6-rpearsonhpe@gmail.com
Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
A previous commit incorrectly added an 'if(!err)' before scheduling the
requester task in rxe_post_send_kernel(). But if there were send wrs
successfully added to the send queue before a bad wr they might never get
executed.
This commit fixes this by scheduling the requester task if any wqes were
successfully posted in rxe_post_send_kernel() in rxe_verbs.c.
Link: https://lore.kernel.org/r/20240329145513.35381-5-rpearsonhpe@gmail.com
Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
Fixes: 5bf944f241 ("RDMA/rxe: Add error messages")
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
In rxe_comp_queue_pkt() an incoming response packet skb is enqueued to the
resp_pkts queue and then a decision is made whether to run the completer
task inline or schedule it. Finally the skb is dereferenced to bump a 'hw'
performance counter. This is wrong because if the completer task is
already running in a separate thread it may have already processed the skb
and freed it which can cause a seg fault. This has been observed
infrequently in testing at high scale.
This patch fixes this by changing the order of enqueuing the packet until
after the counter is accessed.
Link: https://lore.kernel.org/r/20240329145513.35381-4-rpearsonhpe@gmail.com
Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
Fixes: 0b1e5b99a4 ("IB/rxe: Add port protocol stats")
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The existing behavior of ib_umad, which maintains received MAD
packets in an unbounded list, poses a risk of uncontrolled growth.
As user-space applications extract packets from this list, the rate
of extraction may not match the rate of incoming packets, leading
to potential list overflow.
To address this, we introduce a limit to the size of the list. After
considering typical scenarios, such as OpenSM processing, which can
handle approximately 100k packets per second, and the 1-second retry
timeout for most packets, we set the list size limit to 200k. Packets
received beyond this limit are dropped, assuming they are likely timed
out by the time they are handled by user-space.
Notably, packets queued on the receive list due to reasons like
timed-out sends are preserved even when the list is full.
Signed-off-by: Michael Guralnik <michaelgur@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Link: https://lore.kernel.org/r/7197cb58a7d9e78399008f25036205ceab07fbd5.1713268818.git.leon@kernel.org
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Erick Archer says:
====================
mana: Add flex array to struct mana_cfg_rx_steer_req_v2 (part)
The "struct mana_cfg_rx_steer_req_v2" uses a dynamically sized set of
trailing elements. Specifically, it uses a "mana_handle_t" array. So,
use the preferred way in the kernel declaring a flexible array [1].
At the same time, 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 via
CONFIG_UBSAN_BOUNDS (for array indexing) and CONFIG_FORTIFY_SOURCE (for
strcpy/memcpy-family functions).
Also, avoid the open-coded arithmetic in the memory allocator functions
[2] using the "struct_size" macro.
Moreover, use the "offsetof" helper to get the indirect table offset
instead of the "sizeof" operator and avoid the open-coded arithmetic in
pointers using the new flex member. This new structure member also allow
us to remove the "req_indir_tab" variable since it is no longer needed.
Now, it is also possible to use the "flex_array_size" helper to compute
the size of these trailing elements in the "memcpy" function.
Specifically, the first commit adds the flex member and the patches 2 and
3 refactor the consumers of the "struct mana_cfg_rx_steer_req_v2".
This code was detected with the help of Coccinelle, and audited and
modified manually. The Coccinelle script used to detect this code pattern
is the following:
virtual report
@rule1@
type t1;
type t2;
identifier i0;
identifier i1;
identifier i2;
identifier ALLOC =~ "kmalloc|kzalloc|kmalloc_node|kzalloc_node|vmalloc|vzalloc|kvmalloc|kvzalloc";
position p1;
@@
i0 = sizeof(t1) + sizeof(t2) * i1;
...
i2 = ALLOC@p1(..., i0, ...);
@script:python depends on report@
p1 << rule1.p1;
@@
msg = "WARNING: verify allocation on line %s" % (p1[0].line)
coccilib.report.print_report(p1[0],msg)
Link: https://www.kernel.org/doc/html/next/process/deprecated.html#zero-length-and-one-element-arrays [1]
Link: https://www.kernel.org/doc/html/next/process/deprecated.html#open-coded-arithmetic-in-allocator-arguments [2]
v1: https://lore.kernel.org/linux-hardening/AS8PR02MB7237974EF1B9BAFA618166C38B382@AS8PR02MB7237.eurprd02.prod.outlook.com/
v2: https://lore.kernel.org/linux-hardening/AS8PR02MB723729C5A63F24C312FC9CD18B3F2@AS8PR02MB7237.eurprd02.prod.outlook.com/
====================
Link: https://lore.kernel.org/r/AS8PR02MB72374BD1B23728F2E3C3B1A18B022@AS8PR02MB7237.eurprd02.prod.outlook.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In the function __rxe_add_to_pool, the function xa_alloc_cyclic is
called. The return value of the function xa_alloc_cyclic is as below:
"
Return: 0 if the allocation succeeded without wrapping. 1 if the
allocation succeeded after wrapping, -ENOMEM if memory could not be
allocated or -EBUSY if there are no free entries in @limit.
"
But now the function __rxe_add_to_pool only returns -EINVAL. All the
returned error value should be returned to the caller.
Signed-off-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Link: https://lore.kernel.org/r/20240408142142.792413-1-yanjun.zhu@linux.dev
Signed-off-by: Leon Romanovsky <leon@kernel.org>
The "struct mana_cfg_rx_steer_req_v2" uses a dynamically sized set of
trailing elements. Specifically, it uses a "mana_handle_t" array. So,
use the preferred way in the kernel declaring a flexible array [1].
At the same time, 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 via
CONFIG_UBSAN_BOUNDS (for array indexing) and CONFIG_FORTIFY_SOURCE (for
strcpy/memcpy-family functions).
Also, avoid the open-coded arithmetic in the memory allocator functions
[2] using the "struct_size" macro.
Moreover, use the "offsetof" helper to get the indirect table offset
instead of the "sizeof" operator and avoid the open-coded arithmetic in
pointers using the new flex member. This new structure member also allow
us to remove the "req_indir_tab" variable since it is no longer needed.
Now, it is also possible to use the "flex_array_size" helper to compute
the size of these trailing elements in the "memcpy" function.
Specifically, the first commit adds the flex member and the patches 2
and 3 refactor the consumers of the "struct mana_cfg_rx_steer_req_v2".
This code was detected with the help of Coccinelle, and audited and
modified manually. The Coccinelle script used to detect this code
pattern is the following:
virtual report
@rule1@
type t1;
type t2;
identifier i0;
identifier i1;
identifier i2;
identifier ALLOC =~
"kmalloc|kzalloc|kmalloc_node|kzalloc_node|vmalloc|vzalloc|kvmalloc|kvzalloc";
position p1;
@@
i0 = sizeof(t1) + sizeof(t2) * i1;
...
i2 = ALLOC@p1(..., i0, ...);
@script:python depends on report@
p1 << rule1.p1;
@@
msg = "WARNING: verify allocation on line %s" % (p1[0].line)
coccilib.report.print_report(p1[0],msg)
[1] https://www.kernel.org/doc/html/next/process/deprecated.html#zero-length-and-one-element-arrays
[2] https://www.kernel.org/doc/html/next/process/deprecated.html#open-coded-arithmetic-in-allocator-arguments
Link: https://lore.kernel.org/all/AS8PR02MB72374BD1B23728F2E3C3B1A18B022@AS8PR02MB7237.eurprd02.prod.outlook.com
Signed-off-by: Erick Archer <erick.archer@outlook.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
* mana-ib-flex:
net: mana: Avoid open coded arithmetic
RDMA/mana_ib: Prefer struct_size over open coded arithmetic
net: mana: Add flex array to struct mana_cfg_rx_steer_req_v2
This is an effort to get rid of all multiplications from allocation
functions in order to prevent integer overflows [1][2].
As the "req" variable is a pointer to "struct mana_cfg_rx_steer_req_v2"
and this structure ends in a flexible array:
struct mana_cfg_rx_steer_req_v2 {
[...]
mana_handle_t indir_tab[] __counted_by(num_indir_entries);
};
the preferred way in the kernel is to use the struct_size() helper to
do the arithmetic instead of the calculation "size + size * count" in
the kzalloc() function.
Moreover, use the "offsetof" helper to get the indirect table offset
instead of the "sizeof" operator and avoid the open-coded arithmetic in
pointers using the new flex member. This new structure member also allow
us to remove the "req_indir_tab" variable since it is no longer needed.
This way, the code is more readable and safer.
This code was detected with the help of Coccinelle, and audited and
modified manually.
Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#open-coded-arithmetic-in-allocator-arguments [1]
Link: https://github.com/KSPP/linux/issues/160 [2]
Signed-off-by: Erick Archer <erick.archer@outlook.com>
Link: https://lore.kernel.org/r/AS8PR02MB72375EB06EE1A84A67BE722E8B022@AS8PR02MB7237.eurprd02.prod.outlook.com
Reviewed-by: Long Li <longli@microsoft.com>
Reviewed-by: Justin Stitt <justinstitt@google.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Add support for DSCP configuration. For DSCP, get dscp-prio mapping
via hns3 nic driver api .get_dscp_prio() and fill the SL (in WQE for
UD or in QPC for RC) with the priority value. The prio-tc mapping is
configured to HW by hns3 nic driver. HW will select a corresponding
TC according to SL and the prio-tc mapping.
Signed-off-by: Junxian Huang <huangjunxian6@hisilicon.com>
Link: https://lore.kernel.org/r/20240315093551.1650088-1-huangjunxian6@hisilicon.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Add a "scope" parameter to ip_route_output() so that callers don't have
to override the tos parameter with the RTO_ONLINK flag if they want a
local scope.
This will allow converting flowi4_tos to dscp_t in the future, thus
allowing static analysers to flag invalid interactions between
"tos" (the DSCP bits) and ECN.
Only three users ask for local scope (bonding, arp and atm). The others
continue to use RT_SCOPE_UNIVERSE. While there, add a comment to warn
users about the limitations of ip_route_output().
Signed-off-by: Guillaume Nault <gnault@redhat.com>
Acked-by: Leon Romanovsky <leonro@nvidia.com> # infiniband
Signed-off-by: David S. Miller <davem@davemloft.net>
struct rdma_restrack_entry's kern_name was set to KBUILD_MODNAME
in ib_create_cq(), while if the module exited but forgot del this
rdma_restrack_entry, it would cause a invalid address access in
rdma_restrack_clean() when print the owner of this rdma_restrack_entry.
These code is used to help find one forgotten PD release in one of the
ULPs. But it is not needed anymore, so delete them.
Signed-off-by: Wenchao Hao <haowenchao2@huawei.com>
Link: https://lore.kernel.org/r/20240318092320.1215235-1-haowenchao2@huawei.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Pull rdma updates from Jason Gunthorpe:
"Very small update this cycle:
- Minor code improvements in fi, rxe, ipoib, mana, cxgb4, mlx5,
irdma, rxe, rtrs, mana
- Simplify the hns hem mechanism
- Fix EFA's MSI-X allocation in resource constrained configurations
- Fix a KASN splat in srpt
- Narrow hns's congestion control selection to QPs granularity and
allow userspace to select it
- Solve a parallel module loading race between the CM module and a
driver module
- Flexible array cleanup
- Dump hns's SCC Conext to 'rdma res' for debugging
- Make mana build page lists for HW objects that require a 0 offset
correctly
- Stuck CM ID debugging"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma: (29 commits)
RDMA/cm: add timeout to cm_destroy_id wait
RDMA/mana_ib: Use virtual address in dma regions for MRs
RDMA/mana_ib: Fix bug in creation of dma regions
RDMA/hns: Append SCC context to the raw dump of QPC
RDMA/uverbs: Avoid -Wflex-array-member-not-at-end warnings
RDMA/hns: Support userspace configuring congestion control algorithm with QP granularity
RDMA/rtrs-clt: Check strnlen return len in sysfs mpath_policy_store()
RDMA/uverbs: Remove flexible arrays from struct *_filter
RDMA/device: Fix a race between mad_client and cm_client init
RDMA/hns: Fix mis-modifying default congestion control algorithm
RDMA/rxe: Remove unused 'iova' parameter from rxe_mr_init_user
RDMA/srpt: Do not register event handler until srpt device is fully setup
RDMA/irdma: Remove duplicate assignment
RDMA/efa: Limit EQs to available MSI-X vectors
RDMA/mlx5: Delete unused mlx5_ib_copy_pas prototype
RDMA/cxgb4: Delete unused c4iw_ep_redirect prototype
RDMA/mana_ib: Introduce mana_ib_install_cq_cb helper function
RDMA/mana_ib: Introduce mana_ib_get_netdev helper function
RDMA/mana_ib: Introduce mdev_to_gc helper function
RDMA/hns: Simplify 'struct hns_roce_hem' allocation
...
Add timeout to cm_destroy_id, so that userspace can trigger any data
collection that would help in analyzing the cause of delay in destroying
the cm_id.
New noinline function helps dtrace/ebpf programs to hook on to it.
Existing functionality isn't changed except triggering a probe-able new
function at every timeout interval.
We have seen cases where CM messages stuck with MAD layer (either due to
software bug or faulty HCA), leading to cm_id getting stuck in the
following call stack. This patch helps in resolving such issues faster.
kernel: ... INFO: task XXXX:56778 blocked for more than 120 seconds.
...
Call Trace:
__schedule+0x2bc/0x895
schedule+0x36/0x7c
schedule_timeout+0x1f6/0x31f
? __slab_free+0x19c/0x2ba
wait_for_completion+0x12b/0x18a
? wake_up_q+0x80/0x73
cm_destroy_id+0x345/0x610 [ib_cm]
ib_destroy_cm_id+0x10/0x20 [ib_cm]
rdma_destroy_id+0xa8/0x300 [rdma_cm]
ucma_destroy_id+0x13e/0x190 [rdma_ucm]
ucma_write+0xe0/0x160 [rdma_ucm]
__vfs_write+0x3a/0x16d
vfs_write+0xb2/0x1a1
? syscall_trace_enter+0x1ce/0x2b8
SyS_write+0x5c/0xd3
do_syscall_64+0x79/0x1b9
entry_SYSCALL_64_after_hwframe+0x16d/0x0
Signed-off-by: Manjunath Patil <manjunath.b.patil@oracle.com>
Link: https://lore.kernel.org/r/20240309063323.458102-1-manjunath.b.patil@oracle.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Introduce mana_ib_create_dma_region() to create dma regions with iova
for MRs. It allows creating MRs with any page offset. Previously,
only page-aligned addresses worked.
For dma regions that must have a zero dma offset (e.g., for queues),
mana_ib_create_zero_offset_dma_region() is added.
To get the zero offset, ib_umem_find_best_pgoff() is used with zero
pgoff_bitmask.
Fixes: 0266a17763 ("RDMA/mana_ib: Add a driver for Microsoft Azure Network Adapter")
Signed-off-by: Konstantin Taranov <kotaranov@microsoft.com>
Link: https://lore.kernel.org/r/1709560361-26393-3-git-send-email-kotaranov@linux.microsoft.com
Reviewed-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
SCCC (SCC Context) is a context with QP granularity that contains
information about congestion control. Dump SCCC and QPC together
to improve troubleshooting.
When dumping raw QPC with rdmatool, there will be a total of 576 bytes
data output, where the first 512 bytes is QPC and the last 64 bytes is
SCCC. When congestion control is disabled, the 64 byte SCCC will be all 0.
Example:
$rdma res show qp -jpr
[ {
"ifindex": 0,
"ifname": "hns_0",
"data": [ 67,0,0,0... 512bytes
4,0,2... 64bytes]
},...
} ]
Signed-off-by: wenglianfa <wenglianfa@huawei.com>
Signed-off-by: Junxian Huang <huangjunxian6@hisilicon.com>
Link: https://lore.kernel.org/r/20240305055257.823513-1-huangjunxian6@hisilicon.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
