IF YOU WOULD LIKE TO GET AN ACCOUNT, please write an
email to Administrator. User accounts are meant only to access repo
and report issues and/or generate pull requests.
This is a purpose-specific Git hosting for
BaseALT
projects. Thank you for your understanding!
Только зарегистрированные пользователи имеют доступ к сервису!
Для получения аккаунта, обратитесь к администратору.
Current taprio code operates on a very simplistic (and incorrect)
assumption: that egress scheduling for a traffic class can only take
place for the duration of the current interval, or i.o.w., it assumes
that at the end of each schedule entry, there is a "gate close" event
for all traffic classes.
As an example, traffic sent with the schedule below will be jumpy, even
though all 8 TC gates are open, so there is absolutely no "gate close"
event (effectively a transition from BIT(tc)==1 to BIT(tc)==0 in
consecutive schedule entries):
tc qdisc replace dev veth0 parent root taprio \
num_tc 2 \
map 0 1 \
queues 1@0 1@1 \
base-time 0 \
sched-entry S 0xff 4000000000 \
clockid CLOCK_TAI \
flags 0x0
This qdisc simply does not have what it takes in terms of logic to
*actually* compute the durations of traffic classes. Also, it does not
recognize the need to use this information on a per-traffic-class basis:
it always looks at entry->interval and entry->close_time.
This change proposes that each schedule entry has an array called
tc_gate_duration[tc]. This holds the information: "for how long will
this traffic class gate remain open, starting from *this* schedule
entry". If the traffic class gate is always open, that value is equal to
the cycle time of the schedule.
We'll also need to keep track, for the purpose of queueMaxSDU[tc]
calculation, what is the maximum time duration for a traffic class
having an open gate. This gives us directly what is the maximum sized
packet that this traffic class will have to accept. For everything else
it has to qdisc_drop() it in qdisc_enqueue().
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Kurt Kanzenbach <kurt@linutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Current taprio software implementation is haunted by the shadow of the
igb/igc hardware model. It iterates over child qdiscs in increasing
order of TXQ index, therefore giving higher xmit priority to TXQ 0 and
lower to TXQ N. According to discussions with Vinicius, that is the
default (perhaps even unchangeable) prioritization scheme used for the
NICs that taprio was first written for (igb, igc), and we have a case of
two bugs canceling out, resulting in a functional setup on igb/igc, but
a less sane one on other NICs.
To the best of my understanding, taprio should prioritize based on the
traffic class, so it should really dequeue starting with the highest
traffic class and going down from there. We get to the TXQ using the
tc_to_txq[] netdev property.
TXQs within the same TC have the same (strict) priority, so we should
pick from them as fairly as we can. We can achieve that by implementing
something very similar to q->curband from multiq_dequeue().
Since igb/igc really do have TXQ 0 of higher hardware priority than
TXQ 1 etc, we need to preserve the behavior for them as well. We really
have no choice, because in txtime-assist mode, taprio is essentially a
software scheduler towards offloaded child tc-etf qdiscs, so the TXQ
selection really does matter (not all igb TXQs support ETF/SO_TXTIME,
says Kurt Kanzenbach).
To preserve the behavior, we need a capability bit so that taprio can
determine if it's running on igb/igc, or on something else. Because igb
doesn't offload taprio at all, we can't piggyback on the
qdisc_offload_query_caps() call from taprio_enable_offload(), but
instead we need a separate call which is also made for software
scheduling.
Introduce two static keys to minimize the performance penalty on systems
which only have igb/igc NICs, and on systems which only have other NICs.
For mixed systems, taprio will have to dynamically check whether to
dequeue using one prioritization algorithm or using the other.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Simplify taprio_dequeue_from_txq() by noticing that we can goto one call
earlier than the previous skb_found label. This is possible because
we've unified the treatment of the child->ops->dequeue(child) return
call, we always try other TXQs now, instead of abandoning the root
dequeue completely if we failed in the peek() case.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Kurt Kanzenbach <kurt@linutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Future changes will refactor the TXQ selection procedure, and a lot of
stuff will become messy, the indentation of the bulk of the dequeue
procedure would increase, etc.
Break out the bulk of the function into a new one, which knows the TXQ
(child qdisc) we should perform a dequeue from.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Kurt Kanzenbach <kurt@linutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
This changes the handling of an unlikely condition to not stop dequeuing
if taprio failed to dequeue the peeked skb in taprio_dequeue().
I've no idea when this can happen, but the only side effect seems to be
that the atomic_sub_return() call right above will have consumed some
budget. This isn't a big deal, since either that made us remain without
any budget (and therefore, we'd exit on the next peeked skb anyway), or
we could send some packets from other TXQs.
I'm making this change because in a future patch I'll be refactoring the
dequeue procedure to simplify it, and this corner case will have to go
away.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Kurt Kanzenbach <kurt@linutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
There isn't any code in the network stack which calls taprio_peek().
We only see qdisc->ops->peek() being called on child qdiscs of other
classful qdiscs, never from the generic qdisc code. Whereas taprio is
never a child qdisc, it is always root.
This snippet of a comment from qdisc_peek_dequeued() seems to confirm:
/* we can reuse ->gso_skb because peek isn't called for root qdiscs */
Since I've been known to be wrong many times though, I'm not completely
removing it, but leaving a stub function in place which emits a warning.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Kurt Kanzenbach <kurt@linutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Horatiu Vultur says:
====================
net: micrel: Add support for lan8841 PHY
Add support for lan8841 PHY.
The first patch add the support for lan8841 PHY which can run at
10/100/1000Mbit. It also has support for other features, but they are not
added in this series.
The second patch updates the documentation for the dt-bindings which is
similar to the ksz9131.
v3->v4:
- add space between defines and function names
- inside lan8841_config_init use only ret variable
v2->v3:
- reuse ksz9131_config_init
- allow only open-drain configuration
- change from single patch to a patch series
v1->v2:
- Remove hardcoded values
- Fix typo in commit message
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
The lan8841 has the same bindings as ksz9131, so just reuse the entire
section of ksz9131.
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The LAN8841 is completely integrated triple-speed (10BASE-T/ 100BASE-TX/
1000BASE-T) Ethernet physical layer transceivers for transmission and
reception of data on standard CAT-5, as well as CAT-5e and CAT-6,
unshielded twisted pair (UTP) cables.
The LAN8841 offers the industry-standard GMII/MII as well as the RGMII.
Some of the features of the PHY are:
- Wake on LAN
- Auto-MDIX
- IEEE 1588-2008 (V2)
- LinkMD Capable diagnosis
Currently the patch offers support only for link configuration.
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add flower filter packet statistics. This will just read the TCAM
counter of the rule, which mention how many packages were hit by this
rule.
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Tony Nguyen says:
====================
ice: various virtualization cleanups
Jacob Keller says:
This series contains a variety of refactors and cleanups in the VF code for
the ice driver. Its primary focus is cleanup and simplification of the VF
operations and addition of a few new operations that will be required by
Scalable IOV, as well as some other refactors needed for the handling of VF
subfunctions.
* '100GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/next-queue:
ice: remove unnecessary virtchnl_ether_addr struct use
ice: introduce .irq_close VF operation
ice: introduce clear_reset_state operation
ice: convert vf_ops .vsi_rebuild to .create_vsi
ice: introduce ice_vf_init_host_cfg function
ice: add a function to initialize vf entry
ice: Pull common tasks into ice_vf_post_vsi_rebuild
ice: move ice_vf_vsi_release into ice_vf_lib.c
ice: move vsi_type assignment from ice_vsi_alloc to ice_vsi_cfg
ice: refactor VSI setup to use parameter structure
ice: drop unnecessary VF parameter from several VSI functions
ice: fix function comment referring to ice_vsi_alloc
ice: Add more usage of existing function ice_get_vf_vsi(vf)
====================
Link: https://lore.kernel.org/r/20230206214813.20107-1-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The callback devlink_nl_cmd_health_reporter_diagnose_doit() miss
devlink_fmsg_free(), which leads to memory leak.
Fix it by adding devlink_fmsg_free().
Fixes: e994a75fb7 ("devlink: remove reporter reference counting")
Signed-off-by: Moshe Shemesh <moshe@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Link: https://lore.kernel.org/r/1675698976-45993-1-git-send-email-moshe@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Move the nfp_net_get_port_mac_by_hwinfo() check to ahead in the
get/set_eeprom() functions to in order to check for a VF netdev, which
this function does not support.
It is debatable if this is a fix or an enhancement, and we have chosen
to go for the latter. It does address a problem introduced by
commit 74b4f1739d ("nfp: flower: change get/set_eeprom logic and enable for flower reps").
However, the ethtool->len == 0 check avoids the problem manifesting as a
run-time bug (NULL pointer dereference of app).
Signed-off-by: James Hershaw <james.hershaw@corigine.com>
Reviewed-by: Louis Peens <louis.peens@corigine.com>
Signed-off-by: Simon Horman <simon.horman@corigine.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Link: https://lore.kernel.org/r/20230206154836.2803995-1-simon.horman@corigine.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Petr Machata says:
====================
mlxsw: Misc devlink changes
This patchset adjusts mlxsw to recent devlink changes in net-next.
Patch #1 removes a devl_param_driverinit_value_set() call that was
unnecessary, but now additionally triggers a WARN_ON.
Patches #2-#4 are non-functional preparations for the following patches.
Patch #5 fixes a use-after-free that is triggered while changing network
namespaces.
Patch #6 makes mlxsw consistent with netdevsim by having mlxsw register
its devlink instance before its sub-objects. It helps us avoid a warning
described in the commit message.
====================
Link: https://lore.kernel.org/r/cover.1675692666.git.petrm@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Recent changes made it possible to register the devlink instance before
its sub-objects and under the instance lock. Among other things, it
allows us to avoid warnings such as this one [1]. The warning is
generated because a buggy firmware is generating a health event during
driver initialization, before the devlink instance is registered.
Move the registration of the devlink instance to the beginning of the
initialization flow to avoid such problems.
A similar change was implemented in netdevsim in commit 82a3aef2e6
("netdevsim: move devlink registration under the instance lock").
[1]
WARNING: CPU: 3 PID: 49 at net/devlink/leftover.c:7509 devlink_recover_notify.constprop.0+0xaf/0xc0
[...]
Call Trace:
<TASK>
devlink_health_report+0x45/0x1d0
mlxsw_core_health_event_work+0x24/0x30 [mlxsw_core]
process_one_work+0x1db/0x390
worker_thread+0x49/0x3b0
kthread+0xe5/0x110
ret_from_fork+0x1f/0x30
</TASK>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Cited commit added 'DEVLINK_CMD_PARAM_DEL' notifications whenever the
network namespace of the devlink instance is changed. Specifically, the
notifications are generated after calling reload_down(), but before
calling reload_up(). At this stage, the data structures accessed while
reading the value of the "acl_region_rehash_interval" devlink parameter
are uninitialized, resulting in a use-after-free [1].
Fix by moving the registration and unregistration of the devlink
parameter to the TCAM code where it is actually used. This means that
the parameter is unregistered during reload_down() and then
re-registered during reload_up(), avoiding the use-after-free between
these two operations.
Reproducer:
# ip netns add test123
# devlink dev reload pci/0000:06:00.0 netns test123
[1]
BUG: KASAN: use-after-free in mlxsw_sp_acl_tcam_vregion_rehash_intrvl_get+0xb2/0xd0
Read of size 4 at addr ffff888162fd37d8 by task devlink/1323
[...]
Call Trace:
<TASK>
dump_stack_lvl+0x95/0xbd
print_report+0x181/0x4a1
kasan_report+0xdb/0x200
mlxsw_sp_acl_tcam_vregion_rehash_intrvl_get+0xb2/0xd0
mlxsw_sp_params_acl_region_rehash_intrvl_get+0x32/0x80
devlink_nl_param_fill.constprop.0+0x29a/0x11e0
devlink_param_notify.constprop.0+0xb9/0x250
devlink_notify_unregister+0xbc/0x470
devlink_reload+0x1aa/0x440
devlink_nl_cmd_reload+0x559/0x11b0
genl_family_rcv_msg_doit.isra.0+0x1f8/0x2e0
genl_rcv_msg+0x558/0x7f0
netlink_rcv_skb+0x170/0x440
genl_rcv+0x2d/0x40
netlink_unicast+0x53f/0x810
netlink_sendmsg+0x961/0xe80
__sys_sendto+0x2a4/0x420
__x64_sys_sendto+0xe5/0x1c0
do_syscall_64+0x38/0x80
entry_SYSCALL_64_after_hwframe+0x63/0xcd
Fixes: 7d7e9169a3 ("devlink: move devlink reload notifications back in between _down() and _up() calls")
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Move the initialization and de-initialization code further below in
order to avoid forward declarations in the next patch. No functional
changes.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Move mutex_destroy() to the end to make the function symmetric with
mlxsw_sp_acl_tcam_init(). No functional changes.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Pair mutex_init() with a mutex_destroy() in the error path. Found during
code review. No functional changes.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The "acl_region_rehash_interval" devlink parameter is a "runtime"
parameter, making the call to devl_param_driverinit_value_set()
pointless. Before cited commit the function simply returned an error
(that was not checked), but now it emits a WARNING [1].
Fix by removing the function call.
[1]
WARNING: CPU: 0 PID: 7 at net/devlink/leftover.c:10974
devl_param_driverinit_value_set+0x8c/0x90
[...]
Call Trace:
<TASK>
mlxsw_sp2_params_register+0x83/0xb0 [mlxsw_spectrum]
__mlxsw_core_bus_device_register+0x5e5/0x990 [mlxsw_core]
mlxsw_core_bus_device_register+0x42/0x60 [mlxsw_core]
mlxsw_pci_probe+0x1f0/0x230 [mlxsw_pci]
local_pci_probe+0x1a/0x40
work_for_cpu_fn+0xf/0x20
process_one_work+0x1db/0x390
worker_thread+0x1d5/0x3b0
kthread+0xe5/0x110
ret_from_fork+0x1f/0x30
</TASK>
Fixes: 85fe0b324c ("devlink: make devlink_param_driverinit_value_set() return void")
Signed-off-by: Danielle Ratson <danieller@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Add PF driver support for the following:
- Viewing the standardized MAC Merge layer counters.
- Viewing the standardized Ethernet MAC and RMON counters associated
with the pMAC.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://lore.kernel.org/r/20230206094531.444988-2-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Add PF driver support for viewing and changing the MAC Merge sublayer
parameters, and seeing the verification state machine's current state.
The verification handshake with the link partner is driven by hardware.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://lore.kernel.org/r/20230206094531.444988-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Yury Norov says:
====================
sched: cpumask: improve on cpumask_local_spread() locality
cpumask_local_spread() currently checks local node for presence of i'th
CPU, and then if it finds nothing makes a flat search among all non-local
CPUs. We can do it better by checking CPUs per NUMA hops.
This has significant performance implications on NUMA machines, for example
when using NUMA-aware allocated memory together with NUMA-aware IRQ
affinity hints.
Performance tests from patch 8 of this series for mellanox network
driver show:
TCP multi-stream, using 16 iperf3 instances pinned to 16 cores (with aRFS on).
Active cores: 64,65,72,73,80,81,88,89,96,97,104,105,112,113,120,121
+-------------------------+-----------+------------------+------------------+
| | BW (Gbps) | TX side CPU util | RX side CPU util |
+-------------------------+-----------+------------------+------------------+
| Baseline | 52.3 | 6.4 % | 17.9 % |
+-------------------------+-----------+------------------+------------------+
| Applied on TX side only | 52.6 | 5.2 % | 18.5 % |
+-------------------------+-----------+------------------+------------------+
| Applied on RX side only | 94.9 | 11.9 % | 27.2 % |
+-------------------------+-----------+------------------+------------------+
| Applied on both sides | 95.1 | 8.4 % | 27.3 % |
+-------------------------+-----------+------------------+------------------+
Bottleneck in RX side is released, reached linerate (~1.8x speedup).
~30% less cpu util on TX.
====================
Link: https://lore.kernel.org/r/20230121042436.2661843-1-yury.norov@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Now that we have an iterator-based alternative for a very common case
of using cpumask_local_spread for all cpus in a row, it's worth to
mention that in comment to cpumask_local_spread().
Signed-off-by: Yury Norov <yury.norov@gmail.com>
Reviewed-by: Valentin Schneider <vschneid@redhat.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The recently introduced sched_numa_hop_mask() exposes cpumasks of CPUs
reachable within a given distance budget, wrap the logic for iterating over
all (distance, mask) values inside an iterator macro.
Signed-off-by: Valentin Schneider <vschneid@redhat.com>
Reviewed-by: Yury Norov <yury.norov@gmail.com>
Signed-off-by: Yury Norov <yury.norov@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Tariq has pointed out that drivers allocating IRQ vectors would benefit
from having smarter NUMA-awareness - cpumask_local_spread() only knows
about the local node and everything outside is in the same bucket.
sched_domains_numa_masks is pretty much what we want to hand out (a cpumask
of CPUs reachable within a given distance budget), introduce
sched_numa_hop_mask() to export those cpumasks.
Link: http://lore.kernel.org/r/20220728191203.4055-1-tariqt@nvidia.com
Signed-off-by: Valentin Schneider <vschneid@redhat.com>
Reviewed-by: Yury Norov <yury.norov@gmail.com>
Signed-off-by: Yury Norov <yury.norov@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Now after moving all NUMA logic into sched_numa_find_nth_cpu(),
else-branch of cpumask_local_spread() is just a function call, and
we can simplify logic by using ternary operator.
While here, replace BUG() with WARN_ON().
Signed-off-by: Yury Norov <yury.norov@gmail.com>
Acked-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Peter Lafreniere <peter@n8pjl.ca>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The function finds Nth set CPU in a given cpumask starting from a given
node.
Leveraging the fact that each hop in sched_domains_numa_masks includes the
same or greater number of CPUs than the previous one, we can use binary
search on hops instead of linear walk, which makes the overall complexity
of O(log n) in terms of number of cpumask_weight() calls.
Signed-off-by: Yury Norov <yury.norov@gmail.com>
Acked-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Peter Lafreniere <peter@n8pjl.ca>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Introduce cpumask_nth_and_andnot() based on find_nth_and_andnot_bit().
It's used in the following patch to traverse cpumasks without storing
intermediate result in temporary cpumask.
Signed-off-by: Yury Norov <yury.norov@gmail.com>
Acked-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Peter Lafreniere <peter@n8pjl.ca>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In the following patches the function is used to implement in-place bitmaps
traversing without storing intermediate result in temporary bitmaps.
Signed-off-by: Yury Norov <yury.norov@gmail.com>
Acked-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Peter Lafreniere <peter@n8pjl.ca>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Eric Dumazet says:
====================
net: core: use a dedicated kmem_cache for skb head allocs
Our profile data show that using kmalloc(non_const_size)/kfree(ptr)
has a certain cost, because kfree(ptr) has to pull a 'struct page'
in cpu caches.
Using a dedicated kmem_cache for TCP skb->head allocations makes
a difference, both in cpu cycles and memory savings.
This kmem_cache could also be used for GRO skb allocations,
this is left as a future exercise.
====================
Link: https://lore.kernel.org/r/20230206173103.2617121-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Recent removal of ksize() in alloc_skb() increased
performance because we no longer read
the associated struct page.
We have an equivalent cost at kfree_skb() time.
kfree(skb->head) has to access a struct page,
often cold in cpu caches to get the owning
struct kmem_cache.
Considering that many allocations are small (at least for TCP ones)
we can have our own kmem_cache to avoid the cache line miss.
This also saves memory because these small heads
are no longer padded to 1024 bytes.
CONFIG_SLUB=y
$ grep skbuff_small_head /proc/slabinfo
skbuff_small_head 2907 2907 640 51 8 : tunables 0 0 0 : slabdata 57 57 0
CONFIG_SLAB=y
$ grep skbuff_small_head /proc/slabinfo
skbuff_small_head 607 624 640 6 1 : tunables 54 27 8 : slabdata 104 104 5
Notes:
- After Kees Cook patches and this one, we might
be able to revert commit
dbae2b0628 ("net: skb: introduce and use a single page frag cache")
because GRO_MAX_HEAD is also small.
- This patch is a NOP for CONFIG_SLOB=y builds.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
All kmalloc_reserve() callers have to make the same computation,
we can factorize them, to prepare following patch in the series.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This is a cleanup patch, to prepare following change.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
We have many places using this expression:
SKB_DATA_ALIGN(sizeof(struct skb_shared_info))
Use of SKB_HEAD_ALIGN() will allow to clean them.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
-----BEGIN PGP SIGNATURE-----
iQFHBAABCgAxFiEEDs2BvajyNKlf9TJQvlAcSiqKBOgFAmPg+r4THG1rbEBwZW5n
dXRyb25peC5kZQAKCRC+UBxKKooE6A44B/0c8jjyLikVE2fEBKkvPk8iRlFKap48
140HNR+1TuVBBPCJTe2Gq3l8fDfFUgUVnFfVDjAQq6pMWufLkz9jVr5uGN7O94v+
tFCjip8NLRTa6ibgQlGrFXl31X5vqT+jV6V4l29wFgbwgOMRrod7wLDo+bHTtuuY
46Sf4GgIazYjxEeZBBsvjnAbraFq0twZFcEz9SyeOxqjR+9+71X4Y3x0kgQuz72N
CrvRng0yFLqbHOBE60iaDqbjuNnwBGTH476pwX25Kj9gVyIH20T45h8tV4j6z2Sq
A42cHKDlWZRWUiWQjGOHHT4igUf//Sfv8WDaFjqobO/DDY+nBA75EPAv
=hWLT
-----END PGP SIGNATURE-----
Merge tag 'linux-can-next-for-6.3-20230206' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can-next
Marc Kleine-Budde says:
====================
pull-request: can-next 2023-02-06
this is a pull request of 47 patches for net-next/master.
The first two patch is by Oliver Hartkopp. One adds missing error
checking to the CAN_GW protocol, the other adds a missing CAN address
family check to the CAN ISO TP protocol.
Thomas Kopp contributes a performance optimization to the mcp251xfd
driver.
The next 11 patches are by Geert Uytterhoeven and add support for
R-Car V4H systems to the rcar_canfd driver.
Stephane Grosjean and Lukas Magel contribute 8 patches to the peak_usb
driver, which add support for configurable CAN channel ID.
The last 17 patches are by me and target the CAN bit timing
configuration. The bit timing is cleaned up, error messages are
improved and forwarded to user space via NL_SET_ERR_MSG_FMT() instead
of netdev_err(), and the SJW handling is updated, including the
definition of a new default value that will benefit CAN-FD
controllers, by increasing their oscillator tolerance.
* tag 'linux-can-next-for-6.3-20230206' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can-next: (47 commits)
can: bittiming: can_validate_bitrate(): report error via netlink
can: bittiming: can_calc_bittiming(): convert from netdev_err() to NL_SET_ERR_MSG_FMT()
can: bittiming: can_calc_bittiming(): clean up SJW handling
can: bittiming: can_sjw_set_default(): use Phase Seg2 / 2 as default for SJW
can: bittiming: can_sjw_check(): check that SJW is not longer than either Phase Buffer Segment
can: bittiming: can_sjw_check(): report error via netlink and harmonize error value
can: bittiming: can_fixup_bittiming(): report error via netlink and harmonize error value
can: bittiming: factor out can_sjw_set_default() and can_sjw_check()
can: bittiming: can_changelink() pass extack down callstack
can: netlink: can_changelink(): convert from netdev_err() to NL_SET_ERR_MSG_FMT()
can: netlink: can_validate(): validate sample point for CAN and CAN-FD
can: dev: register_candev(): bail out if both fixed bit rates and bit timing constants are provided
can: dev: register_candev(): ensure that bittiming const are valid
can: bittiming: can_get_bittiming(): use direct return and remove unneeded else
can: bittiming: can_fixup_bittiming(): set effective tq
can: bittiming: can_fixup_bittiming(): use CAN_SYNC_SEG instead of 1
can: bittiming(): replace open coded variants of can_bit_time()
can: peak_usb: Reorder include directives alphabetically
can: peak_usb: align CAN channel ID format in log with sysfs attribute
can: peak_usb: export PCAN CAN channel ID as sysfs device attribute
...
====================
Link: https://lore.kernel.org/r/20230206131620.2758724-1-mkl@pengutronix.de
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
If ops->get_mm() returns a non-zero error code, we goto out_complete,
but there, we return 0. Fix that to propagate the "ret" variable to the
caller. If ops->get_mm() succeeds, it will always return 0.
Fixes: 2b30f8291a ("net: ethtool: add support for MAC Merge layer")
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Link: https://lore.kernel.org/r/20230206094932.446379-1-vladimir.oltean@nxp.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Use actual CPU number instead of hardcoded value to decide the size
of 'cpu_used_mask' in 'struct sw_flow'. Below is the reason.
'struct cpumask cpu_used_mask' is embedded in struct sw_flow.
Its size is hardcoded to CONFIG_NR_CPUS bits, which can be
8192 by default, it costs memory and slows down ovs_flow_alloc.
To address this:
Redefine cpu_used_mask to pointer.
Append cpumask_size() bytes after 'stat' to hold cpumask.
Initialization cpu_used_mask right after stats_last_writer.
APIs like cpumask_next and cpumask_set_cpu never access bits
beyond cpu count, cpumask_size() bytes of memory is enough.
Signed-off-by: Eddy Tao <taoyuan_eddy@hotmail.com>
Acked-by: Eelco Chaudron <echaudro@redhat.com>
Link: https://lore.kernel.org/r/OS3P286MB229570CCED618B20355D227AF5D59@OS3P286MB2295.JPNP286.PROD.OUTLOOK.COM
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The forward declaration was introduced with a prototype that does
not match the function definition:
drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c:2166:13: error: conflicting types for 'xgbe_phy_perform_ratechange' due to enum/integer mismatch; have 'void(struct xgbe_prv_data *, enum xgbe_mb_cmd, enum xgbe_mb_subcmd)' [-Werror=enum-int-mismatch]
2166 | static void xgbe_phy_perform_ratechange(struct xgbe_prv_data *pdata,
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c:391:13: note: previous declaration of 'xgbe_phy_perform_ratechange' with type 'void(struct xgbe_prv_data *, unsigned int, unsigned int)'
391 | static void xgbe_phy_perform_ratechange(struct xgbe_prv_data *pdata,
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~
Ideally there should not be any forward declarations here, which
would make it easier to show that there is no unbounded recursion.
I tried fixing this but could not figure out how to avoid the
recursive call.
As a hotfix, address only the broken prototype to fix the build
problem instead.
Fixes: 4f3b20bfbb ("amd-xgbe: add support for rx-adaptation")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Acked-by: Shyam Sundar S K <Shyam-sundar.S-k@amd.com>
Link: https://lore.kernel.org/r/20230203121553.2871598-1-arnd@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
There are no external users of the vsc7514_*_regmap[] symbols or
vsc7514_vcap_* functions. They were exported in commit 32ecd22ba6 ("net:
mscc: ocelot: split register definitions to a separate file") with the
intention of being used, but the actual structure used in commit
2efaca411c ("net: mscc: ocelot: expose vsc7514_regmap definition") ended
up being all that was needed.
Bury these unnecessary symbols.
Signed-off-by: Colin Foster <colin.foster@in-advantage.com>
Suggested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Link: https://lore.kernel.org/r/20230204182056.25502-1-colin.foster@in-advantage.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Ajit Khaparde says:
====================
bnxt: Add Auxiliary driver support
Add auxiliary device driver for Broadcom devices.
The bnxt_en driver will register and initialize an aux device
if RDMA is enabled in the underlying device.
The bnxt_re driver will then probe and initialize the
RoCE interfaces with the infiniband stack.
We got rid of the bnxt_en_ops which the bnxt_re driver used to
communicate with bnxt_en.
Similarly We have tried to clean up most of the bnxt_ulp_ops.
In most of the cases we used the functions and entry points provided
by the auxiliary bus driver framework.
And now these are the minimal functions needed to support the functionality.
We will try to work on getting rid of the remaining if we find any
other viable option in future.
* 'aux-bus-v11' of https://github.com/ajitkhaparde1/linux:
bnxt_en: Remove runtime interrupt vector allocation
RDMA/bnxt_re: Remove the sriov config callback
bnxt_en: Remove struct bnxt access from RoCE driver
bnxt_en: Use auxiliary bus calls over proprietary calls
bnxt_en: Use direct API instead of indirection
bnxt_en: Remove usage of ulp_id
RDMA/bnxt_re: Use auxiliary driver interface
bnxt_en: Add auxiliary driver support
====================
Link: https://lore.kernel.org/r/20230202033809.3989-1-ajit.khaparde@broadcom.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The dev_lan_addr and hw_lan_addr members of ice_vf are used only to store
the MAC address for the VF. They are defined using virtchnl_ether_addr, but
only the .addr sub-member is actually used. Drop the use of
virtchnl_ether_addr and just use a u8 array of length [ETH_ALEN].
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Marek Szlosek <marek.szlosek@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
The Scalable IOV implementation will require notifying the VDCM driver when
an IRQ must be closed. This allows the VDCM to handle releasing stale IRQ
context values and properly reconfigure.
To handle this, introduce a new optional .irq_close callback to the VF
operations structure. This will be implemented by Scalable IOV to handle
the shutdown of the IRQ context.
Since the SR-IOV implementation does not need this, we must check that its
non-NULL before calling it.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Marek Szlosek <marek.szlosek@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
When hardware is reset, the VF relies on the VFGEN_RSTAT register to detect
when the VF is finished resetting. This is a tri-state register where 0
indicates a reset is in progress, 1 indicates the hardware is done
resetting, and 2 indicates that the software is done resetting.
Currently the PF driver relies on the device hardware resetting VFGEN_RSTAT
when a global reset occurs. This works ok, but it does mean that the VF
might not immediately notice a reset when the driver first detects that the
global reset is occurring.
This is also problematic for Scalable IOV, because there is no read/write
equivalent VFGEN_RSTAT register for the Scalable VSI type. Instead, the
Scalable IOV VFs will need to emulate this register.
To support this, introduce a new VF operation, clear_reset_state, which is
called when the PF driver first detects a global reset. The Single Root IOV
implementation can just write to VFGEN_RSTAT to ensure it's cleared
immediately, without waiting for the actual hardware reset to begin. The
Scalable IOV implementation will use this as part of its tracking of the
reset status to allow properly reporting the emulated VFGEN_RSTAT to the VF
driver.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Paul Menzel <pmenzel@molgen.mpg.de>
Tested-by: Marek Szlosek <marek.szlosek@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
The .vsi_rebuild function exists for ice_reset_vf. It is used to release
and re-create the VSI during a single-VF reset.
This function is only called when we need to re-create the VSI, and not
when rebuilding an existing VSI. This makes the single-VF reset process
different from the process used to restore functionality after a
hardware reset such as the PF reset or EMP reset.
When we add support for Scalable IOV VFs, the implementation will be very
similar. The primary difference will be in the fact that each VF type uses
a different underlying VSI type in hardware.
Move the common functionality into a new ice_vf_recreate VSI function. This
will allow the two IOV paths to share this functionality. Rework the
.vsi_rebuild vf_op into .create_vsi, only performing the task of creating a
new VSI.
This creates a nice dichotomy between the ice_vf_rebuild_vsi and
ice_vf_recreate_vsi, and should make it more clear why the two flows atre
distinct.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Marek Szlosek <marek.szlosek@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Introduce a new generic helper ice_vf_init_host_cfg which performs common
host configuration initialization tasks that will need to be done for both
Single Root IOV and the new Scalable IOV implementation.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Marek Szlosek <marek.szlosek@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Some of the initialization code for Single Root IOV VFs will need to be
reused when we introduce Scalable IOV. Pull this code out into a new
ice_initialize_vf_entry helper function.
Co-developed-by: Harshitha Ramamurthy <harshitha.ramamurthy@intel.com>
Signed-off-by: Harshitha Ramamurthy <harshitha.ramamurthy@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Marek Szlosek <marek.szlosek@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
The Single Root IOV implementation of .post_vsi_rebuild performs some tasks
that will ultimately need to be shared with the Scalable IOV implementation
such as rebuilding the host configuration.
Refactor by introducing a new wrapper function, ice_vf_post_vsi_rebuild
which performs the tasks that will be shared between SR-IOV and Scalable
IOV. Move the ice_vf_rebuild_host_cfg and ice_vf_set_initialized calls into
this wrapper. Then call the implementation specific post_vsi_rebuild
handler afterwards.
This ensures that we will properly re-initialize filters and expected
settings for both SR-IOV and Scalable IOV.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Marek Szlosek <marek.szlosek@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
