Attribute access is now serialized in the hardware monitoring core,
so locking in the driver code is no longer necessary. Drop it.
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Attribute access is now serialized in the hardware monitoring core,
so locking in the driver code is no longer necessary. Drop it.
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Attribute access is now serialized in the hardware monitoring core,
so locking in the driver code is no longer necessary. Drop it.
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Attribute access is now serialized in the hardware monitoring core,
so locking in the driver code is no longer necessary. Drop it.
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Attribute access is now serialized in the hardware monitoring core,
so locking in the driver is no longer necessary. Drop it.
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Attribute access is now serialized in the hardware monitoring core,
so locking in the driver code is no longer necessary. Drop it.
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Attribute access is now serialized in the hardware monitoring core,
so locking in the driver code is no longer necessary. Drop it.
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Attribute access is now serialized in the hardware monitoring core,
so locking n the driver code is no longer necessary. Drop it.
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Attribute access is now serialized in the hardware monitoring core,
so locking in the driver code is no longer necessary. Drop it.
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add support for mp9945 device from Monolithic Power Systems, Inc.
(MPS) vendor. This is a single phase digital step down converter.
Signed-off-by: Cosmo Chou <chou.cosmo@gmail.com>
Link: https://lore.kernel.org/r/20251009205458.396368-2-chou.cosmo@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Use the energy64 attribute type instead of a locally defined sysfs
attribute to report the accumulated energy.
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Reviewed-by: Nuno Sá <nuno.sa@analog.com>
Link: https://lore.kernel.org/r/20251012211625.533791-2-linux@roeck-us.net
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Use the energy64 attribute type instead of a locally defined sysfs
attribute to report the accumulated energy.
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Reviewed-by: Nuno Sá <nuno.sa@analog.com>
Link: https://lore.kernel.org/r/20251012211625.533791-1-linux@roeck-us.net
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add support to read DTS for reading Intel Emerald Rapids platform.
Signed-off-by: Ivan Mikhaylov <fr0st61te@gmail.com>
Reviewed-by: Paul Menzel <pmenzel@molgen.mpg.de>
Link: https://lore.kernel.org/r/20251006215321.5036-4-fr0st61te@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Extend the functionality of hwmon (peci/dimmtemp) for Emerald Rapids
platform.
The patch has been tested on a 5S system with 16 DIMMs installed.
Verified read of DIMM temperature thresholds & temperature.
Using Sapphire's callbacks about getting thresholds because it's same
platform/socket.
Signed-off-by: Ivan Mikhaylov <fr0st61te@gmail.com>
Reviewed-by: Paul Menzel <pmenzel@molgen.mpg.de>
Link: https://lore.kernel.org/r/20251006215321.5036-3-fr0st61te@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add support for detection of Intel Emerald Rapids processor based on
CPU model.
Emerald Rapids Xeon processors with the model set to
INTEL_EMERALDRAPIDS_X. The data field for this entry is "emr".
Tested the patch series with AST2600 BMC with 5S Intel Emerald Rapids
processors & verified by reading cpu & dimm temperature which matches
host sensor values from lmsensors.
Signed-off-by: Ivan Mikhaylov <fr0st61te@gmail.com>
Reviewed-by: Paul Menzel <pmenzel@molgen.mpg.de>
Link: https://lore.kernel.org/r/20251006215321.5036-2-fr0st61te@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
The board has a similar sensor configuration as the ROG STRIX X870E-E
GAMING WIFI, with an additional temperature sensor header.
Signed-off-by: Maximilian Luz <luzmaximilian@gmail.com>
Signed-off-by: Eugene Shalygin <eugene.shalygin@gmail.com>
Link: https://lore.kernel.org/r/20251003081002.1013313-3-eugene.shalygin@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
ADPM12200 is a quarter brick DC/DC Power Module. It is a high power
non-isolated converter capable of delivering regulated 12V with
continuous power level of 2000W. Uses PMBus.
Signed-off-by: Alexis Czezar Torreno <alexisczezar.torreno@analog.com>
Link: https://lore.kernel.org/r/20251001-hwmon-next-v1-2-f8ca6a648203@analog.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Since fan properties can include reg, the fan controller should be
able to specify address-cells and size-cells properties and use
an addressed fan child node.
Signed-off-by: Eddie James <eajames@linux.ibm.com>
Reviewed-by: Rob Herring (Arm) <robh@kernel.org>
Link: https://lore.kernel.org/r/20251001144441.310950-6-eajames@linux.ibm.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
The binding for the max31785 is the same as the max31790, so just add
some compatible strings for the max31785 chip.
Signed-off-by: Eddie James <eajames@linux.ibm.com>
Reviewed-by: Rob Herring (Arm) <robh@kernel.org>
Link: https://lore.kernel.org/r/20251001144441.310950-5-eajames@linux.ibm.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add support for the Murata NCP18WM474 NTC.
Compensation table has been constructed by linear interpolation between
well defined points[1] on the resistance vs. temperature graph in the
datasheet[2]. The readouts of the graph has been done to the best of my
abilities, but the compensation table will be subject to inaccuracies
nonetheless.
[1] -40, -25, 0, 25, 50, 75, 100, 125 degrees
[2] https://www.murata.com/en-eu/api/pdfdownloadapi?cate=&partno=NCP18WM474E03RB
Signed-off-by: Emil Dahl Juhl <juhl.emildahl@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Link: https://lore.kernel.org/r/20251001-ntc-thermistor-ncp18wm474-v1-1-2c64b9b509ff@pengutronix.de
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add support for MAX17616/MAX17616A current-limiter with
overvoltage/surge, undervoltage, reverse polarity, loss of ground
protection with PMBus interface. The PMBus interface allows monitoring
of input/output voltages, output current and temperature.
Signed-off-by: Kim Seer Paller <kimseer.paller@analog.com>
Link: https://lore.kernel.org/r/20250930-upstream-max17616-v1-2-1525a85f126c@analog.com
[groeck: Fixed htmldocs 'WARNING: Title underline too short'
as reported by Kriish Sharma <kriish.sharma2006@gmail.com>]
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Modern AMD CPUs do not support segment limit checks in 64-bit mode
(i.e. EFER.LMSLE must be zero). Do not allow a guest to set EFER.LMSLE
on a CPU that requires the bit to be zero.
For backwards compatibility, allow EFER.LMSLE to be set on CPUs that
support segment limit checks in 64-bit mode, even though KVM's
implementation of the feature is incomplete (e.g. KVM's emulator does
not enforce segment limits in 64-bit mode).
Fixes: eec4b140c9 ("KVM: SVM: Allow EFER.LMSLE to be set with nested svm")
Signed-off-by: Jim Mattson <jmattson@google.com>
Reviewed-by: Nikunj A Dadhania <nikunj@amd.com>
Reviewed-by: Yosry Ahmed <yosry.ahmed@linux.dev>
Link: https://lore.kernel.org/r/20251001001529.1119031-3-jmattson@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
This commit adds two tests. The first is the most basic unit test:
make sure an empty queue peeks as empty, and when we put one element
in the queue, make sure peek returns that element.
However, even this simple test is a little complicated by the different
behavior of scx_bpf_dsq_insert in different calling contexts:
- insert is for direct dispatch in enqueue
- insert is delayed when called from select_cpu
In this case we split the insert and the peek that verifies the
result between enqueue/dispatch.
Note: An alternative would be to call `scx_bpf_dsq_move_to_local` on an
empty queue, which in turn calls `flush_dispatch_buf`, in order to flush
the buffered insert. Unfortunately, this is not viable within the
enqueue path, as it attempts a voluntary context switch within an RCU
read-side critical section.
The second test is a stress test that performs many peeks on all DSQs
and records the observed tasks.
Signed-off-by: Ryan Newton <newton@meta.com>
Reviewed-by: Christian Loehle <christian.loehle@arm.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
The builtin DSQ queue data structures are meant to be used by a wide
range of different sched_ext schedulers with different demands on these
data structures. They might be per-cpu with low-contention, or
high-contention shared queues. Unfortunately, DSQs have a coarse-grained
lock around the whole data structure. Without going all the way to a
lock-free, more scalable implementation, a small step we can take to
reduce lock contention is to allow a lockless, small-fixed-cost peek at
the head of the queue.
This change allows certain custom SCX schedulers to cheaply peek at
queues, e.g. during load balancing, before locking them. But it
represents a few extra memory operations to update the pointer each
time the DSQ is modified, including a memory barrier on ARM so the write
appears correctly ordered.
This commit adds a first_task pointer field which is updated
atomically when the DSQ is modified, and allows any thread to peek at
the head of the queue without holding the lock.
Signed-off-by: Ryan Newton <newton@meta.com>
Reviewed-by: Andrea Righi <arighi@nvidia.com>
Reviewed-by: Christian Loehle <christian.loehle@arm.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Jean-François Lessard <jefflessard3@gmail.com> says:
This series modernizes the auxdisplay line display (linedisp) library to
enable seamless integration with auxdisplay parent devices while
maintaining backward compatibility.
The key improvement is adding attach/detach APIs that allow linedisp sysfs
attributes to be bound directly to their parent auxdisplay devices avoiding
child device proliferation and enabling a uniform 7-segment userspace
interface across different driver architectures.
This series introduces attachment infrastructure for linedisp devices.
The first consumer of this API will be the TM16XX driver series.
See the related patch series:
auxdisplay: Add TM16xx 7-segment LED matrix display controllers driver
Changes include:
1. Encapsulate container_of() usage with to_linedisp() helper function for
cleaner context retrieval
2. Improve message display behavior with static padding when message length
is smaller than display width
3. Add 'num_chars' read-only attribute for userspace capability discovery
4. Add attach/detach API for sysfs attributes binding to parent devices
5. Document all linedisp sysfs attributes in ABI documentation
All existing linedisp_register() users remain unaffected. The new APIs
enable drivers like TM16XX to integrate 7-segment functionality within
their LED class device hierarchy while providing a uniform 7-segment API.
Thanks to Andy Shevchenko for early feedback and guidance.
Link: https://lore.kernel.org/r/20250918121321.116248-1-jefflessard3@gmail.com
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Enable linedisp library integration into existing kernel devices (like LED
class) to provide a uniform 7-segment userspace API without creating
separate child devices, meeting the consistent interface while maintaining
coherent device hierarchies.
This allows uniform 7-segment API across all drivers while solving device
proliferation and fragmented userspace interfaces.
The sysfs attributes appear in one of the two locations depending on usage:
1. On linedisp.N child devices (legacy linedisp_register())
2. On the parent auxdisplay device (new linedisp_attach())
Functionality is identical in both modes.
Existing consumers of linedisp_register() are unaffected. The new API
enables drivers like TM16XX to integrate 7-segment display functionality
seamlessly within their LED class device hierarchy.
Signed-off-by: Jean-François Lessard <jefflessard3@gmail.com>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Add a read-only 'num_chars' sysfs attribute to report display digit count.
The num_chars attribute provides essential capability information to
userspace applications that need to know display dimensions before writing
messages, complementing the existing message and scroll controls.
No functional changes to existing behavior.
Signed-off-by: Jean-François Lessard <jefflessard3@gmail.com>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Currently, when a message shorter than the display size is written, the
content wraps around (e.g., "123" on a 4-digit display shows "1231")
without scrolling, which is confusing and unintuitive.
Change behavior to display short messages statically with space padding
(e.g. "123 ") while only scrolling messages longer than the display width.
This provides more natural behavior that aligns with user expectations
and current linedisp_display() kernel-doc.
The scroll logic is also consolidated into a helper function for clarity.
No API changes are introduced.
Signed-off-by: Jean-François Lessard <jefflessard3@gmail.com>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Replace direct container_of() calls with a to_linedisp() helper function
throughout the line-display auxdisplay library module. This abstraction
prepares for upcoming dual-mode support where linedisp context retrieval
will need to handle both dedicated child devices and attached parent
auxdisplay devices.
No functional changes in this patch.
Signed-off-by: Jean-François Lessard <jefflessard3@gmail.com>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Add ABI documentation for sysfs attributes provided by the line-display
auxdisplay library module. These attributes enable text message display and
configuration on character-based auxdisplay devices.
Documents previously undocumented attributes:
- message, scroll_step_ms (introduced in v5.16)
- map_seg7, map_seg14 (introduced in v6.9)
The line-display library is used by multiple auxdisplay drivers and
can expose these attributes either on linedisp.N child devices or
directly on parent auxdisplay devices.
Signed-off-by: Jean-François Lessard <jefflessard3@gmail.com>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
The kernel-doc for struct drm_gpuvm_map_req.map was added as '@op_map'
instead of '@map', leading to this warning during htmldocs build:
WARNING: include/drm/drm_gpuvm.h:1083 struct member 'map' not described in 'drm_gpuvm_map_req'
Fixes: 000a45dce7 ("drm/gpuvm: Pass map arguments through a struct")
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Closes: https://lore.kernel.org/all/20250821133539.03aa298e@canb.auug.org.au/
Signed-off-by: Ankan Biswas <spyjetfayed@gmail.com>
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
We should wait for the rdma_cm to become SMBDIRECT_SOCKET_DISCONNECTED,
it turns out that (at least running some xfstests e.g. cifs/001)
often triggers the case where wait_event_interruptible() returns
with -ERESTARTSYS instead of waiting for SMBDIRECT_SOCKET_DISCONNECTED
to be reached.
Or we are already in SMBDIRECT_SOCKET_DISCONNECTING and never wait
for SMBDIRECT_SOCKET_DISCONNECTED.
Fixes: 050b8c3740 ("smbd: Make upper layer decide when to destroy the transport")
Fixes: e8b3bfe9bc ("cifs: smbd: Don't destroy transport on RDMA disconnect")
Fixes: b0aa92a229 ("smb: client: make sure smbd_disconnect_rdma_work() doesn't run after smbd_destroy() took over")
Cc: Steve French <smfrench@gmail.com>
Cc: Tom Talpey <tom@talpey.com>
Cc: Long Li <longli@microsoft.com>
Cc: Namjae Jeon <linkinjeon@kernel.org>
Cc: linux-cifs@vger.kernel.org
Cc: samba-technical@lists.samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
This effectively reverts b0ba512e25 ("net: bcmgenet: enable driver to
work without a device tree"). There has never been an in-tree user of
struct bcmgenet_platform_data, all devices use OF or ACPI.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
Link: https://patch.msgid.link/108b4e64-55d4-4b4e-9a11-3c810c319d66@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Driver calls skb_tx_timestamp(skb) in rtl8152_start_xmit(), but does not advertise the capability in ethtool.
Advertise software timestamp capabilities on struct ethtool_ops.
Signed-off-by: Abhishek Rawal <rawal.abhishek92@gmail.com>
Reviewed-by: Jamie Bainbridge <jamie.bainbridge@gmail.com>
Reviewed-by: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Link: https://patch.msgid.link/20251014055234.46527-1-rawal.abhishek92@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The device returns a valid bit in the LSB of the low timestamp byte in
the completion descriptor that the driver should check before
setting the SKB's hardware timestamp. If the timestamp is not valid, do not
hardware timestamp the SKB.
Cc: stable@vger.kernel.org
Fixes: b2c7aeb490 ("gve: Implement ndo_hwtstamp_get/set for RX timestamping")
Reviewed-by: Joshua Washington <joshwash@google.com>
Signed-off-by: Tim Hostetler <thostet@google.com>
Signed-off-by: Harshitha Ramamurthy <hramamurthy@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Reviewed-by: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Link: https://patch.msgid.link/20251014004740.2775957-1-hramamurthy@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Eric Dumazet says:
====================
net: deal with sticky tx queues
Back in 2010, Tom Herbert added skb->ooo_okay to TCP flows.
Extend the feature to connected flows for other protocols like UDP.
skb->ooo_okay might never be set for bulk flows that always
have at least one skb in a qdisc queue of NIC queue,
especially if TX completion is delayed because of a stressed cpu
or aggressive interrupt mitigation.
The so-called "strange attractors" has caused many performance
issues, we need to do better now that TCP reacts better to
potential reorders.
Add new net.core.txq_reselection_ms sysctl to let
flows follow XPS and select a more efficient queue.
After this series, we no longer have to make sure threads
are pinned to cpus, they can migrate without adding
too much [spinlock, qdisc, TX completion] pressure anymore.
====================
Link: https://patch.msgid.link/20251013152234.842065-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This is a followup of commit 726e9e8b94 ("tcp: refine
skb->ooo_okay setting") and of prior commit in this series
("net: control skb->ooo_okay from skb_set_owner_w()")
skb->ooo_okay might never be set for bulk flows that always
have at least one skb in a qdisc queue of NIC queue,
especially if TX completion is delayed because of a stressed cpu.
The so-called "strange attractors" has caused many performance
issues (see for instance 9b462d02d6 ("tcp: TCP Small Queues
and strange attractors")), we need to do better.
We have tried very hard to avoid reorders because TCP was
not dealing with them nicely a decade ago.
Use the new net.core.txq_reselection_ms sysctl to let
flows follow XPS and select a more efficient queue.
After this patch, we no longer have to make sure threads
are pinned to cpus, they now can be migrated without
adding too much spinlock/qdisc/TX completion pressure anymore.
TX completion part was problematic, because it added false sharing
on various socket fields, but also added false sharing and spinlock
contention in mm layers. Calling skb_orphan() from ndo_start_xmit()
is not an option unfortunately.
Note for later:
1) move sk->sk_tx_queue_mapping closer
to sk_tx_queue_mapping_jiffies for better cache locality.
2) Study if 9b462d02d6 ("tcp: TCP Small Queues
and strange attractors") could be revised.
Tested:
Used a host with 32 TX queues, shared by groups of 8 cores.
XPS setup :
echo ff >/sys/class/net/eth1/queue/tx-0/xps_cpus
echo ff00 >/sys/class/net/eth1/queue/tx-1/xps_cpus
echo ff0000 >/sys/class/net/eth1/queue/tx-2/xps_cpus
echo ff000000 >/sys/class/net/eth1/queue/tx-3/xps_cpus
echo ff,00000000 >/sys/class/net/eth1/queue/tx-4/xps_cpus
echo ff00,00000000 >/sys/class/net/eth1/queue/tx-5/xps_cpus
echo ff0000,00000000 >/sys/class/net/eth1/queue/tx-6/xps_cpus
echo ff000000,00000000 >/sys/class/net/eth1/queue/tx-7/xps_cpus
...
Launched a tcp_stream with 15 threads and 1000 flows, initially affined to core 0-15
taskset -c 0-15 tcp_stream -T15 -F1000 -l1000 -c -H target_host
Checked that only queues 0 and 1 are used as instructed by XPS :
tc -s qdisc show dev eth1|grep backlog|grep -v "backlog 0b 0p"
backlog 123489410b 1890p
backlog 69809026b 1064p
backlog 52401054b 805p
Then force each thread to run on cpu 1,9,17,25,33,41,49,57,65,73,81,89,97,105,113,121
C=1;PID=`pidof tcp_stream`;for P in `ls /proc/$PID/task`; do taskset -pc $C $P; C=$(($C + 8));done
Set txq_reselection_ms to 1000
echo 1000 > /proc/sys/net/core/txq_reselection_ms
Check that the flows have migrated nicely:
tc -s qdisc show dev eth1|grep backlog|grep -v "backlog 0b 0p"
backlog 130508314b 1916p
backlog 8584380b 126p
backlog 8584380b 126p
backlog 8379990b 123p
backlog 8584380b 126p
backlog 8487484b 125p
backlog 8584380b 126p
backlog 8448120b 124p
backlog 8584380b 126p
backlog 8720640b 128p
backlog 8856900b 130p
backlog 8584380b 126p
backlog 8652510b 127p
backlog 8448120b 124p
backlog 8516250b 125p
backlog 7834950b 115p
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://patch.msgid.link/20251013152234.842065-5-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Add a new sysctl to control how often a queue reselection
can happen even if a flow has a persistent queue of skbs
in a Qdisc or NIC queue.
A value of zero means the feature is disabled.
Default is 1000 (1 second).
This sysctl is used in the following patch.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Neal Cardwell <ncardwell@google.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://patch.msgid.link/20251013152234.842065-4-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
15 years after Tom Herbert added skb->ooo_okay, only TCP transport
benefits from it.
We can support other transports directly from skb_set_owner_w().
If no other TX packet for this socket is in a host queue (qdisc, NIC queue)
there is no risk of self-inflicted reordering, we can set skb->ooo_okay.
This allows netdev_pick_tx() to choose a TX queue based on XPS settings,
instead of reusing the queue chosen at the time the first packet was sent
for connected sockets.
Tested:
500 concurrent UDP_RR connected UDP flows, host with 32 TX queues,
512 cpus, XPS setup.
super_netperf 500 -t UDP_RR -H <host> -l 1000 -- -r 100,100 -Nn &
This patch saves between 10% and 20% of cycles, depending on how
process scheduler migrates threads among cpus.
Using following bpftrace script, we can see the effect on Qdisc/NIC tx queues
being better used (less cache line misses).
bpftrace -e '
k:__dev_queue_xmit { @start[cpu] = nsecs; }
kr:__dev_queue_xmit {
if (@start[cpu]) {
$delay = nsecs - @start[cpu];
delete(@start[cpu]);
@__dev_queue_xmit_ns = hist($delay);
}
}
END { clear(@start); }'
Before:
@__dev_queue_xmit_ns:
[128, 256) 6 | |
[256, 512) 116283 | |
[512, 1K) 1888205 |@@@@@@@@@@@ |
[1K, 2K) 8106167 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ |
[2K, 4K) 8699293 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@|
[4K, 8K) 2600676 |@@@@@@@@@@@@@@@ |
[8K, 16K) 721688 |@@@@ |
[16K, 32K) 122995 | |
[32K, 64K) 10639 | |
[64K, 128K) 119 | |
[128K, 256K) 1 | |
After:
@__dev_queue_xmit_ns:
[128, 256) 3 | |
[256, 512) 651112 |@@ |
[512, 1K) 8109938 |@@@@@@@@@@@@@@@@@@@@@@@@@@ |
[1K, 2K) 16081031 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@|
[2K, 4K) 2411692 |@@@@@@@ |
[4K, 8K) 98994 | |
[8K, 16K) 1536 | |
[16K, 32K) 587 | |
[32K, 64K) 2 | |
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Neal Cardwell <ncardwell@google.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://patch.msgid.link/20251013152234.842065-3-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
sk->sk_wmem_alloc is initialized to 1, and sk_wmem_alloc_get()
takes care of this initial value.
Add SK_WMEM_ALLOC_BIAS define to not spread this magic value.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Neal Cardwell <ncardwell@google.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://patch.msgid.link/20251013152234.842065-2-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Some applications uses TCP_TX_DELAY socket option after TCP flow
is established.
Some metrics need to be updated, otherwise TCP might take time to
adapt to the new (emulated) RTT.
This patch adjusts tp->srtt_us, tp->rtt_min, icsk_rto
and sk->sk_pacing_rate.
This is best effort, and for instance icsk_rto is reset
without taking backoff into account.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20251013145926.833198-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Add optional clock for OSC_IN and fix the below CHECK_DTBS warnings:
arch/arm/boot/dts/nxp/imx/imx6qp-prtwd3.dtb: switch@0 (nxp,sja1105q): Unevaluated properties are not allowed ('clocks' was unexpected)
Signed-off-by: Frank Li <Frank.Li@nxp.com>
Acked-by: Conor Dooley <conor.dooley@microchip.com>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://patch.msgid.link/20251010183418.2179063-1-Frank.Li@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
