Merge branch 'next' into for-linus

Prepare input updates for 6.11 merge window.

Merge tag 'v6.9' into next

Sync up with the mainline to bring in the new cleanup API.

Merge drm/drm-next into drm-intel-gt-next

Work that needs to land in drm-intel-gt-next depends on two patches only
present in drm-intel-next, absence of which is causing a merge conflict:

3b918f4

Merge drm/drm-next into drm-intel-gt-next

Work that needs to land in drm-intel-gt-next depends on two patches only
present in drm-intel-next, absence of which is causing a merge conflict:

3b918f4f0c8b ("drm/i915/pxp: Optimize GET_PARAM:PXP_STATUS")
ac765b7018f6 ("drm/i915/pxp/mtl: intel_pxp_init_hw needs runtime-pm inside pm-complete")

Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>

show more ...

Merge branch 'linus' into smp/core

Pull in upstream to get the fixes so depending changes can be applied.

Merge drm/drm-next into drm-misc-next

Updating drm-misc-next to the state of Linux v6.6-rc2.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>

Merge tag 'v6.6-rc4' into perf/core, to pick up fixes

Signed-off-by: Ingo Molnar <mingo@kernel.org>

Merge tag 'v6.6-rc2' into locking/core, to pick up fixes

Signed-off-by: Ingo Molnar <mingo@kernel.org>

Merge drm/drm-next into drm-intel-next

Sync to v6.6-rc1.

Signed-off-by: Jani Nikula <jani.nikula@intel.com>

Merge drm/drm-fixes into drm-misc-fixes

Forwarding to v6.6-rc1.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>

Merge tag 'docs-6.6' of git://git.lwn.net/linux

Pull documentation updates from Jonathan Corbet:
"Documentation work keeps chugging along; this includes:

- Work from Carlos Bilbao to integrate

Merge tag 'docs-6.6' of git://git.lwn.net/linux

Pull documentation updates from Jonathan Corbet:
"Documentation work keeps chugging along; this includes:

- Work from Carlos Bilbao to integrate rustdoc output into the
generated HTML documentation. This took some work to figure out how
to do it without slowing the docs build and without creating people
who don't have Rust installed, but Carlos got there

- Move the loongarch and mips architecture documentation under
Documentation/arch/

- Some more maintainer documentation from Jakub

... plus the usual assortment of updates, translations, and fixes"

* tag 'docs-6.6' of git://git.lwn.net/linux: (56 commits)
Docu: genericirq.rst: fix irq-example
input: docs: pxrc: remove reference to phoenix-sim
Documentation: serial-console: Fix literal block marker
docs/mm: remove references to hmm_mirror ops and clean typos
docs/zh_CN: correct regi_chg(),regi_add() to region_chg(),region_add()
Documentation: Fix typos
Documentation/ABI: Fix typos
scripts: kernel-doc: fix macro handling in enums
scripts: kernel-doc: parse DEFINE_DMA_UNMAP_[ADDR|LEN]
Documentation: riscv: Update boot image header since EFI stub is supported
Documentation: riscv: Add early boot document
Documentation: arm: Add bootargs to the table of added DT parameters
docs: kernel-parameters: Refer to the correct bitmap function
doc: update params of memhp_default_state=
docs: Add book to process/kernel-docs.rst
docs: sparse: fix invalid link addresses
docs: vfs: clean up after the iterate() removal
docs: Add a section on surveys to the researcher guidelines
docs: move mips under arch
docs: move loongarch under arch
...

show more ...

Documentation/ABI: Fix typos

Fix typos in Documentation/ABI. The changes are in descriptions or
comments where they shouldn't affect use of the ABIs.

Signed-off-by: Bjorn Helgaas <bhelgaas@google.

Documentation/ABI: Fix typos

Fix typos in Documentation/ABI. The changes are in descriptions or
comments where they shouldn't affect use of the ABIs.

Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Randy Dunlap <rdunlap@infradead.org>
Link: https://lore.kernel.org/r/20230814212822.193684-2-helgaas@kernel.org
Signed-off-by: Jonathan Corbet <corbet@lwn.net>

show more ...

Merge branch 'next' into for-linus

Prepare input updates for 5.13 merge window.

Merge tag 'v5.12-rc7' into ecryptfs/next

Required to pick up idmapped mount changes which changed some function
parameters.

Merge tag 'v5.12-rc4' into next

Sync up with the mainline to bring in newest APIs.

Merge drm/drm-next into drm-misc-next

Noralf needs some patches in 5.12-rc3, and we've been delaying the 5.12
merge due to the swap issue so it looks like a good time.

Signed-off-by: Maxime Ripard

Merge drm/drm-next into drm-misc-next

Noralf needs some patches in 5.12-rc3, and we've been delaying the 5.12
merge due to the swap issue so it looks like a good time.

Signed-off-by: Maxime Ripard <maxime@cerno.tech>

show more ...

Merge tag 'irqchip-fixes-5.12-1' of git://git.kernel.org/pub/scm/linux/kernel/git/maz/arm-platforms into irq/urgent

Pull irqchip fixes from Marc Zyngier:

- More compatible strings for the Ingenic

Merge tag 'irqchip-fixes-5.12-1' of git://git.kernel.org/pub/scm/linux/kernel/git/maz/arm-platforms into irq/urgent

Pull irqchip fixes from Marc Zyngier:

- More compatible strings for the Ingenic irqchip (introducing the
JZ4760B SoC)
- Select GENERIC_IRQ_MULTI_HANDLER on the ARM ep93xx platform
- Drop all GENERIC_IRQ_MULTI_HANDLER selections from the irqchip
Kconfig, now relying on the architecture to get it right
- Drop the debugfs_file field from struct irq_domain, now that
debugfs can track things on its own

show more ...

Merge drm/drm-next into drm-intel-next

Sync up with upstream.

Signed-off-by: Jani Nikula <jani.nikula@intel.com>

Merge branch 'fixes-rc2' into fixes

Merge remote-tracking branch 'net/master'

Merge branch 'next' into for-linus

Prepare input updates for 5.12 merge window.

Merge tag 'v5.11' into next

Merge with mainline to get latest APIs and device tree bindings.

Merge commit '3e10585335b7967326ca7b4118cada0d2d00a2ab' into v5.12/vfio/next

Update to new follow_pte() definition

Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next

Pull networking updates from David Miller:
"Here is what we have this merge window:

1) Support SW steering for mlx5 Connect-X

Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next

Pull networking updates from David Miller:
"Here is what we have this merge window:

1) Support SW steering for mlx5 Connect-X6Dx, from Yevgeny Kliteynik.

2) Add RSS multi group support to octeontx2-pf driver, from Geetha
Sowjanya.

3) Add support for KS8851 PHY. From Marek Vasut.

4) Add support for GarfieldPeak bluetooth controller from Kiran K.

5) Add support for half-duplex tcan4x5x can controllers.

6) Add batch skb rx processing to bcrm63xx_enet, from Sieng Piaw
Liew.

7) Rework RX port offload infrastructure, particularly wrt, UDP
tunneling, from Jakub Kicinski.

8) Add BCM72116 PHY support, from Florian Fainelli.

9) Remove Dsa specific notifiers, they are unnecessary. From Vladimir
Oltean.

10) Add support for picosecond rx delay in dwmac-meson8b chips. From
Martin Blumenstingl.

11) Support TSO on xfrm interfaces from Eyal Birger.

12) Add support for MP_PRIO to mptcp stack, from Geliang Tang.

13) Support BCM4908 integrated switch, from Rafał Miłecki.

14) Support for directly accessing kernel module variables via module
BTF info, from Andrii Naryiko.

15) Add DASH (esktop and mobile Architecture for System Hardware)
support to r8169 driver, from Heiner Kallweit.

16) Add rx vlan filtering to dpaa2-eth, from Ionut-robert Aron.

17) Add support for 100 base0x SFP devices, from Bjarni Jonasson.

18) Support link aggregation in DSA, from Tobias Waldekranz.

19) Support for bitwidse atomics in bpf, from Brendan Jackman.

20) SmartEEE support in at803x driver, from Russell King.

21) Add support for flow based tunneling to GTP, from Pravin B Shelar.

22) Allow arbitrary number of interconnrcts in ipa, from Alex Elder.

23) TLS RX offload for bonding, from Tariq Toukan.

24) RX decap offklload support in mac80211, from Felix Fietkou.

25) devlink health saupport in octeontx2-af, from George Cherian.

26) Add TTL attr to SCM_TIMESTAMP_OPT_STATS, from Yousuk Seung

27) Delegated actionss support in mptcp, from Paolo Abeni.

28) Support receive timestamping when doin zerocopy tcp receive. From
Arjun Ray.

29) HTB offload support for mlx5, from Maxim Mikityanskiy.

30) UDP GRO forwarding, from Maxim Mikityanskiy.

31) TAPRIO offloading in dsa hellcreek driver, from Kurt Kanzenbach.

32) Weighted random twos choice algorithm for ipvs, from Darby Payne.

33) Fix netdev registration deadlock, from Johannes Berg.

34) Various conversions to new tasklet api, from EmilRenner Berthing.

35) Bulk skb allocations in veth, from Lorenzo Bianconi.

36) New ethtool interface for lane setting, from Danielle Ratson.

37) Offload failiure notifications for routes, from Amit Cohen.

38) BCM4908 support, from Rafał Miłecki.

39) Support several new iwlwifi chips, from Ihab Zhaika.

40) Flow drector support for ipv6 in i40e, from Przemyslaw Patynowski.

41) Support for mhi prrotocols, from Loic Poulain.

42) Optimize bpf program stats.

43) Implement RFC6056, for better port randomization, from Eric
Dumazet.

44) hsr tag offloading support from George McCollister.

45) Netpoll support in qede, from Bhaskar Upadhaya.

46) 2005/400g speed support in bonding 3ad mode, from Nikolay
Aleksandrov.

47) Netlink event support in mptcp, from Florian Westphal.

48) Better skbuff caching, from Alexander Lobakin.

49) MRP (Media Redundancy Protocol) offloading in DSA and a few
drivers, from Horatiu Vultur.

50) mqprio saupport in mvneta, from Maxime Chevallier.

51) Remove of_phy_attach, no longer needed, from Florian Fainelli"

* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (1766 commits)
octeontx2-pf: Fix otx2_get_fecparam()
cteontx2-pf: cn10k: Prevent harmless double shift bugs
net: stmmac: Add PCI bus info to ethtool driver query output
ptp: ptp_clockmatrix: clean-up - parenthesis around a == b are unnecessary
ptp: ptp_clockmatrix: Simplify code - remove unnecessary `err` variable.
ptp: ptp_clockmatrix: Coding style - tighten vertical spacing.
ptp: ptp_clockmatrix: Clean-up dev_*() messages.
ptp: ptp_clockmatrix: Remove unused header declarations.
ptp: ptp_clockmatrix: Add alignment of 1 PPS to idtcm_perout_enable.
ptp: ptp_clockmatrix: Add wait_for_sys_apll_dpll_lock.
net: stmmac: dwmac-sun8i: Add a shutdown callback
net: stmmac: dwmac-sun8i: Minor probe function cleanup
net: stmmac: dwmac-sun8i: Use reset_control_reset
net: stmmac: dwmac-sun8i: Remove unnecessary PHY power check
net: stmmac: dwmac-sun8i: Return void from PHY unpower
r8169: use macro pm_ptr
net: mdio: Remove of_phy_attach()
net: mscc: ocelot: select PACKING in the Kconfig
net: re-solve some conflicts after net -> net-next merge
net: dsa: tag_rtl4_a: Support also egress tags
...

show more ...

Merge branch 'implement-kthread-based-napi-poll'

Wei Wang says:

====================
implement kthread based napi polle

The idea of moving the napi poll process out of softirq context to a
kernel

Merge branch 'implement-kthread-based-napi-poll'

Wei Wang says:

====================
implement kthread based napi polle

The idea of moving the napi poll process out of softirq context to a
kernel thread based context is not new.
Paolo Abeni and Hannes Frederic Sowa have proposed patches to move napi
poll to kthread back in 2016. And Felix Fietkau has also proposed
patches of similar ideas to use workqueue to process napi poll just a
few weeks ago.

The main reason we'd like to push forward with this idea is that the
scheduler has poor visibility into cpu cycles spent in softirq context,
and is not able to make optimal scheduling decisions of the user threads.
For example, we see in one of the application benchmark where network
load is high, the CPUs handling network softirqs has ~80% cpu util. And
user threads are still scheduled on those CPUs, despite other more idle
cpus available in the system. And we see very high tail latencies. In this
case, we have to explicitly pin away user threads from the CPUs handling
network softirqs to ensure good performance.
With napi poll moved to kthread, scheduler is in charge of scheduling both
the kthreads handling network load, and the user threads, and is able to
make better decisions. In the previous benchmark, if we do this and we
pin the kthreads processing napi poll to specific CPUs, scheduler is
able to schedule user threads away from these CPUs automatically.

And the reason we prefer 1 kthread per napi, instead of 1 workqueue
entity per host, is that kthread is more configurable than workqueue,
and we could leverage existing tuning tools for threads, like taskset,
chrt, etc to tune scheduling class and cpu set, etc. Another reason is
if we eventually want to provide busy poll feature using kernel threads
for napi poll, kthread seems to be more suitable than workqueue.
Furthermore, for large platforms with 2 NICs attached to 2 sockets,
kthread is more flexible to be pinned to different sets of CPUs.

In this patch series, I revived Paolo and Hannes's patch in 2016 and
made modifications. Then there are changes proposed by Felix, Jakub,
Paolo and myself on top of those, with suggestions from Eric Dumazet.

In terms of performance, I ran tcp_rr tests with 1000 flows with
various request/response sizes, with RFS/RPS disabled, and compared
performance between softirq vs kthread vs workqueue (patchset proposed
by Felix Fietkau).
Host has 56 hyper threads and 100Gbps nic, 8 rx queues and only 1 numa
node. All threads are unpinned.

req/resp QPS 50%tile 90%tile 99%tile 99.9%tile
softirq 1B/1B 2.75M 337us 376us 1.04ms 3.69ms
kthread 1B/1B 2.67M 371us 408us 455us 550us
workq 1B/1B 2.56M 384us 435us 673us 822us

softirq 5KB/5KB 1.46M 678us 750us 969us 2.78ms
kthread 5KB/5KB 1.44M 695us 789us 891us 1.06ms
workq 5KB/5KB 1.34M 720us 905us 1.06ms 1.57ms

softirq 1MB/1MB 11.0K 79ms 166ms 306ms 630ms
kthread 1MB/1MB 11.0K 75ms 177ms 303ms 596ms
workq 1MB/1MB 11.0K 79ms 180ms 303ms 587ms

When running workqueue implementation, I found the number of threads
used is usually twice as much as kthread implementation. This probably
introduces higher scheduling cost, which results in higher tail
latencies in most cases.

I also ran an application benchmark, which performs fixed qps remote SSD
read/write operations, with various sizes. Again, both with RFS/RPS
disabled.
The result is as follows:
op_size QPS 50%tile 95%tile 99%tile 99.9%tile
softirq 4K 572.6K 385us 1.5ms 3.16ms 6.41ms
kthread 4K 572.6K 390us 803us 2.21ms 6.83ms
workq 4k 572.6K 384us 763us 3.12ms 6.87ms

softirq 64K 157.9K 736us 1.17ms 3.40ms 13.75ms
kthread 64K 157.9K 745us 1.23ms 2.76ms 9.87ms
workq 64K 157.9K 746us 1.23ms 2.76ms 9.96ms

softirq 1M 10.98K 2.03ms 3.10ms 3.7ms 11.56ms
kthread 1M 10.98K 2.13ms 3.21ms 4.02ms 13.3ms
workq 1M 10.98K 2.13ms 3.20ms 3.99ms 14.12ms

In this set of tests, the latency is predominant by the SSD operation.
Also, the user threads are much busier compared to tcp_rr tests. We have
to pin the kthreads/workqueue threads to limit to a few CPUs, to not
disturb user threads, and provide some isolation.

Changes since v9:
Small change in napi_poll() in patch 1.
Split napi_kthread_stop() functionality to add separately in
napi_disable() and netif_napi_del() in patch 2.
Add description for napi_set_threaded() and return dev->threaded when
dev->napi_list is empty for threaded sysfs in patch 3.

Changes since v8:
Added description for threaded param in struct net_device in patch 2.

Changes since v7:
Break napi_set_threaded() into 2 parts, one to create kthread called
from netif_napi_add(), the other to set threaded bit in napi_enable(),
to get rid of inconsistency through all napi in 1 dev.
Added documentation for /sys/class/net/<dev>/threaded.

Changes since v6:
Added memory barrier in napi_set_threaded().
Changed /sys/class/net/<dev>/thread to a ternary value.
Change dev->threaded to a bit instead of bool.

Changes since v5:
Removed ASSERT_RTNL() from napi_set_threaded() and removed rtnl_lock()
operation from napi_enable().

Changes since v4:
Recorded the threaded setting in dev and restore it in napi_enable().

Changes since v3:
Merged and rearranged patches in a logical order for easier review.
Changed sysfs control to be per device.

Changes since v2:
Corrected typo in patch 1, and updated the cover letter with more
detailed and updated test results.

Changes since v1:
Replaced kthread_create() with kthread_run() in patch 5 as suggested by
Felix Fietkau.

Changes since RFC:
Renamed the kthreads to be napi/<dev>-<napi_id> in patch 5 as suggested
by Hannes Frederic Sowa.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

show more ...

net: add sysfs attribute to control napi threaded mode

This patch adds a new sysfs attribute to the network device class.
Said attribute provides a per-device control to enable/disable the
threaded

net: add sysfs attribute to control napi threaded mode

This patch adds a new sysfs attribute to the network device class.
Said attribute provides a per-device control to enable/disable the
threaded mode for all the napi instances of the given network device,
without the need for a device up/down.
User sets it to 1 or 0 to enable or disable threaded mode.
Note: when switching between threaded and the current softirq based mode
for a napi instance, it will not immediately take effect if the napi is
currently being polled. The mode switch will happen for the next time
napi_schedule() is called.

Co-developed-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Co-developed-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Co-developed-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Wei Wang <weiwan@google.com>
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

show more ...