- ops.enqueue() didn't have a way to tell whether select_task_rq_scx() and
thus ops.select() were skipped. Some schedulers were incorrectly using
SCX_ENQ_WAKEUP. Add SCX_ENQ_CPU_SELECTED and fix scx_qmap using it.
- Remove a spurious WARN_ON_ONCE() in scx_cgroup_exit().
- Fix error information clobbering during load.
- Add missing __weak markers to BPF helper declarations.
- Doc update.
-----BEGIN PGP SIGNATURE-----
iIQEABYKACwWIQTfIjM1kS57o3GsC/uxYfJx3gVYGQUCZwWKkA4cdGpAa2VybmVs
Lm9yZwAKCRCxYfJx3gVYGelnAQDTA8GSIahTEHKM0c3yXE6K1/M56zo8Spp5OOA7
kXHR3AD/Y0RcXgaCvMI13aozmQWq756gyB6/qczN0+X3jx6wZwI=
=6xbe
-----END PGP SIGNATURE-----
Merge tag 'sched_ext-for-6.12-rc2-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext
Pull sched_ext fixes from Tejun Heo:
- ops.enqueue() didn't have a way to tell whether select_task_rq_scx()
and thus ops.select() were skipped. Some schedulers were incorrectly
using SCX_ENQ_WAKEUP. Add SCX_ENQ_CPU_SELECTED and fix scx_qmap using
it.
- Remove a spurious WARN_ON_ONCE() in scx_cgroup_exit()
- Fix error information clobbering during load
- Add missing __weak markers to BPF helper declarations
- Doc update
* tag 'sched_ext-for-6.12-rc2-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext:
sched_ext: Documentation: Update instructions for running example schedulers
sched_ext, scx_qmap: Add and use SCX_ENQ_CPU_SELECTED
sched/core: Add ENQUEUE_RQ_SELECTED to indicate whether ->select_task_rq() was called
sched/core: Make select_task_rq() take the pointer to wake_flags instead of value
sched_ext: scx_cgroup_exit() may be called without successful scx_cgroup_init()
sched_ext: Improve error reporting during loading
sched_ext: Add __weak markers to BPF helper function decalarations
The key_free LSM hook has been removed.
Remove the corresponding BPF hook.
Avoid warnings during the build:
BTFIDS vmlinux
WARN: resolve_btfids: unresolved symbol bpf_lsm_key_free
Fixes: 5f8d28f6d7 ("lsm: infrastructure management of the key security blob")
Signed-off-by: Thomas Weißschuh <linux@weissschuh.net>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Song Liu <song@kernel.org>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: https://lore.kernel.org/bpf/20241005-lsm-key_free-v1-1-42ea801dbd63@weissschuh.net
No longer document drivers should switch to lib80211,
they really should never have done that. While at it,
also remove the recommendation to use cfg80211, if it
switches to mac80211 then it implicitly uses cfg80211
but doesn't need to do anything about that, normally.
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Link: https://patch.msgid.link/20241007202707.87481ddcfc00.I2cfb9940807e9c5017a052efcd3d1f2b6dc15fb1@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
This driver is using lib80211 and any driver that plans to ever
leave staging should never have done that, so remove the driver
to enable cleaning up lib80211 into libipw inside the old Intel
drivers.
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Link: https://patch.msgid.link/20241007202707.d0e59cdd2cdc.I8e4d74a6e1d09eefe1f5e2e208735ba2ccef1d4f@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
In 'cfg80211_free_coalesce', '&coalesce->rules[i]' is a pointer
to VLA member of 'struct cfg80211_coalesce' and should never be NULL,
so redundant check may be dropped.
I think this is correct, but I haven't tested it seriously.
Compile tested only.
Found by Linux Verification Center (linuxtesting.org) with SVACE.
Signed-off-by: Dmitry Kandybka <d.kandybka@gmail.com>
Link: https://patch.msgid.link/20241003095912.218465-1-d.kandybka@gmail.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Reorganize kerneldoc parameter names to match the parameter
order in the function header.
Problems identified using Coccinelle.
Signed-off-by: Julia Lawall <Julia.Lawall@inria.fr>
Reviewed-by: Jeff Johnson <quic_jjohnson@quicinc.com>
Link: https://patch.msgid.link/20240930112121.95324-28-Julia.Lawall@inria.fr
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Currently, wiphy_verify_combinations() fails for the multi-radio per wiphy
due to the condition check on new global interface combination that DFS
only works on one channel. In a multi-radio scenario, new global interface
combination encompasses the capabilities of all radio combinations, so it
supports more than one channel with DFS. For multi-radio per wiphy,
interface combination verification needs to be performed for radio specific
interface combinations. This is necessary as the new global interface
combination combines the capabilities of all radio combinations.
Fixes: a01b1e9f99 ("wifi: mac80211: add support for DFS with multiple radios")
Signed-off-by: Karthikeyan Periyasamy <quic_periyasa@quicinc.com>
Link: https://patch.msgid.link/20240917140239.886083-1-quic_periyasa@quicinc.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
The chandef parameter passed to ieee80211_ie_build_he_oper() and
ieee80211_ie_build_eht_oper is read-only. Since it is never modified,
add the const qualifier to this parameter. This makes these consistent
with ieee80211_ie_build_ht_oper() and ieee80211_ie_build_vht_oper().
Signed-off-by: Jeff Johnson <quic_jjohnson@quicinc.com>
Link: https://patch.msgid.link/20240910-wireless-utils-constify-v1-1-e59947bcb3c3@quicinc.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
After commit 0edb555a65 ("platform: Make platform_driver::remove()
return void") .remove() is (again) the right callback to implement for
platform drivers.
Convert all platform drivers below drivers/video/fbdev to use .remove(),
with the eventual goal to drop struct platform_driver::remove_new(). As
.remove() and .remove_new() have the same prototypes, conversion is done
by just changing the structure member name in the driver initializer.
While touching these files, make indention of the struct initializer
consistent in several files.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@baylibre.com>
Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Helge Deller <deller@gmx.de>
Similar to the MSR RAPL interface, MMIO RAPL supports PL4 too, so add
MMIO RAPL PL4d support to the processor_thermal driver.
As a result, the powercap sysfs for MMIO RAPL will show a new "peak
power" constraint.
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Reviewed-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://patch.msgid.link/20240930081801.28502-7-rui.zhang@intel.com
[ rjw: Subject and changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
CPU0/package0 is always online and the MMIO RAPL driver runs on single
package systems only, so there is no need to handle CPU hotplug in it.
Always register a RAPL package device for package 0 and remove the
unnecessary CPU hotplug support.
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Reviewed-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://patch.msgid.link/20240930081801.28502-6-rui.zhang@intel.com
[ rjw: Subject edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The hardware definition of every TPMI feature contains a major and minor
version. When there is a change in the MMIO offset or change in the
definition of a field, hardware will change major version. For addition
of new fields without modifying existing MMIO offsets or fields, only
the minor version is changed.
If the driver has not been updated to recognize a new hardware major
version, it cannot provide the RAPL interface to users due to possible
register layout incompatibilities. However, the driver does not need to
be updated every time the hardware minor version changes because in that
case it will just miss some new functionality exposed by the hardware.
The current implementation causes the driver to refuse to work for any
hardware version change which is unnecessarily restrictive.
If there is a minor version mismatch, log an information message and
continue, but if there is a major version mismatch, log a warning and
exit (as before).
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Reviewed-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://patch.msgid.link/20240930081801.28502-4-rui.zhang@intel.com
Fixes: 9eef7f9da9 ("powercap: intel_rapl: Introduce RAPL TPMI interface driver")
[ rjw: Changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
After commit dcb0b5575d ("tracing: Remove TRACE_EVENT_FL_USE_CALL_FILTER
logic"), no one's going to set the TRACE_EVENT_FL_FILTERED or change the
call->filter, so remove related logic.
Link: https://lore.kernel.org/20240911010026.2302849-1-zhengyejian@huaweicloud.com
Signed-off-by: Zheng Yejian <zhengyejian@huaweicloud.com>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Use this_cpu_ptr() instead of open coding the equivalent in various
ftrace functions.
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Link: https://lore.kernel.org/87y14t6ofi.wl-me@linux.beauty
Signed-off-by: Li Chen <chenl311@chinatelecom.cn>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
-Wflex-array-member-not-at-end was introduced in GCC-14, and we are
getting ready to enable it, globally.
So, in order to avoid ending up with a flexible-array member in the
middle of multiple other structs, we use the `__struct_group()`
helper to create a new tagged `struct ieee80211_radiotap_header_fixed`.
This structure groups together all the members of the flexible
`struct ieee80211_radiotap_header` except the flexible array.
As a result, the array is effectively separated from the rest of the
members without modifying the memory layout of the flexible structure.
We then change the type of the middle struct members currently causing
trouble from `struct ieee80211_radiotap_header` to `struct
ieee80211_radiotap_header_fixed`.
We also want to ensure that in case new members need to be added to the
flexible structure, they are always included within the newly created
tagged struct. For this, we use `static_assert()`. This ensures that the
memory layout for both the flexible structure and the new tagged struct
is the same after any changes.
This approach avoids having to implement `struct ieee80211_radiotap_header_fixed`
as a completely separate structure, thus preventing having to maintain
two independent but basically identical structures, closing the door
to potential bugs in the future.
So, with these changes, fix the following warnings:
drivers/net/wireless/ath/wil6210/txrx.c:309:50: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
drivers/net/wireless/intel/ipw2x00/ipw2100.c:2521:50: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
drivers/net/wireless/intel/ipw2x00/ipw2200.h:1146:42: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
drivers/net/wireless/intel/ipw2x00/libipw.h:595:36: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
drivers/net/wireless/marvell/libertas/radiotap.h:34:42: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
drivers/net/wireless/marvell/libertas/radiotap.h:5:42: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
drivers/net/wireless/microchip/wilc1000/mon.c:10:42: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
drivers/net/wireless/microchip/wilc1000/mon.c:15:42: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
drivers/net/wireless/virtual/mac80211_hwsim.c:758:42: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
drivers/net/wireless/virtual/mac80211_hwsim.c:767:42: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Link: https://patch.msgid.link/ZwBMtBZKcrzwU7l4@kspp
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Call to ieee80211_color_collision_detection_work() needs wiphy lock to
be held (see lockdep assert in cfg80211_bss_color_notify()). Not locking
wiphy causes the following lockdep error:
WARNING: CPU: 2 PID: 42 at net/wireless/nl80211.c:19505 cfg80211_bss_color_notify+0x1a4/0x25c
Modules linked in:
CPU: 2 PID: 42 Comm: kworker/u8:3 Tainted: G W 6.4.0-02327-g36c6cb260481 #1048
Hardware name:
Workqueue: phy1 ieee80211_color_collision_detection_work
pstate: 60000005 (nZCv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
pc : cfg80211_bss_color_notify+0x1a4/0x25c
lr : cfg80211_bss_color_notify+0x1a0/0x25c
sp : ffff000002947d00
x29: ffff000002947d00 x28: ffff800008e1a000 x27: ffff000002bd4705
x26: ffff00000d034000 x25: ffff80000903cf40 x24: 0000000000000000
x23: ffff00000cb70720 x22: 0000000000800000 x21: ffff800008dfb008
x20: 000000000000008d x19: ffff00000d035fa8 x18: 0000000000000010
x17: 0000000000000001 x16: 000003564b1ce96a x15: 000d69696d057970
x14: 000000000000003b x13: 0000000000000001 x12: 0000000000040000
x11: 0000000000000001 x10: ffff80000978f9c0 x9 : ffff0000028d3174
x8 : ffff800008e30000 x7 : 0000000000000000 x6 : 0000000000000028
x5 : 000000000002f498 x4 : ffff00000d034a80 x3 : 0000000000800000
x2 : ffff800016143000 x1 : 0000000000000000 x0 : 0000000000000000
Call trace:
cfg80211_bss_color_notify+0x1a4/0x25c
ieee80211_color_collision_detection_work+0x20/0x118
process_one_work+0x294/0x554
worker_thread+0x70/0x440
kthread+0xf4/0xf8
ret_from_fork+0x10/0x20
irq event stamp: 77372
hardirqs last enabled at (77371): [<ffff800008a346fc>] _raw_spin_unlock_irq+0x2c/0x4c
hardirqs last disabled at (77372): [<ffff800008a28754>] el1_dbg+0x20/0x48
softirqs last enabled at (77350): [<ffff8000089e120c>] batadv_send_outstanding_bcast_packet+0xb8/0x120
softirqs last disabled at (77348): [<ffff8000089e11d4>] batadv_send_outstanding_bcast_packet+0x80/0x120
The wiphy lock cannot be taken directly from color collision detection
delayed work (ieee80211_color_collision_detection_work()) because this
work is cancel_delayed_work_sync() under this wiphy lock causing a
potential deadlock( see [0] for details).
To fix that ieee80211_color_collision_detection_work() could be
converted to a wiphy work and cancel_delayed_work_sync() can be simply
replaced by wiphy_delayed_work_cancel() serving the same purpose under
wiphy lock.
This could potentially fix [1].
[0]: https://lore.kernel.org/linux-wireless/D4A40Q44OAY2.W3SIF6UEPBUN@freebox.fr/
[1]: https://lore.kernel.org/lkml/000000000000612f290618eee3e5@google.com/
Reported-by: Nicolas Escande <nescande@freebox.fr>
Signed-off-by: Remi Pommarel <repk@triplefau.lt>
Link: https://patch.msgid.link/20240924192805.13859-3-repk@triplefau.lt
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Add wiphy_delayed_work_pending() to check if any delayed work timer is
pending, that can be used to be sure that wiphy_delayed_work_queue()
won't postpone an already pending delayed work.
Signed-off-by: Remi Pommarel <repk@triplefau.lt>
Link: https://patch.msgid.link/20240924192805.13859-2-repk@triplefau.lt
[fix return value kernel-doc]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
For most compatibles, the "brcm,bluetooth.yaml" binding doesn't allow
the 'reset-gpios' property, but there is a 'shutdown-gpios' property.
Page 12 of the AzureWave-CM256SM datasheet (v1.9) has the following wrt
pin 34 'BT_REG_ON' (connected to GPIO0_C4_d on the PineNote):
Used by PMU to power up or power down the internal regulators used
by the Bluetooth section. Also, when deasserted, this pin holds the
Bluetooth section in reset. This pin has an internal 200k ohm pull
down resistor that is enabled by default.
So it is safe to replace 'reset-gpios' with 'shutdown-gpios'.
Fixes: d449121e5e ("arm64: dts: rockchip: Add Pine64 PineNote board")
Signed-off-by: Diederik de Haas <didi.debian@cknow.org>
Link: https://lore.kernel.org/r/20241008113344.23957-5-didi.debian@cknow.org
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
The "brcm,bluetooth.yaml" binding has 'device-wakeup-gpios' and
'host-wakeup-gpios' property names, not '*-wake-gpios'.
Fix the incorrect property names.
Note that the "realtek,bluetooth.yaml" binding does use the
'*-wake-gpios' property names.
Fixes: d449121e5e ("arm64: dts: rockchip: Add Pine64 PineNote board")
Signed-off-by: Diederik de Haas <didi.debian@cknow.org>
Link: https://lore.kernel.org/r/20241008113344.23957-4-didi.debian@cknow.org
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
The "synopsys,dw-hdmi.yaml" binding specifies that the interrupts
property of the hdmi node has 'maxItems: 1', so the hdmi node in
rk3328.dtsi having 2 is incorrect.
Paragraph 1.3 ("System Interrupt connection") of the RK3328 TRM v1.1
page 16 and 17 define the following hdmi related interrupts:
- 67 hdmi_intr
- 103 hdmi_intr_wakeup
The difference of 32 is due to a different base used in the TRM.
The RK3399 (which uses the same binding) has '23: hdmi_irq' and
'24: hdmi_wakeup_irq' according to its TRM (page 19).
The RK3568 (also same binding) has '76: hdmi_wakeup' and '77: hdmi'
according to page 17 of its TRM.
In both cases the non-wakeup IRQ was used, so use that too for rk3328.
Helped-by: Heiko Stuebner <heiko@sntech.de>
Fixes: 725e351c26 ("arm64: dts: rockchip: add rk3328 display nodes")
Signed-off-by: Diederik de Haas <didi.debian@cknow.org>
Link: https://lore.kernel.org/r/20241008113344.23957-3-didi.debian@cknow.org
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Currently, in cfg80211_parse_ml_elem_sta_data(), when RNR element
indicates a BSS that operates in a channel that current regulatory
domain doesn't support, a NULL value is returned by
ieee80211_get_channel_khz() and assigned to this BSS entry's channel
field. Later in cfg80211_inform_single_bss_data(), the reported
BSS entry's channel will be wrongly overridden by transmitted BSS's.
This could result in connection failure that when wpa_supplicant
tries to select this reported BSS entry while it actually resides in
an unsupported channel.
Since this channel is not supported, it is reasonable to skip such
entries instead of reporting wrong information.
Signed-off-by: Chenming Huang <quic_chenhuan@quicinc.com>
Link: https://patch.msgid.link/20240923021644.12885-1-quic_chenhuan@quicinc.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Convert RK3328 clock controller bindings to DT schema
Changes against original bindings:
- Add clocks and clock-names as the device has at least one input clock.
Signed-off-by: Johan Jonker <jbx6244@gmail.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
[add Krzysztof's review from v1, shorten commit description]
Link: https://lore.kernel.org/r/20240930215001.1999212-2-heiko@sntech.de
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Add clocks property to rk3328 cru node to fix warnings like:
'clocks' is a dependency of 'assigned-clocks'
Signed-off-by: Johan Jonker <jbx6244@gmail.com>
Link: https://lore.kernel.org/r/20240930215001.1999212-4-heiko@sntech.de
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
The cru node references undocumented compatibles of "rockchip,cru" and also
marks it as syscon.
A general rockchip,cru is way too generic to ever be used anywhere, so
needs to go away, similarly the cru should not be written to from other
places, instead regular clock routines should be used.
Both mainline Linux as well as the vendor-kernel up to their 6.1 branch
only reference the cru via the normal assigned-clocks, clocks and resets
properties and do not get a syscon from the node.
Similarly, there is no syscon access by compatible both in mainline
nor the vendor-kernel up to their 6.1 branch, through either the
rockchip,rk3328-cru nor rockchip,cru compatibles.
So these two really are unused in all publically visible places.
Sidenote: the vendor-kernel does pretty crazy stuff in the camera interface
and tdm driver, where they map the cru separately and set clock muxes and
gates directly. This should of course never reach mainline anyway.
Signed-off-by: Johan Jonker <jbx6244@gmail.com>
[update commit message, to explain the unused compatibles]
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/20240930215001.1999212-3-heiko@sntech.de
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Add netns cookie test that verifies the helper is now supported and work
in the context of tc programs.
Signed-off-by: Mahe Tardy <mahe.tardy@gmail.com>
Link: https://lore.kernel.org/r/20241007095958.97442-2-mahe.tardy@gmail.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
This is needed in the context of Cilium and Tetragon to retrieve netns
cookie from hostns when traffic leaves Pod, so that we can correlate
skb->sk's netns cookie.
Signed-off-by: Mahe Tardy <mahe.tardy@gmail.com>
Link: https://lore.kernel.org/r/20241007095958.97442-1-mahe.tardy@gmail.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
After a recent LLVM change [1] that deduces __cold on functions that only call
cold code (such as __init functions), there is a section mismatch warning from
__get_mem_config_intel(), which got moved to .text.unlikely. as a result of
that optimization:
WARNING: modpost: vmlinux: section mismatch in reference: \
__get_mem_config_intel+0x77 (section: .text.unlikely.) -> thread_throttle_mode_init (section: .init.text)
Mark __get_mem_config_intel() as __init as well since it is only called
from __init code, which clears up the warning.
While __rdt_get_mem_config_amd() does not exhibit a warning because it
does not call any __init code, it is a similar function that is only
called from __init code like __get_mem_config_intel(), so mark it __init
as well to keep the code symmetrical.
CONFIG_SECTION_MISMATCH_WARN_ONLY=n would turn this into a fatal error.
Fixes: 05b93417ce ("x86/intel_rdt/mba: Add primary support for Memory Bandwidth Allocation (MBA)")
Fixes: 4d05bf71f1 ("x86/resctrl: Introduce AMD QOS feature")
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Reviewed-by: Reinette Chatre <reinette.chatre@intel.com>
Cc: <stable@kernel.org>
Link: 6b11573b8c [1]
Link: https://lore.kernel.org/r/20240917-x86-restctrl-get_mem_config_intel-init-v3-1-10d521256284@kernel.org
Creates a cgroup with a single nice CPU hog process running.
fork() is called to generate the nice process because un-nicing is
not possible (see man nice(3)). If fork() was not used to generate
the CPU hog, we would run the rest of the cgroup selftest suite as a
nice process.
Signed-off-by: Joshua Hahn <joshua.hahnjy@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Cgroup-level CPU statistics currently include time spent on
user/system processes, but do not include niced CPU time (despite
already being tracked). This patch exposes niced CPU time to the
userspace, allowing users to get a better understanding of their
hardware limits and can facilitate more informed workload distribution.
A new field 'ntime' is added to struct cgroup_base_stat as opposed to
struct task_cputime to minimize footprint.
Signed-off-by: Joshua Hahn <joshua.hahnjy@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Since the artifact paths for tools changed, we need to update the documentation to reflect that path.
Signed-off-by: Devaansh-Kumar <devaanshk840@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
WQ_MAX_ACTIVE is currently set to 512, which was established approximately
15 yeas ago. However, with the significant increase in machine sizes and
capabilities, the previous limit of 256 concurrent tasks is no longer
sufficient. Therefore, we propose to increase WQ_MAX_ACTIVE to 2048.
and WQ_DFL_ACTIVE is 1024 now.
Signed-off-by: Chen Ridong <chenridong@huawei.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
If something is expected to generate large number of concurrent works,
it should utilize its own dedicated workqueue rather than system wq.
Because this may saturate system_wq and potentially block other's works.
eg, cgroup release work. Let's document this as a note.
Signed-off-by: Chen Ridong <chenridong@huawei.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
A hung_task problem shown below was found:
INFO: task kworker/0:0:8 blocked for more than 327 seconds.
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
Workqueue: events cgroup_bpf_release
Call Trace:
<TASK>
__schedule+0x5a2/0x2050
? find_held_lock+0x33/0x100
? wq_worker_sleeping+0x9e/0xe0
schedule+0x9f/0x180
schedule_preempt_disabled+0x25/0x50
__mutex_lock+0x512/0x740
? cgroup_bpf_release+0x1e/0x4d0
? cgroup_bpf_release+0xcf/0x4d0
? process_scheduled_works+0x161/0x8a0
? cgroup_bpf_release+0x1e/0x4d0
? mutex_lock_nested+0x2b/0x40
? __pfx_delay_tsc+0x10/0x10
mutex_lock_nested+0x2b/0x40
cgroup_bpf_release+0xcf/0x4d0
? process_scheduled_works+0x161/0x8a0
? trace_event_raw_event_workqueue_execute_start+0x64/0xd0
? process_scheduled_works+0x161/0x8a0
process_scheduled_works+0x23a/0x8a0
worker_thread+0x231/0x5b0
? __pfx_worker_thread+0x10/0x10
kthread+0x14d/0x1c0
? __pfx_kthread+0x10/0x10
ret_from_fork+0x59/0x70
? __pfx_kthread+0x10/0x10
ret_from_fork_asm+0x1b/0x30
</TASK>
This issue can be reproduced by the following pressuse test:
1. A large number of cpuset cgroups are deleted.
2. Set cpu on and off repeatly.
3. Set watchdog_thresh repeatly.
The scripts can be obtained at LINK mentioned above the signature.
The reason for this issue is cgroup_mutex and cpu_hotplug_lock are
acquired in different tasks, which may lead to deadlock.
It can lead to a deadlock through the following steps:
1. A large number of cpusets are deleted asynchronously, which puts a
large number of cgroup_bpf_release works into system_wq. The max_active
of system_wq is WQ_DFL_ACTIVE(256). Consequently, all active works are
cgroup_bpf_release works, and many cgroup_bpf_release works will be put
into inactive queue. As illustrated in the diagram, there are 256 (in
the acvtive queue) + n (in the inactive queue) works.
2. Setting watchdog_thresh will hold cpu_hotplug_lock.read and put
smp_call_on_cpu work into system_wq. However step 1 has already filled
system_wq, 'sscs.work' is put into inactive queue. 'sscs.work' has
to wait until the works that were put into the inacvtive queue earlier
have executed (n cgroup_bpf_release), so it will be blocked for a while.
3. Cpu offline requires cpu_hotplug_lock.write, which is blocked by step 2.
4. Cpusets that were deleted at step 1 put cgroup_release works into
cgroup_destroy_wq. They are competing to get cgroup_mutex all the time.
When cgroup_metux is acqured by work at css_killed_work_fn, it will
call cpuset_css_offline, which needs to acqure cpu_hotplug_lock.read.
However, cpuset_css_offline will be blocked for step 3.
5. At this moment, there are 256 works in active queue that are
cgroup_bpf_release, they are attempting to acquire cgroup_mutex, and as
a result, all of them are blocked. Consequently, sscs.work can not be
executed. Ultimately, this situation leads to four processes being
blocked, forming a deadlock.
system_wq(step1) WatchDog(step2) cpu offline(step3) cgroup_destroy_wq(step4)
...
2000+ cgroups deleted asyn
256 actives + n inactives
__lockup_detector_reconfigure
P(cpu_hotplug_lock.read)
put sscs.work into system_wq
256 + n + 1(sscs.work)
sscs.work wait to be executed
warting sscs.work finish
percpu_down_write
P(cpu_hotplug_lock.write)
...blocking...
css_killed_work_fn
P(cgroup_mutex)
cpuset_css_offline
P(cpu_hotplug_lock.read)
...blocking...
256 cgroup_bpf_release
mutex_lock(&cgroup_mutex);
..blocking...
To fix the problem, place cgroup_bpf_release works on a dedicated
workqueue which can break the loop and solve the problem. System wqs are
for misc things which shouldn't create a large number of concurrent work
items. If something is going to generate >WQ_DFL_ACTIVE(256) concurrent
work items, it should use its own dedicated workqueue.
Fixes: 4bfc0bb2c6 ("bpf: decouple the lifetime of cgroup_bpf from cgroup itself")
Cc: stable@vger.kernel.org # v5.3+
Link: https://lore.kernel.org/cgroups/e90c32d2-2a85-4f28-9154-09c7d320cb60@huawei.com/T/#t
Tested-by: Vishal Chourasia <vishalc@linux.ibm.com>
Signed-off-by: Chen Ridong <chenridong@huawei.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
New:
implement fallocate for compressed files;
add support for the compression attribute;
optimize large writes to sparse files.
Fixed:
fix several potential deadlock scenarios;
fix various internal bugs detected by syzbot;
add checks before accessing NTFS structures during parsing;
correct the format of output messages.
Refactored:
replace fsparam_flag_no with fsparam_flag in options parser;
remove unused functions and macros.
-----BEGIN PGP SIGNATURE-----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=j71V
-----END PGP SIGNATURE-----
Merge tag 'ntfs3_for_6.12' of https://github.com/Paragon-Software-Group/linux-ntfs3
Pull ntfs3 updates from Konstantin Komarov:
"New:
- implement fallocate for compressed files
- add support for the compression attribute
- optimize large writes to sparse files
Fixes:
- fix several potential deadlock scenarios
- fix various internal bugs detected by syzbot
- add checks before accessing NTFS structures during parsing
- correct the format of output messages
Refactoring:
- replace fsparam_flag_no with fsparam_flag in options parser
- remove unused functions and macros"
* tag 'ntfs3_for_6.12' of https://github.com/Paragon-Software-Group/linux-ntfs3: (25 commits)
fs/ntfs3: Format output messages like others fs in kernel
fs/ntfs3: Additional check in ntfs_file_release
fs/ntfs3: Fix general protection fault in run_is_mapped_full
fs/ntfs3: Sequential field availability check in mi_enum_attr()
fs/ntfs3: Additional check in ni_clear()
fs/ntfs3: Fix possible deadlock in mi_read
ntfs3: Change to non-blocking allocation in ntfs_d_hash
fs/ntfs3: Remove unused al_delete_le
fs/ntfs3: Rename ntfs3_setattr into ntfs_setattr
fs/ntfs3: Replace fsparam_flag_no -> fsparam_flag
fs/ntfs3: Add support for the compression attribute
fs/ntfs3: Implement fallocate for compressed files
fs/ntfs3: Make checks in run_unpack more clear
fs/ntfs3: Add rough attr alloc_size check
fs/ntfs3: Stale inode instead of bad
fs/ntfs3: Refactor enum_rstbl to suppress static checker
fs/ntfs3: Fix sparse warning in ni_fiemap
fs/ntfs3: Fix warning possible deadlock in ntfs_set_state
fs/ntfs3: Fix sparse warning for bigendian
fs/ntfs3: Separete common code for file_read/write iter/splice
...
