| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Pull vfs fixes from Christian Brauner:
- Fix a memory leak in cachefiles
- Restrict aio cancellations to I/O submitted through the aio
interfaces as this is otherwise causing issues for I/O submitted
via io_uring
- Increase buffer for afs volume status to avoid overflow
- Fix a missing zero-length check in unbuffered writes in the
netfs library. If generic_write_checks() returns zero make
netfs_unbuffered_write_iter() return right away
- Prevent a leak in i_dio_count caused by netfs_begin_read() operating
past i_size. It will return early and leave i_dio_count incremented
- Account for ipv4 addresses as well as ipv6 addresses when processing
incoming callbacks in afs
* tag 'vfs-6.8-rc6.fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
fs/aio: Restrict kiocb_set_cancel_fn() to I/O submitted via libaio
afs: Increase buffer size in afs_update_volume_status()
afs: Fix ignored callbacks over ipv4
cachefiles: fix memory leak in cachefiles_add_cache()
netfs: Fix missing zero-length check in unbuffered write
netfs: Fix i_dio_count leak on DIO read past i_size
|
|
The timer pull model is in place so we can remove the heuristics which try
to guess the best target CPU at enqueue/modification time.
All non pinned timers are queued on the local CPU in the separate storage
and eventually pulled at expiry time to a remote CPU.
Originally-by: Richard Cochran (linutronix GmbH) <[email protected]>
Signed-off-by: Anna-Maria Behnsen <[email protected]>
Signed-off-by: Thomas Gleixner <[email protected]>
Reviewed-by: Frederic Weisbecker <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
|
|
Placing timers at enqueue time on a target CPU based on dubious heuristics
does not make any sense:
1) Most timer wheel timers are canceled or rearmed before they expire.
2) The heuristics to predict which CPU will be busy when the timer expires
are wrong by definition.
So placing the timers at enqueue wastes precious cycles.
The proper solution to this problem is to always queue the timers on the
local CPU and allow the non pinned timers to be pulled onto a busy CPU at
expiry time.
Therefore split the timer storage into local pinned and global timers:
Local pinned timers are always expired on the CPU on which they have been
queued. Global timers can be expired on any CPU.
As long as a CPU is busy it expires both local and global timers. When a
CPU goes idle it arms for the first expiring local timer. If the first
expiring pinned (local) timer is before the first expiring movable timer,
then no action is required because the CPU will wake up before the first
movable timer expires. If the first expiring movable timer is before the
first expiring pinned (local) timer, then this timer is queued into an idle
timerqueue and eventually expired by another active CPU.
To avoid global locking the timerqueues are implemented as a hierarchy. The
lowest level of the hierarchy holds the CPUs. The CPUs are associated to
groups of 8, which are separated per node. If more than one CPU group
exist, then a second level in the hierarchy collects the groups. Depending
on the size of the system more than 2 levels are required. Each group has a
"migrator" which checks the timerqueue during the tick for remote expirable
timers.
If the last CPU in a group goes idle it reports the first expiring event in
the group up to the next group(s) in the hierarchy. If the last CPU goes
idle it arms its timer for the first system wide expiring timer to ensure
that no timer event is missed.
Signed-off-by: Anna-Maria Behnsen <[email protected]>
Signed-off-by: Thomas Gleixner <[email protected]>
Reviewed-by: Frederic Weisbecker <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
|
|
A timer might be used as a pinned timer (using add_timer_on()) and later on
as non-pinned timer using add_timer(). When the "NOHZ timer pull at expiry
model" is in place, the TIMER_PINNED flag is required to be used whenever a
timer needs to expire on a dedicated CPU. Otherwise the flag must not be
set if expiration on a dedicated CPU is not required.
add_timer_on()'s behavior will be changed during the preparation patches
for the "NOHZ timer pull at expiry model" to unconditionally set the
TIMER_PINNED flag. To be able to clear/ set the flag when queueing a
timer, two variants of add_timer() are introduced.
This is a preparatory step and has no functional change.
Signed-off-by: Anna-Maria Behnsen <[email protected]>
Signed-off-by: Thomas Gleixner <[email protected]>
Reviewed-by: Frederic Weisbecker <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
|
|
https://lore.kernel.org/r/170820083431.6328.16233178852085891453.stgit@91.116.238.104.host.secureserver.net
Pull simple offset series from Chuck Lever
In an effort to address slab fragmentation issues reported a few
months ago, I've replaced the use of xarrays for the directory
offset map in "simple" file systems (including tmpfs).
Thanks to Liam Howlett for helping me get this working with Maple
Trees.
* series 'Use Maple Trees for simple_offset utilities' of https://lore.kernel.org/r/170820083431.6328.16233178852085891453.stgit@91.116.238.104.host.secureserver.net: (6 commits)
libfs: Convert simple directory offsets to use a Maple Tree
test_maple_tree: testing the cyclic allocation
maple_tree: Add mtree_alloc_cyclic()
libfs: Add simple_offset_empty()
libfs: Define a minimum directory offset
libfs: Re-arrange locking in offset_iterate_dir()
Signed-off-by: Christian Brauner <[email protected]>
|
|
Add a common enum to define the standard clock OEM types defined by the
SCMI specification, so as to enable the configuration of such extended
configuration properties with the existent clock protocol operations.
Signed-off-by: Cristian Marussi <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Sudeep Holla <[email protected]>
|
|
SCMI v3.2 added support to set/get clock custom OEM types; such support is
conditionally present, though, depending on an extended config attribute
bit possibly advertised by the platform server on a per-domain base.
Add a check to verify if OEM types are supported before allowing any kind
of OEM-specific get/set operation. Also add a check around all the new
v3.2 clock features.
Signed-off-by: Cristian Marussi <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Sudeep Holla <[email protected]>
|
|
Add four functions to register clk_hw based on the fw_name field in
clk_parent_data, ie the value in the DT property `clock-names`.
There are variants for devm or not and passing an accuracy or not
passing one:
- clk_hw_register_fixed_factor_fwname
- clk_hw_register_fixed_factor_with_accuracy_fwname
- devm_clk_hw_register_fixed_factor_fwname
- devm_clk_hw_register_fixed_factor_with_accuracy_fwname
The `struct clk_parent_data` init is extracted from
__clk_hw_register_fixed_factor to each calling function. It is required
to allow each function to pass whatever field they want, not only index.
Signed-off-by: Théo Lebrun <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Stephen Boyd <[email protected]>
|
|
Fixed factor clock reports the parent clock accuracy. Add flags and acc
fields to `struct clk_fixed_factor` to support setting a fixed
accuracy. The default if no flag is set is not changed: use the parent
clock accuracy.
Signed-off-by: Théo Lebrun <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Stephen Boyd <[email protected]>
|
|
Provide a managed devm_clk_bulk* wrapper to get and enable all
bulk clocks in order to simplify drivers that keeps all clocks
enabled for the time of driver operation.
Suggested-by: Marek Szyprowski <[email protected]>
Reviewed-by: Alim Akhtar <[email protected]>
Reviewed-by: Manivannan Sadhasivam <[email protected]>
Signed-off-by: Shradha Todi <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Stephen Boyd <[email protected]>
|
|
We want to re-organize the struct sock layout. The sk_peek_off
field location is problematic, as most protocols want it in the
RX read area, while UDP wants it on a cacheline different from
sk_receive_queue.
Create a local (inside udp_sock) copy of the 'peek offset is enabled'
flag and place it inside the same cacheline of reader_queue.
Check such flag before reading sk_peek_off. This will save potential
false sharing and cache misses in the fast-path.
Tested under UDP flood with small packets. The struct sock layout
update causes a 4% performance drop, and this patch restores completely
the original tput.
Signed-off-by: Paolo Abeni <[email protected]>
Reviewed-by: Eric Dumazet <[email protected]>
Reviewed-by: Willem de Bruijn <[email protected]>
Link: https://lore.kernel.org/r/67ab679c15fbf49fa05b3ffe05d91c47ab84f147.1708426665.git.pabeni@redhat.com
Signed-off-by: Jakub Kicinski <[email protected]>
|
|
Now all callers of mm_counter_file() have a folio, convert
mm_counter_file() to take a folio. Saves a call to compound_head() hidden
inside PageSwapBacked().
Link: https://lkml.kernel.org/r/[email protected]
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Matthew Wilcox (Oracle) <[email protected]>
Cc: David Hildenbrand <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
|
|
Now all callers of mm_counter() have a folio, convert mm_counter() to take
a folio. Saves a call to compound_head() hidden inside PageAnon().
Link: https://lkml.kernel.org/r/[email protected]
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Matthew Wilcox (Oracle) <[email protected]>
Cc: David Hildenbrand <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
|
|
Patch series "mm: convert mm counter to take a folio", v3.
Make sure all mm_counter() and mm_counter_file() callers have a folio,
then convert mm counter functions to take a folio, which saves some
compound_head() calls.
This patch (of 10):
Thanks to the compound_head() hidden inside PageLocked(), this saves a
call to compound_head() over calling page_folio(pfn_swap_entry_to_page())
Link: https://lkml.kernel.org/r/[email protected]
Link: https://lkml.kernel.org/r/[email protected]
Signed-off-by: Matthew Wilcox (Oracle) <[email protected]>
Cc: David Hildenbrand <[email protected]>
Cc: Kefeng Wang <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
|
|
list_lru_init_key() isn't used by anyone, remove it to clean up.
Link: https://lkml.kernel.org/r/[email protected]
Signed-off-by: Haifeng Xu <[email protected]>
Acked-by: Roman Gushchin <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Michal Hocko <[email protected]>
Cc: Shakeel Butt <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
|
|
Patch series "implement "memmap on memory" feature on s390".
This series provides "memmap on memory" support on s390 platform. "memmap
on memory" allows struct pages array to be allocated from the hotplugged
memory range instead of allocating it from main system memory.
s390 currently preallocates struct pages array for all potentially
possible memory, which ensures memory onlining always succeeds, but with
the cost of significant memory consumption from the available system
memory during boottime. In certain extreme configuration, this could lead
to ipl failure.
"memmap on memory" ensures struct pages array are populated from self
contained hotplugged memory range instead of depleting the available
system memory and this could eliminate ipl failure on s390 platform.
On other platforms, system might go OOM when the physically hotplugged
memory depletes the available memory before it is onlined. Hence, "memmap
on memory" feature was introduced as described in commit a08a2ae34613
("mm,memory_hotplug: allocate memmap from the added memory range").
Unlike other architectures, s390 memory blocks are not physically
accessible until it is online. To make it physically accessible two new
memory notifiers MEM_PREPARE_ONLINE / MEM_FINISH_OFFLINE are added and
this notifier lets the hypervisor inform that the memory should be made
physically accessible. This allows for "memmap on memory" initialization
during memory hotplug onlining phase, which is performed before calling
MEM_GOING_ONLINE notifier.
Patch 1 introduces MEM_PREPARE_ONLINE/MEM_FINISH_OFFLINE memory notifiers
to prepare the transition of memory to and from a physically accessible
state. New mhp_flag MHP_OFFLINE_INACCESSIBLE is introduced to ensure
altmap cannot be written when adding memory - before it is set online.
This enhancement is crucial for implementing the "memmap on memory"
feature for s390 in a subsequent patch.
Patches 2 allocates vmemmap pages from self-contained memory range for
s390. It allocates memory map (struct pages array) from the hotplugged
memory range, rather than using system memory by passing altmap to vmemmap
functions.
Patch 3 removes unhandled memory notifier types on s390.
Patch 4 implements MEM_PREPARE_ONLINE/MEM_FINISH_OFFLINE memory notifiers
on s390. MEM_PREPARE_ONLINE memory notifier makes memory block physical
accessible via sclp assign command. The notifier ensures self-contained
memory maps are accessible and hence enabling the "memmap on memory" on
s390. MEM_FINISH_OFFLINE memory notifier shifts the memory block to an
inaccessible state via sclp unassign command.
Patch 5 finally enables MHP_MEMMAP_ON_MEMORY on s390.
This patch (of 5):
Introduce MEM_PREPARE_ONLINE/MEM_FINISH_OFFLINE memory notifiers to
prepare the transition of memory to and from a physically accessible
state. This enhancement is crucial for implementing the "memmap on
memory" feature for s390 in a subsequent patch.
Platforms such as x86 can support physical memory hotplug via ACPI. When
there is physical memory hotplug, ACPI event leads to the memory addition
with the following callchain:
acpi_memory_device_add()
-> acpi_memory_enable_device()
-> __add_memory()
After this, the hotplugged memory is physically accessible, and altmap
support prepared, before the "memmap on memory" initialization in
memory_block_online() is called.
On s390, memory hotplug works in a different way. The available hotplug
memory has to be defined upfront in the hypervisor, but it is made
physically accessible only when the user sets it online via sysfs,
currently in the MEM_GOING_ONLINE notifier. This is too late and "memmap
on memory" initialization is performed before calling MEM_GOING_ONLINE
notifier.
During the memory hotplug addition phase, altmap support is prepared and
during the memory onlining phase s390 requires memory to be physically
accessible and then subsequently initiate the "memmap on memory"
initialization process.
The memory provider will handle new MEM_PREPARE_ONLINE /
MEM_FINISH_OFFLINE notifications and make the memory accessible.
The mhp_flag MHP_OFFLINE_INACCESSIBLE is introduced and is relevant when
used along with MHP_MEMMAP_ON_MEMORY, because the altmap cannot be written
(e.g., poisoned) when adding memory -- before it is set online. This
allows for adding memory with an altmap that is not currently made
available by a hypervisor. When onlining that memory, the hypervisor can
be instructed to make that memory accessible via the new notifiers and the
onlining phase will not require any memory allocations, which is helpful
in low-memory situations.
All architectures ignore unknown memory notifiers. Therefore, the
introduction of these new notifiers does not result in any functional
modifications across architectures.
Link: https://lkml.kernel.org/r/[email protected]
Link: https://lkml.kernel.org/r/[email protected]
Signed-off-by: Sumanth Korikkar <[email protected]>
Suggested-by: Gerald Schaefer <[email protected]>
Suggested-by: David Hildenbrand <[email protected]>
Acked-by: David Hildenbrand <[email protected]>
Cc: Alexander Gordeev <[email protected]>
Cc: Aneesh Kumar K.V <[email protected]>
Cc: Anshuman Khandual <[email protected]>
Cc: Heiko Carstens <[email protected]>
Cc: Michal Hocko <[email protected]>
Cc: Oscar Salvador <[email protected]>
Cc: Vasily Gorbik <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
|
|
Add a driver for the Marvell 88Q2220. This driver allows to detect the
link, switch between 100BASE-T1 and 1000BASE-T1 and switch between
master and slave mode. Autonegotiation is supported.
Reviewed-by: Andrew Lunn <[email protected]>
Tested-by: Gregor Herburger <[email protected]>
Signed-off-by: Dimitri Fedrau <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
|
|
Extend helper functions mii_t1_adv_m_mod_linkmode_t and
linkmode_adv_to_mii_t1_adv_m_t to support 100BT1 and 1000BT1 linkmode
advertisements.
Reviewed-by: Andrew Lunn <[email protected]>
Signed-off-by: Dimitri Fedrau <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
|
|
Back in 2018 the commit be95a845cc44 ("bpf: avoid false sharing of map refcount
with max_entries") added ____cacheline_aligned to "struct bpf_map" to make sure
that fields like refcnt don't share a cache line with max_entries that is used
to bounds check map access. That was done to make spectre style attacks harder.
The main mitigation is done via code similar to array_index_nospec(), of course.
This was an additional precaution.
It increased the size of "struct bpf_map" a little, but it's affect on all
other maps (like array) is significant, since "struct bpf_map" is typically
the first member in other map types.
Undo this ____cacheline_aligned tag. Instead move freeze_mutex field around, so
that refcnt and max_entries are still in different cache lines.
The main effect is seen in sizeof(struct bpf_array) that reduces from 320
to 248 bytes.
BEFORE:
struct bpf_map {
const struct bpf_map_ops * ops; /* 0 8 */
...
char name[16]; /* 96 16 */
/* XXX 16 bytes hole, try to pack */
/* --- cacheline 2 boundary (128 bytes) --- */
atomic64_t refcnt __attribute__((__aligned__(64))); /* 128 8 */
...
/* size: 256, cachelines: 4, members: 30 */
/* sum members: 232, holes: 1, sum holes: 16 */
/* padding: 8 */
/* paddings: 1, sum paddings: 2 */
} __attribute__((__aligned__(64)));
struct bpf_array {
struct bpf_map map; /* 0 256 */
...
/* size: 320, cachelines: 5, members: 5 */
/* padding: 48 */
/* paddings: 1, sum paddings: 8 */
} __attribute__((__aligned__(64)));
AFTER:
struct bpf_map {
/* size: 232, cachelines: 4, members: 30 */
/* paddings: 1, sum paddings: 2 */
/* last cacheline: 40 bytes */
};
struct bpf_array {
/* size: 248, cachelines: 4, members: 5 */
/* last cacheline: 56 bytes */
};
Signed-off-by: Alexei Starovoitov <[email protected]>
Signed-off-by: Daniel Borkmann <[email protected]>
Acked-by: Yonghong Song <[email protected]>
Link: https://lore.kernel.org/bpf/[email protected]
|
|
If kiocb_set_cancel_fn() is called for I/O submitted via io_uring, the
following kernel warning appears:
WARNING: CPU: 3 PID: 368 at fs/aio.c:598 kiocb_set_cancel_fn+0x9c/0xa8
Call trace:
kiocb_set_cancel_fn+0x9c/0xa8
ffs_epfile_read_iter+0x144/0x1d0
io_read+0x19c/0x498
io_issue_sqe+0x118/0x27c
io_submit_sqes+0x25c/0x5fc
__arm64_sys_io_uring_enter+0x104/0xab0
invoke_syscall+0x58/0x11c
el0_svc_common+0xb4/0xf4
do_el0_svc+0x2c/0xb0
el0_svc+0x2c/0xa4
el0t_64_sync_handler+0x68/0xb4
el0t_64_sync+0x1a4/0x1a8
Fix this by setting the IOCB_AIO_RW flag for read and write I/O that is
submitted by libaio.
Suggested-by: Jens Axboe <[email protected]>
Cc: Christoph Hellwig <[email protected]>
Cc: Avi Kivity <[email protected]>
Cc: Sandeep Dhavale <[email protected]>
Cc: Jens Axboe <[email protected]>
Cc: Greg Kroah-Hartman <[email protected]>
Cc: Kent Overstreet <[email protected]>
Cc: [email protected]
Signed-off-by: Bart Van Assche <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Christian Brauner <[email protected]>
|
|
Use the existing ML element parsing helpers and add a new
one for this (ieee80211_mle_get_mld_id).
Signed-off-by: Johannes Berg <[email protected]>
Signed-off-by: Miri Korenblit <[email protected]>
Link: https://msgid.link/20240216135047.4da47b1f035b.I437a5570ac456449facb0b147851ef24a1e473c2@changeid
Signed-off-by: Johannes Berg <[email protected]>
|
|
Align the prototype of ieee80211_mle_get_bss_param_ch_cnt()
to also take a u8 * like the other functions, and make it
return -1 when the field isn't found, so that mac80211 can
check that instead of explicitly open-coding the check.
Signed-off-by: Johannes Berg <[email protected]>
Signed-off-by: Miri Korenblit <[email protected]>
Link: https://msgid.link/20240216135047.583309181bc3.Ia61cb0b4fc034d5ac8fcfaf6f6fb2e115fadafe7@changeid
Signed-off-by: Johannes Berg <[email protected]>
|
|
The KHZ_PER_GHZ might be used by others (with the name aligned
with similar constants). Define it in units.h and convert
wireless to use it.
Signed-off-by: Andy Shevchenko <[email protected]>
Reviewed-by: Simon Horman <[email protected]>
Link: https://msgid.link/[email protected]
Signed-off-by: Johannes Berg <[email protected]>
|
|
Up until now we have managed not to have the mdio-bcm-unimac manage its
clock except during probe and suspend/resume. This works most of the
time, except where it does not.
With a fully modular build, we can get into a situation whereby the
GENET driver is fully registered, and so is the mdio-bcm-unimac driver,
however the Ethernet PHY driver is not yet, because it depends on a
resource that is not yet available (e.g.: GPIO provider). In that state,
the network device is not usable yet, and so to conserve power, the
GENET driver will have turned off its "main" clock which feeds its MDIO
controller.
When the PHY driver finally probes however, we make an access to the PHY
registers to e.g.: disable interrupts, and this causes a bus error
within the MDIO controller space because the MDIO controller clock(s)
are turned off.
To remedy that, we manage the clock around all of the I/O accesses to
the hardware which are done exclusively during read, write and clock
divider configuration.
This ensures that the register space is accessible, and this also
ensures that there are not unnecessarily elevated reference counts
keeping the clocks active when the network device is administratively
turned off. It would be the case with the previous way of managing the
clock.
Reviewed-by: Jacob Keller <[email protected]>
Signed-off-by: Florian Fainelli <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
Remove documentation of non-existent children field
from the Kernel doc for struct framer_ops.
Introduced by 82c944d05b1a ("net: wan: Add framer framework support")
Signed-off-by: Simon Horman <[email protected]>
Acked-by: Herve Codina <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless-next
Kalle Valo says:
====================
wireless-next patches for v6.9
The second "new features" pull request for v6.9. Lots of iwlwifi and
stack changes this time. And naturally smaller changes to other drivers.
We also twice merged wireless into wireless-next to avoid conflicts
between the trees.
Major changes:
stack
* mac80211: negotiated TTLM request support
* SPP A-MSDU support
* mac80211: wider bandwidth OFDMA config support
iwlwifi
* kunit tests
* bump FW API to 89 for AX/BZ/SC devices
* enable SPP A-MSDUs
* support for new devices
ath12k
* refactoring in preparation for Multi-Link Operation (MLO) support
* 1024 Block Ack window size support
* provide firmware wmi logs via a trace event
ath11k
* 36 bit DMA mask support
* support 6 GHz station power modes: Low Power Indoor (LPI), Standard
Power) SP and Very Low Power (VLP)
rtl8xxxu
* TP-Link TL-WN823N V2 support
====================
Signed-off-by: David S. Miller <[email protected]>
|
|
No need to keep this in the core, move it to the nfnetlink_queue module.
nf_reroute is moved too, there were no other callers.
Signed-off-by: Florian Westphal <[email protected]>
|
|
On a 8-socket server the TSC is wrongly marked as 'unstable' and disabled
during boot time on about one out of 120 boot attempts:
clocksource: timekeeping watchdog on CPU227: wd-tsc-wd excessive read-back delay of 153560ns vs. limit of 125000ns,
wd-wd read-back delay only 11440ns, attempt 3, marking tsc unstable
tsc: Marking TSC unstable due to clocksource watchdog
TSC found unstable after boot, most likely due to broken BIOS. Use 'tsc=unstable'.
sched_clock: Marking unstable (119294969739, 159204297)<-(125446229205, -5992055152)
clocksource: Checking clocksource tsc synchronization from CPU 319 to CPUs 0,99,136,180,210,542,601,896.
clocksource: Switched to clocksource hpet
The reason is that for platform with a large number of CPUs, there are
sporadic big or huge read latencies while reading the watchog/clocksource
during boot or when system is under stress work load, and the frequency and
maximum value of the latency goes up with the number of online CPUs.
The cCurrent code already has logic to detect and filter such high latency
case by reading the watchdog twice and checking the two deltas. Due to the
randomness of the latency, there is a low probabilty that the first delta
(latency) is big, but the second delta is small and looks valid. The
watchdog code retries the readouts by default twice, which is not
necessarily sufficient for systems with a large number of CPUs.
There is a command line parameter 'max_cswd_read_retries' which allows to
increase the number of retries, but that's not user friendly as it needs to
be tweaked per system. As the number of required retries is proportional to
the number of online CPUs, this parameter can be calculated at runtime.
Scale and enlarge the number of retries according to the number of online
CPUs and remove the command line parameter completely.
[ tglx: Massaged change log and comments ]
Signed-off-by: Feng Tang <[email protected]>
Signed-off-by: Thomas Gleixner <[email protected]>
Tested-by: Jin Wang <[email protected]>
Tested-by: Paul E. McKenney <[email protected]>
Reviewed-by: Waiman Long <[email protected]>
Reviewed-by: Paul E. McKenney <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
|
|
Devices sharing a reset GPIO could use the reset framework for
coordinated handling of that shared GPIO line. We have several cases of
such needs, at least for Devicetree-based platforms.
If Devicetree-based device requests a reset line, while "resets"
Devicetree property is missing but there is a "reset-gpios" one,
instantiate a new "reset-gpio" platform device which will handle such
reset line. This allows seamless handling of such shared reset-gpios
without need of changing Devicetree binding [1].
To avoid creating multiple "reset-gpio" platform devices, store the
Devicetree "reset-gpios" GPIO specifiers used for new devices on a
linked list. Later such Devicetree GPIO specifier (phandle to GPIO
controller, GPIO number and GPIO flags) is used to check if reset
controller for given GPIO was already registered.
If two devices have conflicting "reset-gpios" property, e.g. with
different ACTIVE_xxx flags, this would allow to spawn two separate
"reset-gpio" devices, where the second would fail probing on busy GPIO
request.
Link: https://lore.kernel.org/all/[email protected]/ [1]
Cc: Bartosz Golaszewski <[email protected]>
Cc: Chris Packham <[email protected]>
Cc: Sean Anderson <[email protected]>
Reviewed-by: Philipp Zabel <[email protected]>
Signed-off-by: Krzysztof Kozlowski <[email protected]>
Acked-by: Bartosz Golaszewski <[email protected]>
Acked-by: Linus Walleij <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Philipp Zabel <[email protected]>
|
|
Use newly added of_phandle_args_equal() helper to compare two
of_phandle_args.
Acked-by: Viresh Kumar <[email protected]>
Reviewed-by: Philipp Zabel <[email protected]>
Signed-off-by: Krzysztof Kozlowski <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Philipp Zabel <[email protected]>
|
|
Add a helper comparing two "struct of_phandle_args" to avoid
reinventing the wheel.
Reviewed-by: Philipp Zabel <[email protected]>
Acked-by: Rob Herring <[email protected]>
Signed-off-by: Krzysztof Kozlowski <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Philipp Zabel <[email protected]>
|
|
Test robot reports:
> kernel test robot noticed a -19.0% regression of aim9.disk_src.ops_per_sec on:
>
> commit: a2e459555c5f9da3e619b7e47a63f98574dc75f1 ("shmem: stable directory offsets")
> https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master
Feng Tang further clarifies that:
> ... the new simple_offset_add()
> called by shmem_mknod() brings extra cost related with slab,
> specifically the 'radix_tree_node', which cause the regression.
Willy's analysis is that, over time, the test workload causes
xa_alloc_cyclic() to fragment the underlying SLAB cache.
This patch replaces the offset_ctx's xarray with a Maple Tree in the
hope that Maple Tree's dense node mode will handle this scenario
more scalably.
In addition, we can widen the simple directory offset maximum to
signed long (as loff_t is also signed).
Suggested-by: Matthew Wilcox <[email protected]>
Reported-by: kernel test robot <[email protected]>
Closes: https://lore.kernel.org/oe-lkp/[email protected]
Signed-off-by: Chuck Lever <[email protected]>
Link: https://lore.kernel.org/r/170820145616.6328.12620992971699079156.stgit@91.116.238.104.host.secureserver.net
Reviewed-by: Jan Kara <[email protected]>
Signed-off-by: Christian Brauner <[email protected]>
|
|
I need a cyclic allocator for the simple_offset implementation in
fs/libfs.c.
Signed-off-by: Chuck Lever <[email protected]>
Link: https://lore.kernel.org/r/170820144179.6328.12838600511394432325.stgit@91.116.238.104.host.secureserver.net
Signed-off-by: Christian Brauner <[email protected]>
|
|
For simple filesystems that use directory offset mapping, rely
strictly on the directory offset map to tell when a directory has
no children.
After this patch is applied, the emptiness test holds only the RCU
read lock when the directory being tested has no children.
In addition, this adds another layer of confirmation that
simple_offset_add/remove() are working as expected.
Reviewed-by: Jan Kara <[email protected]>
Signed-off-by: Chuck Lever <[email protected]>
Link: https://lore.kernel.org/r/170820143463.6328.7872919188371286951.stgit@91.116.238.104.host.secureserver.net
Signed-off-by: Christian Brauner <[email protected]>
|
|
Pick up CXL CPER notification removal for v6.8-rc6, to return in a later
merge window.
|
|
Initial tests with the CXL CPER implementation identified that error
reports were being duplicated in the log and the trace event [1]. Then
it was discovered that the notification handler took sleeping locks
while the GHES event handling runs in spin_lock_irqsave() context [2]
While the duplicate reporting was fixed in v6.8-rc4, the fix for the
sleeping-lock-vs-atomic collision would enjoy more time to settle and
gain some test cycles. Given how late it is in the development cycle,
remove the CXL hookup for now and try again during the next merge
window.
Note that end result is that v6.8 does not emit CXL CPER payloads to the
kernel log, but this is in line with the CXL trend to move error
reporting to trace events instead of the kernel log.
Cc: Ard Biesheuvel <[email protected]>
Cc: Rafael J. Wysocki <[email protected]>
Cc: Jonathan Cameron <[email protected]>
Reviewed-by: Ira Weiny <[email protected]>
Link: http://lore.kernel.org/r/[email protected] [1]
Closes: http://lore.kernel.org/r/[email protected] [2]
Signed-off-by: Dan Williams <[email protected]>
|
|
shmem_aops really should not be exported to the world. Move
shmem_mapping and export it as internal for the one semi-legitimate
modular user in udmabuf.
This effectively reverts commit 30e6a51dbb05 ("mm/shmem.c: make
shmem_mapping() inline"). which added a bogus shmem_aops non-GPL export
for no reason whatsoever as there as no shmem_mapping call outside of
core MM code at that point.
Signed-off-by: Christoph Hellwig <[email protected]>
Reviewed-by: "Matthew Wilcox (Oracle)" <[email protected]>
Signed-off-by: Chandan Babu R <[email protected]>
|
|
mapping_set_update is only used inside mm/. Move mapping_set_update to
mm/internal.h and turn it into an inline function instead of a macro.
Signed-off-by: Christoph Hellwig <[email protected]>
Reviewed-by: "Matthew Wilcox (Oracle)" <[email protected]>
Signed-off-by: Chandan Babu R <[email protected]>
|
|
This reverts commit 3316ab2b45f6bf4797d8d65b22fda3cc13318890.
The MHI spec owner pointed out that the SOC_HW_VERSION register is part
of the BHIe segment, and only valid on devices which implement BHIe.
Only a small subset of MHI devices implement BHIe so blindly accessing
the register for all devices is not correct. Also, since the BHIe
segment offset is not used when accessing the register, any
implementation which moves the BHIe segment will result in accessing
some other register. We've seen that accessing this register on AIC100
which does not support BHIe can result in initialization failures.
We could try to put checks into the code to address these issues, but in
the roughly 4 years this functionality has existed, no one has used it.
Easier to drop this dead code and address the issues if anyone comes up
with a real world use for it.
Signed-off-by: Jeffrey Hugo <[email protected]>
Reviewed-by: Manivannan Sadhasivam <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Manivannan Sadhasivam <[email protected]>
|
|
The bits of work->data are used for a few different purposes. How the bits
are used is determined by enum work_bits. The planned disable/enable support
will add another use, so let's clean it up a bit in preparation.
- Let WORK_STRUCT_*_BIT's values be determined by enum definition order.
- Deliminate different bit sections the same way using SHIFT and BITS
values.
- Rename __WORK_OFFQ_CANCELING to WORK_OFFQ_CANCELING_BIT for consistency.
- Introduce WORK_STRUCT_PWQ_SHIFT and replace WORK_STRUCT_FLAG_MASK and
WORK_STRUCT_WQ_DATA_MASK with WQ_STRUCT_PWQ_MASK for clarity.
- Improve documentation.
No functional changes.
Signed-off-by: Tejun Heo <[email protected]>
Reviewed-by: Lai Jiangshan <[email protected]>
|
|
Similar to strscpy(), update strscpy_pad()'s 3rd argument to be
optional when the destination is a compile-time known size array.
Cc: Andy Shevchenko <[email protected]>
Cc: <[email protected]>
Reviewed-by: Justin Stitt <[email protected]>
Signed-off-by: Kees Cook <[email protected]>
|
|
Using sizeof(dst) for the "size" argument in strscpy() is the
overwhelmingly common case. Instead of requiring this everywhere, allow a
2-argument version to be used that will use the sizeof() internally. There
are other functions in the kernel with optional arguments[1], so this
isn't unprecedented, and improves readability. Update and relocate the
kern-doc for strscpy() too, and drop __HAVE_ARCH_STRSCPY as it is unused.
Adjust ARCH=um build to notice the changed export name, as it doesn't
do full header includes for the string helpers.
This could additionally let us save a few hundred lines of code:
1177 files changed, 2455 insertions(+), 3026 deletions(-)
with a treewide cleanup using Coccinelle:
@needless_arg@
expression DST, SRC;
@@
strscpy(DST, SRC
-, sizeof(DST)
)
Link: https://elixir.bootlin.com/linux/v6.7/source/include/linux/pci.h#L1517 [1]
Reviewed-by: Justin Stitt <[email protected]>
Cc: Andy Shevchenko <[email protected]>
Cc: [email protected]
Signed-off-by: Kees Cook <[email protected]>
|
|
In preparation for making strscpy_pad()'s 3rd argument optional, redefine
it as a macro. This also has the benefit of allowing greater FORITFY
introspection, as it couldn't see into the strscpy() nor the memset()
within strscpy_pad().
Cc: Andy Shevchenko <[email protected]>
Cc: Andrew Morton <[email protected]>
Cc: <[email protected]>
Reviewed-by: Justin Stitt <[email protected]>
Signed-off-by: Kees Cook <[email protected]>
|
|
In order to mitigate unexpected signed wrap-around[1], bring back the
signed integer overflow sanitizer. It was removed in commit 6aaa31aeb9cf
("ubsan: remove overflow checks") because it was effectively a no-op
when combined with -fno-strict-overflow (which correctly changes signed
overflow from being "undefined" to being explicitly "wrap around").
Compilers are adjusting their sanitizers to trap wrap-around and to
detecting common code patterns that should not be instrumented
(e.g. "var + offset < var"). Prepare for this and explicitly rename
the option from "OVERFLOW" to "WRAP" to more accurately describe the
behavior.
To annotate intentional wrap-around arithmetic, the helpers
wrapping_add/sub/mul_wrap() can be used for individual statements. At
the function level, the __signed_wrap attribute can be used to mark an
entire function as expecting its signed arithmetic to wrap around. For a
single object file the Makefile can use "UBSAN_SIGNED_WRAP_target.o := n"
to mark it as wrapping, and for an entire directory, "UBSAN_SIGNED_WRAP :=
n" can be used.
Additionally keep these disabled under CONFIG_COMPILE_TEST for now.
Link: https://github.com/KSPP/linux/issues/26 [1]
Cc: Miguel Ojeda <[email protected]>
Cc: Nathan Chancellor <[email protected]>
Cc: Peter Zijlstra <[email protected]>
Cc: Hao Luo <[email protected]>
Reviewed-by: Marco Elver <[email protected]>
Reviewed-by: Justin Stitt <[email protected]>
Signed-off-by: Kees Cook <[email protected]>
|
|
The xlate callbacks are supposed to translate of_phandle_args to proper
provider without modifying the of_phandle_args. Make the argument
pointer to const for code safety and readability.
Signed-off-by: Krzysztof Kozlowski <[email protected]>
Reviewed-by: Simon Horman <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
|
|
Pull rdma fixes from Jason Gunthorpe:
"Mostly irdma and bnxt_re fixes:
- Missing error unwind in hf1
- For bnxt - fix fenching behavior to work on new chips, fail
unsupported SRQ resize back to userspace, propogate SRQ FW failure
back to userspace.
- Correctly fail unsupported SRQ resize back to userspace in bnxt
- Adjust a memcpy in mlx5 to not overflow a struct field.
- Prevent userspace from triggering mlx5 fw syndrome logging from
sysfs
- Use the correct access mode for MLX5_IB_METHOD_DEVX_OBJ_MODIFY to
avoid a userspace failure on modify
- For irdma - Don't UAF a concurrent tasklet during destroy, prevent
userspace from issuing invalid QP attrs, fix a possible CQ
overflow, capture a missing HW async error event
- sendmsg() triggerable memory access crash in hfi1
- Fix the srpt_service_guid parameter to not crash due to missing
function pointer
- Don't leak objects in error unwind in qedr
- Don't weirdly cast function pointers in srpt"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
RDMA/srpt: fix function pointer cast warnings
RDMA/qedr: Fix qedr_create_user_qp error flow
RDMA/srpt: Support specifying the srpt_service_guid parameter
IB/hfi1: Fix sdma.h tx->num_descs off-by-one error
RDMA/irdma: Add AE for too many RNRS
RDMA/irdma: Set the CQ read threshold for GEN 1
RDMA/irdma: Validate max_send_wr and max_recv_wr
RDMA/irdma: Fix KASAN issue with tasklet
RDMA/mlx5: Relax DEVX access upon modify commands
IB/mlx5: Don't expose debugfs entries for RRoCE general parameters if not supported
RDMA/mlx5: Fix fortify source warning while accessing Eth segment
RDMA/bnxt_re: Add a missing check in bnxt_qplib_query_srq
RDMA/bnxt_re: Return error for SRQ resize
RDMA/bnxt_re: Fix unconditional fence for newer adapters
RDMA/bnxt_re: Remove a redundant check inside bnxt_re_vf_res_config
RDMA/bnxt_re: Avoid creating fence MR for newer adapters
IB/hfi1: Fix a memleak in init_credit_return
|
|
Merge series from Cezary Rojewski <[email protected]>:
The avs-driver continues to be utilized on more recent Intel machines.
As TGL-based (cAVS 2.5) e.g.: RPL, inherit most of the functionality
from previous platforms:
SKL <- APL <- CNL <- ICL <- TGL
rather than putting everything into a single file, the platform-specific
bits are split into cnl/icl/tgl.c files instead. Makes the division clear
and code easier to maintain.
Layout of the patchset:
First are two changes combined together address the sound-clipping
problem, present when only one stream is running - specifically one
CAPTURE stream.
Follow up is naming-scheme adjustment for some of the existing functions
what improves code incohesiveness. As existing IPC/IRQ code operates
solely on cAVS 1.5 architecture, it needs no abstraction. The situation
changes when newer platforms come into the picture. Thus the next two
patches abstract the existing IPC/IRQ handlers so that majority of the
common code can be re-used.
The ICCMAX change stands out a bit - the AudioDSP firmware loading
procedure differs on ICL-based platforms (and onwards) and having a
separate commit makes the situation clear to the developers who are
going to support the solution from LTS perspective. For that reason
I decided not to merge it into the commit introducing the icl.c file.
|
|
When skipping swapcache for SWP_SYNCHRONOUS_IO, if two or more threads
swapin the same entry at the same time, they get different pages (A, B).
Before one thread (T0) finishes the swapin and installs page (A) to the
PTE, another thread (T1) could finish swapin of page (B), swap_free the
entry, then swap out the possibly modified page reusing the same entry.
It breaks the pte_same check in (T0) because PTE value is unchanged,
causing ABA problem. Thread (T0) will install a stalled page (A) into the
PTE and cause data corruption.
One possible callstack is like this:
CPU0 CPU1
---- ----
do_swap_page() do_swap_page() with same entry
<direct swapin path> <direct swapin path>
<alloc page A> <alloc page B>
swap_read_folio() <- read to page A swap_read_folio() <- read to page B
<slow on later locks or interrupt> <finished swapin first>
... set_pte_at()
swap_free() <- entry is free
<write to page B, now page A stalled>
<swap out page B to same swap entry>
pte_same() <- Check pass, PTE seems
unchanged, but page A
is stalled!
swap_free() <- page B content lost!
set_pte_at() <- staled page A installed!
And besides, for ZRAM, swap_free() allows the swap device to discard the
entry content, so even if page (B) is not modified, if swap_read_folio()
on CPU0 happens later than swap_free() on CPU1, it may also cause data
loss.
To fix this, reuse swapcache_prepare which will pin the swap entry using
the cache flag, and allow only one thread to swap it in, also prevent any
parallel code from putting the entry in the cache. Release the pin after
PT unlocked.
Racers just loop and wait since it's a rare and very short event. A
schedule_timeout_uninterruptible(1) call is added to avoid repeated page
faults wasting too much CPU, causing livelock or adding too much noise to
perf statistics. A similar livelock issue was described in commit
029c4628b2eb ("mm: swap: get rid of livelock in swapin readahead")
Reproducer:
This race issue can be triggered easily using a well constructed
reproducer and patched brd (with a delay in read path) [1]:
With latest 6.8 mainline, race caused data loss can be observed easily:
$ gcc -g -lpthread test-thread-swap-race.c && ./a.out
Polulating 32MB of memory region...
Keep swapping out...
Starting round 0...
Spawning 65536 workers...
32746 workers spawned, wait for done...
Round 0: Error on 0x5aa00, expected 32746, got 32743, 3 data loss!
Round 0: Error on 0x395200, expected 32746, got 32743, 3 data loss!
Round 0: Error on 0x3fd000, expected 32746, got 32737, 9 data loss!
Round 0 Failed, 15 data loss!
This reproducer spawns multiple threads sharing the same memory region
using a small swap device. Every two threads updates mapped pages one by
one in opposite direction trying to create a race, with one dedicated
thread keep swapping out the data out using madvise.
The reproducer created a reproduce rate of about once every 5 minutes, so
the race should be totally possible in production.
After this patch, I ran the reproducer for over a few hundred rounds and
no data loss observed.
Performance overhead is minimal, microbenchmark swapin 10G from 32G
zram:
Before: 10934698 us
After: 11157121 us
Cached: 13155355 us (Dropping SWP_SYNCHRONOUS_IO flag)
[[email protected]: v4]
Link: https://lkml.kernel.org/r/[email protected]
Link: https://lkml.kernel.org/r/[email protected]
Fixes: 0bcac06f27d7 ("mm, swap: skip swapcache for swapin of synchronous device")
Reported-by: "Huang, Ying" <[email protected]>
Closes: https://lore.kernel.org/lkml/[email protected]/
Link: https://github.com/ryncsn/emm-test-project/tree/master/swap-stress-race [1]
Signed-off-by: Kairui Song <[email protected]>
Reviewed-by: "Huang, Ying" <[email protected]>
Acked-by: Yu Zhao <[email protected]>
Acked-by: David Hildenbrand <[email protected]>
Acked-by: Chris Li <[email protected]>
Cc: Hugh Dickins <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Matthew Wilcox (Oracle) <[email protected]>
Cc: Michal Hocko <[email protected]>
Cc: Minchan Kim <[email protected]>
Cc: Yosry Ahmed <[email protected]>
Cc: Yu Zhao <[email protected]>
Cc: Barry Song <[email protected]>
Cc: SeongJae Park <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
|
|
Some IO will dispatch from kworker with different io_context settings
than the submitting task, we may need to specify a priority to avoid
losing priority.
Add dm_bufio_read_with_ioprio() and dm_bufio_prefetch_with_ioprio()
for use by bufio users to pass an ioprio other than IOPRIO_DEFAULT.
Co-developed-by: Yibin Ding <[email protected]>
Signed-off-by: Yibin Ding <[email protected]>
Signed-off-by: Hongyu Jin <[email protected]>
Reviewed-by: Eric Biggers <[email protected]>
Reviewed-by: Mikulas Patocka <[email protected]>
[snitzer: introduced _with_ioprio() wrappers to reduce churn]
Signed-off-by: Mike Snitzer <[email protected]>
|
|
Some IO will dispatch from kworker with different io_context settings
than the submitting task, we may need to specify a priority to avoid
losing priority.
Add IO priority parameter to dm_io() and update all callers.
Co-developed-by: Yibin Ding <[email protected]>
Signed-off-by: Yibin Ding <[email protected]>
Signed-off-by: Hongyu Jin <[email protected]>
Reviewed-by: Eric Biggers <[email protected]>
Reviewed-by: Mikulas Patocka <[email protected]>
Signed-off-by: Mike Snitzer <[email protected]>
|