It will return all zero data when DIO reading from inline_data inode, it
is because f2fs_iomap_begin() assign iomap->type w/ IOMAP_HOLE incorrectly
for this case.
We can let iomap framework handle inline data via assigning iomap->type
and iomap->inline_data correctly, however, it will be a little bit
complicated when handling race case in between direct IO and buffered IO.
So, let's force to use buffered IO to fix this issue.
Cc: stable@vger.kernel.org
Reported-by: Barry Song <v-songbaohua@oppo.com>
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Subsequent f2fs_stop_checkpoint will set cp_err, so this
SBI_NEED_FSCK flag set action is invalid.
Signed-off-by: Zhiguo Niu <zhiguo.niu@unisoc.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
If curseg is not the first segment in its zone, the zone is not empty.
A new section should be allocated and avoid resetting the old zone.
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Sheng Yong <shengyong@oppo.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
FS_IOC_GETFSSYSFSPATH ioctl expects sysfs sub-path of a filesystem, the
format can be "$FSTYP/$SYSFS_IDENTIFIER" under /sys/fs, it can helps to
standardizes exporting sysfs datas across filesystems.
This patch wires up FS_IOC_GETFSSYSFSPATH for f2fs, it will output
"f2fs/<dev>".
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
inode can be fuzzed, so it can has F2FS_INLINE_DATA flag and valid
i_blocks/i_nid value, this patch supports to do extra sanity check
to detect such corrupted state.
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
syzbot reports a f2fs bug as below:
------------[ cut here ]------------
kernel BUG at fs/f2fs/inline.c:258!
CPU: 1 PID: 34 Comm: kworker/u8:2 Not tainted 6.9.0-rc6-syzkaller-00012-g9e4bc4bcae01 #0
RIP: 0010:f2fs_write_inline_data+0x781/0x790 fs/f2fs/inline.c:258
Call Trace:
f2fs_write_single_data_page+0xb65/0x1d60 fs/f2fs/data.c:2834
f2fs_write_cache_pages fs/f2fs/data.c:3133 [inline]
__f2fs_write_data_pages fs/f2fs/data.c:3288 [inline]
f2fs_write_data_pages+0x1efe/0x3a90 fs/f2fs/data.c:3315
do_writepages+0x35b/0x870 mm/page-writeback.c:2612
__writeback_single_inode+0x165/0x10b0 fs/fs-writeback.c:1650
writeback_sb_inodes+0x905/0x1260 fs/fs-writeback.c:1941
wb_writeback+0x457/0xce0 fs/fs-writeback.c:2117
wb_do_writeback fs/fs-writeback.c:2264 [inline]
wb_workfn+0x410/0x1090 fs/fs-writeback.c:2304
process_one_work kernel/workqueue.c:3254 [inline]
process_scheduled_works+0xa12/0x17c0 kernel/workqueue.c:3335
worker_thread+0x86d/0xd70 kernel/workqueue.c:3416
kthread+0x2f2/0x390 kernel/kthread.c:388
ret_from_fork+0x4d/0x80 arch/x86/kernel/process.c:147
ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:244
The root cause is: inline_data inode can be fuzzed, so that there may
be valid blkaddr in its direct node, once f2fs triggers background GC
to migrate the block, it will hit f2fs_bug_on() during dirty page
writeback.
Let's add sanity check on F2FS_INLINE_DATA flag in inode during GC,
so that, it can forbid migrating inline_data inode's data block for
fixing.
Reported-by: syzbot+848062ba19c8782ca5c8@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/linux-f2fs-devel/000000000000d103ce06174d7ec3@google.com
Signed-off-by: Chao Yu <chao@kernel.org>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
5 minutes is too short for a regular user to search and understand
what he needs to do to report capture devcoredump and report a bug to
us, so here increasing this timeout to 1 hour.
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Jonathan Cavitt <jonathan.cavitt@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Jonathan Cavitt <jonathan.cavitt@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240611174716.72660-2-jose.souza@intel.com
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Add function to set a custom coredump timeout.
For Xe driver usage, current 5 minutes timeout may be too short for
users to search and understand what needs to be done to capture
coredump to report bugs.
We have plans to automate(distribute a udev script) it but at the end
will be up to distros and users to pack it so having a option to
increase the timeout is a safer option.
v2:
- replace dev_coredump_timeout_set() by dev_coredumpm_timeout() (Mukesh)
v3:
- make dev_coredumpm() static inline (Johannes)
v5:
- rename DEVCOREDUMP_TIMEOUT -> DEVCD_TIMEOUT to avoid redefinition
in include/net/bluetooth/coredump.h
v6:
- fix definition of dev_coredumpm_timeout() when CONFIG_DEV_COREDUMP
is disabled
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Mukesh Ojha <quic_mojha@quicinc.com>
Cc: Johannes Berg <johannes@sipsolutions.net>
Cc: Jonathan Cavitt <jonathan.cavitt@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Jonathan Cavitt <jonathan.cavitt@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Acked-by: Johannes Berg <johannes@sipsolutions.net>
Link: https://patchwork.freedesktop.org/patch/msgid/20240611174716.72660-1-jose.souza@intel.com
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Add spicc loopback mode for debugging convenience.
Signed-off-by: Sunny Luo <sunny.luo@amlogic.com>
Signed-off-by: Xianwei Zhao <xianwei.zhao@amlogic.com>
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Link: https://lore.kernel.org/r/20240612-spi_lbc-v1-1-d52e8c8011bd@amlogic.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Ramp values are inverted. This caused wrong values written to register
when ramp values were defined in device tree.
Invert values in table to fix this.
Signed-off-by: Kalle Niemi <kaleposti@gmail.com>
Fixes: 1aad39001e ("regulator: Support ROHM BD71815 regulators")
Reviewed-by: Matti Vaittinen <mazziesaccount@gmail.com>
Link: https://lore.kernel.org/r/ZmmJXtuVJU6RgQAH@latitude5580
Signed-off-by: Mark Brown <broonie@kernel.org>
On x86, make allmodconfig && make W=1 C=1 reports:
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/pmdomain/amlogic/meson-gx-pwrc-vpu.o
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/pmdomain/amlogic/meson-ee-pwrc.o
Add the missing invocation of the MODULE_DESCRIPTION() macro to all
files which have a MODULE_LICENSE().
This includes meson-secure-pwrc.c which, although it did not produce a
warning with the x86 allmodconfig configuration, may cause this
warning with other configurations where CONFIG_MESON_SM is enabled.
Signed-off-by: Jeff Johnson <quic_jjohnson@quicinc.com>
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Link: https://lore.kernel.org/r/20240610-md-drivers-pmdomain-amlogic-v1-1-b49ddb1a8bdf@quicinc.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Cortex-X4 erratum 3194386 and Neoverse-V3 erratum 3312417 are identical,
with duplicate Kconfig text and some unsightly ifdeffery. While we try
to share code behind CONFIG_ARM64_WORKAROUND_SPECULATIVE_SSBS, having
separate options results in a fair amount of boilerplate code, and this
will only get worse as we expand the set of affected CPUs.
To reduce this boilerplate, unify the two behind a common Kconfig
option. This removes the duplicate text and Kconfig logic, and removes
the need for the intermediate ARM64_WORKAROUND_SPECULATIVE_SSBS option.
The set of affected CPUs is described as a list so that this can easily
be extended.
I've used ARM64_ERRATUM_3194386 (matching the Neoverse-V3 erratum ID) as
the common option, matching the way we use ARM64_ERRATUM_1319367 to
cover Cortex-A57 erratum 1319537 and Cortex-A72 erratum 1319367.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Will Deacon <wilL@kernel.org>
Link: https://lore.kernel.org/r/20240603111812.1514101-5-mark.rutland@arm.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Add cputype definitions for Cortex-X925. These will be used for errata
detection in subsequent patches.
These values can be found in Table A-285 ("MIDR_EL1 bit descriptions")
in issue 0001-05 of the Cortex-X925 TRM, which can be found at:
https://developer.arm.com/documentation/102807/0001/?lang=en
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Will Deacon <will@kernel.org>
Link: https://lore.kernel.org/r/20240603111812.1514101-4-mark.rutland@arm.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Add cputype definitions for Cortex-A720. These will be used for errata
detection in subsequent patches.
These values can be found in Table A-186 ("MIDR_EL1 bit descriptions")
in issue 0002-05 of the Cortex-A720 TRM, which can be found at:
https://developer.arm.com/documentation/102530/0002/?lang=en
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Will Deacon <will@kernel.org>
Link: https://lore.kernel.org/r/20240603111812.1514101-3-mark.rutland@arm.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Add cputype definitions for Cortex-X3. These will be used for errata
detection in subsequent patches.
These values can be found in Table A-263 ("MIDR_EL1 bit descriptions")
in issue 07 of the Cortex-X3 TRM, which can be found at:
https://developer.arm.com/documentation/101593/0102/?lang=en
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Will Deacon <will@kernel.org>
Link: https://lore.kernel.org/r/20240603111812.1514101-2-mark.rutland@arm.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
The check in mte_check_tfsr_el1() is only necessary if HW tag
based KASAN is enabled. However, we were also executing the check
if MTE is enabled and KASAN is enabled at build time but disabled
at runtime. This turned out to cause a measurable increase in
power consumption on a specific microarchitecture after enabling
MTE. Moreover, on the same system, an increase in invalid syscall
latency (as measured by [1]) of around 20-30% (depending on the
cluster) was observed after enabling MTE; this almost entirely goes
away after removing this check. Therefore, make the check conditional
on whether KASAN is enabled rather than on whether MTE is enabled.
[1] https://lore.kernel.org/all/CAMn1gO4MwRV8bmFJ_SeY5tsYNPn2ZP56LjAhafygjFaKuu5ouw@mail.gmail.com/
Signed-off-by: Peter Collingbourne <pcc@google.com>
Link: https://linux-review.googlesource.com/id/I22d98d1483dd400a95595946552b769a5a1ad7bd
Reviewed-by: Alexandru Elisei <alexandru.elisei@arm.com>
Link: https://lore.kernel.org/r/20240528225131.3577704-1-pcc@google.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Currently, we are writing the same value as we read into the TLS
register, hence we cannot confirm update of the register, making the
testcase "verify_tpidr_one" redundant. Fix this.
Signed-off-by: Dev Jain <dev.jain@arm.com>
Reviewed-by: Mark Brown <broonie@kernel.org>
Link: https://lore.kernel.org/r/20240605115448.640717-1-dev.jain@arm.com
[catalin.marinas@arm.com: remove the increment style change]
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Define a maximum allowable number of pids that can be livepatched in
test-syscall.sh as with extremely large machines the output from a
large number of processes overflows the dev/kmsg "expect" buffer in
the "check_result" function and causes a false error.
Reported-by: CKI Project <cki-project@redhat.com>
Signed-off-by: Ryan Sullivan <rysulliv@redhat.com>
Acked-by: Miroslav Benes <mbenes@suse.cz>
Tested-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Reviewed-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Reviewed-by: Petr Mladek <pmladek@suse.com>
Acked-by: Joel Savitz <jsavitz@redhat.com>
Link: https://lore.kernel.org/r/20240606135348.4708-1-rysulliv@redhat.com
Signed-off-by: Petr Mladek <pmladek@suse.com>
Currently fp-stress only covers userspace use of floating point, it does
not cover any kernel mode uses. Since currently kernel mode floating
point usage can't be preempted and there are explicit preemption points in
the existing implementations this isn't so important for fp-stress but
when we readd preemption it will be good to try to exercise it.
When the arm64 accelerated crypto operations are implemented we can
relatively straightforwardly trigger kernel mode floating point usage by
using the crypto userspace API to hash data, using the splice() support
in an effort to minimise copying. We use /proc/crypto to check which
accelerated implementations are available, picking the first symmetric
hash we find. We run the kernel mode test unconditionally, replacing the
second copy of the FPSIMD testcase for systems with FPSIMD only. If we
don't think there are any suitable kernel mode implementations we fall back
to running another copy of fpsimd-stress.
There are a number issues with this approach, we don't actually verify
that we are using an accelerated (or even CPU) implementation of the
algorithm being tested and even with attempting to use splice() to
minimise copying there are sizing limits on how much data gets spliced
at once.
Signed-off-by: Mark Brown <broonie@kernel.org>
Link: https://lore.kernel.org/r/20240521-arm64-fp-stress-kernel-v1-1-e38f107baad4@kernel.org
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Historically, arm64 implemented raw_smp_processor_id() as a read of
current_thread_info()->cpu. This changed when arm64 moved thread_info into
task struct, as at the time CONFIG_THREAD_INFO_IN_TASK made core code use
thread_struct::cpu for the cpu number, and due to header dependencies
prevented using this in raw_smp_processor_id(). As a workaround, we moved to
using a percpu variable in commit:
57c82954e7 ("arm64: make cpu number a percpu variable")
Since then, thread_info::cpu was reintroduced, and core code was made to use
this in commits:
001430c191 ("arm64: add CPU field to struct thread_info")
bcf9033e54 ("sched: move CPU field back into thread_info if THREAD_INFO_IN_TASK=y")
Consequently it is possible to use current_thread_info()->cpu again.
This decreases the number of emitted instructions like in the following
example:
Dump of assembler code for function bpf_get_smp_processor_id:
0xffff8000802cd608 <+0>: nop
0xffff8000802cd60c <+4>: nop
0xffff8000802cd610 <+8>: adrp x0, 0xffff800082138000
0xffff8000802cd614 <+12>: mrs x1, tpidr_el1
0xffff8000802cd618 <+16>: add x0, x0, #0x8
0xffff8000802cd61c <+20>: ldrsw x0, [x0, x1]
0xffff8000802cd620 <+24>: ret
After this patch:
Dump of assembler code for function bpf_get_smp_processor_id:
0xffff8000802c9130 <+0>: nop
0xffff8000802c9134 <+4>: nop
0xffff8000802c9138 <+8>: mrs x0, sp_el0
0xffff8000802c913c <+12>: ldr w0, [x0, #24]
0xffff8000802c9140 <+16>: ret
A microbenchmark[1] was built to measure the performance improvement
provided by this change. It calls the following function given number of
times and finds the runtime overhead:
static noinline int get_cpu_id(void)
{
return smp_processor_id();
}
Run the benchmark like:
modprobe smp_processor_id nr_function_calls=1000000000
+--------------------------+------------------------+
| | Number of Calls | Time taken |
+--------+-----------------+------------------------+
| Before | 1000000000 | 1602888401ns |
+--------+-----------------+------------------------+
| After | 1000000000 | 1206212658ns |
+--------+-----------------+------------------------+
| Difference (decrease) | 396675743ns (24.74%) |
+---------------------------------------------------+
Remove the percpu variable cpu_number as it is used only in
set_smp_ipi_range() as a dummy variable to be passed to ipi_handler().
Use irq_stat in place of cpu_number here like arm32.
[1] 77d3fdd
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Acked-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Link: https://lore.kernel.org/r/20240503171847.68267-2-puranjay@kernel.org
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
arch_timer.h includes linux/smp.h since the commit:
6acc71ccac ("arm64: arch_timer: Allows a CPU-specific erratum to only affect a subset of CPUs")
It was included to use DEFINE_PER_CPU(), etc. But It should have
included <linux/percpu.h> rather than <linux/smp.h>. It worked because
smp.h includes percpu.h.
The next commit will remove percpu.h from smp.h and it will break this
usage.
Explicitly include percpu.h and remove smp.h
Signed-off-by: Puranjay Mohan <puranjay@kernel.org>
Acked-by: Mark Rutland <mark.rutland@arm.com>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Reviewed-by: Anshuman Khandual <anshuman.khandual@arm.com>
Link: https://lore.kernel.org/r/20240503171847.68267-1-puranjay@kernel.org
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
The INA230 has an Alert pin which is asserted when the alert
function selected in the Mask/Enable register exceeds the
value programmed into the Alert Limit register. Assertion is based
on the Alert Polarity Bit (APOL, bit 1 of the Mask/Enable register).
It is default set to value 0 i.e Normal (active-low open collector).
However, hardware can be designed in such a way that expects Alert pin
to become active high if a user-defined threshold in Alert limit
register has been exceeded. This patch adds a way to pass alert polarity
value to the driver via device tree.
Signed-off-by: Amna Waseem <Amna.Waseem@axis.com>
Link: https://lore.kernel.org/r/20240611-apol-ina2xx-fix-v4-2-8df1d2282fc5@axis.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Add a property to the binding to configure the Alert Polarity.
Alert pin is asserted based on the value of Alert Polarity bit of
Mask/Enable register. It is by default 0 which means Alert pin is
configured to be active low open collector. Value of 1 maps to
Inverted (active high open collector).
Signed-off-by: Amna Waseem <Amna.Waseem@axis.com>
Reviewed-by: Rob Herring (Arm) <robh@kernel.org>
Link: https://lore.kernel.org/r/20240603-apol-ina2xx-fix-v3-1-b9eff3158e4e@axis.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Enable ECBHB bits in ID_AA64MMFR1 register as per ARM DDI 0487K.a
specification.
When guest OS read ID_AA64MMFR1_EL1, kvm emulate this reg using
ftr_id_aa64mmfr1 and always return ID_AA64MMFR1_EL1.ECBHB=0 to guest.
It results in guest syscall jump to tramp ventry, which is not needed
in implementation with ID_AA64MMFR1_EL1.ECBHB=1.
Let's make the guest syscall process the same as the host.
Signed-off-by: Nianyao Tang <tangnianyao@huawei.com>
Link: https://lore.kernel.org/r/20240611122049.2758600-1-tangnianyao@huawei.com
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Instead of inventing a custom way to conditionally enable debugging,
just make use of pr_debug(), which also has dynamic debugging facilities
and is more likely known to someone who hunts a problem in the netfs
code. Also drop the module parameter netfs_debug which didn't have any
effect without further source changes. (The variable netfs_debug was
only used in #ifdef blocks for cpp vars that don't exist; Note that
CONFIG_NETFS_DEBUG isn't settable via kconfig, a variable with that name
never existed in the mainline and is probably just taken over (and
renamed) from similar custom debug logging implementations.)
Signed-off-by: Uwe Kleine-König <ukleinek@kernel.org>
Link: https://lore.kernel.org/r/20240608151352.22860-2-ukleinek@kernel.org
Signed-off-by: Christian Brauner <brauner@kernel.org>
After recent changes in intel_pstate, global.turbo_disabled is only set
at the initialization time and never changed. However, it turns out
that on some systems the "turbo disabled" bit in MSR_IA32_MISC_ENABLE,
the initial state of which is reflected by global.turbo_disabled, can be
flipped later and there should be a way to take that into account (other
than checking that MSR every time the driver runs which is costly and
useless overhead on the vast majority of systems).
For this purpose, notice that before the changes in question,
store_no_turbo() contained a turbo_is_disabled() check that was used
for updating global.turbo_disabled if the "turbo disabled" bit in
MSR_IA32_MISC_ENABLE had been flipped and that functionality can be
restored. Then, users will be able to reset global.turbo_disabled
by writing 0 to no_turbo which used to work before on systems with
flipping "turbo disabled" bit.
This guarantees the driver state to remain in sync, but READ_ONCE()
annotations need to be added in two places where global.turbo_disabled
is accessed locklessly, so modify the driver to make that happen.
Fixes: 0940f1a801 ("cpufreq: intel_pstate: Do not update global.turbo_disabled after initialization")
Closes: https://lore.kernel.org/linux-pm/bf3ebf1571a4788e97daf861eb493c12d42639a3.camel@xry111.site
Suggested-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Reported-by: Xi Ruoyao <xry111@xry111.site>
Tested-by: Xi Ruoyao <xry111@xry111.site>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
With 20 threads each walking a dedicated 1000 dirs * 1000 files
directory tree to stat(2) on a 32 core + 24GB ram vm:
before: 3.54s user 892.30s system 1966% cpu 45.549 total
after: 3.28s user 738.66s system 1955% cpu 37.932 total (-16.7%)
Benchmark can be found here: https://people.freebsd.org/~mjg/fstree.tgz
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Mateusz Guzik <mjguzik@gmail.com>
Link: https://lore.kernel.org/r/20240611173824.535995-3-mjguzik@gmail.com
Acked-by: David Sterba <dsterba@suse.com>
Signed-off-by: Christian Brauner <brauner@kernel.org>
This avoids one inode hash lock acquire in the common case on inode
creation, in effect significantly reducing contention.
On the stock kernel said lock is typically taken twice:
1. once to check if the inode happens to already be present
2. once to add it to the hash
The back-to-back lock/unlock pattern is known to degrade performance
significantly, which is further exacerbated if the hash is heavily
populated (long chains to walk, extending hold time). Arguably hash
sizing and hashing algo need to be revisited, but that's beyond the
scope of this patch.
With the acquire from step 1 eliminated with RCU lookup throughput
increases significantly at the scale of 20 cores (benchmark results at
the bottom).
So happens the hash already supports RCU-based operation, but lookups on
inode insertions didn't take advantage of it.
This of course has its limits as the global lock is still a bottleneck.
There was a patchset posted which introduced fine-grained locking[1] but
it appears staled. Apart from that doubt was expressed whether a
handrolled hash implementation is appropriate to begin with, suggesting
replacement with rhashtables. Nobody committed to carrying [1] across
the finish line or implementing anything better, thus the bandaid below.
iget_locked consumers (notably ext4) get away without any changes
because inode comparison method is built-in.
iget5_locked consumers pass a custom callback. Since removal of locking
adds more problems (inode can be changing) it's not safe to assume all
filesystems happen to cope. Thus iget5_locked_rcu gets added, requiring
manual conversion of interested filesystems.
In order to reduce code duplication find_inode and find_inode_fast grow
an argument indicating whether inode hash lock is held, which is passed
down in case sleeping is necessary. They always rcu_read_lock, which is
redundant but harmless. Doing it conditionally reduces readability for
no real gain that I can see. RCU-alike restrictions were already put on
callbacks due to the hash spinlock being held.
Benchmarking:
There is a real cache-busting workload scanning millions of files in
parallel (it's a backup appliance), where the initial lookup is
guaranteed to fail resulting in the two lock acquires on stock kernel
(and one with the patch at hand).
Implemented below is a synthetic benchmark providing the same behavior.
[I shall note the workload is not running on Linux, instead it was
causing trouble elsewhere. Benchmark below was used while addressing
said problems and was found to adequately represent the real workload.]
Total real time fluctuates by 1-2s.
With 20 threads each walking a dedicated 1000 dirs * 1000 files
directory tree to stat(2) on a 32 core + 24GB RAM vm:
ext4 (needed mkfs.ext4 -N 24000000):
before: 3.77s user 890.90s system 1939% cpu 46.118 total
after: 3.24s user 397.73s system 1858% cpu 21.581 total (-53%)
That's 20 million files to visit, while the machine can only cache about
15 million at a time (obtained from ext4_inode_cache object count in
/proc/slabinfo). Since each terminal inode is only visited once per run
this amounts to 0% hit ratio for the dentry cache and the hash table
(there are however hits for the intermediate directories).
On repeated runs the kernel caches the last ~15 mln, meaning there is ~5
mln of uncached inodes which are going to be visited first, evicting the
previously cached state as it happens.
Lack of hits can be trivially verified with bpftrace, like so:
bpftrace -e 'kretprobe:find_inode_fast { @[kstack(), retval != 0] = count(); }'\
-c "/bin/sh walktrees /testfs 20"
Best ran more than once.
Expected results after "warmup":
[snip]
@[
__ext4_iget+275
ext4_lookup+224
__lookup_slow+130
walk_component+219
link_path_walk.part.0.constprop.0+614
path_lookupat+62
filename_lookup+204
vfs_statx+128
vfs_fstatat+131
__do_sys_newfstatat+38
do_syscall_64+87
entry_SYSCALL_64_after_hwframe+118
, 1]: 20000
@[
__ext4_iget+275
ext4_lookup+224
__lookup_slow+130
walk_component+219
path_lookupat+106
filename_lookup+204
vfs_statx+128
vfs_fstatat+131
__do_sys_newfstatat+38
do_syscall_64+87
entry_SYSCALL_64_after_hwframe+118
, 1]: 20000000
That is 20 million calls for the initial lookup and 20 million after
allocating a new inode, all of them failing to return a value != 0
(i.e., they are returning NULL -- no match found).
Of course aborting the benchmark in the middle and starting it again (or
messing with the state in other ways) is going to alter these results.
Benchmark can be found here: https://people.freebsd.org/~mjg/fstree.tgz
[1] https://lore.kernel.org/all/20231206060629.2827226-1-david@fromorbit.com/
Signed-off-by: Mateusz Guzik <mjguzik@gmail.com>
Link: https://lore.kernel.org/r/20240611173824.535995-2-mjguzik@gmail.com
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Christian Brauner <brauner@kernel.org>
The documentation does not make sufficiently clear that edge polarity is
based on changes to the logical line values, and that the physical
polarity of edges is dependent on the active low flag.
Clarify the relationship between the active low flag and edge polarity
for the functions that read edge events.
Suggested-by: David C. Rankin <drankinatty@gmail.com>
Signed-off-by: Kent Gibson <warthog618@gmail.com>
Link: https://lore.kernel.org/r/20240610092157.9147-3-warthog618@gmail.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
The documentation does not make sufficiently clear that the uAPI deals with
logical line values, nor does it describe the influence of the active low
flag on the mapping between physical and logical line values.
Clarify the relationship between physical and logical line values and the
effect of the active low flag for ioctls passing line values.
Suggested-by: David C. Rankin <drankinatty@gmail.com>
Signed-off-by: Kent Gibson <warthog618@gmail.com>
Link: https://lore.kernel.org/r/20240610092157.9147-2-warthog618@gmail.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
There is a confusing pattern in the kernel to use a variable named 'timeout' to
store the result of wait_for_completion_timeout() causing patterns like:
timeout = wait_for_completion_timeout(...)
if (!timeout) return -ETIMEDOUT;
with all kinds of permutations. Use 'time_left' as a variable to make the code
self explaining.
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://msgid.link/20240603091541.8367-7-wsa+renesas@sang-engineering.com
There is a confusing pattern in the kernel to use a variable named 'timeout' to
store the result of wait_for_completion_interruptible_timeout() causing patterns like:
timeout = wait_for_completion_interruptible_timeout(...)
if (!timeout) return -ETIMEDOUT;
with all kinds of permutations. Use 'time_left' as a variable to make the code
self explaining.
Fix to the proper variable type 'long' while here.
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://msgid.link/20240603091541.8367-5-wsa+renesas@sang-engineering.com
There is a confusing pattern in the kernel to use a variable named 'timeout' to
store the result of wait_for_completion_timeout() causing patterns like:
timeout = wait_for_completion_timeout(...)
if (!timeout) return -ETIMEDOUT;
with all kinds of permutations. Use 'time_left' as a variable to make the code
self explaining.
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://msgid.link/20240603091541.8367-4-wsa+renesas@sang-engineering.com
There is a confusing pattern in the kernel to use a variable named 'timeout' to
store the result of wait_event_timeout() causing patterns like:
timeout = wait_event_timeout(...)
if (!timeout) return -ETIMEDOUT;
with all kinds of permutations. Use 'time_left' as a variable to make the code
self explaining.
Fix to the proper variable type 'long' while here.
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Acked-by: Arend van Spriel <arend.vanspriel@broadcom.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://msgid.link/20240603091541.8367-3-wsa+renesas@sang-engineering.com
This is an effort to get rid of all multiplications from allocation
functions in order to prevent integer overflows [1][2].
As the "cmd_buf" variable is a pointer to "struct at76_command" and
this structure ends in a flexible array:
struct at76_command {
[...]
u8 data[];
} __packed;
the preferred way in the kernel is to use the struct_size() helper to
do the arithmetic instead of the calculation "size + count" in the
kmalloc() function.
Also, declare a new variable (total_size) since the return value of the
struct_size() helper is used several times.
At the same time, prepare for the coming implementation by GCC and Clang
of the __counted_by attribute. Flexible array members annotated with
__counted_by can have their accesses bounds-checked at run-time via
CONFIG_UBSAN_BOUNDS (for array indexing) and CONFIG_FORTIFY_SOURCE (for
strcpy/memcpy-family functions). In this case, it is important to note
that the attribute used is "__counted_by_le" since the counter type is
"__le16".
This way, the code is more readable and safer.
Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#open-coded-arithmetic-in-allocator-arguments [1]
Link: https://github.com/KSPP/linux/issues/160 [2]
Signed-off-by: Erick Archer <erick.archer@outlook.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://msgid.link/AS8PR02MB7237578654CEDDFE5F8C17BA8BFE2@AS8PR02MB7237.eurprd02.prod.outlook.com
It is preferred to use sizeof(*pointer) instead of sizeof(type)
due to the type of the variable can change and one needs not
change the former (unlike the latter). This patch has no effect
on runtime behavior.
At the same time remove some redundant NULL initializations.
Signed-off-by: Erick Archer <erick.archer@outlook.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://msgid.link/AS8PR02MB7237C784C14DBC943CB719F88BFE2@AS8PR02MB7237.eurprd02.prod.outlook.com
make allmodconfig && make W=1 C=1 reports:
WARNING: modpost: missing MODULE_DESCRIPTION() in lib/test_printf.o
WARNING: modpost: missing MODULE_DESCRIPTION() in lib/test_scanf.o
Add the missing invocations of the MODULE_DESCRIPTION() macro.
Signed-off-by: Jeff Johnson <quic_jjohnson@quicinc.com>
Link: https://lore.kernel.org/r/20240531-md-vsprintf-v1-1-d8bc7e21539a@quicinc.com
Reviewed-by: Petr Mladek <pmladek@suse.com>
Signed-off-by: Petr Mladek <pmladek@suse.com>
zap_modalias_env() wrongly calculates size of memory block to move, so
will cause OOB memory access issue if variable MODALIAS is not the last
one within its @env parameter, fixed by correcting size to memmove.
Fixes: 9b3fa47d4a ("kobject: fix suppressing modalias in uevents delivered over netlink")
Cc: stable@vger.kernel.org
Signed-off-by: Zijun Hu <quic_zijuhu@quicinc.com>
Reviewed-by: Lk Sii <lk_sii@163.com>
Link: https://lore.kernel.org/r/1717074877-11352-1-git-send-email-quic_zijuhu@quicinc.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
make allmodconfig && make W=1 C=1 reports:
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/phy/broadcom/phy-bcm-ns-usb2.o
WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/phy/broadcom/phy-bcm-ns-usb3.o
Add the missing invocations of the MODULE_DESCRIPTION() macro.
Signed-off-by: Jeff Johnson <quic_jjohnson@quicinc.com>
Link: https://lore.kernel.org/r/20240608-md-drivers-phy-broadcom-v1-1-f070f84cc3f0@quicinc.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Armada CP110 UTMI supports swapping D+ and D- signals.
usb251xb.yaml already describes a suitable device-tree property for the
same purpose but as child usb controller node.
Add optional swap-dx-lanes device-tree property to armada cp110 utmi phy
with same semantics as usb251xb:
The property lists all ports that swap D+ and D-, unlisted ports are
considered correct.
Signed-off-by: Josua Mayer <josua@solid-run.com>
Reviewed-by: Rob Herring (Arm) <robh@kernel.org>
Link: https://lore.kernel.org/r/20240602-cn9130-som-v6-4-89393e86d4c7@solid-run.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
