ublk_ch_release currently quiesces the device's request_queue while
setting force_abort/fail_io. This avoids data races by preventing
concurrent reads from the I/O path, but is not strictly needed - at this
point, canceling is already set and guaranteed to be observed by any
concurrently executing I/Os, so they will be handled properly even if
the changes to force_abort/fail_io propagate to the I/O path later.
Remove the quiesce/unquiesce calls from ublk_ch_release. This makes the
writes to force_abort/fail_io concurrent with the reads in the I/O path,
so make the accesses atomic.
Before this change, the call to blk_mq_quiesce_queue was responsible for
most (90%) of the runtime of ublk_ch_release. With that call eliminated,
ublk_ch_release runs much faster. Here is a comparison of the total time
spent in calls to ublk_ch_release when a server handling 128 devices
exits, before and after this change:
before: 1.11s
after: 0.09s
Signed-off-by: Uday Shankar <ushankar@purestorage.com>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20250808-ublk_quiesce2-v1-1-f87ade33fa3d@purestorage.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If the network stack keeps a reference for too long, DRBD keeps
references on a higher number of pages as a consequence.
Fix all that by no longer relying on page reference counts dropping to
an expected value. Instead, DRBD gives up its reference and lets the
system handle everything else. While at it, remove the open-coded
custom page pool mechanism and use the page_pool included in the
kernel.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Christoph Böhmwalder <christoph.boehmwalder@linbit.com>
Tested-by: Eric Hagberg <ehagberg@janestreet.com>
Link: https://lore.kernel.org/r/20250605103852.23029-1-christoph.boehmwalder@linbit.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Refactor the gpio-aggregator module as a prerequisite for merging the
pin controller driver for AAEON UP boards.
-----BEGIN PGP SIGNATURE-----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=lL6Q
-----END PGP SIGNATURE-----
Merge tag 'gpio-aggregator-refactoring-for-v6.18-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux into gpio/for-next
Immutable branch for the pinctrl tree to pull from
Refactor the gpio-aggregator module as a prerequisite for merging the
pin controller driver for AAEON UP boards.
Commit 8b17e54096 ("vfs: add initial support for CONFIG_DEBUG_VFS") added
dump_inode(), but dump_inode() currently reports only raw pointer address.
Comment says that adding a proper inode dumping routine is a TODO.
However, syzkaller concurrently tests multiple filesystems, and several
filesystems started calling dump_inode() due to hitting VFS_BUG_ON_INODE()
added by commit af153bb63a ("vfs: catch invalid modes in may_open()")
before a proper inode dumping routine is implemented.
Show filesystem name at dump_inode() so that we can find which filesystem
has passed an invalid mode to may_open() from syzkaller's crash reports.
Link: https://syzkaller.appspot.com/bug?extid=895c23f6917da440ed0d
Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Link: https://lore.kernel.org/ceaf4021-65cc-422e-9d0e-6afa18dd8276@I-love.SAKURA.ne.jp
Signed-off-by: Christian Brauner <brauner@kernel.org>
USB_CLK Clock and Pin Control DT binding definitions for the Renesas
RZ/T2H (R9A09G077) and RZ/N2H (R9A09G087) SoCs, shared by driver and DT
source files.
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQQ9qaHoIs/1I4cXmEiKwlD9ZEnxcAUCaJn0uQAKCRCKwlD9ZEnx
cP16AP9bhBcnWHfForGpSJAfuceMJY5en2uoU07Qmb2PGMdf0AD8DLhHVwW11UVN
VGlRLKMPZ8MDuVQzDUomawY1gZbiNgc=
=EVIx
-----END PGP SIGNATURE-----
Merge tag 'renesas-r9a09g077-dt-binding-defs-tag3' into renesas-pinctrl-for-v6.18
Renesas RZ/T2H and RZ/N2H USB_CLK and Pin Control DT Binding Definitions
USB_CLK Clock and Pin Control DT binding definitions for the Renesas
RZ/T2H (R9A09G077) and RZ/N2H (R9A09G087) SoCs, shared by driver and DT
source files.
Document the pin and GPIO controller IP for the Renesas RZ/T2H
(R9A09G077) and RZ/N2H (R9A09G087) SoCs, and add the shared DTSI header
file used by both the bindings and the driver.
The RZ/T2H SoC supports 729 pins, while RZ/N2H supports 576 pins.
Both share the same controller architecture; separate compatible strings
are added for each SoC to distinguish them.
Co-developed-by: Thierry Bultel <thierry.bultel.yh@bp.renesas.com>
Signed-off-by: Thierry Bultel <thierry.bultel.yh@bp.renesas.com>
Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Reviewed-by: "Rob Herring (Arm)" <robh@kernel.org>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/20250808133017.2053637-2-prabhakar.mahadev-lad.rj@bp.renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Add the USB clock (USB_CLK) definition for the Renesas RZ/T2H
(R9A09G077) and RZ/N2H (R9A09G087) SoCs. USB_CLK is used as the
reference clock for USB PHY layer.
Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/20250804202643.3967484-2-prabhakar.mahadev-lad.rj@bp.renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Resolve a conflict between
commit 6a68d28066 ("selftests/coredump: Fix "socket_detect_userspace_client" test failure")
and
commit 994dc26302 ("selftests/coredump: fix build")
The first commit adds a read() to wait for write() from another thread to
finish. But the second commit removes the write().
Now that the two commits are in the same tree, the read() now gets EOF and
the test fails.
Remove this read() so that the test passes.
Signed-off-by: Nam Cao <namcao@linutronix.de>
Link: https://lore.kernel.org/20250811074957.4079616-1-namcao@linutronix.de
Signed-off-by: Christian Brauner <brauner@kernel.org>
Add str_input_output() helper to return 'input' or 'output' string literal.
Also add the inversed variant str_output_input().
Suggested-by: Andy Shevchenko <andy@kernel.org>
Reviewed-by: Andy Shevchenko <andy@kernel.org>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Thomas Richard <thomas.richard@bootlin.com>
Link: https://lore.kernel.org/r/20250811-aaeon-up-board-pinctrl-support-v9-9-29f0cbbdfb30@bootlin.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Add a data pointer to store private data in the forwarder.
Reviewed-by: Andy Shevchenko <andy@kernel.org>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Thomas Richard <thomas.richard@bootlin.com>
Link: https://lore.kernel.org/r/20250811-aaeon-up-board-pinctrl-support-v9-8-29f0cbbdfb30@bootlin.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Add request() callback to check if the GPIO descriptor was well registered
in the gpiochip_fwd before using it. This is done to handle the case where
GPIO descriptor is added at runtime in the forwarder.
If at least one GPIO descriptor was not added before the forwarder
registration, we assume the forwarder can sleep as if a GPIO is added at
runtime it may sleep.
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Thomas Richard <thomas.richard@bootlin.com>
Link: https://lore.kernel.org/r/20250811-aaeon-up-board-pinctrl-support-v9-7-29f0cbbdfb30@bootlin.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Export all symbols and create header file for the GPIO forwarder library.
It will be used in the next changes.
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Thomas Richard <thomas.richard@bootlin.com>
Link: https://lore.kernel.org/r/20250811-aaeon-up-board-pinctrl-support-v9-6-29f0cbbdfb30@bootlin.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Add a new function gpiochip_fwd_register(), which finalizes the
initialization of the forwarder and registers the corresponding gpiochip.
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Thomas Richard <thomas.richard@bootlin.com>
Link: https://lore.kernel.org/r/20250811-aaeon-up-board-pinctrl-support-v9-4-29f0cbbdfb30@bootlin.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Create a dedicated function to add a GPIO desc in the forwarder. Instead of
saving a GPIO descs array pointer, now the GPIO descs are passed one by one
to the forwarder which registers them in its own array. So after the call
of gpiochip_fwd_create(), the passed array can be free.
Reviewed-by: Andy Shevchenko <andy@kernel.org>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Thomas Richard <thomas.richard@bootlin.com>
Link: https://lore.kernel.org/r/20250811-aaeon-up-board-pinctrl-support-v9-3-29f0cbbdfb30@bootlin.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Move the GPIO forwarder allocation and static initialization in a dedicated
function.
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Thomas Richard <thomas.richard@bootlin.com>
Link: https://lore.kernel.org/r/20250811-aaeon-up-board-pinctrl-support-v9-2-29f0cbbdfb30@bootlin.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Add support to register for GPIO<->pin mapping using a list of non
consecutive pins. The core already supports sparse pin range (pins member
of struct pinctrl_gpio_range), but it was not possible to register one. If
pins is not NULL the core uses it, otherwise it assumes that a consecutive
pin range was registered and it uses pin_base.
The function gpiochip_add_pin_range() which allocates and fills the struct
pinctrl_gpio_range was renamed to gpiochip_add_pin_range_with_pins() and
the pins parameter was added.
Two new functions were added, gpiochip_add_pin_range() and
gpiochip_add_sparse_pin_range() to register a consecutive or sparse pins
range. Both use gpiochip_add_pin_range_with_pins().
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Thomas Richard <thomas.richard@bootlin.com>
Link: https://lore.kernel.org/r/20250811-aaeon-up-board-pinctrl-support-v9-1-29f0cbbdfb30@bootlin.com
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
When I authored these bindings I had no idea what "AEM" stood
for. Now I know: it means "Architecture Envelope Model".
Detail this in the bindings.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Acked-by: Conor Dooley <conor.dooley@microchip.com>
Reviewed-by: Sudeep Holla <sudeep.holla@arm.com>
Link: https://lore.kernel.org/r/20250806-aem-dt-bind-v1-1-d14676dfb027@linaro.org
Signed-off-by: Rob Herring (Arm) <robh@kernel.org>
Convert the ASpeed Coldfire offloaded GPIO FSI master.
Drop the "fsi-master" compatible as it has not be used consistently and
doesn't represent anything.
Acked-by: Eddie James <eajames@linux.ibm.com>
Link: https://lore.kernel.org/r/20250731-dt-fsi-cleanups-v1-6-e7b695a29fc3@kernel.org
Signed-off-by: Rob Herring (Arm) <robh@kernel.org>
Convert the GPIO-based FSI master binding to DT schema format.
Drop the "fsi-master" compatible as it has not be used consistently and
doesn't represent anything.
Acked-by: Eddie James <eajames@linux.ibm.com>
Link: https://lore.kernel.org/r/20250731-dt-fsi-cleanups-v1-5-e7b695a29fc3@kernel.org
Signed-off-by: Rob Herring (Arm) <robh@kernel.org>
Maintainers of a subsystem should also be the maintainer for
corresponding DT bindings. Add the FSI bindings to the FSI subsystem
entry.
Acked-by: Eddie James <eajames@linux.ibm.com>
Link: https://lore.kernel.org/r/20250731-dt-fsi-cleanups-v1-4-e7b695a29fc3@kernel.org
Signed-off-by: Rob Herring (Arm) <robh@kernel.org>
The active-semi,act8945a binding is documented in multiple places. The
charger child node is documented in regulator/active-semi,act8945a.yaml
and power/supply/active-semi,act8945a-charger.yaml. An old text binding
is in mfd/act8945a.txt.
Update the regulator/active-semi,act8945a.yaml with the additional
descriptions and constraints from
power/supply/active-semi,act8945a-charger.yaml, and then remove it and
mfd/act8945a.txt.
Signed-off-by: Rob Herring (Arm) <robh@kernel.org>
Link: https://patch.msgid.link/20250807214459.4173892-1-robh@kernel.org
Signed-off-by: Mark Brown <broonie@kernel.org>
devm_kzalloc() has already been initialized to full 0 space, there is no
need to use memset() to initialize again.
Signed-off-by: Liao Yuanhong <liaoyuanhong@vivo.com>
Link: https://patch.msgid.link/20250807123949.495193-1-liaoyuanhong@vivo.com
Signed-off-by: Mark Brown <broonie@kernel.org>
This adds the following commits from upstream:
52f07dcca47c dtc: Add informative error for stray identifier
9cabae6b0351 checks: Fix detection of 'i2c-bus' node
605dc044c3fe New helper to add markers
7da5d106c740 fdtput: Fix documentation about existing nodes
53c63dd421d7 dtdiff: Use input format dtb for dtbo files
84d9dd2fcbc8 dtc: Add data_insert_data function
97011d1f4e98 meson: use override_find_program/override_dependency
b841391bbd08 srcpos: Define srcpos_free
e0b7749c26a9 Add alloc_marker
ecb21febfdd3 meson: port python bindings to build natively via meson and meson-python
7ebfcac8520e Makefile: deprecate in favor of Meson
f4c53f4ebf78 Use __ASSEMBLER__ instead of __ASSEMBLY__
205fbef17b7b Fix some typos
da85f91931e5 Remove duplicated words in documentation and comments
dd1b3e532d22 meson: support building libfdt without static library
1ccd232709d4 meson: don't build test programs by default
ce1d8588880a tests: When building .so from -O asm output mark as non-executable stack
915daadbb62d Start with empty __local_fixups__ and __fixups__ nodes
4ea851f5a44d Let get_subnode() not return deleted nodes
175d2a564c47 Use build_root_node() instead of open-coding it
18f4f305fdd7 build: fix -Dtools=false build
267efc7d4694 checks: Warn about missing #address-cells for interrupt parents
755db115355b libfdt: Add fdt_setprop_namelen_string()
bdca8612009e libfdt: Add fdt_setprop_namelen()
0f69cedc08fc libfdt_internal: fdt_find_string_len_()
56b2b30c5bd0 libfdt: add fdt_get_property_namelen_w()
1e8c5f60e127 Add clang-format config
6f183c7d9246 checks: Relax avoid_unnecessary_addr_size check to allow child ranges properties
66c7d0e6f4f3 tests/sw_tree1.c: fix unitialized saveptr
9a969f3b70b0 pylibfdt/libfdt.i: fix backwards compatibility of return values
4292b072a23a .github/workflows: update ubuntu runner to supported version
1c745a9bd169 libfdt: Remove fdt parameter from overlay_fixup_one_phandle
b3bbee6b1242 libfdt: Move the SBOM authors section
d1656730abfb Add a SBOM file in CycloneDX format
b75515af4576 libfdt: Remove extra semi-colons outside functions
2d10aa2afe35 Bump version to v1.7.2
48795c82bdb6 pylibfdt: Don't emit warnings from swig generate C code
838f11e830e3 fdtoverlay: provide better error message for missing `/__symbols__`
d1e2384185c5 pylibfdt/libfdt.i: Use SWIG_AppendOutput
18aa49a9f68d Escape spaces in depfile with backslashes.
f9968fa06921 libfdt.h: whitespace consistency fixups
9b5f65fb3d8d libfdt.h: typo and consistency fixes
99031e3a4a6e Bump version to v1.7.1
3d5e376925fd setup: Move setting of srcdir down to the bottom
e277553b9880 setup: Collect top-level code together
7e5a88984081 setup: Move version and full_description into a function
78b6a85c113b Tidy up some pylint warnings
3501d373f0a2 Require Python 3
The added include of string.h in libfdt_internal.h breaks the kernel
overriding libfdt_env.h with its own string functions, so it is dropped.
An upstream fix is pending.
Signed-off-by: Rob Herring (Arm) <robh@kernel.org>
Remove unnecessary semicolons reported by Coccinelle/coccicheck and the
semantic patch at scripts/coccinelle/misc/semicolon.cocci.
Signed-off-by: Chen Ni <nichen@iscas.ac.cn>
Link: https://patch.msgid.link/20250729040044.1851988-1-nichen@iscas.ac.cn
Signed-off-by: Mark Brown <broonie@kernel.org>
If "st,omm-req2ack-ns" property is found and its value is not 0,
the current test doesn't allow to compute and set req2ack value,
Fix this test.
Fixes: 8181d061dc ("memory: Add STM32 Octo Memory Manager driver")
Signed-off-by: Patrice Chotard <patrice.chotard@foss.st.com>
Link: https://lore.kernel.org/r/20250807-upstream_omm_fix_req2ack_test_condition-v2-1-d7df4af2b48b@foss.st.com
Cc: <stable@vger.kernel.org>
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Aleksa Sarai <cyphar@cyphar.com> says:
Userspace generally expects APIs that return -EMSGSIZE to allow for them
to adjust their buffer size and retry the operation. However, the
fscontext log would previously clear the message even in the -EMSGSIZE
case.
Given that it is very cheap for us to check whether the buffer is too
small before we remove the message from the ring buffer, let's just do
that instead. While we're at it, refactor some fscontext_read() into a
separate helper to make the ring buffer logic a bit easier to read.
* patches from https://lore.kernel.org/20250807-fscontext-log-cleanups-v3-0-8d91d6242dc3@cyphar.com:
selftests/filesystems: add basic fscontext log tests
fscontext: do not consume log entries when returning -EMSGSIZE
Link: https://lore.kernel.org/20250807-fscontext-log-cleanups-v3-0-8d91d6242dc3@cyphar.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
Aleksa Sarai <cyphar@cyphar.com> says:
It makes little sense for fsmount() to output the warning message when
mount_too_revealing() is violated to kmsg. Instead, the warning should
be output (with a "VFS" prefix) to the fscontext log. In addition,
include the same log message for mount_too_revealing() when doing a
regular mount for consistency.
With the newest fsopen()-based mount(8) from util-linux, the error
messages now look like
# mount -t proc proc /tmp
mount: /tmp: fsmount() failed: VFS: Mount too revealing.
dmesg(1) may have more information after failed mount system call.
which could finally result in mount_too_revealing() errors being easier
for users to detect and understand.
* patches from https://lore.kernel.org/20250806-errorfc-mount-too-revealing-v2-0-534b9b4d45bb@cyphar.com:
vfs: output mount_too_revealing() errors to fscontext
fscontext: add custom-prefix log helpers
Link: https://lore.kernel.org/20250806-errorfc-mount-too-revealing-v2-0-534b9b4d45bb@cyphar.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
Userspace generally expects APIs that return -EMSGSIZE to allow for them
to adjust their buffer size and retry the operation. However, the
fscontext log would previously clear the message even in the -EMSGSIZE
case.
Given that it is very cheap for us to check whether the buffer is too
small before we remove the message from the ring buffer, let's just do
that instead. While we're at it, refactor some fscontext_read() into a
separate helper to make the ring buffer logic a bit easier to read.
Fixes: 007ec26cdc ("vfs: Implement logging through fs_context")
Cc: David Howells <dhowells@redhat.com>
Cc: stable@vger.kernel.org # v5.2+
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/20250807-fscontext-log-cleanups-v3-1-8d91d6242dc3@cyphar.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
Pedro Falcato <pfalcato@suse.de> says:
This patchset contains a very small cleanup, where we remove mount_nodev
(unused since 6.15) and mount_bdev (unused since f2fs converted their
mount API usage, still in -next). While we're at it, we also get to
removing some stale mount API docs.
* patches from https://lore.kernel.org/20250723132156.225410-1-pfalcato@suse.de:
docs/vfs: Remove mentions to the old mount API helpers
fs: Remove mount_bdev
fs: Remove mount_nodev
Link: https://lore.kernel.org/20250723132156.225410-1-pfalcato@suse.de
Signed-off-by: Christian Brauner <brauner@kernel.org>
It makes little sense for fsmount() to output the warning message when
mount_too_revealing() is violated to kmsg. Instead, the warning should
be output (with a "VFS" prefix) to the fscontext log. In addition,
include the same log message for mount_too_revealing() when doing a
regular mount for consistency.
With the newest fsopen()-based mount(8) from util-linux, the error
messages now look like
# mount -t proc proc /tmp
mount: /tmp: fsmount() failed: VFS: Mount too revealing.
dmesg(1) may have more information after failed mount system call.
which could finally result in mount_too_revealing() errors being easier
for users to detect and understand.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/20250806-errorfc-mount-too-revealing-v2-2-534b9b4d45bb@cyphar.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
Now that mount_bdev(), mount_nodev() and mount_single() have all been
removed, remove mentions to them in vfs.rst.
While we're at it, redirect people looking for mount API docs to
mount_api.rst (which documents the newer API).
Signed-off-by: Pedro Falcato <pfalcato@suse.de>
Link: https://lore.kernel.org/20250723132156.225410-4-pfalcato@suse.de
Signed-off-by: Christian Brauner <brauner@kernel.org>
Sometimes, errors associated with an fscontext come from the VFS or
otherwise outside of the filesystem driver itself. However, the default
logging of errorfc will always prefix the message with the filesystem
name.
So, add some *fcp() wrappers that allow for custom prefixes to be used
when emitting information to the fscontext log.
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/20250806-errorfc-mount-too-revealing-v2-1-534b9b4d45bb@cyphar.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
mount_bdev has no in-tree users ever since f2fs adopted the new mount
API. Remove it.
Signed-off-by: Pedro Falcato <pfalcato@suse.de>
Link: https://lore.kernel.org/20250723132156.225410-3-pfalcato@suse.de
Signed-off-by: Christian Brauner <brauner@kernel.org>
replace_fd() returns the number of the new file descriptor through the
return value of do_dup2(). However its callers never care about the
specific returned number. In fact the caller in receive_fd_replace() treats
any non-zero return value as an error and therefore never calls
__receive_sock() for most file descriptors, which is a bug.
To fix the bug in receive_fd_replace() and to avoid the same issue
happening in future callers, signal success through a plain zero.
Suggested-by: Al Viro <viro@zeniv.linux.org.uk>
Link: https://lore.kernel.org/lkml/20250801220215.GS222315@ZenIV/
Fixes: 173817151b ("fs: Expand __receive_fd() to accept existing fd")
Fixes: 42eb0d54c0 ("fs: split receive_fd_replace from __receive_fd")
Cc: stable@vger.kernel.org
Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de>
Link: https://lore.kernel.org/20250805-fix-receive_fd_replace-v3-1-b72ba8b34bac@linutronix.de
Signed-off-by: Christian Brauner <brauner@kernel.org>
The kernel-doc for name_contains_dotdot() was missing the @name
parameter description, leading to a warning during make htmldocs.
Add the missing documentation to resolve this warning.
Signed-off-by: Kriish Sharma <kriish.sharma2006@gmail.com>
Link: https://lore.kernel.org/20250730201853.8436-1-kriish.sharma2006@gmail.com
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Christian Brauner <brauner@kernel.org>
file_remove_privs_flags is only used inside of inode.c, mark it static.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/20250724074854.3316911-1-hch@lst.de
Signed-off-by: Christian Brauner <brauner@kernel.org>
Commit b875bd5b38 ("exportfs: Remove EXPORT_OP_ASYNC_LOCK") removed
all references to EXPORT_OP_ASYNC_LOCK, but one lasted in the
comments for fs/locks.c. Remove it.
Signed-off-by: Thiago Becker <tbecker@redhat.com>
Link: https://lore.kernel.org/20250724203516.153616-1-tbecker@redhat.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
With fuse now using iomap for writeback handling, inode blkbits changes
are problematic because iomap relies on inode->i_blkbits for its
internal bitmap logic. Currently we change inode->i_blkbits in fuse to
match the attr->blksize value passed in by the server.
This commit keeps inode->i_blkbits constant in fuse. Any attr->blksize
values passed in by the server will not update inode->i_blkbits. The
client-side behavior for stat is unaffected, stat will still reflect the
blocksize passed in by the server.
Signed-off-by: Joanne Koong <joannelkoong@gmail.com>
Link: https://lore.kernel.org/20250807175015.515192-1-joannelkoong@gmail.com
Fixes: ef7e7cbb32 ("fuse: use iomap for writeback")
Signed-off-by: Christian Brauner <brauner@kernel.org>
Aleksa Sarai <cyphar@cyphar.com> says:
As described in commit 7a54947e72 ('Merge patch series "fs: allow
changing idmappings"'), open_tree_attr(2) was necessary in order to
allow for a detached mount to be created and have its idmappings changed
without the risk of any racing threads operating on it. For this reason,
mount_setattr(2) still does not allow for id-mappings to be changed.
However, there was a bug in commit 2462651ffa ("fs: allow changing
idmappings") which allowed users to bypass this restriction by calling
open_tree_attr(2) *without* OPEN_TREE_CLONE.
can_idmap_mount() prevented this bug from allowing an attached
mountpoint's id-mapping from being modified (thanks to an is_anon_ns()
check), but this still allows for detached (but visible) mounts to have
their be id-mapping changed. This risks the same UAF and locking issues
as described in the merge commit, and was likely unintentional.
For what it's worth, I found this while working on the open_tree_attr(2)
man page, and was trying to figure out what open_tree_attr(2)'s
behaviour was in the (slightly fruity) ~OPEN_TREE_CLONE case.
* patches from https://lore.kernel.org/20250808-open_tree_attr-bugfix-idmap-v1-0-0ec7bc05646c@cyphar.com:
selftests/mount_setattr: add smoke tests for open_tree_attr(2) bug
open_tree_attr: do not allow id-mapping changes without OPEN_TREE_CLONE
Link: https://lore.kernel.org/20250808-open_tree_attr-bugfix-idmap-v1-0-0ec7bc05646c@cyphar.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
Commit d279c80e0b ("iomap: inline iomap_dio_bio_opflags()") has broken
the logic in iomap_dio_bio_iter() in a way that when the device does
support FUA (or has no writeback cache) and the direct IO happens to
freshly allocated or unwritten extents, we will *not* issue fsync after
completing direct IO O_SYNC / O_DSYNC write because the
IOMAP_DIO_WRITE_THROUGH flag stays mistakenly set. Fix the problem by
clearing IOMAP_DIO_WRITE_THROUGH whenever we do not perform FUA write as
it was originally intended.
CC: John Garry <john.g.garry@oracle.com>
CC: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
Fixes: d279c80e0b ("iomap: inline iomap_dio_bio_opflags()")
CC: stable@vger.kernel.org
Signed-off-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/20250730102840.20470-2-jack@suse.cz
Reviewed-by: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
Reviewed-by: John Garry <john.g.garry@oracle.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Christian Brauner <brauner@kernel.org>
There appear to be no other open_tree_attr(2) tests at the moment, but
as a minimal solution just add some additional checks in the existing
MOUNT_ATTR_IDMAP tests to make sure that open_tree_attr(2) cannot be
used to bypass the tested restrictions that apply to mount_setattr(2).
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/20250808-open_tree_attr-bugfix-idmap-v1-2-0ec7bc05646c@cyphar.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
As described in commit 7a54947e72 ('Merge patch series "fs: allow
changing idmappings"'), open_tree_attr(2) was necessary in order to
allow for a detached mount to be created and have its idmappings changed
without the risk of any racing threads operating on it. For this reason,
mount_setattr(2) still does not allow for id-mappings to be changed.
However, there was a bug in commit 2462651ffa ("fs: allow changing
idmappings") which allowed users to bypass this restriction by calling
open_tree_attr(2) *without* OPEN_TREE_CLONE.
can_idmap_mount() prevented this bug from allowing an attached
mountpoint's id-mapping from being modified (thanks to an is_anon_ns()
check), but this still allows for detached (but visible) mounts to have
their be id-mapping changed. This risks the same UAF and locking issues
as described in the merge commit, and was likely unintentional.
Fixes: 2462651ffa ("fs: allow changing idmappings")
Cc: stable@vger.kernel.org # v6.15+
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Link: https://lore.kernel.org/20250808-open_tree_attr-bugfix-idmap-v1-1-0ec7bc05646c@cyphar.com
Signed-off-by: Christian Brauner <brauner@kernel.org>
