Shin'ichiro reported that when he's running fstests' test-case
btrfs/167 on emulated zoned devices, he's seeing the following NULL
pointer dereference in 'btrfs_zone_finish_endio()':
Oops: general protection fault, probably for non-canonical address 0xdffffc0000000011: 0000 [#1] PREEMPT SMP KASAN NOPTI
KASAN: null-ptr-deref in range [0x0000000000000088-0x000000000000008f]
CPU: 4 PID: 2332440 Comm: kworker/u80:15 Tainted: G W 6.10.0-rc2-kts+ #4
Hardware name: Supermicro Super Server/X11SPi-TF, BIOS 3.3 02/21/2020
Workqueue: btrfs-endio-write btrfs_work_helper [btrfs]
RIP: 0010:btrfs_zone_finish_endio.part.0+0x34/0x160 [btrfs]
As zoned emulation mode simulates conventional zones on regular devices,
we cannot use zone-append for writing. But we're only attaching dummy
checksums if we're doing a zone-append write.
So for NOCOW zoned data writes on conventional zones, also attach a
dummy checksum.
Reported-by: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Fixes: cbfce4c7fbde ("btrfs: optimize the logical to physical mapping for zoned writes") CC: Naohiro Aota <Naohiro.Aota@wdc.com> # 6.6+ Tested-by: Shin'ichiro Kawasaki <shinichiro.kawasaki@wdc.com> Reviewed-by: Naohiro Aota <naohiro.aota@wdc.com> Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Juan reported that after doing some changes to buzzer [0] and implementing
a new fuzzing strategy guided by coverage, they noticed the following in
one of the probes:
What can be seen here is a register invariant violation on line 19. After
the binary-or in line 18, the verifier knows that bit 2 is set but knows
nothing about the rest of the content which was loaded from a map value,
meaning, range is [2,0x7fffffff] with var_off=(0x2; 0x7ffffffd). When in
line 19 the verifier analyzes the branch, it splits the register states
in reg_set_min_max() into the registers of the true branch (true_reg1,
true_reg2) and the registers of the false branch (false_reg1, false_reg2).
Since the test is w6 != 0x7ffffffd, the src_reg is a known constant.
Internally, the verifier creates a "fake" register initialized as scalar
to the value of 0x7ffffffd, and then passes it onto reg_set_min_max(). Now,
for line 19, it is mathematically impossible to take the false branch of
this program, yet the verifier analyzes it. It is impossible because the
second bit of r6 will be set due to the prior or operation and the
constant in the condition has that bit unset (hex(fd) == binary(1111 1101).
When the verifier first analyzes the false / fall-through branch, it will
compute an intersection between the var_off of r6 and of the constant. This
is because the verifier creates a "fake" register initialized to the value
of the constant. The intersection result later refines both registers in
regs_refine_cond_op():
Since the verifier is analyzing the false branch of the conditional jump,
reg1 is equal to false_reg1 and reg2 is equal to false_reg2, i.e. the reg2
is the "fake" register that was meant to hold a constant value. The resulting
var_off of the intersection says that both registers now hold a known value
of var_off=(0x7fffffff, 0x0) or in other words: this operation manages to
make the verifier think that the "constant" value that was passed in the
jump operation now holds a different value.
Normally this would not be an issue since it should not influence the true
branch, however, false_reg2 and true_reg2 are pointers to the same "fake"
register. Meaning, the false branch can influence the results of the true
branch. In line 24, the verifier assumes R6_w=0, but the actual runtime
value in this case is 1. The fix is simply not passing in the same "fake"
register location as inputs to reg_set_min_max(), but instead making a
copy. Moving the fake_reg into the env also reduces stack consumption by
120 bytes. With this, the verifier successfully rejects invalid accesses
from the test program.
[0] https://github.com/google/buzzer
Fixes: 67420501e868 ("bpf: generalize reg_set_min_max() to handle non-const register comparisons") Reported-by: Juan José López Jaimez <jjlopezjaimez@google.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Reviewed-by: John Fastabend <john.fastabend@gmail.com> Link: https://lore.kernel.org/r/20240613115310.25383-1-daniel@iogearbox.net Signed-off-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
ice_aqc_opc_download_pkg (0x0C40) AQ sporadically returns error due
to FW issue. Fix this by retrying five times before moving to
Safe Mode. Sleep for 20 ms before retrying. This was tested with the
4.40 firmware.
Fixes: c76488109616 ("ice: Implement Dynamic Device Personalization (DDP) download") Reviewed-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com> Signed-off-by: Wojciech Drewek <wojciech.drewek@intel.com> Reviewed-by: Brett Creeley <brett.creeley@amd.com> Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com> Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel) Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Commit 24407a01e57c ("ice: Add 200G speed/phy type use") added support
for 200G PHY speeds, but did not include 200G link speed message
support. As a result the driver incorrectly reports Unknown for 200G
link speed.
Fix this by adding 200G support to ice_print_link_msg().
Fixes: 24407a01e57c ("ice: Add 200G speed/phy type use") Reviewed-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com> Signed-off-by: Paul Greenwalt <paul.greenwalt@intel.com> Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel) Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
A bug in https://bugzilla.kernel.org/show_bug.cgi?id=218906 describes
that irdma would break and report hardware initialization failed after
suspend/resume with Intel E810 NIC (tested on 6.9.0-rc5).
The problem is caused due to the collision between the irq numbers
requested in irdma and the irq numbers requested in other drivers
after suspend/resume.
The irq numbers used by irdma are derived from ice's ice_pf->msix_entries
which stores mappings between MSI-X index and Linux interrupt number.
It's supposed to be cleaned up when suspend and rebuilt in resume but
it's not, causing irdma using the old irq numbers stored in the old
ice_pf->msix_entries to request_irq() when resume. And eventually
collide with other drivers.
This patch fixes this problem. On suspend, we call ice_deinit_rdma() to
clean up the ice_pf->msix_entries (and free the MSI-X vectors used by
irdma if we've dynamically allocated them). On resume, we call
ice_init_rdma() to rebuild the ice_pf->msix_entries (and allocate the
MSI-X vectors if we would like to dynamically allocate them).
Fixes: f9f5301e7e2d ("ice: Register auxiliary device to provide RDMA") Tested-by: Cyrus Lien <cyrus.lien@canonical.com> Signed-off-by: En-Wei Wu <en-wei.wu@canonical.com> Reviewed-by: Wojciech Drewek <wojciech.drewek@intel.com> Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel) Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
The Framework Laptop 16 does not have a combination headphone/headset
3.5mm jack; however, applying the pincfg from the Laptop 13 (nid=0x19)
erroneously informs hda that the node is present.
When a monitor interface is started, ieee80211_recalc_offload() is
called and 802.11 encapsulation offloading support get disabled so
monitor interface could get native wifi frames directly. But when
this interface is stopped there is no need to keep the 802.11
encpasulation offloading off.
This call ieee80211_recalc_offload() when monitor interface is stopped
so 802.11 encapsulation offloading gets re-activated if possible.
For using the ROC command, check that the ROC version
is *greater or equal* to 3, rather than *equal* to 3.
The ROC version was added to the TLV starting from
version 3.
The kprobes and synth event generation test modules add events and lock
(get a reference) those event file reference in module init function,
and unlock and delete it in module exit function. This is because those
are designed for playing as modules.
If we make those modules as built-in, those events are left locked in the
kernel, and never be removed. This causes kprobe event self-test failure
as below.
It was discovered that some device have CBR address set to 0 causing
kernel panic when arch_sync_dma_for_cpu_all is called.
This was notice in situation where the system is booted from TP1 and
BMIPS_GET_CBR() returns 0 instead of a valid address and
!!(read_c0_brcm_cmt_local() & (1 << 31)); not failing.
The current check whether RAC flush should be disabled or not are not
enough hence lets check if CBR is a valid address or not.
Fixes: ab327f8acdf8 ("mips: bmips: BCM6358: disable RAC flush for TP1") Signed-off-by: Christian Marangi <ansuelsmth@gmail.com> Acked-by: Florian Fainelli <florian.fainelli@broadcom.com> Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Sasha Levin <sashal@kernel.org>
read_config_dword() contains strange condition checking ret for a
number of values. The ret variable, however, is always zero because
config_access() never returns anything else. Thus, the retry is always
taken until number of tries is exceeded.
The code looks like it wants to check *val instead of ret to see if the
read gave an error response.
Fixes: 73b4390fb234 ("[MIPS] Routerboard 532: Support for base system") Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com> Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Sasha Levin <sashal@kernel.org>
The cs35l41_hda_unbind() function clears the hda_component entry
matching it's index and then dereferences the codec pointer held in the
first element of the hda_component array, this is an issue when the
device index was 0.
Instead use the codec pointer stashed in the cs35l41_hda structure as it
will still be valid.
Fixes: 7cf5ce66dfda ("ALSA: hda: cs35l41: Add device_link between HDA and cs35l41_hda") Signed-off-by: Simon Trimmer <simont@opensource.cirrus.com> Link: https://lore.kernel.org/r/20240531120820.35367-1-simont@opensource.cirrus.com Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Sasha Levin <sashal@kernel.org>
The required_opp_tables parsing is not perfect, as the OPP core does the
parsing solely based on the DT node pointers.
The core sets the required_opp_tables entry to the first OPP table in
the "opp_tables" list, that matches with the node pointer.
If the target DT OPP table is used by multiple devices and they all
create separate instances of 'struct opp_table' from it, then it is
possible that the required_opp_tables entry may be set to the incorrect
sibling device.
Unfortunately, there is no clear way to initialize the right values
during the initial parsing and we need to do this at a later point of
time.
Cross check the OPP table again while the genpds are attached and fix
them if required.
Also add a new API for the genpd core to fetch the device pointer for
the genpd.
While a device is runtime suspended along with its PCI hierarchy, the
device could get disconnected. In such case, the link will not come up no
matter how long pci_dev_wait() waits for it.
Besides the above mentioned case, there could be other ways to get the
device disconnected while pci_dev_wait() is waiting for the link to come
up.
Make pci_dev_wait() exit if the device is already disconnected to avoid
unnecessary delay.
The use cases of pci_dev_wait() boil down to two:
1. Waiting for the device after reset
2. pci_bridge_wait_for_secondary_bus()
The callers in both cases seem to benefit from propagating the
disconnection as error even if device disconnection would be more
analoguous to the case where there is no device in the first place which
return 0 from pci_dev_wait(). In the case 2, it results in unnecessary
marking of the devices disconnected again but that is just harmless extra
work.
Also make sure compiler does not become too clever with dev->error_state
and use READ_ONCE() to force a fetch for the up-to-date value.
Link: https://lore.kernel.org/r/20240208132322.4811-1-ilpo.jarvinen@linux.intel.com Reported-by: Mika Westerberg <mika.westerberg@linux.intel.com> Tested-by: Mika Westerberg <mika.westerberg@linux.intel.com> Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com> Signed-off-by: Bjorn Helgaas <bhelgaas@google.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
It is reported that _DSM evaluation fails in ucsi_acpi_dsm() on Lenovo
IdeaPad Pro 5 due to a missing address space handler for the EC address
space:
ACPI Error: No handler for Region [ECSI] (000000007b8176ee) [EmbeddedControl] (20230628/evregion-130)
This happens because if there is no ECDT, the EC driver only registers
the EC address space handler for operation regions defined in the EC
device scope of the ACPI namespace while the operation region being
accessed by the _DSM in question is located beyond that scope.
To address this, modify the ACPI EC driver to install the EC address
space handler at the root of the ACPI namespace for the first EC that
can be found regardless of whether or not an ECDT is present.
Note that this change is consistent with some examples in the ACPI
specification in which EC operation regions located outside the EC
device scope are used (for example, see Section 9.17.15 in ACPI 6.5),
so the current behavior of the EC driver is arguably questionable.
The cpudata memory from kzalloc() in amd_pstate_epp_cpu_init() is
not freed in the analogous exit function, so fix that.
Signed-off-by: Peng Ma <andypma@tencent.com> Acked-by: Mario Limonciello <mario.limonciello@amd.com> Reviewed-by: Perry Yuan <Perry.Yuan@amd.com>
[ rjw: Subject and changelog edits ] Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
If a probe function returns -EPROBE_DEFER after creating another device
there is a change of ending up in a probe deferral loop, (see commit fbc35b45f9f6 ("Add documentation on meaning of -EPROBE_DEFER"). In case
of the qcom-pmic-typec driver the tcpm_register_port() function looks up
external resources (USB role switch and inherently via called
typec_register_port() USB-C muxes, switches and retimers).
In order to prevent such probe-defer loops caused by qcom-pmic-typec
driver, use the API added by Johan Hovold and move HPD bridge
registration to the end of the probe function.
The devm_drm_dp_hpd_bridge_add() is called at the end of the probe
function after all TCPM start functions. This is done as a way to
overcome a different problem, the DRM subsystem can not properly cope
with the DRM bridges being destroyed once the bridge is attached. Having
this function call at the end of the probe function prevents possible
DRM bridge device creation followed by destruction in case one of the
TCPM start functions returns an error.
Lockdep reports the below circular locking dependency issue. The
mmap_lock acquisition while holding pci_bus_sem is due to the use of
copy_to_user() from within a pci_walk_bus() callback.
Building the devices array directly into the user buffer is only for
convenience. Instead we can allocate a local buffer for the array,
bounded by the number of devices on the bus/slot, fill the device
information into this local buffer, then copy it into the user buffer
outside the bus walk callback.
======================================================
WARNING: possible circular locking dependency detected
6.9.0-rc5+ #39 Not tainted
------------------------------------------------------
CPU 0/KVM/4113 is trying to acquire lock: ffff99a609ee18a8 (&vdev->vma_lock){+.+.}-{4:4}, at: vfio_pci_mmap_fault+0x35/0x1a0 [vfio_pci_core]
but task is already holding lock: ffff99a243a052a0 (&mm->mmap_lock){++++}-{4:4}, at: vaddr_get_pfns+0x3f/0x170 [vfio_iommu_type1]
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
Secondary Bus Reset (SBR) is equivalent to a device being hot removed and
inserted again. Doing a SBR on a CXL type 3 device is problematic if the
exported device memory is part of system memory that cannot be offlined.
The event is equivalent to violently ripping out that range of memory from
the kernel. While the hardware requires the "Unmask SBR" bit set in the
Port Control Extensions register and the kernel currently does not unmask
it, user can unmask this bit via setpci or similar tool.
The driver does not have a way to detect whether a reset coming from the
PCI subsystem is a Function Level Reset (FLR) or SBR. The only way to
detect is to note if a decoder is marked as enabled in software but the
decoder control register indicates it's not committed.
Add a helper function to find discrepancy between the decoder software
state versus the hardware register state.
Suggested-by: Dan Williams <dan.j.williams@intel.com> Link: https://lore.kernel.org/r/20240502165851.1948523-6-dave.jiang@intel.com Signed-off-by: Dave Jiang <dave.jiang@intel.com> Signed-off-by: Bjorn Helgaas <bhelgaas@google.com> Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com> Reviewed-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Instead of repeatedly calling clk_get_rate for each transfer, lock
the clock rate and cache the value.
A deadlock has been observed while adding tlv320aic32x4 audio codec to
the system. When this clock provider adds its clock, the clk mutex is
locked already, it needs to access i2c, which in return needs the mutex
for clk_get_rate as well.
... and use it to limit the virtual terminals to just N_TTY. They are
kind of special, and in particular, the "con_write()" routine violates
the "writes cannot sleep" rule that some ldiscs rely on.
This avoids the
BUG: sleeping function called from invalid context at kernel/printk/printk.c:2659
when N_GSM has been attached to a virtual console, and gsmld_write()
calls con_write() while holding a spinlock, and con_write() then tries
to get the console lock.
If this quirk was set then driver would treat transfer events with
'Success' completion code as 'Short packet' if there were untransferred
bytes left.
This is so common that turn it into default behavior.
xhci_warn_ratelimited() is no longer used after this, so remove it.
A success event with untransferred bytes left doesn't always mean a
misbehaving controller. If there was an error mid a multi-TRB TD it's
allowed to issue a success event for the last TRB in that TD.
See xhci 1.2 spec 4.9.1 Transfer Descriptors
"Note: If an error is detected while processing a multi-TRB TD, the xHC
shall generate a Transfer Event for the TRB that the error was detected
on with the appropriate error Condition Code, then may advance to the
next TD. If in the process of advancing to the next TD, a Transfer TRB
is encountered with its IOC flag set, then the Condition Code of the
Transfer Event generated for that Transfer TRB should be Success,
because there was no error actually associated with the TRB that
generated the Event. However, an xHC implementation may redundantly
assert the original error Condition Code."
The standard PCIe configuration read-write interface is used to
access the configuration space of the peripheral PCIe devices
of the mips processor after the PCIe link surprise down, it can
generate kernel panic caused by "Data bus error". So it is
necessary to add PCIe link status check for system protection.
When the PCIe link is down or in training, assigning a value
of 0 to the configuration address can prevent read-write behavior
to the configuration space of peripheral PCIe devices, thereby
preventing kernel panic.
Signed-off-by: Songyang Li <leesongyang@outlook.com> Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Sasha Levin <sashal@kernel.org>
Hewlett-Packard HP Pavilion 17 Notebook PC/1972 is an Intel Ivy Bridge
system with a muxless AMD Radeon dGPU. Attempting to use the dGPU fails
with the following sequence:
ACPI Error: Aborting method \AMD3._ON due to previous error (AE_AML_LOOP_TIMEOUT) (20230628/psparse-529)
radeon 0000:01:00.0: not ready 1023ms after resume; waiting
radeon 0000:01:00.0: not ready 2047ms after resume; waiting
radeon 0000:01:00.0: not ready 4095ms after resume; waiting
radeon 0000:01:00.0: not ready 8191ms after resume; waiting
radeon 0000:01:00.0: not ready 16383ms after resume; waiting
radeon 0000:01:00.0: not ready 32767ms after resume; waiting
radeon 0000:01:00.0: not ready 65535ms after resume; giving up
radeon 0000:01:00.0: Unable to change power state from D3cold to D0, device inaccessible
The issue is that the Root Port the dGPU is connected to can't handle the
transition from D3cold to D0 so the dGPU can't properly exit runtime PM.
The existing logic in pci_bridge_d3_possible() checks for systems that are
newer than 2015 to decide that D3 is safe. This would nominally work for
an Ivy Bridge system (which was discontinued in 2015), but this system
appears to have continued to receive BIOS updates until 2017 and so this
existing logic doesn't appropriately capture it.
Add the system to bridge_d3_blacklist to prevent D3cold from being used.
Link: https://lore.kernel.org/r/20240307163709.323-1-mario.limonciello@amd.com Reported-by: Eric Heintzmann <heintzmann.eric@free.fr> Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3229 Signed-off-by: Mario Limonciello <mario.limonciello@amd.com> Signed-off-by: Bjorn Helgaas <bhelgaas@google.com> Tested-by: Eric Heintzmann <heintzmann.eric@free.fr> Signed-off-by: Sasha Levin <sashal@kernel.org>
An overflow can occur in a situation where src.centiseconds
takes the value of 255. This situation is unlikely, but there
is no validation check anywere in the code.
Found by Linux Verification Center (linuxtesting.org) with Svace.
Suggested-by: Jan Kara <jack@suse.cz> Signed-off-by: Roman Smirnov <r.smirnov@omp.ru> Reviewed-by: Sergey Shtylyov <s.shtylyov@omp.ru> Signed-off-by: Jan Kara <jack@suse.cz>
Message-Id: <20240327132755.13945-1-r.smirnov@omp.ru> Signed-off-by: Sasha Levin <sashal@kernel.org>
Newer Qualcomm platforms (sm8450+) successfully handle busy state and
send the Command Completion after sending the Busy state. Older devices
have firmware bug and can not continue after sending the CCI_BUSY state,
but the command that leads to CCI_BUSY is already forbidden by the
NO_PARTNER_PDOS quirk.
Follow other UCSI glue drivers and drop special handling for CCI_BUSY
event. Let the UCSI core properly handle this state.
The Lenovo Yoga Tablet 2 Pro 1380 model is the exception to the rule that
devices which use the Crystal Cove PMIC without using ACPI for battery and
AC power_supply class support use the USB-phy for charger detection.
Unlike the Lenovo Yoga Tablet 2 830 / 1050 models this model has an extra
LC824206XA Micro USB switch which does the charger detection.
Add a DMI quirk to not set the "linux,phy_charger_detect" property on
the 1380 model. This quirk matches on the BIOS version to differentiate
the 1380 model from the 830 and 1050 models which otherwise have
the same DMI strings.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Acked-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com> Link: https://lore.kernel.org/r/20240406140127.17885-1-hdegoede@redhat.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
I have a use case where nr_buffers = 3 and in which each descriptor is composed by 3
segments, resulting in the DMA channel descs_allocated to be 9. Since axi_desc_put()
handles the hw_desc considering the descs_allocated, this scenario would result in a
kernel panic (hw_desc array will be overrun).
To fix this, the proposal is to add a new member to the axi_dma_desc structure,
where we keep the number of allocated hw_descs (axi_desc_alloc()) and use it in
axi_desc_put() to handle the hw_desc array correctly.
Additionally I propose to remove the axi_chan_start_first_queued() call after completing
the transfer, since it was identified that unbalance can occur (started descriptors can
be interrupted and transfer ignored due to DMA channel not being enabled).
The incompatible device in my possession has a sticker that says
"F5U002 Rev 2" and "P80453-B", and lsusb identifies it as
"050d:0002 Belkin Components IEEE-1284 Controller". There is a bug
report from 2007 from Michael Trausch who was seeing the exact same
errors that I saw in 2024 trying to use this cable.
In preparation to adding more quirks, extract quirks to the static
variables and reference them through match->data. Otherwise adding
more quirks will add the table really cumbersome.
In f2fs_remount, SB_INLINECRYPT flag will be clear and re-set.
If create new file or open file during this gap, these files
will not use inlinecrypt. Worse case, it may lead to data
corruption if wrappedkey_v0 is enable.
Thread A: Thread B:
-f2fs_remount -f2fs_file_open or f2fs_new_inode
-default_options
<- clear SB_INLINECRYPT flag
-fscrypt_select_encryption_impl
-parse_options
<- set SB_INLINECRYPT again
Signed-off-by: Yunlei He <heyunlei@oppo.com> Reviewed-by: Chao Yu <chao@kernel.org> Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
"
There is a possible bug in f2fs_truncate_inode_blocks():
if (err < 0 && err != -ENOENT)
goto fail;
...
offset[1] = 0;
offset[0]++;
nofs += err;
If err = -ENOENT then nofs will sum with an error code,
which is strange behaviour. Also if nofs < ENOENT this will
cause an overflow. err will be equal to -ENOENT with the
following call stack:
When setting the guid via configfs it is possible to test if
its value is one of the kernel supported ones by calling
uvc_format_by_guid on it. If the result is NULL, we know the
guid is unsupported and can be ignored.
If an error happens in ftrace, ftrace_kill() will prevent disarming
kprobes. Eventually, the ftrace_ops associated with the kprobes will be
freed, yet the kprobes will still be active, and when triggered, they
will use the freed memory, likely resulting in a page fault and panic.
This behavior can be reproduced quite easily, by creating a kprobe and
then triggering a ftrace_kill(). For simplicity, we can simulate an
ftrace error with a kernel module like [1]:
sudo perf probe --add commit_creds
sudo perf trace -e probe:commit_creds
# In another terminal
make
sudo insmod ftrace_killer.ko # calls ftrace_kill(), simulating bug
# Back to perf terminal
# ctrl-c
sudo perf probe --del commit_creds
After a short period, a page fault and panic would occur as the kprobe
continues to execute and uses the freed ftrace_ops. While ftrace_kill()
is supposed to be used only in extreme circumstances, it is invoked in
FTRACE_WARN_ON() and so there are many places where an unexpected bug
could be triggered, yet the system may continue operating, possibly
without the administrator noticing. If ftrace_kill() does not panic the
system, then we should do everything we can to continue operating,
rather than leave a ticking time bomb.
In the following concurrency we will access the uninitialized rs->lock:
ext4_fill_super
ext4_register_sysfs
// sysfs registered msg_ratelimit_interval_ms
// Other processes modify rs->interval to
// non-zero via msg_ratelimit_interval_ms
ext4_orphan_cleanup
ext4_msg(sb, KERN_INFO, "Errors on filesystem, "
__ext4_msg
___ratelimit(&(EXT4_SB(sb)->s_msg_ratelimit_state)
if (!rs->interval) // do nothing if interval is 0
return 1;
raw_spin_trylock_irqsave(&rs->lock, flags)
raw_spin_trylock(lock)
_raw_spin_trylock
__raw_spin_trylock
spin_acquire(&lock->dep_map, 0, 1, _RET_IP_)
lock_acquire
__lock_acquire
register_lock_class
assign_lock_key
dump_stack();
ratelimit_state_init(&sbi->s_msg_ratelimit_state, 5 * HZ, 10);
raw_spin_lock_init(&rs->lock);
// init rs->lock here
and get the following dump_stack:
=========================================================
INFO: trying to register non-static key.
The code is fine but needs lockdep annotation, or maybe
you didn't initialize this object before use?
turning off the locking correctness validator.
CPU: 12 PID: 753 Comm: mount Tainted: G E 6.7.0-rc6-next-20231222 #504
[...]
Call Trace:
dump_stack_lvl+0xc5/0x170
dump_stack+0x18/0x30
register_lock_class+0x740/0x7c0
__lock_acquire+0x69/0x13a0
lock_acquire+0x120/0x450
_raw_spin_trylock+0x98/0xd0
___ratelimit+0xf6/0x220
__ext4_msg+0x7f/0x160 [ext4]
ext4_orphan_cleanup+0x665/0x740 [ext4]
__ext4_fill_super+0x21ea/0x2b10 [ext4]
ext4_fill_super+0x14d/0x360 [ext4]
[...]
=========================================================
Normally interval is 0 until s_msg_ratelimit_state is initialized, so
___ratelimit() does nothing. But registering sysfs precedes initializing
rs->lock, so it is possible to change rs->interval to a non-zero value
via the msg_ratelimit_interval_ms interface of sysfs while rs->lock is
uninitialized, and then a call to ext4_msg triggers the problem by
accessing an uninitialized rs->lock. Therefore register sysfs after all
initializations are complete to avoid such problems.
ext4_xattr_set_entry() creates new EA inodes while holding buffer lock
on the external xattr block. This is problematic as it nests all the
allocation locking (which acquires locks on other buffers) under the
buffer lock. This can even deadlock when the filesystem is corrupted and
e.g. quota file is setup to contain xattr block as data block. Move the
allocation of EA inode out of ext4_xattr_set_entry() into the callers.
Reported-by: syzbot+a43d4f48b8397d0e41a9@syzkaller.appspotmail.com Signed-off-by: Jan Kara <jack@suse.cz> Link: https://lore.kernel.org/r/20240321162657.27420-2-jack@suse.cz Signed-off-by: Theodore Ts'o <tytso@mit.edu> Signed-off-by: Sasha Levin <sashal@kernel.org>
If devm_add_action() returns -ENOMEM, then MSIs are allocated but not
not freed on teardown. Use devm_add_action_or_reset() instead to keep
the static analyser happy.
Found by Linux Verification Center (linuxtesting.org) with SVACE.
With -Wextra clang warns about pointer arithmetic using a null pointer.
When building with CONFIG_PCI=n, that triggers a warning in the IO
accessors, eg:
In file included from linux/arch/powerpc/include/asm/io.h:672:
linux/arch/powerpc/include/asm/io-defs.h:23:1: warning: performing pointer arithmetic on a null pointer has undefined behavior [-Wnull-pointer-arithmetic]
23 | DEF_PCI_AC_RET(inb, u8, (unsigned long port), (port), pio, port)
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
...
linux/arch/powerpc/include/asm/io.h:591:53: note: expanded from macro '__do_inb'
591 | #define __do_inb(port) readb((PCI_IO_ADDR)_IO_BASE + port);
| ~~~~~~~~~~~~~~~~~~~~~ ^
That is because when CONFIG_PCI=n, _IO_BASE is defined as 0.
Although _IO_BASE is defined as plain 0, the cast (PCI_IO_ADDR) converts
it to void * before the addition with port happens.
Instead the addition can be done first, and then the cast. The resulting
value will be the same, but avoids the warning, and also avoids void
pointer arithmetic which is apparently non-standard.
Modify the code so it can be compiled tested in configurations that do
not have ACPI enabled.
It fixes the following errors:
drivers/media/pci/intel/ipu-bridge.c:103:30: error: implicit declaration of function ‘acpi_device_handle’; did you mean ‘acpi_fwnode_handle’? [-Werror=implicit-function-declaration]
drivers/media/pci/intel/ipu-bridge.c:103:30: warning: initialization of ‘acpi_handle’ {aka ‘void *’} from ‘int’ makes pointer from integer without a cast [-Wint-conversion]
drivers/media/pci/intel/ipu-bridge.c:110:17: error: implicit declaration of function ‘for_each_acpi_dev_match’ [-Werror=implicit-function-declaration]
drivers/media/pci/intel/ipu-bridge.c:110:74: error: expected ‘;’ before ‘for_each_acpi_consumer_dev’
drivers/media/pci/intel/ipu-bridge.c:104:29: warning: unused variable ‘consumer’ [-Wunused-variable]
drivers/media/pci/intel/ipu-bridge.c:103:21: warning: unused variable ‘handle’ [-Wunused-variable]
drivers/media/pci/intel/ipu-bridge.c:166:38: error: invalid use of undefined type ‘struct acpi_device’
drivers/media/pci/intel/ipu-bridge.c:185:43: error: invalid use of undefined type ‘struct acpi_device’
drivers/media/pci/intel/ipu-bridge.c:191:30: error: invalid use of undefined type ‘struct acpi_device’
drivers/media/pci/intel/ipu-bridge.c:196:30: error: invalid use of undefined type ‘struct acpi_device’
drivers/media/pci/intel/ipu-bridge.c:202:30: error: invalid use of undefined type ‘struct acpi_device’
drivers/media/pci/intel/ipu-bridge.c:223:31: error: invalid use of undefined type ‘struct acpi_device’
drivers/media/pci/intel/ipu-bridge.c:236:18: error: implicit declaration of function ‘acpi_get_physical_device_location’ [-Werror=implicit-function-declaration]
drivers/media/pci/intel/ipu-bridge.c:236:56: error: invalid use of undefined type ‘struct acpi_device’
drivers/media/pci/intel/ipu-bridge.c:238:31: error: invalid use of undefined type ‘struct acpi_device’
drivers/media/pci/intel/ipu-bridge.c:256:31: error: invalid use of undefined type ‘struct acpi_device’
drivers/media/pci/intel/ipu-bridge.c:275:31: error: invalid use of undefined type ‘struct acpi_device’
drivers/media/pci/intel/ipu-bridge.c:280:30: error: invalid use of undefined type ‘struct acpi_device’
drivers/media/pci/intel/ipu-bridge.c:469:26: error: implicit declaration of function ‘acpi_device_hid’; did you mean ‘dmi_device_id’? [-Werror=implicit-function-declaration]
drivers/media/pci/intel/ipu-bridge.c:468:74: warning: format ‘%s’ expects argument of type ‘char *’, but argument 4 has type ‘int’ [-Wformat=]
drivers/media/pci/intel/ipu-bridge.c:637:58: error: expected ‘;’ before ‘{’ token
drivers/media/pci/intel/ipu-bridge.c:696:1: warning: label ‘err_put_adev’ defined but not used [-Wunused-label]
drivers/media/pci/intel/ipu-bridge.c:693:1: warning: label ‘err_put_ivsc’ defined but not used [-Wunused-label]
drivers/media/pci/intel/ipu-bridge.c:691:1: warning: label ‘err_free_swnodes’ defined but not used [-Wunused-label]
drivers/media/pci/intel/ipu-bridge.c:632:40: warning: unused variable ‘primary’ [-Wunused-variable]
drivers/media/pci/intel/ipu-bridge.c:632:31: warning: unused variable ‘fwnode’ [-Wunused-variable]
drivers/media/pci/intel/ipu-bridge.c:733:73: error: expected ‘;’ before ‘{’ token
drivers/media/pci/intel/ipu-bridge.c:725:24: warning: unused variable ‘csi_dev’ [-Wunused-variable]
drivers/media/pci/intel/ipu-bridge.c:724:43: warning: unused variable ‘adev’ [-Wunused-variable]
drivers/media/pci/intel/ipu-bridge.c:599:12: warning: ‘ipu_bridge_instantiate_ivsc’ defined but not used [-Wunused-function]
drivers/media/pci/intel/ipu-bridge.c:444:13: warning: ‘ipu_bridge_create_connection_swnodes’ defined but not used [-Wunused-function]
drivers/media/pci/intel/ipu-bridge.c:297:13: warning: ‘ipu_bridge_create_fwnode_properties’ defined but not used [-Wunused-function]
drivers/media/pci/intel/ipu-bridge.c:155:12: warning: ‘ipu_bridge_check_ivsc_dev’ defined but not used [-Wunused-function]
plpar_hcall(), plpar_hcall9(), and related functions expect callers to
provide valid result buffers of certain minimum size. Currently this
is communicated only through comments in the code and the compiler has
no idea.
For example, if I write a bug like this:
long retbuf[PLPAR_HCALL_BUFSIZE]; // should be PLPAR_HCALL9_BUFSIZE
plpar_hcall9(H_ALLOCATE_VAS_WINDOW, retbuf, ...);
This compiles with no diagnostics emitted, but likely results in stack
corruption at runtime when plpar_hcall9() stores results past the end
of the array. (To be clear this is a contrived example and I have not
found a real instance yet.)
To make this class of error less likely, we can use explicitly-sized
array parameters instead of pointers in the declarations for the hcall
APIs. When compiled with -Warray-bounds[1], the code above now
provokes a diagnostic like this:
error: array argument is too small;
is of size 32, callee requires at least 72 [-Werror,-Warray-bounds]
60 | plpar_hcall9(H_ALLOCATE_VAS_WINDOW, retbuf,
| ^ ~~~~~~
[1] Enabled for LLVM builds but not GCC for now. See commit 0da6e5fd6c37 ("gcc: disable '-Warray-bounds' for gcc-13 too") and
related changes.
The Lenovo Yoga Tablet 2 Pro 1380F/L is a x86 ACPI tablet which ships with
Android x86 as factory OS. Its DSDT contains a bunch of I2C devices which
are not actually there, causing various resource conflicts. Enumeration of
these is skipped through the acpi_quirk_skip_i2c_client_enumeration().
Add support for manually instantiating the I2C + other devices which are
actually present on this tablet by adding the necessary device info to
the x86-android-tablets module.
Not all subsystems support a device getting removed while there are
still consumers of the device with a reference to the device.
One example of this is the regulator subsystem. If a regulator gets
unregistered while there are still drivers holding a reference
a WARN() at drivers/regulator/core.c:5829 triggers, e.g.:
On the Lenovo Yoga Tablet 2 series the bq24190 charger chip also provides
a 5V boost converter output for powering USB devices connected to the micro
USB port, the bq24190-charger driver exports this as a Vbus regulator.
On the 830 (8") and 1050 ("10") models this regulator is controlled by
a platform_device and x86_android_tablet_remove() removes platform_device-s
before i2c_clients so the consumer gets removed first.
But on the 1380 (13") model there is a lc824206xa micro-USB switch
connected over I2C and the extcon driver for that controls the regulator.
The bq24190 i2c-client *must* be registered first, because that creates
the regulator with the lc824206xa listed as its consumer. If the regulator
has not been registered yet the lc824206xa driver will end up getting
a dummy regulator.
Since in this case both the regulator provider and consumer are I2C
devices, the only way to ensure that the consumer is unregistered first
is to unregister the I2C devices in reverse order of in which they were
created.
For consistency and to avoid similar problems in the future change
x86_android_tablet_remove() to unregister all device types in reverse
order.
There is a race condition in which a rendering job might take just long
enough to trigger the drm sched job timeout handler but also still
complete before the hard reset is done by the timeout handler.
This runs into race conditions not expected by the timeout handler.
In some very specific cases it currently may result in a refcount
imbalance on lima_pm_idle, with a stack dump such as:
We can prevent that race condition entirely by masking the irqs at the
beginning of the timeout handler, at which point we give up on waiting
for that job entirely.
The irqs will be enabled again at the next hard reset which is already
done as a recovery by the timeout handler.
In commit 53cb55b20208 ("drm/lima: handle spurious timeouts due to high
irq latency") a check was added to detect an unexpectedly high interrupt
latency timeout.
With further investigation it was noted that on Mali-450 the pp bcast
irq may also be a trigger of race conditions against the timeout
handler, so add it to this check too.
This is needed because we want to reset those devices in device-agnostic
code such as lima_sched.
In particular, masking irqs will be useful before a hard reset to
prevent race conditions.
Since commit 3a5a6d0c2b03("cpuset: don't nest cgroup_mutex inside
get_online_cpus()"), cpuset hotplug was done asynchronously via a work
function. This is to avoid recursive locking of cgroup_mutex.
Since then, the cgroup locking scheme has changed quite a bit. A
cpuset_mutex was introduced to protect cpuset specific operations.
The cpuset_mutex is then replaced by a cpuset_rwsem. With commit d74b27d63a8b ("cgroup/cpuset: Change cpuset_rwsem and hotplug lock
order"), cpu_hotplug_lock is acquired before cpuset_rwsem. Later on,
cpuset_rwsem is reverted back to cpuset_mutex. All these locking changes
allow the hotplug code to call into cpuset core directly.
The following commits were also merged due to the asynchronous nature
of cpuset hotplug processing.
- commit b22afcdf04c9 ("cpu/hotplug: Cure the cpusets trainwreck")
- commit 50e76632339d ("sched/cpuset/pm: Fix cpuset vs. suspend-resume
bugs")
- commit 28b89b9e6f7b ("cpuset: handle race between CPU hotplug and
cpuset_hotplug_work")
Clean up all these bandages by making cpuset hotplug
processing synchronous again with the exception that the call to
cgroup_transfer_tasks() to transfer tasks out of an empty cgroup v1
cpuset, if necessary, will still be done via a work function due to the
existing cgroup_mutex -> cpu_hotplug_lock dependency. It is possible
to reverse that dependency, but that will require updating a number of
different cgroup controllers. This special hotplug code path should be
rarely taken anyway.
As all the cpuset states will be updated by the end of the hotplug
operation, we can revert most the above commits except commit 50e76632339d ("sched/cpuset/pm: Fix cpuset vs. suspend-resume bugs")
which is partially reverted. Also removing some cpus_read_lock trylock
attempts in the cpuset partition code as they are no longer necessary
since the cpu_hotplug_lock is now held for the whole duration of the
cpuset hotplug code path.
The Z830 has some buttons that will only work properly as "quickstart"
buttons. To enable them in that mode, a value between 1 and 7 must be
used for HCI_HOTKEY_EVENT. Windows uses 0x5 on this laptop so use that for
maximum predictability and compatibility.
As there is not yet a known way of auto detection, this patch uses a DMI
quirk table. A module parameter is exposed to allow setting this on other
models for testing.
Signed-off-by: Arvid Norlander <lkml@vorpal.se> Tested-by: Hans de Goede <hdegoede@redhat.com> Reviewed-by: Hans de Goede <hdegoede@redhat.com> Link: https://lore.kernel.org/r/20240131111641.4418-3-W_Armin@gmx.de Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
[Why]
Cursor update can be pre-empted by a request for setting target flip
submission.
This causes an issue where we're in the middle of the exit sequence
trying to log to DM, but the pre-emption starts another DMCUB
command submission that requires being out of idle.
The DC lock aqusition can fail, and depending on the DM/OS interface
it's possible that the function inserted into this thread must not fail.
This means that lock aqusition must be skipped and exit *must* occur.
[How]
Modify when we consider idle as active. Consider it exited only once
the exit has fully finished.
Consider it as entered prior to actual notification.
Since we're on the same core/thread the cached values are coherent
and we'll see that we still need to exit. Once the cursor update resumes
it'll continue doing the double exit but this won't cause a functional
issue, just a (potential) redundant operation.
Reviewed-by: Duncan Ma <duncan.ma@amd.com> Acked-by: Wayne Lin <wayne.lin@amd.com> Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
[WHY]
PSP can access DCN registers during command submission and we need
to ensure that DCN is not in PG before doing so.
[HOW]
Add a callback to DM to lock and notify DC for idle optimization exit.
It can't be DC directly because of a potential race condition with the
link protection thread and the rest of DM operation.
Cc: Mario Limonciello <mario.limonciello@amd.com> Cc: Alex Deucher <alexander.deucher@amd.com> Reviewed-by: Charlene Liu <charlene.liu@amd.com> Acked-by: Alex Hung <alex.hung@amd.com> Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
This fixes some CHECKs reported by the checkpatch script.
Issues reported in ath3k.c:
-------
ath3k.c
-------
CHECK: Please don't use multiple blank lines
+
+
CHECK: Blank lines aren't necessary after an open brace '{'
+static const struct usb_device_id ath3k_blist_tbl[] = {
+
CHECK: Alignment should match open parenthesis
+static int ath3k_load_firmware(struct usb_device *udev,
+ const struct firmware *firmware)
CHECK: Alignment should match open parenthesis
+ err = usb_bulk_msg(udev, pipe, send_buf, size,
+ &len, 3000);
CHECK: Unnecessary parentheses around 'len != size'
+ if (err || (len != size)) {
CHECK: Alignment should match open parenthesis
+static int ath3k_get_version(struct usb_device *udev,
+ struct ath3k_version *version)
CHECK: Alignment should match open parenthesis
+static int ath3k_load_fwfile(struct usb_device *udev,
+ const struct firmware *firmware)
CHECK: Alignment should match open parenthesis
+ err = usb_bulk_msg(udev, pipe, send_buf, size,
+ &len, 3000);
CHECK: Unnecessary parentheses around 'len != size'
+ if (err || (len != size)) {
CHECK: Blank lines aren't necessary after an open brace '{'
+ switch (fw_version.ref_clock) {
+
CHECK: Alignment should match open parenthesis
+ snprintf(filename, ATH3K_NAME_LEN, "ar3k/ramps_0x%08x_%d%s",
+ le32_to_cpu(fw_version.rom_version), clk_value, ".dfu");
CHECK: Alignment should match open parenthesis
+static int ath3k_probe(struct usb_interface *intf,
+ const struct usb_device_id *id)
CHECK: Alignment should match open parenthesis
+ BT_ERR("Firmware file \"%s\" not found",
+ ATH3K_FIRMWARE);
CHECK: Alignment should match open parenthesis
+ BT_ERR("Firmware file \"%s\" request failed (err=%d)",
+ ATH3K_FIRMWARE, ret);
Signed-off-by: Uri Arev <me@wantyapps.xyz> Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Validate that `first` flag is set only for the first
descriptor in multi-buffer packets.
In case of an invalid descriptor, a reset will occur.
A new reset reason for RX data corruption has been added.
Signed-off-by: Shahar Itzko <itzko@amazon.com> Signed-off-by: David Arinzon <darinzon@amazon.com> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://lore.kernel.org/r/20240512134637.25299-4-darinzon@amazon.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
On slow machines the SND timestamp sometimes doesn't arrive before
we quit. The test only waits as long as the packet delay, so it's
easy for a race condition to happen.
Double the wait but do a bit of polling, once the SND timestamp
arrives there's no point to wait any longer.
This fixes the "TXTIME abs" failures on debug kernels, like:
Case ICMPv4 - TXTIME abs returned '', expected 'OK'
[Changes from V1:
- The warning to disable is -Wmaybe-uninitialized, not -Wuninitialized.
- This warning is only supported in GCC.]
The BPF selftest verifier_global_subprogs.c contains code that
purposedly performs out of bounds access to memory, to check whether
the kernel verifier is able to catch them. For example:
__noinline int global_unsupp(const int *mem)
{
if (!mem)
return 0;
return mem[100]; /* BOOM */
}
With -O1 and higher and no inlining, GCC notices this fact and emits a
"maybe uninitialized" warning. This is by design. Note that the
emission of these warnings is highly dependent on the precise
optimizations that are performed.
This patch adds a compiler pragma to verifier_global_subprogs.c to
ignore these warnings.
Tested in bpf-next master.
No regressions.
Signed-off-by: Jose E. Marchesi <jose.marchesi@oracle.com> Cc: david.faust@oracle.com Cc: cupertino.miranda@oracle.com Cc: Yonghong Song <yonghong.song@linux.dev> Cc: Eduard Zingerman <eddyz87@gmail.com> Acked-by: Yonghong Song <yonghong.song@linux.dev> Link: https://lore.kernel.org/r/20240507184756.1772-1-jose.marchesi@oracle.com Signed-off-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Lenovo Slim 7 16ARH7 is a machine with switchable graphics between AMD
and Nvidia, and the backlight can't be adjusted properly unless
acpi_backlight=native is passed. Although nvidia-wmi-backlight is
present and loaded, this doesn't work as expected at all.
For making it working as default, add the corresponding quirk entry
with a DMI matching "LENOVO" "82UX".
Link: https://bugzilla.suse.com/show_bug.cgi?id=1217750 Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
This device sometimes doesn't send touch release signals when moving
from >=4 fingers to <4 fingers. Using MT_QUIRK_NOT_SEEN_MEANS_UP instead
of MT_QUIRK_ALWAYS_VALID makes sure that no touches become stuck.
MT_QUIRK_FORCE_MULTI_INPUT is not necessary for this device, but does no
harm.
During chip recovery (e.g. chip reset), there is a possible situation that
kernel worker reset_work is holding the lock and waiting for kernel thread
stat_worker to be parked, while stat_worker is waiting for the release of
the same lock.
It causes a deadlock resulting in the dumping of hung tasks messages and
possible rebooting of the device.
This patch prevents the execution of stat_worker during the chip recovery.
Signed-off-by: Leon Yen <leon.yen@mediatek.com> Signed-off-by: Ming Yen Hsieh <MingYen.Hsieh@mediatek.com> Signed-off-by: Felix Fietkau <nbd@nbd.name> Signed-off-by: Sasha Levin <sashal@kernel.org>
Currently, when using WCN7850 or QCN9274 as AP, ath12k always performs down
grade phy mode operation regardless of whether the firmware supports EHT
capability or not and then vdev will start in HE mode. When stations that
support EHT capability try to connect to the AP, the AP will set phy mode
to EHT after receiving the association request packet, and then send
WMI_PEER_ASSOC_CMDID command to firmware, AP’s firmware will crash.
This is because when the ath12k_mac_copy_sband_iftype_data() function
handles EHT capability, it does not copy the EHT capability into the
iftype[band][type] array according to the interface type. So, interface
type should not be used as an index to get eht_cap in
ath12k_mac_check_down_grade_phy_mode() function.
To address this issue, use types_mask to select the eht_cap in
ath12k_mac_check_down_grade_phy_mode() function.
This patch affects QCN9274 and WCN7850 because they have the same issue.
Hostapd log:
wlo1: STA 02:03:7f:37:12:34 IEEE 802.11: Could not set STA to kernel driver
Kernel log:
[270894.816076] ath12k_pci 0000:03:00.0: failed to send WMI_PEER_SET_PARAM cmd
[270894.816111] ath12k_pci 0000:03:00.0: failed to setup peer SMPS for vdev 0: -108
[270894.816122] ath12k_pci 0000:03:00.0: Failed to associate station: 02:03:7f:37:12:34
[270894.843389] ieee80211 phy5: Hardware restart was requested
[270894.843517] ath12k_pci 0000:03:00.0: failed to lookup peer 02:03:7f:37:12:34 on vdev 0
[270894.843616] ath12k_pci 0000:03:00.0: failed to send WMI_PEER_DELETE cmd
[270894.843650] ath12k_pci 0000:03:00.0: failed to delete peer vdev_id 0 addr 02:03:7f:37:12:34 ret -108
[270894.843663] ath12k_pci 0000:03:00.0: Failed to delete peer: 02:03:7f:37:12:34 for VDEV: 0
Signed-off-by: Lingbo Kong <quic_lingbok@quicinc.com> Acked-by: Jeff Johnson <quic_jjohnson@quicinc.com> Acked-by: Jeff Johnson <quic_jjohnson@quicinc.com> Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com> Link: https://msgid.link/20240425083837.5340-1-quic_lingbok@quicinc.com Signed-off-by: Sasha Levin <sashal@kernel.org>
BUG: KCSAN: data-race in net_rx_action / netpoll_send_skb
write (marked) to 0xffff8881164168b0 of 4 bytes by interrupt on cpu 10:
net_rx_action (./include/linux/netpoll.h:90 net/core/dev.c:6712 net/core/dev.c:6822)
<snip>
read to 0xffff8881164168b0 of 4 bytes by task 1 on cpu 2:
netpoll_send_skb (net/core/netpoll.c:319 net/core/netpoll.c:345 net/core/netpoll.c:393)
netpoll_send_udp (net/core/netpoll.c:?)
<snip>
value changed: 0x0000000a -> 0xffffffff
This happens because netpoll_owner_active() needs to check if the
current CPU is the owner of the lock, touching napi->poll_owner
non atomically. The ->poll_owner field contains the current CPU holding
the lock.
Use an atomic read to check if the poll owner is the current CPU.
Like various other Asus Vivobook and Expertbook laptops, the Asus
Vivobook Pro N6506MV has a DSDT table that describes IRQ 1 as ActiveLow
while the kernel is overriding it to Edge_High. This prevents the internal
keyboard from working. This patch prevents this issue by adding this laptop
to the override table that prevents the kernel from overriding this IRQ
Link: https://bugzilla.kernel.org/show_bug.cgi?id=218745 Tested-by: Gianni <gianni.casagrande.mail@gmail.com> Signed-off-by: Tamim Khan <tamim@fusetak.com> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
The necessity of asserting the reset on removal was previously
questioned, as DSA's own cleanup methods should suffice to prevent
traffic leakage[1].
When a driver has subdrivers controlled by devres, they will be
unregistered after the main driver's .remove is executed. If it asserts
a reset, the subdrivers will be unable to communicate with the hardware
during their cleanup. For LEDs, this means that they will fail to turn
off, resulting in a timeout error.
Signed-off-by: Luiz Angelo Daros de Luca <luizluca@gmail.com> Reviewed-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
This switch family supports four LEDs for each of its six ports. Each
LED group is composed of one of these four LEDs from all six ports. LED
groups can be configured to display hardware information, such as link
activity, or manually controlled through a bitmap in registers
RTL8366RB_LED_0_1_CTRL_REG and RTL8366RB_LED_2_3_CTRL_REG.
After a reset, the default LED group configuration for groups 0 to 3
indicates, respectively, link activity, link at 1000M, 100M, and 10M, or
RTL8366RB_LED_CTRL_REG as 0x5432. These configurations are commonly used
for LED indications. However, the driver was replacing that
configuration to use manually controlled LEDs (RTL8366RB_LED_FORCE)
without providing a way for the OS to control them. The default
configuration is deemed more useful than fixed, uncontrollable turned-on
LEDs.
The driver was enabling/disabling LEDs during port_enable/disable.
However, these events occur when the port is administratively controlled
(up or down) and are not related to link presence. Additionally, when a
port N was disabled, the driver was turning off all LEDs for group N,
not only the corresponding LED for port N in any of those 4 groups. In
such cases, if port 0 was brought down, the LEDs for all ports in LED
group 0 would be turned off. As another side effect, the driver was
wrongly warning that port 5 didn't have an LED ("no LED for port 5").
Since showing the administrative state of ports is not an orthodox way
to use LEDs, it was not worth it to fix it and all this code was
dropped.
The code to disable LEDs was simplified only changing each LED group to
the RTL8366RB_LED_OFF state. Registers RTL8366RB_LED_0_1_CTRL_REG and
RTL8366RB_LED_2_3_CTRL_REG are only used when the corresponding LED
group is configured with RTL8366RB_LED_FORCE and they don't need to be
cleaned. The code still references an LED controlled by
RTL8366RB_INTERRUPT_CONTROL_REG, but as of now, no test device has
actually used it. Also, some magic numbers were replaced by macros.
Signed-off-by: Luiz Angelo Daros de Luca <luizluca@gmail.com> Reviewed-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
There is a 'malloc' call, which can be unsuccessful.
This patch will add the malloc failure checking
to avoid possible null dereference and give more information
about test fail reasons.
Signed-off-by: Kunwu Chan <chentao@kylinos.cn> Reviewed-by: Muhammad Usama Anjum <usama.anjum@collabora.com> Link: https://lore.kernel.org/r/20240423082102.2018886-1-chentao@kylinos.cn Signed-off-by: Will Deacon <will@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
This happens when TC does a mirred egress redirect from the root qdisc of
device A to the root qdisc of device B. As long as these two locks aren't
protecting the same qdisc, they can be acquired in chain: add a per-qdisc
lockdep key to silence false warnings.
This dynamic key should safely replace the static key we have in sch_htb:
it was added to allow enqueueing to the device "direct qdisc" while still
holding the qdisc root lock.
v2: don't use static keys anymore in HTB direct qdiscs (thanks Eric Dumazet)
CC: Maxim Mikityanskiy <maxim@isovalent.com> CC: Xiumei Mu <xmu@redhat.com> Reported-by: Christoph Paasch <cpaasch@apple.com> Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/451 Signed-off-by: Davide Caratti <dcaratti@redhat.com> Link: https://lore.kernel.org/r/7dc06d6158f72053cf877a82e2a7a5bd23692faa.1713448007.git.dcaratti@redhat.com Signed-off-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
This copper module comes with broken TX_FAULT indicator which must be
ignored for it to work.
Co-authored-by: Josef Schlehofer <pepe.schlehofer@gmail.com> Signed-off-by: Daniel Golle <daniel@makrotopia.org>
[ rebased on top of net-next ] Signed-off-by: Marek Behún <kabel@kernel.org> Link: https://lore.kernel.org/r/20240423090025.29231-1-kabel@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Enhance the quirk for Fibrestore 2.5G copper SFP module. The original
commit e27aca3760c0 ("net: sfp: add quirk for FS's 2.5G copper SFP")
introducing the quirk says that the PHY is inaccessible, but that is
not true.
The module uses Rollball protocol to talk to the PHY, and needs a 4
second wait before probing it, same as FS 10G module.
The PHY inside the module is Realtek RTL8221B-VB-CG PHY. The realtek
driver recently gained support to set it up via clause 45 accesses.
Currently during resume, QMI target memory is not properly handled, resulting
in kernel crash in case DMA remap is not supported:
BUG: Bad page state in process kworker/u16:54 pfn:36e80
page: refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x36e80
page dumped because: nonzero _refcount
Call Trace:
bad_page
free_page_is_bad_report
__free_pages_ok
__free_pages
dma_direct_free
dma_free_attrs
ath12k_qmi_free_target_mem_chunk
ath12k_qmi_msg_mem_request_cb
The reason is:
Once ath12k module is loaded, firmware sends memory request to host. In case
DMA remap not supported, ath12k refuses the first request due to failure in
allocating with large segment size:
ath12k_pci 0000:04:00.0: qmi firmware request memory request
ath12k_pci 0000:04:00.0: qmi mem seg type 1 size 7077888
ath12k_pci 0000:04:00.0: qmi mem seg type 4 size 8454144
ath12k_pci 0000:04:00.0: qmi dma allocation failed (7077888 B type 1), will try later with small size
ath12k_pci 0000:04:00.0: qmi delays mem_request 2
ath12k_pci 0000:04:00.0: qmi firmware request memory request
Later firmware comes back with more but small segments and allocation
succeeds:
ath12k_pci 0000:04:00.0: qmi mem seg type 1 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 1 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 1 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 1 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 1 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 1 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 1 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 1 size 262144
ath12k_pci 0000:04:00.0: qmi mem seg type 1 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 1 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 1 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 1 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 1 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 4 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 4 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 4 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 4 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 4 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 4 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 4 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 4 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 4 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 4 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 4 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 4 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 4 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 4 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 4 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 4 size 524288
ath12k_pci 0000:04:00.0: qmi mem seg type 4 size 65536
ath12k_pci 0000:04:00.0: qmi mem seg type 1 size 524288
Now ath12k is working. If suspend is triggered, firmware will be reloaded
during resume. As same as before, firmware requests two large segments at
first. In ath12k_qmi_msg_mem_request_cb() segment count and size are
assigned:
Then allocation failed like before and ath12k_qmi_free_target_mem_chunk()
is called to free all allocated segments. Note the first segment is skipped
because its v.addr is cleared due to allocation failure:
chunk->v.addr = dma_alloc_coherent()
Also note that this leaks that segment because it has not been freed.
While freeing the second segment, a size of 8454144 is passed to
dma_free_coherent(). However remember that this segment is allocated at
the first time firmware is loaded, before suspend. So its real size is
524288, much smaller than 8454144. As a result kernel found we are freeing
some memory which is in use and thus crashed.
So one possible fix would be to free those segments during suspend. This
works because with them freed, ath12k_qmi_free_target_mem_chunk() does
nothing: all segment addresses are NULL so dma_free_coherent() is not called.
But note that ath11k has similar logic but never hits this issue. Reviewing
code there shows the luck comes from QMI memory reuse logic. So the decision
is to port it to ath12k. Like in ath11k, the crash is avoided by adding
prev_size to target_mem_chunk structure and caching real segment size in it,
then prev_size instead of current size is passed to dma_free_coherent(),
no unexpected memory is freed now.
trace_drop_common() is called with preemption disabled, and it acquires
a spin_lock. This is problematic for RT kernels because spin_locks are
sleeping locks in this configuration, which causes the following splat:
BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:48
in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 449, name: rcuc/47
preempt_count: 1, expected: 0
RCU nest depth: 2, expected: 2
5 locks held by rcuc/47/449:
#0: ff1100086ec30a60 ((softirq_ctrl.lock)){+.+.}-{2:2}, at: __local_bh_disable_ip+0x105/0x210
#1: ffffffffb394a280 (rcu_read_lock){....}-{1:2}, at: rt_spin_lock+0xbf/0x130
#2: ffffffffb394a280 (rcu_read_lock){....}-{1:2}, at: __local_bh_disable_ip+0x11c/0x210
#3: ffffffffb394a160 (rcu_callback){....}-{0:0}, at: rcu_do_batch+0x360/0xc70
#4: ff1100086ee07520 (&data->lock){+.+.}-{2:2}, at: trace_drop_common.constprop.0+0xb5/0x290
irq event stamp: 139909
hardirqs last enabled at (139908): [<ffffffffb1df2b33>] _raw_spin_unlock_irqrestore+0x63/0x80
hardirqs last disabled at (139909): [<ffffffffb19bd03d>] trace_drop_common.constprop.0+0x26d/0x290
softirqs last enabled at (139892): [<ffffffffb07a1083>] __local_bh_enable_ip+0x103/0x170
softirqs last disabled at (139898): [<ffffffffb0909b33>] rcu_cpu_kthread+0x93/0x1f0
Preemption disabled at:
[<ffffffffb1de786b>] rt_mutex_slowunlock+0xab/0x2e0
CPU: 47 PID: 449 Comm: rcuc/47 Not tainted 6.9.0-rc2-rt1+ #7
Hardware name: Dell Inc. PowerEdge R650/0Y2G81, BIOS 1.6.5 04/15/2022
Call Trace:
<TASK>
dump_stack_lvl+0x8c/0xd0
dump_stack+0x14/0x20
__might_resched+0x21e/0x2f0
rt_spin_lock+0x5e/0x130
? trace_drop_common.constprop.0+0xb5/0x290
? skb_queue_purge_reason.part.0+0x1bf/0x230
trace_drop_common.constprop.0+0xb5/0x290
? preempt_count_sub+0x1c/0xd0
? _raw_spin_unlock_irqrestore+0x4a/0x80
? __pfx_trace_drop_common.constprop.0+0x10/0x10
? rt_mutex_slowunlock+0x26a/0x2e0
? skb_queue_purge_reason.part.0+0x1bf/0x230
? __pfx_rt_mutex_slowunlock+0x10/0x10
? skb_queue_purge_reason.part.0+0x1bf/0x230
trace_kfree_skb_hit+0x15/0x20
trace_kfree_skb+0xe9/0x150
kfree_skb_reason+0x7b/0x110
skb_queue_purge_reason.part.0+0x1bf/0x230
? __pfx_skb_queue_purge_reason.part.0+0x10/0x10
? mark_lock.part.0+0x8a/0x520
...
trace_drop_common() also disables interrupts, but this is a minor issue
because we could easily replace it with a local_lock.
Replace the spin_lock with raw_spin_lock to avoid sleeping in atomic
context.
Signed-off-by: Wander Lairson Costa <wander@redhat.com> Reported-by: Hu Chunyu <chuhu@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
The x86 Android tablets on which quirks to skip looking for a matching
UartSerialBus resource and instead unconditionally create a serial bus
device (serdev) are necessary there are 2 sorts of serialports:
ACPI enumerated highspeed designware UARTs, these are the ones which
typcially need to be skipped since they need a serdev for the attached
BT HCI.
A PNP enumerated UART which is part of the PCU. So far the existing
quirks have ignored this. But on the Lenovo Yoga Tablet 2 Pro 1380
models this is used for a custom fastcharging protocol. There is
a Micro USB switch which can switch the USB data lines to this uart
and then a 600 baud protocol is used to configure the charger for
a voltage higher then 5V.
Add a new ACPI_QUIRK_PNP_UART1_SKIP quirk type and set this for
the existing entry for the Lenovo Yoga Tablet 2 830 / 1050 models.
Note this will lead to unnecessarily also creating a serdev for
the PCU UART on the 830 / 1050 which don't need this, but the UART
is not used otherwise there so that is not a problem.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Kernel timekeeping is designed to keep the change in cycles (since the last
timer interrupt) below max_cycles, which prevents multiplication overflow
when converting cycles to nanoseconds. However, if timer interrupts stop,
the clocksource_cyc2ns() calculation will eventually overflow.
Add protection against that. Simplify by folding together
clocksource_delta() and clocksource_cyc2ns() into cycles_to_nsec_safe().
Check against max_cycles, falling back to a slower higher precision
calculation.
memcpy: detected field-spanning write (size 28) of single field "dst" at include/linux/sockptr.h:49 (size 16)
Refactor the code to be more explicit.
Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Kees Cook <keescook@chromium.org> Cc: Willem de Bruijn <willemdebruijn.kernel@gmail.com> Reviewed-by: Kees Cook <keescook@chromium.org> Reviewed-by: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
gcc-9 and some other older versions produce a false-positive warning
for zeroing two fields
In file included from include/linux/string.h:369,
from drivers/net/wireless/ath/ath9k/main.c:18:
In function 'fortify_memset_chk',
inlined from 'ath9k_ps_wakeup' at drivers/net/wireless/ath/ath9k/main.c:140:3:
include/linux/fortify-string.h:462:25: error: call to '__write_overflow_field' declared with attribute warning: detected write beyond size of field (1st parameter); maybe use struct_group()? [-Werror=attribute-warning]
462 | __write_overflow_field(p_size_field, size);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Using a struct_group seems to reliably avoid the warning and
not make the code much uglier. The combined memset() should even
save a couple of cpu cycles.
syzbot reported the following lock sequence:
cpu 2:
grabs timer_base lock
spins on bpf_lpm lock
cpu 1:
grab rcu krcp lock
spins on timer_base lock
cpu 0:
grab bpf_lpm lock
spins on rcu krcp lock
bpf_lpm lock can be the same.
timer_base lock can also be the same due to timer migration.
but rcu krcp lock is always per-cpu, so it cannot be the same lock.
Hence it's a false positive.
To avoid lockdep complaining move kfree_rcu() after spin_unlock.
In the test, after map is closed and then after two rcu grace periods,
it is assumed that map_id is not available to user space.
But the above assumption cannot be guaranteed. After zero or one
or two rcu grace periods in different siturations, the actual
freeing-map-work is put into a workqueue. Later on, when the work
is dequeued, the map will be actually freed.
See bpf_map_put() in kernel/bpf/syscall.c.
By using workqueue, there is no ganrantee that map will be actually
freed after a couple of rcu grace periods. This patch removed
such map leak detection and then the test can pass consistently.
In some systems, the netcat server can incur in delay to start listening.
When this happens, the test can randomly fail in various points.
This is an example error message:
# ip gre none gso
# encap 192.168.1.1 to 192.168.1.2, type gre, mac none len 2000
# test basic connectivity
# Ncat: Connection refused.
The issue stems from a race condition between the netcat client and server.
The test author had addressed this problem by implementing a sleep, which
I have removed in this patch.
This patch introduces a function capable of sleeping for up to two seconds.
However, it can terminate the waiting period early if the port is reported
to be listening.
The ssb_device_uevent() function first attempts to convert the 'dev' pointer
to 'struct ssb_device *'. However, it mistakenly dereferences 'dev' before
performing the NULL check, potentially leading to a NULL pointer
dereference if 'dev' is NULL.
To fix this issue, move the NULL check before dereferencing the 'dev' pointer,
ensuring that the pointer is valid before attempting to use it.
Found by Linux Verification Center (linuxtesting.org) with SVACE.
Historically, the signed integer overflow sanitizer did not work in the
kernel due to its interaction with `-fwrapv` but this has since been
changed [1] in the newest version of Clang; It was re-enabled in the
kernel with Commit 557f8c582a9ba8ab ("ubsan: Reintroduce signed overflow
sanitizer").
Let's rework this overflow checking logic to not actually perform an
overflow during the check itself, thus avoiding the UBSAN splat.
Enable CONFIG_INTERCONNECT_QCOM_SM6115 as built-in to enable the
interconnect driver for the SoC used on Qualcomm Robotics RB2 board.
Building as built-in is required as on this platform interconnects are
required to bring up the console.
When generating Runtime Calls, Clang doesn't respect the -mregparm=3
option used on i386. Hopefully this will be fixed correctly in Clang 19:
https://github.com/llvm/llvm-project/pull/89707
but we need to fix this for earlier Clang versions today. Force the
calling convention to use non-register arguments.
projects / thirdparty / kernel / stable.git / log
