commit 6b1e0651e9ce8ce418ad4ff360e7b9925dc5da79 upstream.
It is possible that an interrupt is disabled and masked at the same time.
When the interrupt is enabled again by enable_irq(), only plic_irq_enable()
is called, not plic_irq_unmask(). The interrupt remains masked and never
raises.
An example where interrupt is both disabled and masked is when
handle_fasteoi_irq() is the handler, and IRQS_ONESHOT is set. The interrupt
handler:
1. Mask the interrupt
2. Handle the interrupt
3. Check if interrupt is still enabled, and unmask it (see
cond_unmask_eoi_irq())
If another task disables the interrupt in the middle of the above steps,
the interrupt will not get unmasked, and will remain masked when it is
enabled in the future.
The problem is occasionally observed when PREEMPT_RT is enabled, because
PREEMPT_RT adds the IRQS_ONESHOT flag. But PREEMPT_RT only makes the problem
more likely to appear, the bug has been around since commit a1706a1c50
("irqchip/sifive-plic: Separate the enable and mask operations").
Fix it by unmasking interrupt in plic_irq_enable().
Fixes: a1706a1c50 ("irqchip/sifive-plic: Separate the enable and mask operations")
Signed-off-by: Nam Cao <namcao@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/all/20241003084152.2422969-1-namcao@linutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 93b8ddc54507a227087c60a0013ed833b6ae7d3c upstream.
The current implementation only calls chained_irq_enter() and
chained_irq_exit() if it detects pending interrupts.
```
for (i = 0; i < info->stride; i++) {
uregmap_read(info->map, id_reg + 4 * i, ®);
if (!reg)
continue;
chained_irq_enter(parent_chip, desc);
```
However, in case of GPIO pin configured in level mode and the parent
controller configured in edge mode, GPIO interrupt might be lowered by the
hardware. In the result, if the interrupt is short enough, the parent
interrupt is still pending while the GPIO interrupt is cleared;
chained_irq_enter() never gets called and the system hangs trying to
service the parent interrupt.
Moving chained_irq_enter() and chained_irq_exit() outside the for loop
ensures that they are called even when GPIO interrupt is lowered by the
hardware.
The similar code with chained_irq_enter() / chained_irq_exit() functions
wrapping interrupt checking loop may be found in many other drivers:
```
grep -r -A 10 chained_irq_enter drivers/pinctrl
```
Cc: stable@vger.kernel.org
Signed-off-by: Sergey Matsievskiy <matsievskiysv@gmail.com>
Reviewed-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Link: https://lore.kernel.org/20241012105743.12450-2-matsievskiysv@gmail.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit ffd95846c6ec6cf1f93da411ea10d504036cab42 upstream.
New processors have become pickier about the local APIC timer state
before entering low power modes. These low power modes are used (for
example) when you close your laptop lid and suspend. If you put your
laptop in a bag and it is not in this low power mode, it is likely
to get quite toasty while it quickly sucks the battery dry.
The problem boils down to some CPUs' inability to power down until the
CPU recognizes that the local APIC timer is shut down. The current
kernel code works in one-shot and periodic modes but does not work for
deadline mode. Deadline mode has been the supported and preferred mode
on Intel CPUs for over a decade and uses an MSR to drive the timer
instead of an APIC register.
Disable the TSC Deadline timer in lapic_timer_shutdown() by writing to
MSR_IA32_TSC_DEADLINE when in TSC-deadline mode. Also avoid writing
to the initial-count register (APIC_TMICT) which is ignored in
TSC-deadline mode.
Note: The APIC_LVTT|=APIC_LVT_MASKED operation should theoretically be
enough to tell the hardware that the timer will not fire in any of the
timer modes. But mitigating AMD erratum 411[1] also requires clearing
out APIC_TMICT. Solely setting APIC_LVT_MASKED is also ineffective in
practice on Intel Lunar Lake systems, which is the motivation for this
change.
1. 411 Processor May Exit Message-Triggered C1E State Without an Interrupt if Local APIC Timer Reaches Zero - https://www.amd.com/content/dam/amd/en/documents/archived-tech-docs/revision-guides/41322_10h_Rev_Gd.pdf
Fixes: 279f146143 ("x86: apic: Use tsc deadline for oneshot when available")
Suggested-by: Dave Hansen <dave.hansen@intel.com>
Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Tested-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Tested-by: Todd Brandt <todd.e.brandt@intel.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/all/20241015061522.25288-1-rui.zhang%40intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit d5fd042bf4cfb557981d65628e1779a492cd8cfa upstream.
After a recent LLVM change [1] that deduces __cold on functions that only call
cold code (such as __init functions), there is a section mismatch warning from
__get_mem_config_intel(), which got moved to .text.unlikely. as a result of
that optimization:
WARNING: modpost: vmlinux: section mismatch in reference: \
__get_mem_config_intel+0x77 (section: .text.unlikely.) -> thread_throttle_mode_init (section: .init.text)
Mark __get_mem_config_intel() as __init as well since it is only called
from __init code, which clears up the warning.
While __rdt_get_mem_config_amd() does not exhibit a warning because it
does not call any __init code, it is a similar function that is only
called from __init code like __get_mem_config_intel(), so mark it __init
as well to keep the code symmetrical.
CONFIG_SECTION_MISMATCH_WARN_ONLY=n would turn this into a fatal error.
Fixes: 05b93417ce ("x86/intel_rdt/mba: Add primary support for Memory Bandwidth Allocation (MBA)")
Fixes: 4d05bf71f1 ("x86/resctrl: Introduce AMD QOS feature")
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Reviewed-by: Reinette Chatre <reinette.chatre@intel.com>
Cc: <stable@kernel.org>
Link: 6b11573b8c [1]
Link: https://lore.kernel.org/r/20240917-x86-restctrl-get_mem_config_intel-init-v3-1-10d521256284@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 02ac3a9ef3a18b58d8f3ea2b6e46de657bf6c4f9 upstream.
The recent fix for array out-of-bounds accesses replaced sprintf()
calls blindly with snprintf(). However, since snprintf() returns the
would-be-printed size, not the actually output size, the length
calculation can still go over the given limit.
Use scnprintf() instead of snprintf(), which returns the actually
output letters, for addressing the potential out-of-bounds access
properly.
Fixes: ab11dac93d2d ("dev/parport: fix the array out-of-bounds risk")
Cc: stable@vger.kernel.org
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Link: https://lore.kernel.org/r/20240920103318.19271-1-tiwai@suse.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit c96e31252110a84dcc44412e8a7b456b33c3e298 upstream.
DWC3 programming guide mentions that when operating in USB2.0 speeds,
if GUSB2PHYCFG[6] or GUSB2PHYCFG[8] is set, it must be cleared prior
to issuing commands and may be set again after the command completes.
But currently while issuing EndXfer command without CmdIOC set, we
wait for 1ms after GUSB2PHYCFG is restored. This results in cases
where EndXfer command doesn't get completed and causes SMMU faults
since requests are unmapped afterwards. Hence restore GUSB2PHYCFG
after waiting for EndXfer command completion.
Cc: stable@vger.kernel.org
Fixes: 1d26ba0944d3 ("usb: dwc3: Wait unconditionally after issuing EndXfer command")
Signed-off-by: Prashanth K <quic_prashk@quicinc.com>
Acked-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
Link: https://lore.kernel.org/r/20240924093208.2524531-1-quic_prashk@quicinc.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit fe49df60cdb7c2975aa743dc295f8786e4b7db10 upstream.
Avoid xHC host from processing a cancelled URB by always turning
cancelled URB TDs into no-op TRBs before queuing a 'Set TR Deq' command.
If the command fails then xHC will start processing the cancelled TD
instead of skipping it once endpoint is restarted, causing issues like
Babble error.
This is not a complete solution as a failed 'Set TR Deq' command does not
guarantee xHC TRB caches are cleared.
Fixes: 4db356924a ("xhci: turn cancelled td cleanup to its own function")
Cc: stable@vger.kernel.org
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20241016140000.783905-3-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 6599b6a6fa8060145046d0744456b6abdb3122a7 upstream.
The stream contex type (SCT) bitfield is used both in the stream context
data structure, and in the 'Set TR Dequeue pointer' command TRB.
In both cases it uses bits 3:1
The SCT_FOR_TRB(p) macro used to set the stream context type (SCT) field
for the 'Set TR Dequeue pointer' command TRB incorrectly shifts the value
1 bit left before masking the three bits.
Fix this by first masking and rshifting, just like the similar
SCT_FOR_CTX(p) macro does
This issue has not been visibile as the lost bit 3 is only used with
secondary stream arrays (SSA). Xhci driver currently only supports using
a primary stream array with Linear stream addressing.
Fixes: 95241dbdf8 ("xhci: Set SCT field for Set TR dequeue on streams")
Cc: stable@vger.kernel.org
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20241016140000.783905-2-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 2c1dda2acc4192d826e84008d963b528e24d12bc upstream.
Fake CSR controllers don't seem to handle short-transfer properly which
cause command to time out:
kernel: usb 1-1: new full-speed USB device number 19 using xhci_hcd
kernel: usb 1-1: New USB device found, idVendor=0a12, idProduct=0001, bcdDevice=88.91
kernel: usb 1-1: New USB device strings: Mfr=0, Product=2, SerialNumber=0
kernel: usb 1-1: Product: BT DONGLE10
...
Bluetooth: hci1: Opcode 0x1004 failed: -110
kernel: Bluetooth: hci1: command 0x1004 tx timeout
According to USB Spec 2.0 Section 5.7.3 Interrupt Transfer Packet Size
Constraints a interrupt transfer is considered complete when the size is 0
(ZPL) or < wMaxPacketSize:
'When an interrupt transfer involves more data than can fit in one
data payload of the currently established maximum size, all data
payloads are required to be maximum-sized except for the last data
payload, which will contain the remaining data. An interrupt transfer
is complete when the endpoint does one of the following:
• Has transferred exactly the amount of data expected
• Transfers a packet with a payload size less than wMaxPacketSize or
transfers a zero-length packet'
Link: https://bugzilla.kernel.org/show_bug.cgi?id=219365
Fixes: 7b05933340f4 ("Bluetooth: btusb: Fix not handling ZPL/short-transfer")
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit a9b7b535ba192c6b77e6c15a4c82d853163eab8c upstream.
If bt_debugfs is not created successfully, which happens if either
CONFIG_DEBUG_FS or CONFIG_DEBUG_FS_ALLOW_ALL is unset, then iso_init()
returns early and does not set iso_inited to true. This means that a
subsequent call to iso_init() will result in duplicate calls to
proto_register(), bt_sock_register(), etc.
With CONFIG_LIST_HARDENED and CONFIG_BUG_ON_DATA_CORRUPTION enabled, the
duplicate call to proto_register() triggers this BUG():
list_add double add: new=ffffffffc0b280d0, prev=ffffffffbab56250,
next=ffffffffc0b280d0.
------------[ cut here ]------------
kernel BUG at lib/list_debug.c:35!
Oops: invalid opcode: 0000 [#1] PREEMPT SMP PTI
CPU: 2 PID: 887 Comm: bluetoothd Not tainted 6.10.11-1-ao-desktop #1
RIP: 0010:__list_add_valid_or_report+0x9a/0xa0
...
__list_add_valid_or_report+0x9a/0xa0
proto_register+0x2b5/0x340
iso_init+0x23/0x150 [bluetooth]
set_iso_socket_func+0x68/0x1b0 [bluetooth]
kmem_cache_free+0x308/0x330
hci_sock_sendmsg+0x990/0x9e0 [bluetooth]
__sock_sendmsg+0x7b/0x80
sock_write_iter+0x9a/0x110
do_iter_readv_writev+0x11d/0x220
vfs_writev+0x180/0x3e0
do_writev+0xca/0x100
...
This change removes the early return. The check for iso_debugfs being
NULL was unnecessary, it is always NULL when iso_inited is false.
Cc: stable@vger.kernel.org
Fixes: ccf74f2390 ("Bluetooth: Add BTPROTO_ISO socket type")
Signed-off-by: Aaron Thompson <dev@aaront.org>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit d458cd1221e9e56da3b2cc5518ad3225caa91f20 upstream.
If iso_init() has been called, iso_exit() must be called on module
unload. Without that, the struct proto that iso_init() registered with
proto_register() becomes invalid, which could cause unpredictable
problems later. In my case, with CONFIG_LIST_HARDENED and
CONFIG_BUG_ON_DATA_CORRUPTION enabled, loading the module again usually
triggers this BUG():
list_add corruption. next->prev should be prev (ffffffffb5355fd0),
but was 0000000000000068. (next=ffffffffc0a010d0).
------------[ cut here ]------------
kernel BUG at lib/list_debug.c:29!
Oops: invalid opcode: 0000 [#1] PREEMPT SMP PTI
CPU: 1 PID: 4159 Comm: modprobe Not tainted 6.10.11-4+bt2-ao-desktop #1
RIP: 0010:__list_add_valid_or_report+0x61/0xa0
...
__list_add_valid_or_report+0x61/0xa0
proto_register+0x299/0x320
hci_sock_init+0x16/0xc0 [bluetooth]
bt_init+0x68/0xd0 [bluetooth]
__pfx_bt_init+0x10/0x10 [bluetooth]
do_one_initcall+0x80/0x2f0
do_init_module+0x8b/0x230
__do_sys_init_module+0x15f/0x190
do_syscall_64+0x68/0x110
...
Cc: stable@vger.kernel.org
Fixes: ccf74f2390 ("Bluetooth: Add BTPROTO_ISO socket type")
Signed-off-by: Aaron Thompson <dev@aaront.org>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 530688e39c644543b71bdd9cb45fdfb458a28eaa upstream.
The opt3001 driver uses predetermined full-scale range values to
determine what exponent to use for event trigger threshold values.
The problem is that one of the values specified in the datasheet is
missing from the implementation. This causes larger values to be
scaled down to an incorrect exponent, effectively reducing the
maximum settable threshold value by a factor of 2.
Add missing full-scale range array value.
Fixes: 94a9b7b180 ("iio: light: add support for TI's opt3001 light sensor")
Signed-off-by: Emil Gedenryd <emil.gedenryd@axis.com>
Cc: <Stable@vger.kernel.org>
Link: https://patch.msgid.link/20240913-add_opt3002-v2-1-69e04f840360@axis.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit c7c44e57750c31de43906d97813273fdffcf7d02 upstream.
The dev pointer that is received as an argument in the
in_illuminance_period_available_show function references the device
embedded in the IIO device, not in the i2c client.
dev_to_iio_dev() must be used to accessthe right data. The current
implementation leads to a segmentation fault on every attempt to read
the attribute because indio_dev gets a NULL assignment.
This bug has been present since the first appearance of the driver,
apparently since the last version (V6) before getting applied. A
constant attribute was used until then, and the last modifications might
have not been tested again.
Cc: stable@vger.kernel.org
Fixes: 7b779f573c ("iio: light: add driver for veml6030 ambient light sensor")
Signed-off-by: Javier Carrasco <javier.carrasco.cruz@gmail.com>
Link: https://patch.msgid.link/20240913-veml6035-v1-3-0b09c0c90418@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit c9e9746f275c45108f2b0633a4855d65d9ae0736 upstream.
The driver still uses the sensor resolution provided in the datasheet
until Rev. 1.6, 28-Apr-2022, which was updated with Rev 1.7,
28-Nov-2023. The original ambient light resolution has been updated from
0.0036 lx/ct to 0.0042 lx/ct, which is the value that can be found in
the current device datasheet.
Update the default resolution for IT = 100 ms and GAIN = 1/8 from the
original 4608 mlux/cnt to the current value from the "Resolution and
maximum detection range" table (Application Note 84367, page 5), 5376
mlux/cnt.
Cc: <stable@vger.kernel.org>
Fixes: 7b779f573c ("iio: light: add driver for veml6030 ambient light sensor")
Signed-off-by: Javier Carrasco <javier.carrasco.cruz@gmail.com>
Link: https://patch.msgid.link/20240923-veml6035-v2-1-58c72a0df31c@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit c0ec082f10b7a1fd25e8c1e2a686440da913b7a3 upstream.
Before this patch, if multiple BO_HANDLES chunks were submitted,
the error -EINVAL would be correctly set but could be overwritten
by the return value from amdgpu_cs_p1_bo_handles(). This patch
ensures that if there are multiple BO_HANDLES, we stop.
Fixes: fec5f8e8c6bc ("drm/amdgpu: disallow multiple BO_HANDLES chunks in one submit")
Signed-off-by: Mohammed Anees <pvmohammedanees2003@gmail.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 40f2cd98828f454bdc5006ad3d94330a5ea164b7)
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 28127dba64d8ae1a0b737b973d6d029908599611 upstream.
Include the encoder itself in its possible_clones bitmask.
In the past nothing validated that drivers were populating
possible_clones correctly, but that changed in commit
74d2aacbe8 ("drm: Validate encoder->possible_clones").
Looks like radeon never got the memo and is still not
following the rules 100% correctly.
This results in some warnings during driver initialization:
Bogus possible_clones: [ENCODER:46:TV-46] possible_clones=0x4 (full encoder mask=0x7)
WARNING: CPU: 0 PID: 170 at drivers/gpu/drm/drm_mode_config.c:615 drm_mode_config_validate+0x113/0x39c
...
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: amd-gfx@lists.freedesktop.org
Fixes: 74d2aacbe8 ("drm: Validate encoder->possible_clones")
Reported-by: Erhard Furtner <erhard_f@mailbox.org>
Closes: https://lore.kernel.org/dri-devel/20241009000321.418e4294@yea/
Tested-by: Erhard Furtner <erhard_f@mailbox.org>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 3b6e7d40649c0d75572039aff9d0911864c689db)
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 28aabffae6be54284869a91cd8bccd3720041129 upstream.
When an application uses SQPOLL, it must wait for the SQPOLL thread to
consume SQE entries, if it fails to get an sqe when calling
io_uring_get_sqe(). It can do so by calling io_uring_enter(2) with the
flag value of IORING_ENTER_SQ_WAIT. In liburing, this is generally done
with io_uring_sqring_wait(). There's a natural expectation that once
this call returns, a new SQE entry can be retrieved, filled out, and
submitted. However, the kernel uses the cached sq head to determine if
the SQRING is full or not. If the SQPOLL thread is currently in the
process of submitting SQE entries, it may have updated the cached sq
head, but not yet committed it to the SQ ring. Hence the kernel may find
that there are SQE entries ready to be consumed, and return successfully
to the application. If the SQPOLL thread hasn't yet committed the SQ
ring entries by the time the application returns to userspace and
attempts to get a new SQE, it will fail getting a new SQE.
Fix this by having io_sqring_full() always use the user visible SQ ring
head entry, rather than the internally cached one.
Cc: stable@vger.kernel.org # 5.10+
Link: https://github.com/axboe/liburing/discussions/1267
Reported-by: Benedek Thaler <thaler@thaler.hu>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit e972b08b91ef48488bae9789f03cfedb148667fb upstream.
We're seeing crashes from rq_qos_wake_function that look like this:
BUG: unable to handle page fault for address: ffffafe180a40084
#PF: supervisor write access in kernel mode
#PF: error_code(0x0002) - not-present page
PGD 100000067 P4D 100000067 PUD 10027c067 PMD 10115d067 PTE 0
Oops: Oops: 0002 [#1] PREEMPT SMP PTI
CPU: 17 UID: 0 PID: 0 Comm: swapper/17 Not tainted 6.12.0-rc3-00013-geca631b8fe80 #11
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.0-0-gd239552ce722-prebuilt.qemu.org 04/01/2014
RIP: 0010:_raw_spin_lock_irqsave+0x1d/0x40
Code: 90 90 90 90 90 90 90 90 90 90 90 90 90 f3 0f 1e fa 0f 1f 44 00 00 41 54 9c 41 5c fa 65 ff 05 62 97 30 4c 31 c0 ba 01 00 00 00 <f0> 0f b1 17 75 0a 4c 89 e0 41 5c c3 cc cc cc cc 89 c6 e8 2c 0b 00
RSP: 0018:ffffafe180580ca0 EFLAGS: 00010046
RAX: 0000000000000000 RBX: ffffafe180a3f7a8 RCX: 0000000000000011
RDX: 0000000000000001 RSI: 0000000000000003 RDI: ffffafe180a40084
RBP: 0000000000000000 R08: 00000000001e7240 R09: 0000000000000011
R10: 0000000000000028 R11: 0000000000000888 R12: 0000000000000002
R13: ffffafe180a40084 R14: 0000000000000000 R15: 0000000000000003
FS: 0000000000000000(0000) GS:ffff9aaf1f280000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: ffffafe180a40084 CR3: 000000010e428002 CR4: 0000000000770ef0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
PKRU: 55555554
Call Trace:
<IRQ>
try_to_wake_up+0x5a/0x6a0
rq_qos_wake_function+0x71/0x80
__wake_up_common+0x75/0xa0
__wake_up+0x36/0x60
scale_up.part.0+0x50/0x110
wb_timer_fn+0x227/0x450
...
So rq_qos_wake_function() calls wake_up_process(data->task), which calls
try_to_wake_up(), which faults in raw_spin_lock_irqsave(&p->pi_lock).
p comes from data->task, and data comes from the waitqueue entry, which
is stored on the waiter's stack in rq_qos_wait(). Analyzing the core
dump with drgn, I found that the waiter had already woken up and moved
on to a completely unrelated code path, clobbering what was previously
data->task. Meanwhile, the waker was passing the clobbered garbage in
data->task to wake_up_process(), leading to the crash.
What's happening is that in between rq_qos_wake_function() deleting the
waitqueue entry and calling wake_up_process(), rq_qos_wait() is finding
that it already got a token and returning. The race looks like this:
rq_qos_wait() rq_qos_wake_function()
==============================================================
prepare_to_wait_exclusive()
data->got_token = true;
list_del_init(&curr->entry);
if (data.got_token)
break;
finish_wait(&rqw->wait, &data.wq);
^- returns immediately because
list_empty_careful(&wq_entry->entry)
is true
... return, go do something else ...
wake_up_process(data->task)
(NO LONGER VALID!)-^
Normally, finish_wait() is supposed to synchronize against the waker.
But, as noted above, it is returning immediately because the waitqueue
entry has already been removed from the waitqueue.
The bug is that rq_qos_wake_function() is accessing the waitqueue entry
AFTER deleting it. Note that autoremove_wake_function() wakes the waiter
and THEN deletes the waitqueue entry, which is the proper order.
Fix it by swapping the order. We also need to use
list_del_init_careful() to match the list_empty_careful() in
finish_wait().
Fixes: 38cfb5a45e ("blk-wbt: improve waking of tasks")
Cc: stable@vger.kernel.org
Signed-off-by: Omar Sandoval <osandov@fb.com>
Acked-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Link: https://lore.kernel.org/r/d3bee2463a67b1ee597211823bf7ad3721c26e41.1729014591.git.osandov@fb.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit c62fa117c32bd1abed9304c58e0da6940f8c7fc2 upstream.
Since X86_FEATURE_ENTRY_IBPB will invalidate all harmful predictions
with IBPB, no software-based untraining of returns is needed anymore.
Currently, this change affects retbleed and SRSO mitigations so if
either of the mitigations is doing IBPB and the other one does the
software sequence, the latter is not needed anymore.
[ bp: Massage commit message. ]
Suggested-by: Borislav Petkov <bp@alien8.de>
Signed-off-by: Johannes Wikner <kwikner@ethz.ch>
Cc: <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 0fad2878642ec46225af2054564932745ac5c765 upstream.
entry_ibpb() is designed to follow Intel's IBPB specification regardless
of CPU. This includes invalidating RSB entries.
Hence, if IBPB on VMEXIT has been selected, entry_ibpb() as part of the
RET untraining in the VMEXIT path will take care of all BTB and RSB
clearing so there's no need to explicitly fill the RSB anymore.
[ bp: Massage commit message. ]
Suggested-by: Borislav Petkov <bp@alien8.de>
Signed-off-by: Johannes Wikner <kwikner@ethz.ch>
Cc: <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 50e4b3b94090babe8d4bb85c95f0d3e6b07ea86e upstream.
entry_ibpb() should invalidate all indirect predictions, including return
target predictions. Not all IBPB implementations do this, in which case the
fallback is RSB filling.
Prevent SRSO-style hijacks of return predictions following IBPB, as the return
target predictor can be corrupted before the IBPB completes.
[ bp: Massage. ]
Signed-off-by: Johannes Wikner <kwikner@ethz.ch>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Cc: <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 3ea87dfa31a7b0bb0ff1675e67b9e54883013074 upstream.
Set this flag if the CPU has an IBPB implementation that does not
invalidate return target predictions. Zen generations < 4 do not flush
the RSB when executing an IBPB and this bug flag denotes that.
[ bp: Massage. ]
Signed-off-by: Johannes Wikner <kwikner@ethz.ch>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Cc: <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit ff898623af2ed564300752bba83a680a1e4fec8d upstream.
AMD's initial implementation of IBPB did not clear the return address
predictor. Beginning with Zen4, AMD's IBPB *does* clear the return address
predictor. This behavior is enumerated by CPUID.80000008H:EBX.IBPB_RET[30].
Define X86_FEATURE_AMD_IBPB_RET for use in KVM_GET_SUPPORTED_CPUID,
when determining cross-vendor capabilities.
Suggested-by: Venkatesh Srinivas <venkateshs@chromium.org>
Signed-off-by: Jim Mattson <jmattson@google.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Reviewed-by: Tom Lendacky <thomas.lendacky@amd.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Cc: <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit e8061f06185be0a06a73760d6526b8b0feadfe52 upstream.
Previously, access_guest_page() did not check whether the given guest
address is inside of a memslot. This is not a problem, since
kvm_write_guest_page/kvm_read_guest_page return -EFAULT in this case.
However, -EFAULT is also returned when copy_to/from_user fails.
When emulating a guest instruction, the address being outside a memslot
usually means that an addressing exception should be injected into the
guest.
Failure in copy_to/from_user however indicates that something is wrong
in userspace and hence should be handled there.
To be able to distinguish these two cases, return PGM_ADDRESSING in
access_guest_page() when the guest address is outside guest memory. In
access_guest_real(), populate vcpu->arch.pgm.code such that
kvm_s390_inject_prog_cond() can be used in the caller for injecting into
the guest (if applicable).
Since this adds a new return value to access_guest_page(), we need to make
sure that other callers are not confused by the new positive return value.
There are the following users of access_guest_page():
- access_guest_with_key() does the checking itself (in
guest_range_to_gpas()), so this case should never happen. Even if, the
handling is set up properly.
- access_guest_real() just passes the return code to its callers, which
are:
- read_guest_real() - see below
- write_guest_real() - see below
There are the following users of read_guest_real():
- ar_translation() in gaccess.c which already returns PGM_*
- setup_apcb10(), setup_apcb00(), setup_apcb11() in vsie.c which always
return -EFAULT on read_guest_read() nonzero return - no change
- shadow_crycb(), handle_stfle() always present this as validity, this
could be handled better but doesn't change current behaviour - no change
There are the following users of write_guest_real():
- kvm_s390_store_status_unloaded() always returns -EFAULT on
write_guest_real() failure.
Fixes: 2293897805 ("KVM: s390: add architecture compliant guest access functions")
Cc: stable@vger.kernel.org
Signed-off-by: Nico Boehr <nrb@linux.ibm.com>
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Link: https://lore.kernel.org/r/20240917151904.74314-2-nrb@linux.ibm.com
Acked-by: Janosch Frank <frankja@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 0d9dc27df22d9b5c8dc7185c8dddbc14f5468518 upstream.
On reboot the SCLP interface is deactivated through a reboot notifier.
This happens before other components using SCLP have the chance to run
their own reboot notifiers.
Two of those components are the SCLP console and tty drivers which try
to flush the last outstanding messages.
At that point the SCLP interface is already unusable and the messages
are discarded.
Execute sclp_deactivate() as late as possible to avoid this issue.
Fixes: 4ae46db99c ("s390/consoles: improve panic notifiers reliability")
Cc: stable@vger.kernel.org
Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de>
Reviewed-by: Sven Schnelle <svens@linux.ibm.com>
Link: https://lore.kernel.org/r/20241014-s390-kunit-v1-1-941defa765a6@linutronix.de
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
