commit 91e87045a5 upstream.
Currently, the .port_set_rgmii_delay hook is missing for the 88E6320
family, which causes failure to retrieve an IP address via DHCP.
Add mv88e6320_port_set_rgmii_delay() that allows applying the RGMII
delay for ports 2, 5, and 6, which are the only ports that can be used
in RGMII mode.
Tested on a custom i.MX8MN board connected to an 88E6320 switch.
This change also applies safely to the 88E6321 variant.
The only difference between 88E6320 versus 88E6321 is the temperature
grade and pinout.
They share exactly the same MDIO register map for ports 2, 5, and 6,
which are the only ports that can be used in RGMII mode.
Signed-off-by: Steffen Bätz <steffen@innosonix.de>
[fabio: Improved commit log and extended it to mv88e6321_ops]
Signed-off-by: Fabio Estevam <festevam@denx.de>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Link: https://lore.kernel.org/r/20221028163158.198108-1-festevam@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Cc: Fabio Estevam <festevam@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 4eda19cc8a upstream.
The watchdog countdown is supposed to begin when the device file is
opened. Instead, it would begin countdown upon the first write to or
close of the device file. Now, the ping operation is called within the
start operation which ensures the countdown begins. From experimenation,
it does not appear possible to do this with a single write including
both the start bit and the trigger bit. So, it is done as two distinct
writes.
Signed-off-by: Gregory Oakes <gregory.oakes@amd.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lore.kernel.org/r/20230316201312.17538-1-gregory.oakes@amd.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
Cc: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 99d4645062 ]
Set TPM_CHIP_FLAG_SUSPENDED in tpm_pm_suspend() and reset in
tpm_pm_resume(). While the flag is set, tpm_hwrng() gives back zero
bytes. This prevents hwrng from racing during resume.
Cc: stable@vger.kernel.org
Fixes: 6e592a065d ("tpm: Move Linux RNG connection to hwrng")
Reviewed-by: Jerry Snitselaar <jsnitsel@redhat.com>
Signed-off-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 0c8862de05 ]
TPM chip bootstrapping was removed from tpm_chip_register(), and it
was relocated to tpm_tis_core. This breaks all drivers which are not
based on tpm_tis because the chip will not get properly initialized.
Take the corrective steps:
1. Rename tpm_chip_startup() as tpm_chip_bootstrap() and make it one-shot.
2. Call tpm_chip_bootstrap() in tpm_chip_register(), which reverts the
things as tehy used to be.
Cc: Lino Sanfilippo <l.sanfilippo@kunbus.com>
Fixes: 548eb516ec ("tpm, tpm_tis: startup chip before testing for interrupts")
Reported-by: Pengfei Xu <pengfei.xu@intel.com>
Link: https://lore.kernel.org/all/ZEjqhwHWBnxcaRV5@xpf.sh.intel.com/
Tested-by: Pengfei Xu <pengfei.xu@intel.com>
Signed-off-by: Jarkko Sakkinen <jarkko@kernel.org>
Stable-dep-of: 99d4645062 ("tpm: Prevent hwrng from activating during resume")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 548eb516ec ]
In tpm_tis_gen_interrupt() a request for a property value is sent to the
TPM to test if interrupts are generated. However after a power cycle the
TPM responds with TPM_RC_INITIALIZE which indicates that the TPM is not
yet properly initialized.
Fix this by first starting the TPM up before the request is sent. For this
the startup implementation is removed from tpm_chip_register() and put
into the new function tpm_chip_startup() which is called before the
interrupts are tested.
Signed-off-by: Lino Sanfilippo <l.sanfilippo@kunbus.com>
Tested-by: Jarkko Sakkinen <jarkko@kernel.org>
Reviewed-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Jarkko Sakkinen <jarkko@kernel.org>
Stable-dep-of: 99d4645062 ("tpm: Prevent hwrng from activating during resume")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1398aa803f ]
Before sending a TPM command, CLKRUN protocol must be disabled. This is not
done in the case of tpm1_do_selftest() call site inside tpm_tis_resume().
Address this by decorating the calls with tpm_chip_{start,stop}, which
should be always used to arm and disarm the TPM chip for transmission.
Finally, move the call to the main TPM driver callback as the last step
because it should arm the chip by itself, if it needs that type of
functionality.
Cc: stable@vger.kernel.org
Reported-by: Jason A. Donenfeld <Jason@zx2c4.com>
Closes: https://lore.kernel.org/linux-integrity/CS68AWILHXS4.3M36M1EKZLUMS@suppilovahvero/
Fixes: a3fbfae82b ("tpm: take TPM chip power gating out of tpm_transmit()")
Reviewed-by: Jerry Snitselaar <jsnitsel@redhat.com>
Signed-off-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit e87fcf0dc2 ]
According to the TPM Interface Specification (TIS) support for "stsValid"
and "commandReady" interrupts is only optional.
This has to be taken into account when handling the interrupts in functions
like wait_for_tpm_stat(). To determine the supported interrupts use the
capability query.
Also adjust wait_for_tpm_stat() to only wait for interrupt reported status
changes. After that process all the remaining status changes by polling
the status register.
Signed-off-by: Lino Sanfilippo <l.sanfilippo@kunbus.com>
Tested-by: Michael Niewöhner <linux@mniewoehner.de>
Tested-by: Jarkko Sakkinen <jarkko@kernel.org>
Reviewed-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Jarkko Sakkinen <jarkko@kernel.org>
Stable-dep-of: 1398aa803f ("tpm_tis: Use tpm_chip_{start,stop} decoration inside tpm_tis_resume")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 858e8b792d ]
The interrupt handler that sets the boolean variable irq_tested may run on
another CPU as the thread that checks irq_tested as part of the irq test in
tpm_tis_send().
Since nothing guarantees cache coherency between CPUs for unsynchronized
accesses to boolean variables the testing thread might not perceive the
value change done in the interrupt handler.
Avoid this issue by setting the bit TPM_TIS_IRQ_TESTED in the flags field
of the tpm_tis_data struct and by accessing this field with the bit
manipulating functions that provide cache coherency.
Also convert all other existing sites to use the proper macros when
accessing this bitfield.
Signed-off-by: Lino Sanfilippo <l.sanfilippo@kunbus.com>
Tested-by: Michael Niewöhner <linux@mniewoehner.de>
Tested-by: Jarkko Sakkinen <jarkko@kernel.org>
Reviewed-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Jarkko Sakkinen <jarkko@kernel.org>
Stable-dep-of: 1398aa803f ("tpm_tis: Use tpm_chip_{start,stop} decoration inside tpm_tis_resume")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4e8ef34e36 ]
When work in gadget mode, currently driver doesn't update software level
link_state correctly as link state change event is not enabled for most
devices, in function dwc3_gadget_suspend_interrupt(), it will only pass
suspend event to UDC core when software level link state changes, so when
interrupt generated in sequences of suspend -> reset -> conndone ->
suspend, link state is not updated during reset and conndone, so second
suspend interrupt event will not pass to UDC core.
Remove link_state compare in dwc3_gadget_suspend_interrupt() and add a
suspended flag to replace the compare function.
Fixes: 799e9dc829 ("usb: dwc3: gadget: conditionally disable Link State change events")
Cc: stable <stable@kernel.org>
Acked-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
Signed-off-by: Linyu Yuan <quic_linyyuan@quicinc.com>
Link: https://lore.kernel.org/r/20230512004524.31950-1-quic_linyyuan@quicinc.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit f900fde288 upstream.
The performance of the crypto fuzz tests has greatly regressed since
v5.18. When booting a kernel on an arm64 dev board with all software
crypto algorithms and CONFIG_CRYPTO_MANAGER_EXTRA_TESTS enabled, the
fuzz tests now take about 200 seconds to run, or about 325 seconds with
lockdep enabled, compared to about 5 seconds before.
The root cause is that the random number generation has become much
slower due to commit d4150779e6 ("random32: use real rng for
non-deterministic randomness"). On my same arm64 dev board, at the time
the fuzz tests are run, get_random_u8() is about 345x slower than
prandom_u32_state(), or about 469x if lockdep is enabled.
Lockdep makes a big difference, but much of the rest comes from the
get_random_*() functions taking a *very* slow path when the CRNG is not
yet initialized. Since the crypto self-tests run early during boot,
even having a hardware RNG driver enabled (CONFIG_CRYPTO_DEV_QCOM_RNG in
my case) doesn't prevent this. x86 systems don't have this issue, but
they still see a significant regression if lockdep is enabled.
Converting the "Fully random bytes" case in generate_random_bytes() to
use get_random_bytes() helps significantly, improving the test time to
about 27 seconds. But that's still over 5x slower than before.
This is all a bit silly, though, since the fuzz tests don't actually
need cryptographically secure random numbers. So let's just make them
use a non-cryptographically-secure RNG as they did before. The original
prandom_u32() is gone now, so let's use prandom_u32_state() instead,
with an explicitly managed state, like various other self-tests in the
kernel source tree (rbtree_test.c, test_scanf.c, etc.) already do. This
also has the benefit that no locking is required anymore, so performance
should be even better than the original version that used prandom_u32().
Fixes: d4150779e6 ("random32: use real rng for non-deterministic randomness")
Cc: stable@vger.kernel.org
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit c4c597f1b3 upstream.
The mte_sync_page_tags() function sets PG_mte_tagged if it initializes
page tags. Then we return to mte_sync_tags(), which sets PG_mte_tagged
again. At best, this is redundant. However, it is possible for
mte_sync_page_tags() to return without having initialized tags for the
page, i.e. in the case where check_swap is true (non-compound page),
is_swap_pte(old_pte) is false and pte_is_tagged is false. So at worst,
we set PG_mte_tagged on a page with uninitialized tags. This can happen
if, for example, page migration causes a PTE for an untagged page to
be replaced. If the userspace program subsequently uses mprotect() to
enable PROT_MTE for that page, the uninitialized tags will be exposed
to userspace.
Fix it by removing the redundant call to set_page_mte_tagged().
Fixes: e059853d14 ("arm64: mte: Fix/clarify the PG_mte_tagged semantics")
Signed-off-by: Peter Collingbourne <pcc@google.com>
Cc: <stable@vger.kernel.org> # 6.1
Link: https://linux-review.googlesource.com/id/Ib02d004d435b2ed87603b858ef7480f7b1463052
Reviewed-by: Catalin Marinas <catalin.marinas@arm.com>
Reviewed-by: Alexandru Elisei <alexandru.elisei@arm.com>
Link: https://lore.kernel.org/r/20230420214327.2357985-1-pcc@google.com
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 2862a2fdfa upstream.
Use "a" constraint instead of "d" constraint to pass the state parameter to
the do_sqbs() inline assembly. This prevents that general purpose register
zero is used for the state parameter.
If the compiler would select general purpose register zero this would be
problematic for the used instruction in rsy format: the register used for
the state parameter is a base register. If the base register is general
purpose register zero the contents of the register are unexpectedly ignored
when the instruction is executed.
This only applies to z/VM guests using QIOASSIST with dedicated (pass through)
QDIO-based devices such as FCP [zfcp driver] as well as real OSA or
HiperSockets [qeth driver].
A possible symptom for this case using zfcp is the following repeating kernel
message pattern:
zfcp <devbusid>: A QDIO problem occurred
zfcp <devbusid>: A QDIO problem occurred
zfcp <devbusid>: qdio: ZFCP on SC <sc> using AI:1 QEBSM:1 PRI:1 TDD:1 SIGA: W
zfcp <devbusid>: A QDIO problem occurred
zfcp <devbusid>: A QDIO problem occurred
Each of the qdio problem message can be accompanied by the following entries
for the affected subchannel <sc> in
/sys/kernel/debug/s390dbf/qdio_error/hex_ascii for zfcp or qeth:
<sc> ccq: 69....
<sc> SQBS ERROR.
Reviewed-by: Benjamin Block <bblock@linux.ibm.com>
Cc: Steffen Maier <maier@linux.ibm.com>
Fixes: 8129ee1642 ("[PATCH] s390: qdio V=V pass-through")
Cc: <stable@vger.kernel.org>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit c99bff3429 upstream.
Formatting a thin-provisioned (ESE) device that is part of a PPRC copy
relation might fail with the following error:
dasd-eckd 0.0.f500: An error occurred in the DASD device driver, reason=09
[...]
24 Byte: 0 MSG 4, no MSGb to SYSOP
During format of an ESE disk the Release Allocated Space command is used.
A bit in the payload of the command is set that is not allowed to be set
for devices in a copy relation. This bit is set to allow the partial
release of an extent.
Check for the existence of a copy relation before setting the respective
bit.
Fixes: 91dc4a1975 ("s390/dasd: Add new ioctl to release space")
Cc: stable@kernel.org # 5.3+
Signed-off-by: Stefan Haberland <sth@linux.ibm.com>
Reviewed-by: Jan Hoeppner <hoeppner@linux.ibm.com>
Link: https://lore.kernel.org/r/20230519102340.3854819-2-sth@linux.ibm.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 66b2ca0862 upstream.
It was reported that soft dirty tracking doesn't work when using the
Radix MMU.
The tracking is supposed to work by clearing the soft dirty bit for a
mapping and then write protecting the PTE. If/when the page is written
to, a page fault occurs and the soft dirty bit is added back via
pte_mkdirty(). For example in wp_page_reuse():
entry = maybe_mkwrite(pte_mkdirty(entry), vma);
if (ptep_set_access_flags(vma, vmf->address, vmf->pte, entry, 1))
update_mmu_cache(vma, vmf->address, vmf->pte);
Unfortunately on radix _PAGE_SOFTDIRTY is being dropped by
radix__ptep_set_access_flags(), called from ptep_set_access_flags(),
meaning the soft dirty bit is not set even though the page has been
written to.
Fix it by adding _PAGE_SOFTDIRTY to the set of bits that are able to be
changed in radix__ptep_set_access_flags().
Fixes: b0b5e9b130 ("powerpc/mm/radix: Add radix pte #defines")
Cc: stable@vger.kernel.org # v4.7+
Reported-by: Dan Horák <dan@danny.cz>
Link: https://lore.kernel.org/r/20230511095558.56663a50f86bdc4cd97700b7@danny.cz
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://msgid.link/20230511114224.977423-1-mpe@ellerman.id.au
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 1f7aacc5eb upstream.
For an SR-IOV device, while enabling DDW, a new table is created and
added at index 1 in the group. In the below 2 scenarios, the table is
incorrectly referenced at index 0 (which is where the table is for
default DMA window).
1. When adding DDW
This issue is exposed with "slub_debug". Error thrown out from
dma_iommu_dma_supported()
Warning: IOMMU offset too big for device mask
mask: 0xffffffff, table offset: 0x800000000000000
2. During Dynamic removal of the PCI device.
Error is from iommu_tce_table_put() since a NULL table pointer is
passed in.
Fixes: 381ceda88c ("powerpc/pseries/iommu: Make use of DDW for indirect mapping")
Cc: stable@vger.kernel.org # v5.15+
Signed-off-by: Gaurav Batra <gbatra@linux.vnet.ibm.com>
Reviewed-by: Brian King <brking@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://msgid.link/20230505184701.91613-1-gbatra@linux.vnet.ibm.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit a784452872 upstream.
Current only one entry is enabled but IP itself is using 4 different IDs
which are already listed in zynqmp.dtsi.
sata: ahci@fd0c0000 {
compatible = "ceva,ahci-1v84";
...
iommus = <&smmu 0x4c0>, <&smmu 0x4c1>,
<&smmu 0x4c2>, <&smmu 0x4c3>;
};
Fixes: 8ac47837f0 ("arm64: dts: zynqmp: Add missing iommu IDs")
Cc: stable@vger.kernel.org # v5.12+
Signed-off-by: Michal Simek <michal.simek@amd.com>
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit d5aa417808 upstream.
This code was written prior to previous updates to this
logic for other chips. The RSC registers are part of
SMUIO which is an always on block so there is no need
to disable gfxoff. Additionally add the carryover and
preemption checks.
v2: rebase
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org # 6.1.y: 5591a051b8: drm/amdgpu: refine get gpu clock counter method
Cc: stable@vger.kernel.org # 6.2.y: 5591a051b8: drm/amdgpu: refine get gpu clock counter method
Cc: stable@vger.kernel.org # 6.3.y: 5591a051b8: drm/amdgpu: refine get gpu clock counter method
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 5591a051b8 upstream.
[why]
regGOLDEN_TSC_COUNT_LOWER/regGOLDEN_TSC_COUNT_UPPER are protected and
unaccessible under sriov.
The clock counter high bit may update during reading process.
[How]
Replace regGOLDEN_TSC_COUNT_LOWER/regGOLDEN_TSC_COUNT_UPPER with
regCP_MES_MTIME_LO/regCP_MES_MTIME_HI to get gpu clock under sriov.
Refine get gpu clock counter method to make the result more precise.
Signed-off-by: Tong Liu01 <Tong.Liu01@amd.com>
Acked-by: Luben Tuikov <luben.tuikov@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 68518294d0 upstream.
Implement get_vbios_fb_size() so we can properly reserve
the vbios splash screen to avoid potential artifacts on the
screen during the transition from the pre-OS console to the
OS console.
Acked-by: Sunil Khatri <sunil.khatri@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org # 6.1.x
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 5f949f140f upstream.
The driver have a race, experienced only with PREEMPT_RT patchset:
CPU0 | CPU1
==================================================================
qcom_geni_serial_probe |
uart_add_one_port |
| serdev_drv_probe
| qca_serdev_probe
| serdev_device_open
| uart_open
| uart_startup
| qcom_geni_serial_startup
| enable_irq
| __irq_startup
| WARN_ON()
| IRQ not activated
request_threaded_irq |
irq_domain_activate_irq |
The warning:
894000.serial: ttyHS1 at MMIO 0x894000 (irq = 144, base_baud = 0) is a MSM
serial serial0: tty port ttyHS1 registered
WARNING: CPU: 7 PID: 107 at kernel/irq/chip.c:241 __irq_startup+0x78/0xd8
...
qcom_geni_serial 894000.serial: serial engine reports 0 RX bytes in!
Adding UART port triggers probe of child serial devices - serdev and
eventually Qualcomm Bluetooth hci_qca driver. This opens UART port
which enables the interrupt before it got activated in
request_threaded_irq(). The issue originates in commit f3974413cf
("tty: serial: qcom_geni_serial: Wakeup IRQ cleanup") and discussion on
mailing list [1]. However the above commit does not explain why the
uart_add_one_port() is moved above requesting interrupt.
[1] https://lore.kernel.org/all/5d9f3dfa.1c69fb81.84c4b.30bf@mx.google.com/
Fixes: f3974413cf ("tty: serial: qcom_geni_serial: Wakeup IRQ cleanup")
Cc: <stable@vger.kernel.org>
Cc: Stephen Boyd <swboyd@chromium.org>
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Link: https://lore.kernel.org/r/20230505152301.2181270-1-krzysztof.kozlowski@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 95d698869b upstream.
Possibly the last PCI controller-based (i.e. not a soft/winmodem)
dial-up modem one can still buy.
Looks to have a stock XR17C154 PCI UART chip for communication, but for
some reason when provisioning the PCI IDs they swapped the vendor and
subvendor IDs. Otherwise this card would have worked out of the box.
Searching online, some folks seem to not have this issue and others do,
so it is possible only some batches of cards have this error.
Create a new macro to handle the switched IDs and add support here.
Signed-off-by: Andrew Davis <afd@ti.com>
Cc: stable <stable@kernel.org>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20230420160209.28221-1-afd@ti.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 04fc781608 upstream.
The zswap writeback mechanism can cause a race condition resulting in
memory corruption, where a swapped out page gets swapped in with data that
was written to a different page.
The race unfolds like this:
1. a page with data A and swap offset X is stored in zswap
2. page A is removed off the LRU by zpool driver for writeback in
zswap-shrink work, data for A is mapped by zpool driver
3. user space program faults and invalidates page entry A, offset X is
considered free
4. kswapd stores page B at offset X in zswap (zswap could also be
full, if so, page B would then be IOed to X, then skip step 5.)
5. entry A is replaced by B in tree->rbroot, this doesn't affect the
local reference held by zswap-shrink work
6. zswap-shrink work writes back A at X, and frees zswap entry A
7. swapin of slot X brings A in memory instead of B
The fix:
Once the swap page cache has been allocated (case ZSWAP_SWAPCACHE_NEW),
zswap-shrink work just checks that the local zswap_entry reference is
still the same as the one in the tree. If it's not the same it means that
it's either been invalidated or replaced, in both cases the writeback is
aborted because the local entry contains stale data.
Reproducer:
I originally found this by running `stress` overnight to validate my work
on the zswap writeback mechanism, it manifested after hours on my test
machine. The key to make it happen is having zswap writebacks, so
whatever setup pumps /sys/kernel/debug/zswap/written_back_pages should do
the trick.
In order to reproduce this faster on a vm, I setup a system with ~100M of
available memory and a 500M swap file, then running `stress --vm 1
--vm-bytes 300000000 --vm-stride 4000` makes it happen in matter of tens
of minutes. One can speed things up even more by swinging
/sys/module/zswap/parameters/max_pool_percent up and down between, say, 20
and 1; this makes it reproduce in tens of seconds. It's crucial to set
`--vm-stride` to something other than 4096 otherwise `stress` won't
realize that memory has been corrupted because all pages would have the
same data.
Link: https://lkml.kernel.org/r/20230503151200.19707-1-cerasuolodomenico@gmail.com
Signed-off-by: Domenico Cerasuolo <cerasuolodomenico@gmail.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Chris Li (Google) <chrisl@kernel.org>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Nitin Gupta <ngupta@vflare.org>
Cc: Seth Jennings <sjenning@redhat.com>
Cc: Vitaly Wool <vitaly.wool@konsulko.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit ed40866ec7 upstream.
s390's struct statfs and struct statfs64 contain padding, which
field-by-field copying does not set. Initialize the respective structs
with zeros before filling them and copying them to userspace, like it's
already done for the compat versions of these structs.
Found by KMSAN.
[agordeev@linux.ibm.com: fixed typo in patch description]
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Cc: stable@vger.kernel.org # v4.14+
Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Link: https://lore.kernel.org/r/20230504144021.808932-2-iii@linux.ibm.com
Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit afb2acb2e3 upstream.
In kvm_vm_ioctl_create_vcpu(), add vcpu to vcpu_array iff it's safe to
access vcpu via kvm_get_vcpu() and kvm_for_each_vcpu(), i.e. when there's
no failure path requiring vcpu removal and destruction. Such order is
important because vcpu_array accessors may end up referencing vcpu at
vcpu_array[0] even before online_vcpus is set to 1.
When online_vcpus=0, any call to kvm_get_vcpu() goes through
array_index_nospec() and ends with an attempt to xa_load(vcpu_array, 0):
int num_vcpus = atomic_read(&kvm->online_vcpus);
i = array_index_nospec(i, num_vcpus);
return xa_load(&kvm->vcpu_array, i);
Similarly, when online_vcpus=0, a kvm_for_each_vcpu() does not iterate over
an "empty" range, but actually [0, ULONG_MAX]:
xa_for_each_range(&kvm->vcpu_array, idx, vcpup, 0, \
(atomic_read(&kvm->online_vcpus) - 1))
In both cases, such online_vcpus=0 edge case, even if leading to
unnecessary calls to XArray API, should not be an issue; requesting
unpopulated indexes/ranges is handled by xa_load() and xa_for_each_range().
However, this means that when the first vCPU is created and inserted in
vcpu_array *and* before online_vcpus is incremented, code calling
kvm_get_vcpu()/kvm_for_each_vcpu() already has access to that first vCPU.
This should not pose a problem assuming that once a vcpu is stored in
vcpu_array, it will remain there, but that's not the case:
kvm_vm_ioctl_create_vcpu() first inserts to vcpu_array, then requests a
file descriptor. If create_vcpu_fd() fails, newly inserted vcpu is removed
from the vcpu_array, then destroyed:
vcpu->vcpu_idx = atomic_read(&kvm->online_vcpus);
r = xa_insert(&kvm->vcpu_array, vcpu->vcpu_idx, vcpu, GFP_KERNEL_ACCOUNT);
kvm_get_kvm(kvm);
r = create_vcpu_fd(vcpu);
if (r < 0) {
xa_erase(&kvm->vcpu_array, vcpu->vcpu_idx);
kvm_put_kvm_no_destroy(kvm);
goto unlock_vcpu_destroy;
}
atomic_inc(&kvm->online_vcpus);
This results in a possible race condition when a reference to a vcpu is
acquired (via kvm_get_vcpu() or kvm_for_each_vcpu()) moments before said
vcpu is destroyed.
Signed-off-by: Michal Luczaj <mhal@rbox.co>
Message-Id: <20230510140410.1093987-2-mhal@rbox.co>
Cc: stable@vger.kernel.org
Fixes: c5b0775491 ("KVM: Convert the kvm->vcpus array to a xarray", 2021-12-08)
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
