clang doesn't like printing a 32-bit integer using %hX format string:
drivers/hid/i2c-hid/i2c-hid-core.c:994:18: error: format specifies type 'unsigned short' but the argument has type '__u32' (aka 'unsigned int') [-Werror,-Wformat]
client->name, hid->vendor, hid->product);
^~~~~~~~~~~
drivers/hid/i2c-hid/i2c-hid-core.c:994:31: error: format specifies type 'unsigned short' but the argument has type '__u32' (aka 'unsigned int') [-Werror,-Wformat]
client->name, hid->vendor, hid->product);
^~~~~~~~~~~~
Use an explicit cast to truncate it to the low 16 bits instead.
Fixes: 9ee3e06610 ("HID: i2c-hid: override HID descriptors for certain devices")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Replace kzalloc with devm_kzalloc in driver initialization sequence. The
allocation can be tied to the lifetime of the amd_sfh driver. This cleans
up an exit & error paths, since the objects does not need to be
explicitly freed anymore.
Fixes: 4b2c53d93a ("SFH:Transport Driver to add support of AMD Sensor Fusion Hub (SFH)")
Reviewed-by: Shyam Sundar S K <Shyam-sundar.S-k@amd.com>
Signed-off-by: Basavaraj Natikar <Basavaraj.Natikar@amd.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
The ft260_hid_feature_report_get() checks if the return size matches the
requested size. But the function can also fail with at least -ENOMEM. Add the
< 0 checks.
In ft260_hid_feature_report_get(), do not do the memcpy to the caller's buffer
if there is an error.
Fixes: 6a82582d9f ("HID: ft260: add usb hid to i2c host bridge driver")
Signed-off-by: Tom Rix <trix@redhat.com>
Signed-off-by: Michael Zaidman <michael.zaidman@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
When the Apple Magic Trackpad 2 is connected over USB it registers four
hid_device report descriptors, however, the driver only handles the one
with type HID_TYPE_USBMOUSE and ignores the other three, thus, no driver
data is attached to them.
When the device is disconnected, the remove callback is called for the
four hid_device report descriptors, crashing when the driver data is
NULL.
Check that the driver data is not NULL before using it in the remove
callback.
Signed-off-by: José Expósito <jose.exposito89@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
This patch adds missing MODULE_DEVICE_TABLE definition which generates
correct modalias for automatic loading of this driver when it is built
as an external module.
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Bixuan Cui <cuibixuan@huawei.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Fix to return a negative error code from the error handling
case instead of 0, as done elsewhere in this function.
Fixes: 224ee88fe3 ("Input: add force feedback driver for PID devices")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Static analysis reports this representative problem
hid-logitech-hidpp.c:1356:23: warning: Assigned value is
garbage or undefined
hidpp->battery.level = level;
^ ~~~~~
In some cases, 'level' is never set in hidpp20_battery_map_status_voltage()
Since level is not available on all hw, initialize level to unknown.
Fixes: be281368f2 ("hid-logitech-hidpp: read battery voltage from newer devices")
Signed-off-by: Tom Rix <trix@redhat.com>
Reviewed-by: Filipe Laíns <lains@riseup.net>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
The Asus T101HA has a problem with spurious wakeups when the lid is
closed, this is caused by the screen sitting so close to the touchpad
that the touchpad ends up reporting touch events, causing these wakeups.
Add a quirk which disables event reporting on suspend when set, and
enable this quirk for the Asus T101HA touchpad fixing the spurious
wakeups, while still allowing the device to be woken by pressing a
key on the keyboard (which is part of the same USB device).
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Normally the EXPORT_SYMBOL of a function immediately follows the
declaration of the function and all the other functions in hid-core.c
follow this pattern, drop the extraneous empty line before the
EXPORT_SYMBOL_GPL(hid_check_keys_pressed); line.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
- fix a memory leak in nvme_cdev_add (Guoqing Jiang)
- fix inline data size comparison in nvmet_tcp_queue_response (Hou Pu)
- fix false keep-alive timeout when a controller is torn down
(Sagi Grimberg)
- fix a nvme-tcp Kconfig dependency (Sagi Grimberg)
- short-circuit reconnect retries for FC (Hannes Reinecke)
- decode host pathing error for connect (Hannes Reinecke)
-----BEGIN PGP SIGNATURE-----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=JT0/
-----END PGP SIGNATURE-----
Merge tag 'nvme-5.13-2021-05-27' of git://git.infradead.org/nvme into block-5.13
Pull NVMe fixes from Christoph:
"nvme fixes for Linux 5.13
- fix a memory leak in nvme_cdev_add (Guoqing Jiang)
- fix inline data size comparison in nvmet_tcp_queue_response (Hou Pu)
- fix false keep-alive timeout when a controller is torn down
(Sagi Grimberg)
- fix a nvme-tcp Kconfig dependency (Sagi Grimberg)
- short-circuit reconnect retries for FC (Hannes Reinecke)
- decode host pathing error for connect (Hannes Reinecke)"
* tag 'nvme-5.13-2021-05-27' of git://git.infradead.org/nvme:
nvmet: fix false keep-alive timeout when a controller is torn down
nvmet-tcp: fix inline data size comparison in nvmet_tcp_queue_response
nvme-tcp: remove incorrect Kconfig dep in BLK_DEV_NVME
nvme-fabrics: decode host pathing error for connect
nvme-fc: short-circuit reconnect retries
nvme: fix potential memory leaks in nvme_cdev_add
In commit 8428413b1d ("serial: 8250_pci: Implement MSI(-X) support")
the way the irq gets allocated was changed. With that change the
handling FL_NOIRQ got lost. Restore the old behaviour.
Fixes: 8428413b1d ("serial: 8250_pci: Implement MSI(-X) support")
Cc: <stable@vger.kernel.org>
Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Link: https://lore.kernel.org/r/20210527095529.26281-1-christian.gmeiner@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Macros should not use a trailing semicolon.
Signed-off-by: Huilong Deng <denghuilong@cdjrlc.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
A rx flow control waiting in the control queue may block autosuspend.
Re-request autosuspend after flow control been sent to unblock
the transition to the low power state.
Cc: <stable@vger.kernel.org>
Signed-off-by: Alexander Usyskin <alexander.usyskin@intel.com>
Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Link: https://lore.kernel.org/r/20210526193334.445759-1-tomas.winkler@intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This comment was left over from a previous version of the patch that
introduced wrprot_gfn_range, when skip_4k was passed in instead of
min_level.
Signed-off-by: David Matlack <dmatlack@google.com>
Message-Id: <20210526163227.3113557-1-dmatlack@google.com>
Reviewed-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
If an error occurs after a successful 'kfifo_alloc()' call, it must be
undone by a corresponding 'kfifo_free()' call, as already done in the
remove function.
While at it, move the 'platform_device_put()' call to this new error
handling path and explicitly return 0 in the success path.
Fixes: b5dc75c915 ("firmware: stratix10-svc: extend svc to support new RSU features")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/0ca3f3ab139c53e846804455a1e7599ee8ae896a.1621621271.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Moving the call to ee1004_set_current_page() to ee1004_eeprom_read()
allows to simplify the code.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Link: https://lore.kernel.org/r/2829a131-51e3-8865-462a-564080158b0b@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The value of ee1004_current_page applies to all SPD eeproms connected
to the adapter. Therefore it's sufficient if we set ee1004_current_page
when the first device is added.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Link: https://lore.kernel.org/r/b9240e58-08bb-3d71-7a9c-9a323b470ab6@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
i2c_new_dummy_device() calls i2c_new_client_device() that complains
if it fails to create the device. Therefore we don't have to emit an
error message in case of failure. In addition ensure that
ee1004_set_page is only set if creating the device succeeded.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Link: https://lore.kernel.org/r/d38df5ac-6ecb-7d5f-b5c3-39bfc6a1e8a1@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
We have to read 512 bytes only, therefore read performance isn't really
a concern. Don't bother the user if i2c block read isn't supported.
For i2c_smbus_read_i2c_block_data_or_emulated() to work it's sufficient
if I2C_FUNC_SMBUS_READ_I2C_BLOCK or I2C_FUNC_SMBUS_READ_BYTE_DATA is
supported. Therefore remove the check for I2C_FUNC_SMBUS_READ_WORD_DATA.
In addition check for I2C_FUNC_SMBUS_WRITE_BYTE (included in
I2C_FUNC_SMBUS_BYTE) which is needed for setting the page.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Link: https://lore.kernel.org/r/840c668e-6310-e933-e50e-5abeaecfb39c@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
i2c_smbus_read_i2c_block_data_or_emulated() checks its length argument,
so we don't have to do it. In addition remove the unlikely hint from
the checks, we do i2c reads and therefore are in a slow path.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Link: https://lore.kernel.org/r/eb2a8bff-43ec-c763-a417-9d741e6f0034@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
sysfs_kf_bin_read() checks this for us already. In addition
the function works correctly also w/o this check.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Link: https://lore.kernel.org/r/33889bff-3614-4b73-5010-701635e1edab@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Simplify code and turn 'pvpanic_probe()' into a managed resource version.
This simplify callers that don't need to do some clean-up on error in the
probe and on remove.
Update pvpanic-mmio.c and pvpanic-pci.c accordingly.
'pvpanic_remove()' don't need to be exported anymore.
Suggested-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com>
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/9212cdc8c1e5c187a2f1129a6190085c2a10d28a.1621665058.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is no error handling path in the probe function.
Switch to managed resource so that errors in the probe are handled easily
and simplify the remove function accordingly.
Fixes: b3c0f87746 ("misc/pvpanic: probe multiple instances")
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/2a5dab18f10db783b27e0579ba66cc38d610734a.1621665058.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is no error handling path in the probe function.
Switch to managed resource so that errors in the probe are handled easily
and simplify the remove function accordingly.
Fixes: db3a4f0abe ("misc/pvpanic: add PCI driver")
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/ab071b1f4ed6e1174f9199095fb16a58bb406090.1621665058.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 9ed5af268e ("SUNRPC: Clean up the handling of page padding
in rpc_prepare_reply_pages()") [Dec 2020] affects RPC Replies that
have a data payload (i.e., Write chunks).
rpcrdma_prepare_readch(), as its name suggests, sets up Read chunks
which are data payloads within RPC Calls. Those payloads are
constructed by xdr_write_pages(), which continues to stuff the call
buffer's tail kvec with the payload's XDR roundup. Thus removing
the tail buffer logic in rpcrdma_prepare_readch() was the wrong
thing to do.
Fixes: 586a0787ce ("xprtrdma: Clean up rpcrdma_prepare_readch()")
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Since commit bdcc2cd14e ("NFSv4.2: handle NFS-specific llseek errors"),
nfs42_proc_llseek would return -EOPNOTSUPP rather than -ENOTSUPP when
SEEK_DATA on NFSv4.0/v4.1.
This will lead xfstests generic/285 not run on NFSv4.0/v4.1 when set the
CONFIG_NFS_V4_2, rather than run failed.
Fixes: bdcc2cd14e ("NFSv4.2: handle NFS-specific llseek errors")
Cc: <stable.vger.kernel.org> # 4.2
Signed-off-by: Zhang Xiaoxu <zhangxiaoxu5@huawei.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Theoretically, it will cause index out of bounds error if
'num_bytes_read' is greater than 4. As we expect it(and was tested)
never to be greater than 4, error out if it happens.
Fixes: c1986ee9be ("[PATCH] New Omnikey Cardman 4000 driver")
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Link: https://lore.kernel.org/r/20210521120617.138396-1-yukuai3@huawei.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If booted with verbose>=1, dyndbg prints the memory usage in bytes,
of builtin modules' prdebugs. KiB reads better.
no functional changes.
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Link: https://lore.kernel.org/r/20210525033240.35260-1-jim.cromie@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The opening comment mark '/**' is used for highlighting the beginning of
kernel-doc comments.
The header for drivers/ipack/carriers/tpci200 follows this syntax, but the
content inside does not comply with kernel-doc.
This line was probably not meant for kernel-doc parsing, but is parsed
due to the presence of kernel-doc like comment syntax(i.e, '/**'), which
causes unexpected warning from kernel-doc.
For e.g., running scripts/kernel-doc -none on
drivers/ipack/carriers/tpci200.c emits:
warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
* tpci200.c
Provide a simple fix by replacing this occurrence with general comment
format, i.e. '/*', to prevent kernel-doc from parsing it.
Also remove the redundant file name from the comment headers.
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Acked-by: Samuel Iglesias Gonsalvez <siglesias@igalia.com>
Signed-off-by: Aditya Srivastava <yashsri421@gmail.com>
Link: https://lore.kernel.org/r/20210522124051.12540-1-yashsri421@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The opening comment mark '/**' is used for highlighting the beginning of
kernel-doc comments.
The header for drivers/ipack/devices/ipoctal follows this syntax, but the
content inside does not comply with kernel-doc.
This line was probably not meant for kernel-doc parsing, but is parsed
due to the presence of kernel-doc like comment syntax(i.e, '/**'), which
causes unexpected warning from kernel-doc.
For e.g., running scripts/kernel-doc -none on
drivers/ipack/devices/ipoctal.h emits:
warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
* ipoctal.h
Provide a simple fix by replacing this occurrence with general comment
format, i.e. '/*', to prevent kernel-doc from parsing it.
Also remove the redundant file name from the comment headers.
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Acked-by: Samuel Iglesias Gonsalvez <siglesias@igalia.com>
Signed-off-by: Aditya Srivastava <yashsri421@gmail.com>
Link: https://lore.kernel.org/r/20210522121944.11182-1-yashsri421@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In the out_err_bus_register error branch of tpci200_pci_probe,
tpci200->info->cfg_regs is freed by tpci200_uninstall()->
tpci200_unregister()->pci_iounmap(..,tpci200->info->cfg_regs)
in the first time.
But later, iounmap() is called to free tpci200->info->cfg_regs
again.
My patch sets tpci200->info->cfg_regs to NULL after tpci200_uninstall()
to avoid the double free.
Fixes: cea2f7cdff ("Staging: ipack/bridges/tpci200: Use the TPCI200 in big endian mode")
Cc: stable <stable@vger.kernel.org>
Acked-by: Samuel Iglesias Gonsalvez <siglesias@igalia.com>
Signed-off-by: Lv Yunlong <lyl2019@mail.ustc.edu.cn>
Link: https://lore.kernel.org/r/20210524093205.8333-1-lyl2019@mail.ustc.edu.cn
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The XillyUSB driver is the USB variant for the Xillybus FPGA IP core.
Even though it presents a nearly identical API on the FPGA and host,
it's almost a complete rewrite of the driver: The framework for exchanging
data on a USB bus is fundamentally different from doing the same with a
PCIe interface, which leaves very little in common between the existing
driver and the new one for XillyUSB.
Signed-off-by: Eli Billauer <eli.billauer@gmail.com>
Link: https://lore.kernel.org/r/20210526100311.56327-3-eli.billauer@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch is a preparation for adding another related driver, XillyUSB.
In order to share some code between the existing Xillybus driver and the
one to be added, some functions are moved to xillybus_class.c
XILLYBUS_CLASS is added to Kconfig and is common to all drivers in this
group. The relation with the existing XILLYBUS symbol is "select" rather
than "depends on" XILLYBUS_CLASS, or else "make olddefconfig" will silently
turn off XILLYBUS, which is currently enabled in several distributions.
XILLYBUS_CLASS doesn't depend on anything else, hence using it with
"select" poses no risk for a broken configuration.
After the future addition of the XillyUSB module, the tree of symbols
will be as follows:
XILLYBUS_CLASS --+-- XILLYBUS --+-- XILLYBUS_PCIE
| |
| +-- XILLYBUS_OF
|
+-- XILLYUSB
XILLYBUS is for drivers based upon memory registers + DMA-based interfaces,
and it's combined with XILLYBUS_PCIE and/or XILLYBUS_OF.
XILLYUSB is for the USB variant only.
Or a more detailed, bottom-up outline:
* CONFIG_XILLYBUS_PCIE -> xillybus_pcie.c: Functions related to PCIe.
* CONFIG_XILLYBUS_OF -> xillybus_of.c: Functions related to Xillybus as a
peripheral on an FPGA / Processor combo chip.
* CONFIG_XILLYBUS -> xillybus_core.c: Functions that are common to the two
above, mainly access to the peripheral with memory-mapped registers and
DMA.
* CONFIG_XILLYUSB -> xillyusb.c: The driver for the USB variant, accesses
the peripheral through the USB framework.
* CONFIG_XILLYBUS_CLASS -> xillybus_class.c: The new module, which contains
the class and API parts that would otherwise appear both in
xillybus_core.c and xillyusb.c. Contains utility functions for the two
latter.
And since I'm at it, comments on the module names are added in the
Kconfig's help part.
The functions are exported with the non-GPL EXPORT_SYMBOL (a matter of
taste).
Signed-off-by: Eli Billauer <eli.billauer@gmail.com>
Link: https://lore.kernel.org/r/20210526100311.56327-2-eli.billauer@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
For VMX, when a vcpu enters HLT emulation, pi_post_block will:
1) Add vcpu to per-cpu list of blocked vcpus.
2) Program the posted-interrupt descriptor "notification vector"
to POSTED_INTR_WAKEUP_VECTOR
With interrupt remapping, an interrupt will set the PIR bit for the
vector programmed for the device on the CPU, test-and-set the
ON bit on the posted interrupt descriptor, and if the ON bit is clear
generate an interrupt for the notification vector.
This way, the target CPU wakes upon a device interrupt and wakes up
the target vcpu.
Problem is that pi_post_block only programs the notification vector
if kvm_arch_has_assigned_device() is true. Its possible for the
following to happen:
1) vcpu V HLTs on pcpu P, kvm_arch_has_assigned_device is false,
notification vector is not programmed
2) device is assigned to VM
3) device interrupts vcpu V, sets ON bit
(notification vector not programmed, so pcpu P remains in idle)
4) vcpu 0 IPIs vcpu V (in guest), but since pi descriptor ON bit is set,
kvm_vcpu_kick is skipped
5) vcpu 0 busy spins on vcpu V's response for several seconds, until
RCU watchdog NMIs all vCPUs.
To fix this, use the start_assignment kvm_x86_ops callback to kick
vcpus out of the halt loop, so the notification vector is
properly reprogrammed to the wakeup vector.
Reported-by: Pei Zhang <pezhang@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Message-Id: <20210526172014.GA29007@fuller.cnet>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
KVM_REQ_UNBLOCK will be used to exit a vcpu from
its inner vcpu halt emulation loop.
Rename KVM_REQ_PENDING_TIMER to KVM_REQ_UNBLOCK, switch
PowerPC to arch specific request bit.
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Message-Id: <20210525134321.303768132@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Add a start_assignment hook to kvm_x86_ops, which is called when
kvm_arch_start_assignment is done.
The hook is required to update the wakeup vector of a sleeping vCPU
when a device is assigned to the guest.
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Message-Id: <20210525134321.254128742@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Let's treat lapic_timer_advance_ns automatic tuning logic as hypervisor
overhead, move it before wait_lapic_expire instead of between wait_lapic_expire
and the world switch, the wait duration should be calculated by the
up-to-date guest_tsc after the overhead of automatic tuning logic. This
patch reduces ~30+ cycles for kvm-unit-tests/tscdeadline-latency when testing
busy waits.
Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
Message-Id: <1621339235-11131-5-git-send-email-wanpengli@tencent.com>
Reviewed-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
