There is no reason for stopping all TX queues from dev_watchdog()
Not only this stops feeding the NIC, it also migrates all qdiscs
to be serviced on the cpu calling netif_tx_unlock(), causing
a potential latency artifact.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
These are not fast path, there is no point in inlining them.
Also provide netif_freeze_queues()/netif_unfreeze_queues()
so that we can use them from dev_watchdog() in the following patch.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In following patches, dev_watchdog() will no longer stop all queues.
It will read queue->trans_start locklessly.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
tx_timeout_show() assumed dev_watchdog() would stop all
the queues, to fetch queue->trans_timeout under protection
of the queue->_xmit_lock.
As we want to no longer disrupt transmits, we use an
atomic_long_t instead.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: david decotigny <david.decotigny@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Change the values of EVENT_* constants from bit masks to bit numbers as
accepted by {clear,set,test}_bit() functions.
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Łukasz Stelmach <l.stelmach@samsung.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
virtio_net_hdr_to_skb does not set the skb's gso_size and gso_type
correctly for UFO packets received via virtio-net that are a little over
the GSO size. This can lead to problems elsewhere in the networking
stack, e.g. ovs_vport_send dropping over-sized packets if gso_size is
not set.
This is due to the comparison
if (skb->len - p_off > gso_size)
not properly accounting for the transport layer header.
p_off includes the size of the transport layer header (thlen), so
skb->len - p_off is the size of the TCP/UDP payload.
gso_size is read from the virtio-net header. For UFO, fragmentation
happens at the IP level so does not need to include the UDP header.
Hence the calculation could be comparing a TCP/UDP payload length with
an IP payload length, causing legitimate virtio-net packets to have
lack gso_type/gso_size information.
Example: a UDP packet with payload size 1473 has IP payload size 1481.
If the guest used UFO, it is not fragmented and the virtio-net header's
flags indicate that it is a GSO frame (VIRTIO_NET_HDR_GSO_UDP), with
gso_size = 1480 for an MTU of 1500. skb->len will be 1515 and p_off
will be 42, so skb->len - p_off = 1473. Hence the comparison fails, and
shinfo->gso_size and gso_type are not set as they should be.
Instead, add the UDP header length before comparing to gso_size when
using UFO. In this way, it is the size of the IP payload that is
compared to gso_size.
Fixes: 6dd912f826 ("net: check untrusted gso_size at kernel entry")
Signed-off-by: Jonathan Davies <jonathan.davies@nutanix.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Access to netdev after free_netdev() will cause use-after-free bug.
Move debug log before free_netdev() call to avoid it.
Fixes: 7472dd9f64 ("staging: fsl-dpaa2/eth: Move print message")
Signed-off-by: Pavel Skripkin <paskripkin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch will surround the AF_INET6 case in sk_error_report() of dlm
with a #if IS_ENABLED(CONFIG_IPV6). The field sk->sk_v6_daddr is not
defined when CONFIG_IPV6 is disabled. If CONFIG_IPV6 is disabled, the
socket creation with AF_INET6 should already fail because a runtime
check if AF_INET6 is registered. However if there is the possibility
that AF_INET6 is set as sk_family the sk_error_report() callback will
print then an invalid family type error.
Reported-by: kernel test robot <lkp@intel.com>
Fixes: 4c3d90570b ("fs: dlm: don't call kernel_getpeername() in error_report()")
Signed-off-by: Alexander Aring <aahringo@redhat.com>
Signed-off-by: David Teigland <teigland@redhat.com>
Like ThinkaPad Thunderbolt 4 Dock, more Lenovo docks start to use the original
Realtek USB ethernet chip ID 0bda:8153.
Lenovo Docks always use their own IDs for usb hub, even for older Docks.
If parent hub is from Lenovo, then r8152 should try MAC passthrough.
Verified on Lenovo TBT3 dock too.
Signed-off-by: Aaron Ma <aaron.ma@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
elevator_init_mq() is only called before adding disk, when there isn't
any FS I/O, only passthrough requests can be queued, so freezing queue
plus canceling dispatch work is enough to drain any dispatch activities,
then we can avoid synchronize_srcu() in blk_mq_quiesce_queue().
Long boot latency issue can be fixed in case of lots of disks added
during booting.
Fixes: 737eb78e82 ("block: Delay default elevator initialization")
Reported-by: yangerkun <yangerkun@huawei.com>
Cc: Damien Le Moal <damien.lemoal@wdc.com>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20211117115502.1600950-1-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
The arch_prepare_bpf_dispatcher function does not have a prototype, and
yields the following warning when W=1 is enabled for the kernel build.
>> arch/x86/net/bpf_jit_comp.c:2188:5: warning: no previous \
prototype for 'arch_prepare_bpf_dispatcher' [-Wmissing-prototypes]
2188 | int arch_prepare_bpf_dispatcher(void *image, s64 *funcs, \
int num_funcs)
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~
Remove the warning by adding a function declaration to include/linux/bpf.h.
Fixes: 75ccbef636 ("bpf: Introduce BPF dispatcher")
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Björn Töpel <bjorn@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20211117125708.769168-1-bjorn@kernel.org
The SGX documentation has a few repeated or one-off issues:
* Remove capitalization from regular words in the middle of a sentence.
* Remove punctuation found in the middle of a sentence.
* Fix name of SGX daemon to consistently be ksgxd.
* Fix typo of SGX instruction: ENIT -> EINIT
[ dhansen: tweaked subject and changelog ]
Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Jarkko Sakkinen <jarkko@kernel.org>
Link: https://lkml.kernel.org/r/ab99a87368eef69e3fb96f073368becff3eff874.1635529506.git.reinette.chatre@intel.com
If the used PHYs also support hardware timestamping, all configuration requests
should be forwared to the PHYs instead of being processed by the MAC driver
itself.
This enables PHY timestamping in combination with the cpsw driver.
Tested with an am335x based board with two DP83640 PHYs connected to the cpsw
switch.
Signed-off-by: Kurt Kanzenbach <kurt@linutronix.de>
Acked-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Get rid of cpu_missing because
7bb39313cd ("x86/mce: Make mce_timed_out() identify holdout CPUs")
provides a more detailed message about which CPUs are missing.
Suggested-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Zhaolong Zhang <zhangzl2013@126.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20211109112345.2673403-1-zhangzl2013@126.com
If the first call to devm_usb_get_phy_by_phandle(dev, "fsl,usbphy", 0)
fails with something other than -ENODEV then it leads to an error
pointer dereference. For those errors we should just jump directly to
the error handling.
Fixes: 8253a34bfa ("usb: chipidea: ci_hdrc_imx: Also search for 'phys' phandle")
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Link: https://lore.kernel.org/r/20211117074923.GF5237@kili
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
xHC hardware can only have one slot in default state with address 0
waiting for a unique address at a time, otherwise "undefined behavior
may occur" according to xhci spec 5.4.3.4
The address0_mutex exists to prevent this across both xhci roothubs.
If hub_port_init() fails, it may unlock the mutex and exit with a xhci
slot in default state. If the other xhci roothub calls hub_port_init()
at this point we end up with two slots in default state.
Make sure the address0_mutex protects the slot default state across
hub_port_init() retries, until slot is addressed or disabled.
Note, one known minor case is not fixed by this patch.
If device needs to be reset during resume, but fails all hub_port_init()
retries in usb_reset_and_verify_device(), then it's possible the slot is
still left in default state when address0_mutex is unlocked.
Cc: <stable@vger.kernel.org>
Fixes: 638139eb95 ("usb: hub: allow to process more usb hub events in parallel")
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20211115221630.871204-1-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The code that enables either BC_LVL or COMP_CHNG interrupt in tcpm_set_cc
wrongly assumes that the interrupt is unmasked by writing 1 to the apropriate
bit in the mask register. In fact, interrupts are enabled when the mask
is 0, so the tcpm_set_cc enables interrupt for COMP_CHNG when it expects
BC_LVL interrupt to be enabled.
This causes inability of the driver to recognize cable unplug events
in host mode (unplug is recognized only via a COMP_CHNG interrupt).
In device mode this bug was masked by simultaneous triggering of the VBUS
change interrupt, because of loss of VBUS when the port peer is providing
power.
Fixes: 48242e3053 ("usb: typec: fusb302: Revert "Resolve fixed power role contract setup"")
Cc: stable <stable@vger.kernel.org>
Cc: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Signed-off-by: Ondrej Jirman <megous@megous.com>
Link: https://lore.kernel.org/r/20211108102833.2793803-1-megous@megous.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
in case of a PCI dwc3 controller, leave the default DMA
mask. Calling of a 64 bit DMA mask breaks the driver on
cherrytrail based tablets like Cyberbook T116.
Fixes: 45d39448b4 ("usb: dwc3: support 64 bit DMA in platform driver")
Cc: stable <stable@vger.kernel.org>
Reported-by: Hans De Goede <hdegoede@redhat.com>
Tested-by: Fabio Aiuto <fabioaiuto83@gmail.com>
Tested-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Fabio Aiuto <fabioaiuto83@gmail.com>
Link: https://lore.kernel.org/r/20211113142959.27191-1-fabioaiuto83@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
A new commit in LLVM causes an error on the use of 'long double' when
'-mno-x87' is used, which the kernel does through an alias,
'-mno-80387' (see the LLVM commit below for more details around why it
does this).
drivers/usb/dwc2/hcd_queue.c:1744:25: error: expression requires 'long double' type support, but target 'x86_64-unknown-linux-gnu' does not support it
delay = ktime_set(0, DWC2_RETRY_WAIT_DELAY);
^
drivers/usb/dwc2/hcd_queue.c:62:34: note: expanded from macro 'DWC2_RETRY_WAIT_DELAY'
#define DWC2_RETRY_WAIT_DELAY (1 * 1E6L)
^
1 error generated.
This happens due to the use of a 'long double' literal. The 'E6' part of
'1E6L' causes the literal to be a 'double' then the 'L' suffix promotes
it to 'long double'.
There is no visible reason for a floating point value in this driver, as
the value is only used as a parameter to a function that expects an
integer type. Use NSEC_PER_MSEC, which is the same integer value as
'1E6L', to avoid changing functionality but fix the error.
Link: https://github.com/ClangBuiltLinux/linux/issues/1497
Link: a8083d42b1
Fixes: 6ed30a7d8e ("usb: dwc2: host: use hrtimer for NAK retries")
Cc: stable <stable@vger.kernel.org>
Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
Reviewed-by: John Keeping <john@metanate.com>
Acked-by: Minas Harutyunyan <Minas.Harutyunyan@synopsys.com>
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Link: https://lore.kernel.org/r/20211105145802.2520658-1-nathan@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In the endpoint interrupt functions
dwc3_gadget_endpoint_transfer_in_progress() and
dwc3_gadget_endpoint_trbs_complete() will dereference the endpoint
descriptor. But it could be cleared in __dwc3_gadget_ep_disable()
when accessory disconnected. So we need to check whether it is null
or not before dereferencing it.
Fixes: f09ddcfcb8 ("usb: dwc3: gadget: Prevent EP queuing while stopping transfers")
Cc: stable <stable@vger.kernel.org>
Reviewed-by: Jack Pham <quic_jackp@quicinc.com>
Signed-off-by: Albert Wang <albertccwang@google.com>
Link: https://lore.kernel.org/r/20211109092642.3507692-1-albertccwang@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
A successful 'clk_prepare_enable()' call should be balanced by a
corresponding 'clk_disable_unprepare()' call in the error handling path
of the probe, as already done in the remove function.
Fixes: 2474922921 ("usb: gadget: udc-xilinx: Add clock support")
Reviewed-by: Shubhrajyoti Datta <shubhraj@xilinx.com>
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/ec61a89b83ce34b53a3bdaacfd1413a9869cc608.1636302246.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Added updating of request frame number for elapsed frames,
otherwise frame number will remain as previous use of request.
This will allow function driver to correctly track frames in
case of Missed ISOC occurs.
Added setting request actual length to 0 for elapsed frames.
In Slave mode when pushing data to RxFIFO by dwords, request
actual length incrementing accordingly. But before whole packet
will be pushed into RxFIFO and send to host can occurs Missed
ISOC and data will not send to host. So, in this case request
actual length should be reset to 0.
Fixes: 91bb163e1e ("usb: dwc2: gadget: Fix ISOC flow for BDMA and Slave")
Cc: stable <stable@vger.kernel.org>
Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Minas Harutyunyan <Minas.Harutyunyan@synopsys.com>
Link: https://lore.kernel.org/r/c356baade6e9716d312d43df08d53ae557cb8037.1636011277.git.Minas.Harutyunyan@synopsys.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The programming guide noted that the driver needs to verify if the link
state is in U0 before executing the Start Transfer command. If it's not
in U0, the driver needs to perform remote wakeup. This is not accurate.
If the link state is in U1/U2, then the controller will not respond to
link recovery request from DCTL.ULSTCHNGREQ. The Start Transfer command
will trigger a link recovery if it is in U1/U2. A clarification will be
added to the programming guide for all controller versions.
The current implementation shouldn't cause any functional issue. It may
occasionally report an invalid time out warning from failed link
recovery request. The driver will still go ahead with the Start Transfer
command if the remote wakeup fails. The new change only initiates remote
wakeup where it is needed, which is when the link state is in L1/L2/U3.
Fixes: c36d8e947a ("usb: dwc3: gadget: put link to U0 before Start Transfer")
Cc: <stable@vger.kernel.org>
Signed-off-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
Link: https://lore.kernel.org/r/05b4a5fbfbd0863fc9b1d7af934a366219e3d0b4.1635204761.git.Thinh.Nguyen@synopsys.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The End Transfer command from a stream endpoint will generate a NoStream
event, and we should ignore it. Currently we set the flag
DWC3_EP_IGNORE_NEXT_NOSTREAM to track this prior to sending the command,
and it will be cleared on the next stream event. However, a stream event
may be generated before the End Transfer command completion and
prematurely clear the flag. Fix this by setting the flag on End Transfer
completion instead.
Fixes: 140ca4cfea ("usb: dwc3: gadget: Handle stream transfers")
Cc: <stable@vger.kernel.org>
Signed-off-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
Link: https://lore.kernel.org/r/cee1253af4c3600edb878d11c9c08b040817ae23.1635203975.git.Thinh.Nguyen@synopsys.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
During our predesign phase for DWC_usb32, the GHWPARAMS9 register offset
was 0xc680. We revised our final design, and the GHWPARAMS9 offset is
now moved to 0xc6e8 on release.
Fixes: 16710380d3 ("usb: dwc3: Capture new capability register GHWPARAMS9")
Cc: <stable@vger.kernel.org>
Signed-off-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
Link: https://lore.kernel.org/r/1541737108266a97208ff827805be1f32852590c.1635202893.git.Thinh.Nguyen@synopsys.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Update net_failover documentation with missing and incomplete
details to get a proper working setup.
Signed-off-by: Vasudev Kamath <vasudev@copyninja.info>
Reviewed-by: Krishna Kumar <krikku@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Type-c class provides an option to expose the number of alternate
modes in the port partner as a sysfs interface. Use that to
expose the number of alternate modes in the port partner from
the ucsi driver.
Reviewed-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Signed-off-by: Rajaram Regupathy <rajaram.regupathy@intel.com>
Signed-off-by: Saranya Gopal <saranya.gopal@intel.com>
Link: https://lore.kernel.org/r/20211116103613.31349-1-saranya.gopal@intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Some platforms, e.g. 8183/8192, use low level latch way to keep
wakeup signal, it may latch a wrong signal if debounce more time,
and enable wakeup function earlier.
____________________
ip_sleep ____/ \_____________
___________________
wakeup_signal ____________/ \______
_______________________________
wakeup_en _______/
^ ^
|(1) |(2)
latch wakeup_signal mistakenly at (1), should latch it at (2);
Workaround: delay about 100us to enable wakeup, meanwhile decrease
debounce time.
Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com>
Link: https://lore.kernel.org/r/20211102025004.29156-2-chunfeng.yun@mediatek.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Drop perf's stubs for (un)registering guest callbacks now that KVM
registration of callbacks is hidden behind GUEST_PERF_EVENTS=y. The only
other user is x86 XEN_PV, and x86 unconditionally selects PERF_EVENTS.
No functional change intended.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Link: https://lore.kernel.org/r/20211111020738.2512932-18-seanjc@google.com
Move the definition of kvm_arm_pmu_available to pmu-emul.c and, out of
"necessity", hide it behind CONFIG_HW_PERF_EVENTS. Provide a stub for
the key's wrapper, kvm_arm_support_pmu_v3(). Moving the key's definition
out of perf.c will allow a future commit to delete perf.c entirely.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20211111020738.2512932-16-seanjc@google.com
Drop arm64's version of the callbacks in favor of the callbacks provided
by generic KVM, which are semantically identical.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20211111020738.2512932-15-seanjc@google.com
Now that all state needed for VMX's PT interrupt handler is exposed to
vmx.c (specifically the currently running vCPU), move the handler into
vmx.c where it belongs.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20211111020738.2512932-14-seanjc@google.com
Move x86's perf guest callbacks into common KVM, as they are semantically
identical to arm64's callbacks (the only other such KVM callbacks).
arm64 will convert to the common versions in a future patch.
Implement the necessary arm64 arch hooks now to avoid having to provide
stubs or a temporary #define (from x86) to avoid arm64 compilation errors
when CONFIG_GUEST_PERF_EVENTS=y.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Acked-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20211111020738.2512932-13-seanjc@google.com
Differentiate between IRQ and NMI for KVM's PMC overflow callback, which
was originally invoked in response to an NMI that arrived while the guest
was running, but was inadvertantly changed to fire on IRQs as well when
support for perf without PMU/NMI was added to KVM. In practice, this
should be a nop as the PMC overflow callback shouldn't be reached, but
it's a cheap and easy fix that also better documents the situation.
Note, this also doesn't completely prevent false positives if perf
somehow ends up calling into KVM, e.g. an NMI can arrive in host after
KVM sets its flag.
Fixes: dd60d21706 ("KVM: x86: Fix perf timer mode IP reporting")
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Link: https://lore.kernel.org/r/20211111020738.2512932-12-seanjc@google.com
Use the generic kvm_running_vcpu plus a new 'handling_intr_from_guest'
variable in kvm_arch_vcpu instead of the semi-redundant current_vcpu.
kvm_before/after_interrupt() must be called while the vCPU is loaded,
(which protects against preemption), thus kvm_running_vcpu is guaranteed
to be non-NULL when handling_intr_from_guest is non-zero.
Switching to kvm_get_running_vcpu() will allows moving KVM's perf
callbacks to generic code, and the new flag will be used in a future
patch to more precisely identify the "NMI from guest" case.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Link: https://lore.kernel.org/r/20211111020738.2512932-11-seanjc@google.com
Use static_call to optimize perf's guest callbacks on arm64 and x86,
which are now the only architectures that define the callbacks. Use
DEFINE_STATIC_CALL_RET0 as the default/NULL for all guest callbacks, as
the callback semantics are that a return value '0' means "not in guest".
static_call obviously avoids the overhead of CONFIG_RETPOLINE=y, but is
also advantageous versus other solutions, e.g. per-cpu callbacks, in that
a per-cpu memory load is not needed to detect the !guest case.
Based on code from Peter and Like.
Suggested-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Link: https://lore.kernel.org/r/20211111020738.2512932-10-seanjc@google.com
Introduce GUEST_PERF_EVENTS and require architectures to select it to
allow registering and using guest callbacks in perf. This will hopefully
make it more difficult for new architectures to add useless "support" for
guest callbacks, e.g. via copy+paste.
Stubbing out the helpers has the happy bonus of avoiding a load of
perf_guest_cbs when GUEST_PERF_EVENTS=n on arm64/x86.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Link: https://lore.kernel.org/r/20211111020738.2512932-9-seanjc@google.com
Add helpers for the guest callbacks to prepare for burying the callbacks
behind a Kconfig (it's a lot easier to provide a few stubs than to #ifdef
piles of code), and also to prepare for converting the callbacks to
static_call(). perf_instruction_pointer() in particular will have subtle
semantics with static_call(), as the "no callbacks" case will return 0 if
the callbacks are unregistered between querying guest state and getting
the IP. Implement the change now to avoid a functional change when adding
static_call() support, and because the new helper needs to return
_something_ in this case.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Link: https://lore.kernel.org/r/20211111020738.2512932-8-seanjc@google.com
To prepare for using static_calls to optimize perf's guest callbacks,
replace ->is_in_guest and ->is_user_mode with a new multiplexed hook
->state, tweak ->handle_intel_pt_intr to play nice with being called when
there is no active guest, and drop "guest" from ->get_guest_ip.
Return '0' from ->state and ->handle_intel_pt_intr to indicate "not in
guest" so that DEFINE_STATIC_CALL_RET0 can be used to define the static
calls, i.e. no callback == !guest.
[sean: extracted from static_call patch, fixed get_ip() bug, wrote changelog]
Suggested-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Originally-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Like Xu <like.xu@linux.intel.com>
Signed-off-by: Zhu Lingshan <lingshan.zhu@intel.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Link: https://lore.kernel.org/r/20211111020738.2512932-7-seanjc@google.com
Drop "support" for guest callbacks from architectures that don't implement
the guest callbacks. Future patches will convert the callbacks to
static_call; rather than churn a bunch of arch code (that was presumably
copy+pasted from x86), remove it wholesale as it's useless and at best
wasting cycles.
A future patch will also add a Kconfig to force architcture to opt into
the callbacks to make it more difficult for uses "support" to sneak in in
the future.
No functional change intended.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Link: https://lore.kernel.org/r/20211111020738.2512932-6-seanjc@google.com
Drop the 'int' return value from the perf (un)register callbacks helpers
and stop pretending perf can support multiple callbacks. The 'int'
returns are not future proofing anything as none of the callers take
action on an error. It's also not obvious that there will ever be
co-tenant hypervisors, and if there are, that allowing multiple callbacks
to be registered is desirable or even correct.
Opportunistically rename callbacks=>cbs in the affected declarations to
match their definitions.
No functional change intended.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Link: https://lore.kernel.org/r/20211111020738.2512932-5-seanjc@google.com
Override the Processor Trace (PT) interrupt handler for guest mode if and
only if PT is configured for host+guest mode, i.e. is being used
independently by both host and guest. If PT is configured for system
mode, the host fully controls PT and must handle all events.
Fixes: 8479e04e7d ("KVM: x86: Inject PMI for KVM guest")
Reported-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Reported-by: Artem Kashkanov <artem.kashkanov@intel.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20211111020738.2512932-4-seanjc@google.com
Wait to register perf callbacks until after doing vendor hardaware setup.
VMX's hardware_setup() configures Intel Processor Trace (PT) mode, and a
future fix to register the Intel PT guest interrupt hook if and only if
Intel PT is exposed to the guest will consume the configured PT mode.
Delaying registration to hardware setup is effectively a nop as KVM's perf
hooks all pivot on the per-CPU current_vcpu, which is non-NULL only when
KVM is handling an IRQ/NMI in a VM-Exit path. I.e. current_vcpu will be
NULL throughout both kvm_arch_init() and kvm_arch_hardware_setup().
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20211111020738.2512932-3-seanjc@google.com
