The CLDEMOTE instruction hints to hardware that the cache line that
contains the linear address should be moved("demoted") from
the cache(s) closest to the processor core to a level more distant
from the processor core. This may accelerate subsequent accesses
to the line by other cores in the same coherence domain,
especially if the line was written by the core that demotes the line.
This patch exposes the cldemote feature to the guest.
The release document ref below link:
https://software.intel.com/sites/default/files/managed/c5/15/\
architecture-instruction-set-extensions-programming-reference.pdf
This patch has a dependency on https://lkml.org/lkml/2018/4/23/928
Signed-off-by: Jingqi Liu <jingqi.liu@intel.com>
Reviewed-by: Wei Wang <wei.w.wang@intel.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
When vmcs12 uses VPID, all TLB entries populated by L2 are tagged with
vmx->nested.vpid02. Currently, INVVPID executed by L1 is emulated by L0
by using INVVPID single/global-context to flush all TLB entries
tagged with vmx->nested.vpid02 regardless of INVVPID type executed by
L1.
However, we can easily optimize the case of L1 INVVPID on an
individual-address. Just INVVPID given individual-address tagged with
vmx->nested.vpid02.
Reviewed-by: Liam Merwick <liam.merwick@oracle.com>
Signed-off-by: Liran Alon <liran.alon@oracle.com>
Reviewed-by: Jim Mattson <jmattson@google.com>
[Squashed with a preparatory patch that added the !operand.vpid line.]
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Since commit 5c614b3583 ("KVM: nVMX: nested VPID emulation"),
vmcs01 and vmcs02 don't share the same VPID. vmcs01 uses vmx->vpid
while vmcs02 uses vmx->nested.vpid02. This was done such that TLB
flush could be avoided when switching between L1 and L2.
However, the above mentioned commit only changed L2 VMEntry logic to
not flush TLB when switching from L1 to L2. It forgot to also remove
the TLB flush which is done when simulating a VMExit from L2 to L1.
To fix this issue, on VMExit from L2 to L1 we flush TLB only in case
vmcs01 enables VPID and vmcs01->vpid==vmcs02->vpid. This happens when
vmcs01 enables VPID and vmcs12 does not.
Fixes: 5c614b3583 ("KVM: nVMX: nested VPID emulation")
Reviewed-by: Liam Merwick <liam.merwick@oracle.com>
Reviewed-by: Krish Sadhukhan <krish.sadhukhan@oracle.com>
Signed-off-by: Liran Alon <liran.alon@oracle.com>
Reviewed-by: Jim Mattson <jmattson@google.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
This is a fix from reviewing the code, but it looks like it might be
able to lead to an Oops. It affects 32bit systems.
The KVM_MEMORY_ENCRYPT_REG_REGION ioctl uses a u64 for range->addr and
range->size but the high 32 bits would be truncated away on a 32 bit
system. This is harmless but it's also harmless to prevent it.
Then in sev_pin_memory() the "uaddr + ulen" calculation can wrap around.
The wrap around can happen on 32 bit or 64 bit systems, but I was only
able to figure out a problem for 32 bit systems. We would pick a number
which results in "npages" being zero. The sev_pin_memory() would then
return ZERO_SIZE_PTR without allocating anything.
I made it illegal to call sev_pin_memory() with "ulen" set to zero.
Hopefully, that doesn't cause any problems. I also changed the type of
"first" and "last" to long, just for cosmetic reasons. Otherwise on a
64 bit system you're saving "uaddr >> 12" in an int and it truncates the
high 20 bits away. The math works in the current code so far as I can
see but it's just weird.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
[Brijesh noted that the code is only reachable on X86_64.]
Reviewed-by: Brijesh Singh <brijesh.singh@amd.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
handle_mmio_page_fault() was recently moved to be an internal-only
MMU function, i.e. it's static and no longer defined in kvm_host.h.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Enforce the invariant that existing VMCS12 field offsets must not
change. Experience has shown that without strict enforcement, this
invariant will not be maintained.
Signed-off-by: Jim Mattson <jmattson@google.com>
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
[Changed the code to use BUILD_BUG_ON_MSG instead of better, but GCC 4.6
requiring _Static_assert. - Radim.]
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Changing the VMCS12 layout will break save/restore compatibility with
older kvm releases once the KVM_{GET,SET}_NESTED_STATE ioctls are
accepted upstream. Google has already been using these ioctls for some
time, and we implore the community not to disturb the existing layout.
Move the four most recently added fields to preserve the offsets of
the previously defined fields and reserve locations for the vmread and
vmwrite bitmaps, which will be used in the virtualization of VMCS
shadowing (to improve the performance of double-nesting).
Signed-off-by: Jim Mattson <jmattson@google.com>
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
[Kept the SDM order in vmcs_field_to_offset_table. - Radim]
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
The hypercall was added using a struct timespec based implementation,
but we should not use timespec in new code.
This changes it to timespec64. There is no functional change
here since the implementation is only used in 64-bit kernels
that use the same definition for timespec and timespec64.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
When saving a vCPU's nested state, the vmcs02 is discarded. Only the
shadow vmcs12 is saved. The shadow vmcs12 contains all of the
information needed to reconstruct an equivalent vmcs02 on restore, but
we have to be able to deal with two contexts:
1. The nested state was saved immediately after an emulated VM-entry,
before the vmcs02 was ever launched.
2. The nested state was saved some time after the first successful
launch of the vmcs02.
Though it's an implementation detail rather than an architected bit,
vmx->nested_run_pending serves to distinguish between these two
cases. Hence, we save it as part of the vCPU's nested state. (Yes,
this is ugly.)
Even when restoring from a checkpoint, it may be necessary to build
the vmcs02 as if prepare_vmcs02 was called from nested_vmx_run. So,
the 'from_vmentry' argument should be dropped, and
vmx->nested_run_pending should be consulted instead. The nested state
restoration code then has to set vmx->nested_run_pending prior to
calling prepare_vmcs02. It's important that the restoration code set
vmx->nested_run_pending anyway, since the flag impacts things like
interrupt delivery as well.
Fixes: cf8b84f48a ("kvm: nVMX: Prepare for checkpointing L2 state")
Signed-off-by: Jim Mattson <jmattson@google.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
These private pages have special purposes in the virtualization of L1,
but not in the virtualization of L2. In particular, L1's APIC access
page should never be entered into L2's page tables, because this
causes a great deal of confusion when the APIC virtualization hardware
is being used to accelerate L2's accesses to its own APIC.
Signed-off-by: Jim Mattson <jmattson@google.com>
Signed-off-by: Krish Sadhukhan <krish.sadhukhan@oracle.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
L1 and L2 need to have disjoint mappings, so that L1's APIC access
page (under VMX) can be omitted from L2's mappings.
Signed-off-by: Jim Mattson <jmattson@google.com>
Signed-off-by: Krish Sadhukhan <krish.sadhukhan@oracle.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
It is only possible to share the APIC access page between L1 and L2 if
they also share the virtual-APIC page. If L2 has its own virtual-APIC
page, then MMIO accesses to L1's TPR from L2 will access L2's TPR
instead. Moreover, L1's local APIC has to be in xAPIC mode, which is
another condition that hasn't been checked.
Signed-off-by: Jim Mattson <jmattson@google.com>
Signed-off-by: Krish Sadhukhan <krish.sadhukhan@oracle.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Previously, we toggled between SECONDARY_EXEC_VIRTUALIZE_X2APIC_MODE
and SECONDARY_EXEC_VIRTUALIZE_APIC_ACCESSES, depending on whether or
not the EXTD bit was set in MSR_IA32_APICBASE. However, if the local
APIC is disabled, we should not set either of these APIC
virtualization control bits.
Signed-off-by: Jim Mattson <jmattson@google.com>
Signed-off-by: Krish Sadhukhan <krish.sadhukhan@oracle.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The local APIC can be in one of three modes: disabled, xAPIC or
x2APIC. (A fourth mode, "invalid," is included for completeness.)
Using the new enumeration can make some of the APIC mode logic easier
to read. In kvm_set_apic_base, for instance, it is clear that one
cannot transition directly from x2APIC mode to xAPIC mode or directly
from APIC disabled to x2APIC mode.
Signed-off-by: Jim Mattson <jmattson@google.com>
Signed-off-by: Krish Sadhukhan <krish.sadhukhan@oracle.com>
[Check invalid bits even if msr_info->host_initiated. Reported by
Wanpeng Li. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Enlightened MSR-Bitmap is a natural extension of Enlightened VMCS:
Hyper-V Top Level Functional Specification states:
"The L1 hypervisor may collaborate with the L0 hypervisor to make MSR
accesses more efficient. It can enable enlightened MSR bitmaps by setting
the corresponding field in the enlightened VMCS to 1. When enabled, the L0
hypervisor does not monitor the MSR bitmaps for changes. Instead, the L1
hypervisor must invalidate the corresponding clean field after making
changes to one of the MSR bitmaps."
I reached out to Hyper-V team for additional details and I got the
following information:
"Current Hyper-V implementation works as following:
If the enlightened MSR bitmap is not enabled:
- All MSR accesses of L2 guests cause physical VM-Exits
If the enlightened MSR bitmap is enabled:
- Physical VM-Exits for L2 accesses to certain MSRs (currently FS_BASE,
GS_BASE and KERNEL_GS_BASE) are avoided, thus making these MSR accesses
faster."
I tested my series with a tight rdmsrl loop in L2, for KERNEL_GS_BASE the
results are:
Without Enlightened MSR-Bitmap: 1300 cycles/read
With Enlightened MSR-Bitmap: 120 cycles/read
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Lan Tianyu <Tianyu.Lan@microsoft.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Extract the logic to free a root page in a separate function to avoid code
duplication in mmu_free_roots(). Also, change it to an exported function
i.e. kvm_mmu_free_roots().
Signed-off-by: Junaid Shahid <junaids@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
MSB of CR3 is a reserved bit if the PCIDE bit is not set in CR4.
It should be checked when PCIDE bit is not set, however commit
'd1cd3ce900441 ("KVM: MMU: check guest CR3 reserved bits based on
its physical address width")' removes the bit 63 checking
unconditionally. This patch fixes it by checking bit 63 of CR3
when PCIDE bit is not set in CR4.
Fixes: d1cd3ce900 (KVM: MMU: check guest CR3 reserved bits based on its physical address width)
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Radim Krčmář <rkrcmar@redhat.com>
Cc: Liran Alon <liran.alon@oracle.com>
Cc: stable@vger.kernel.org
Reviewed-by: Junaid Shahid <junaids@google.com>
Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Update SECONDARY_EXEC_DESC for UMIP emulation if and only UMIP
is actually being emulated. Skipping the VMCS update eliminates
unnecessary VMREAD/VMWRITE when UMIP is supported in hardware,
and on platforms that don't have SECONDARY_VM_EXEC_CONTROL. The
latter case resolves a bug where KVM would fill the kernel log
with warnings due to failed VMWRITEs on older platforms.
Fixes: 0367f205a3 ("KVM: vmx: add support for emulating UMIP")
Cc: stable@vger.kernel.org #4.16
Reported-by: Paolo Zeppegno <pzeppegno@gmail.com>
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Suggested-by: Radim KrÄmář <rkrcmar@redhat.com>
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
If the PCIDE bit is not set in CR4, then the MSb of CR3 is a reserved
bit. If the guest tries to set it, that should cause a #GP fault. So
mask out the bit only when the PCIDE bit is set.
Signed-off-by: Junaid Shahid <junaids@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Even though the eventfd is released after the KVM SRCU grace period
elapses, the conn_to_evt data structure itself is not; it uses RCU
internally, instead. Fix the read-side critical section to happen
under rcu_read_lock/unlock; the result is still protected by
vcpu->kvm->srcu.
Reviewed-by: Roman Kagan <rkagan@virtuozzo.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The IP increment should be done after the hypercall emulation, after
calling the various handlers. In this way, these handlers can accurately
identify the the IP of the VMCALL if they need it.
This patch keeps the same functionality for the Hyper-V handler which does
not use the return code of the standard kvm_skip_emulated_instruction()
call.
Signed-off-by: Marian Rotariu <mrotariu@bitdefender.com>
[Hyper-V hypercalls also need kvm_skip_emulated_instruction() - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
ARM:
- Fix proxying of GICv2 CPU interface accesses
- Fix crash when switching to BE
- Track source vcpu git GICv2 SGIs
- Fix an outdated bit of documentation
x86:
- Speed up injection of expired timers (for stable)
-----BEGIN PGP SIGNATURE-----
iQEcBAABCAAGBQJa7s3kAAoJEED/6hsPKofojlwIAKQLt6GuAHTj7Gl+G4/EOOSk
bXSXracJSi3aQHTVdKaREo6iTpAB7/E/4yf+KLkljiqPO/YrWrlTnAbqfCfjqX6b
pWyXgoxPy4v+SEbhP+qiV/yC/HiuPJ4WZVmf5cCDXD4kPF03b7DvImGbZRbEwJNV
qlaO1QqmbbMU5m1I5oZCKn5/BLM3fwAFMn1RERFDOPyn3+HPwANRbnLsZ4q49KHw
W41Rj6i88qDQ3PrbgUCmSvuzboKwTiVUBltPlStk9A04F2toLytcMoo9fdzQOjwD
ZmvjYlLqNlxXFLUHHvhgOECvBTp879AWUrgoManQ16O1s/gApTcHdXpzKQBKjtE=
=I6Dd
-----END PGP SIGNATURE-----
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Pll KVM fixes from Radim Krčmář:
"ARM:
- Fix proxying of GICv2 CPU interface accesses
- Fix crash when switching to BE
- Track source vcpu git GICv2 SGIs
- Fix an outdated bit of documentation
x86:
- Speed up injection of expired timers (for stable)"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
KVM: x86: remove APIC Timer periodic/oneshot spikes
arm64: vgic-v2: Fix proxying of cpuif access
KVM: arm/arm64: vgic_init: Cleanup reference to process_maintenance
KVM: arm64: Fix order of vcpu_write_sys_reg() arguments
KVM: arm/arm64: vgic: Fix source vcpu issues for GICv2 SGI
Pull x86 fix from Thomas Gleixner:
"Unbreak the CPUID CPUID_8000_0008_EBX reload which got dropped when
the evaluation of physical and virtual bits which uses the same CPUID
leaf was moved out of get_cpu_cap()"
* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/cpu: Restore CPUID_8000_0008_EBX reload
Pull clocksource fixes from Thomas Gleixner:
"The recent addition of the early TSC clocksource breaks on machines
which have an unstable TSC because in case that TSC is disabled, then
the clocksource selection logic falls back to the early TSC which is
obviously bogus.
That also unearthed a few robustness issues in the clocksource
derating code which are addressed as well"
* 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
clocksource: Rework stale comment
clocksource: Consistent de-rate when marking unstable
x86/tsc: Fix mark_tsc_unstable()
clocksource: Initialize cs->wd_list
clocksource: Allow clocksource_mark_unstable() on unregistered clocksources
x86/tsc: Always unregister clocksource_tsc_early
Since the commit "8003c9ae204e: add APIC Timer periodic/oneshot mode VMX
preemption timer support", a Windows 10 guest has some erratic timer
spikes.
Here the results on a 150000 times 1ms timer without any load:
Before 8003c9ae20 | After 8003c9ae20
Max 1834us | 86000us
Mean 1100us | 1021us
Deviation 59us | 149us
Here the results on a 150000 times 1ms timer with a cpu-z stress test:
Before 8003c9ae20 | After 8003c9ae20
Max 32000us | 140000us
Mean 1006us | 1997us
Deviation 140us | 11095us
The root cause of the problem is starting hrtimer with an expiry time
already in the past can take more than 20 milliseconds to trigger the
timer function. It can be solved by forward such past timers
immediately, rather than submitting them to hrtimer_start().
In case the timer is periodic, update the target expiration and call
hrtimer_start with it.
v2: Check if the tsc deadline is already expired. Thank you Mika.
v3: Execute the past timers immediately rather than submitting them to
hrtimer_start().
v4: Rearm the periodic timer with advance_periodic_target_expiration() a
simpler version of set_target_expiration(). Thank you Paolo.
Cc: Mika Penttilä <mika.penttila@nextfour.com>
Cc: Wanpeng Li <kernellwp@gmail.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: stable@vger.kernel.org
Signed-off-by: Anthoine Bourgeois <anthoine.bourgeois@blade-group.com>
8003c9ae20 ("KVM: LAPIC: add APIC Timer periodic/oneshot mode VMX preemption timer support")
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
-----BEGIN PGP SIGNATURE-----
iHUEABYIAB0WIQRTLbB6QfY48x44uB6AXGG7T9hjvgUCWuwoogAKCRCAXGG7T9hj
vr23AP4vj3yoii3mihZYjDahwyE+3fILUWECl/d/cMXGxq5tbgD9Esvb6DgtKHJr
Hi/lPMVM0XmN/DIXhY9x7SqO2cKvEAU=
=XwLB
-----END PGP SIGNATURE-----
Merge tag 'for-linus-4.17-rc4-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip
Pull xen cleanup from Juergen Gross:
"One cleanup to remove VLAs from the kernel"
* tag 'for-linus-4.17-rc4-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
x86/xen: Remove use of VLAs
Pull networking fixes from David Miller:
1) Various sockmap fixes from John Fastabend (pinned map handling,
blocking in recvmsg, double page put, error handling during redirect
failures, etc.)
2) Fix dead code handling in x86-64 JIT, from Gianluca Borello.
3) Missing device put in RDS IB code, from Dag Moxnes.
4) Don't process fast open during repair mode in TCP< from Yuchung
Cheng.
5) Move address/port comparison fixes in SCTP, from Xin Long.
6) Handle add a bond slave's master into a bridge properly, from
Hangbin Liu.
7) IPv6 multipath code can operate on unitialized memory due to an
assumption that the icmp header is in the linear SKB area. Fix from
Eric Dumazet.
8) Don't invoke do_tcp_sendpages() recursively via TLS, from Dave
Watson.
9) Fix memory leaks in x86-64 JIT, from Daniel Borkmann.
10) RDS leaks kernel memory to userspace, from Eric Dumazet.
11) DCCP can invoke a tasklet on a freed socket, take a refcount. Also
from Eric Dumazet.
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (78 commits)
dccp: fix tasklet usage
smc: fix sendpage() call
net/smc: handle unregistered buffers
net/smc: call consolidation
qed: fix spelling mistake: "offloded" -> "offloaded"
net/mlx5e: fix spelling mistake: "loobpack" -> "loopback"
tcp: restore autocorking
rds: do not leak kernel memory to user land
qmi_wwan: do not steal interfaces from class drivers
ipv4: fix fnhe usage by non-cached routes
bpf: sockmap, fix error handling in redirect failures
bpf: sockmap, zero sg_size on error when buffer is released
bpf: sockmap, fix scatterlist update on error path in send with apply
net_sched: fq: take care of throttled flows before reuse
ipv6: Revert "ipv6: Allow non-gateway ECMP for IPv6"
bpf, x64: fix memleak when not converging on calls
bpf, x64: fix memleak when not converging after image
net/smc: restrict non-blocking connect finish
8139too: Use disable_irq_nosync() in rtl8139_poll_controller()
sctp: fix the issue that the cookie-ack with auth can't get processed
...
The JIT logic in jit_subprogs() is as follows: for all subprogs we
allocate a bpf_prog_alloc(), populate it (prog->is_func = 1 here),
and pass it to bpf_int_jit_compile(). If a failure occurred during
JIT and prog->jited is not set, then we bail out from attempting to
JIT the whole program, and punt to the interpreter instead. In case
JITing went successful, we fixup BPF call offsets and do another
pass to bpf_int_jit_compile() (extra_pass is true at that point) to
complete JITing calls. Given that requires to pass JIT context around
addrs and jit_data from x86 JIT are freed in the extra_pass in
bpf_int_jit_compile() when calls are involved (if not, they can
be freed immediately). However, if in the original pass, the JIT
image didn't converge then we leak addrs and jit_data since image
itself is NULL, the prog->is_func is set and extra_pass is false
in that case, meaning both will become unreachable and are never
cleaned up, therefore we need to free as well on !image. Only x64
JIT is affected.
Fixes: 1c2a088a66 ("bpf: x64: add JIT support for multi-function programs")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
While reviewing x64 JIT code, I noticed that we leak the prior allocated
JIT image in the case where proglen != oldproglen during the JIT passes.
Prior to the commit e0ee9c1215 ("x86: bpf_jit: fix two bugs in eBPF JIT
compiler") we would just break out of the loop, and using the image as the
JITed prog since it could only shrink in size anyway. After e0ee9c1215,
we would bail out to out_addrs label where we free addrs and jit_data but
not the image coming from bpf_jit_binary_alloc().
Fixes: e0ee9c1215 ("x86: bpf_jit: fix two bugs in eBPF JIT compiler")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
The recent commt which addresses the x86_phys_bits corruption with
encrypted memory on CPUID reload after a microcode update lost the reload
of CPUID_8000_0008_EBX as well.
As a consequence IBRS and IBRS_FW are not longer detected
Restore the behaviour by bringing the reload of CPUID_8000_0008_EBX
back. This restore has a twist due to the convoluted way the cpuid analysis
works:
CPUID_8000_0008_EBX is used by AMD to enumerate IBRB, IBRS, STIBP. On Intel
EBX is not used. But the speculation control code sets the AMD bits when
running on Intel depending on the Intel specific speculation control
bits. This was done to use the same bits for alternatives.
The change which moved the 8000_0008 evaluation out of get_cpu_cap() broke
this nasty scheme due to ordering. So that on Intel the store to
CPUID_8000_0008_EBX clears the IBRB, IBRS, STIBP bits which had been set
before by software.
So the actual CPUID_8000_0008_EBX needs to go back to the place where it
was and the phys/virt address space calculation cannot touch it.
In hindsight this should have used completely synthetic bits for IBRB,
IBRS, STIBP instead of reusing the AMD bits, but that's for 4.18.
/me needs to find time to cleanup that steaming pile of ...
Fixes: d94a155c59 ("x86/cpu: Prevent cpuinfo_x86::x86_phys_bits adjustment corruption")
Reported-by: Jörg Otte <jrg.otte@gmail.com>
Reported-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Jörg Otte <jrg.otte@gmail.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: kirill.shutemov@linux.intel.com
Cc: Borislav Petkov <bp@alien8.de
Link: https://lkml.kernel.org/r/alpine.DEB.2.21.1805021043510.1668@nanos.tec.linutronix.de
mark_tsc_unstable() also needs to affect tsc_early, Now that
clocksource_mark_unstable() can be used on a clocksource irrespective of
its registration state, use it on both tsc_early and tsc.
This does however require cs->list to be initialized empty, otherwise it
cannot tell the registation state before registation.
Fixes: aa83c45762 ("x86/tsc: Introduce early tsc clocksource")
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Diego Viola <diego.viola@gmail.com>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: len.brown@intel.com
Cc: rjw@rjwysocki.net
Cc: rui.zhang@intel.com
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20180430100344.533326547@infradead.org
Don't leave the tsc-early clocksource registered if it errors out
early.
This was reported by Diego, who on his Core2 era machine got TSC
invalidated while it was running with tsc-early (due to C-states).
This results in keeping tsc-early with very bad effects.
Reported-and-Tested-by: Diego Viola <diego.viola@gmail.com>
Fixes: aa83c45762 ("x86/tsc: Introduce early tsc clocksource")
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: len.brown@intel.com
Cc: rjw@rjwysocki.net
Cc: diego.viola@gmail.com
Cc: rui.zhang@intel.com
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20180430100344.350507853@infradead.org
Pull x86 fixes from Thomas Gleixner:
"Another set of x86 related updates:
- Fix the long broken x32 version of the IPC user space headers which
was noticed by Arnd Bergman in course of his ongoing y2038 work.
GLIBC seems to have non broken private copies of these headers so
this went unnoticed.
- Two microcode fixlets which address some more fallout from the
recent modifications in that area:
- Unconditionally save the microcode patch, which was only saved
when CPU_HOTPLUG was enabled causing failures in the late
loading mechanism
- Make the later loader synchronization finally work under all
circumstances. It was exiting early and causing timeout failures
due to a missing synchronization point.
- Do not use mwait_play_dead() on AMD systems to prevent excessive
power consumption as the CPU cannot go into deep power states from
there.
- Address an annoying sparse warning due to lost type qualifiers of
the vmemmap and vmalloc base address constants.
- Prevent reserving crash kernel region on Xen PV as this leads to
the wrong perception that crash kernels actually work there which
is not the case. Xen PV has its own crash mechanism handled by the
hypervisor.
- Add missing TLB cpuid values to the table to make the printout on
certain machines correct.
- Enumerate the new CLDEMOTE instruction
- Fix an incorrect SPDX identifier
- Remove stale macros"
* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/ipc: Fix x32 version of shmid64_ds and msqid64_ds
x86/setup: Do not reserve a crash kernel region if booted on Xen PV
x86/cpu/intel: Add missing TLB cpuid values
x86/smpboot: Don't use mwait_play_dead() on AMD systems
x86/mm: Make vmemmap and vmalloc base address constants unsigned long
x86/vector: Remove the unused macro FPU_IRQ
x86/vector: Remove the macro VECTOR_OFFSET_START
x86/cpufeatures: Enumerate cldemote instruction
x86/microcode: Do not exit early from __reload_late()
x86/microcode/intel: Save microcode patch unconditionally
x86/jailhouse: Fix incorrect SPDX identifier
Pull x86 pti fixes from Thomas Gleixner:
"A set of updates for the x86/pti related code:
- Preserve r8-r11 in int $0x80. r8-r11 need to be preserved, but the
int$80 entry code removed that quite some time ago. Make it correct
again.
- A set of fixes for the Global Bit work which went into 4.17 and
caused a bunch of interesting regressions:
- Triggering a BUG in the page attribute code due to a missing
check for early boot stage
- Warnings in the page attribute code about holes in the kernel
text mapping which are caused by the freeing of the init code.
Handle such holes gracefully.
- Reduce the amount of kernel memory which is set global to the
actual text and do not incidentally overlap with data.
- Disable the global bit when RANDSTRUCT is enabled as it
partially defeats the hardening.
- Make the page protection setup correct for vma->page_prot
population again. The adjustment of the protections fell through
the crack during the Global bit rework and triggers warnings on
machines which do not support certain features, e.g. NX"
* 'x86-pti-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/entry/64/compat: Preserve r8-r11 in int $0x80
x86/pti: Filter at vma->vm_page_prot population
x86/pti: Disallow global kernel text with RANDSTRUCT
x86/pti: Reduce amount of kernel text allowed to be Global
x86/pti: Fix boot warning from Global-bit setting
x86/pti: Fix boot problems from Global-bit setting
Pull perf fixes from Thomas Gleixner:
"The perf update contains the following bits:
x86:
- Prevent setting freeze_on_smi on PerfMon V1 CPUs to avoid #GP
perf stat:
- Keep the '/' event modifier separator in fallback, for example when
fallbacking from 'cpu/cpu-cycles/' to user level only, where it
should become 'cpu/cpu-cycles/u' and not 'cpu/cpu-cycles/:u' (Jiri
Olsa)
- Fix PMU events parsing rule, improving error reporting for invalid
events (Jiri Olsa)
- Disable write_backward and other event attributes for !group events
in a group, fixing, for instance this group: '{cycles,msr/aperf/}:S'
that has leader sampling (:S) and where just the 'cycles', the
leader event, should have the write_backward attribute set, in this
case it all fails because the PMU where 'msr/aperf/' lives doesn't
accepts write_backward style sampling (Jiri Olsa)
- Only fall back group read for leader (Kan Liang)
- Fix core PMU alias list for x86 platform (Kan Liang)
- Print out hint for mixed PMU group error (Kan Liang)
- Fix duplicate PMU name for interval print (Kan Liang)
Core:
- Set main kernel end address properly when reading kernel and module
maps (Namhyung Kim)
perf mem:
- Fix incorrect entries and add missing man options (Sangwon Hong)
s/390:
- Remove s390 specific strcmp_cpuid_cmp function (Thomas Richter)
- Adapt 'perf test' case record+probe_libc_inet_pton.sh for s390
- Fix s390 undefined record__auxtrace_init() return value in 'perf
record' (Thomas Richter)"
* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
perf/x86/intel: Don't enable freeze-on-smi for PerfMon V1
perf stat: Fix duplicate PMU name for interval print
perf evsel: Only fall back group read for leader
perf stat: Print out hint for mixed PMU group error
perf pmu: Fix core PMU alias list for X86 platform
perf record: Fix s390 undefined record__auxtrace_init() return value
perf mem: Document incorrect and missing options
perf evsel: Disable write_backward for leader sampling group events
perf pmu: Fix pmu events parsing rule
perf stat: Keep the / modifier separator in fallback
perf test: Adapt test case record+probe_libc_inet_pton.sh for s390
perf list: Remove s390 specific strcmp_cpuid_cmp function
perf machine: Set main kernel end address properly
ARM:
- PSCI selection API, a leftover from 4.16 (for stable)
- Kick vcpu on active interrupt affinity change
- Plug a VMID allocation race on oversubscribed systems
- Silence debug messages
- Update Christoffer's email address (linaro -> arm)
x86:
- Expose userspace-relevant bits of a newly added feature
- Fix TLB flushing on VMX with VPID, but without EPT
-----BEGIN PGP SIGNATURE-----
iQEcBAABCAAGBQJa44lQAAoJEED/6hsPKofo1dIH/3n9AZSWvavgL2V3j6agT8Yy
hxF4nHCFEJd5aqDNwbG9QEzivKw88r3o3mdB2XAQESB2MlCYR1jkTONm7yvVJTs/
/P9gj+DEQbCj2AgT//u3BGsAsZDKFhB9JwfmV2Mp4zDIqWFa6oCOGeq/iPVAGDcN
vUpuYeIicuH9SRoxH7de3z+BEXW0O+gCABXQtvA93FKTMz35yFTgmbDVCnvaV0zL
3B+3/4/jdbTRICW8EX6Li43+gEBUMtnVNkdqxLPTuCtDG8iuPUGfgF02gH99/9gj
hliV3Q4VUZKkSABW5AqKPe4+9rbsHCh9eL0LpHFGI9y+6LeUIOXAX4CtohR8gWE=
=W9Vz
-----END PGP SIGNATURE-----
rMerge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Pull KVM fixes from Radim Krčmář:
"ARM:
- PSCI selection API, a leftover from 4.16 (for stable)
- Kick vcpu on active interrupt affinity change
- Plug a VMID allocation race on oversubscribed systems
- Silence debug messages
- Update Christoffer's email address (linaro -> arm)
x86:
- Expose userspace-relevant bits of a newly added feature
- Fix TLB flushing on VMX with VPID, but without EPT"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
x86/headers/UAPI: Move DISABLE_EXITS KVM capability bits to the UAPI
kvm: apic: Flush TLB after APIC mode/address change if VPIDs are in use
arm/arm64: KVM: Add PSCI version selection API
KVM: arm/arm64: vgic: Kick new VCPU on interrupt migration
arm64: KVM: Demote SVE and LORegion warnings to debug only
MAINTAINERS: Update e-mail address for Christoffer Dall
KVM: arm/arm64: Close VMID generation race
Move DISABLE_EXITS KVM capability bits to the UAPI just like the rest of
capabilities.
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Radim Krčmář <rkrcmar@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: x86@kernel.org
Cc: kvm@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: KarimAllah Ahmed <karahmed@amazon.de>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Currently, KVM flushes the TLB after a change to the APIC access page
address or the APIC mode when EPT mode is enabled. However, even in
shadow paging mode, a TLB flush is needed if VPIDs are being used, as
specified in the Intel SDM Section 29.4.5.
So replace vmx_flush_tlb_ept_only() with vmx_flush_tlb(), which will
flush if either EPT or VPIDs are in use.
Signed-off-by: Junaid Shahid <junaids@google.com>
Reviewed-by: Jim Mattson <jmattson@google.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
32-bit user code that uses int $80 doesn't care about r8-r11. There is,
however, some 64-bit user code that intentionally uses int $0x80 to invoke
32-bit system calls. From what I've seen, basically all such code assumes
that r8-r15 are all preserved, but the kernel clobbers r8-r11. Since I
doubt that there's any code that depends on int $0x80 zeroing r8-r11,
change the kernel to preserve them.
I suspect that very little user code is broken by the old clobber, since
r8-r11 are only rarely allocated by gcc, and they're clobbered by function
calls, so they only way we'd see a problem is if the same function that
invokes int $0x80 also spills something important to one of these
registers.
The current behavior seems to date back to the historical commit
"[PATCH] x86-64 merge for 2.6.4". Before that, all regs were
preserved. I can't find any explanation of why this change was made.
Update the test_syscall_vdso_32 testcase as well to verify the new
behavior, and it strengthens the test to make sure that the kernel doesn't
accidentally permute r8..r15.
Suggested-by: Denys Vlasenko <dvlasenk@redhat.com>
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Dominik Brodowski <linux@dominikbrodowski.net>
Link: https://lkml.kernel.org/r/d4c4d9985fbe64f8c9e19291886453914b48caee.1523975710.git.luto@kernel.org
A bugfix broke the x32 shmid64_ds and msqid64_ds data structure layout
(as seen from user space) a few years ago: Originally, __BITS_PER_LONG
was defined as 64 on x32, so we did not have padding after the 64-bit
__kernel_time_t fields, After __BITS_PER_LONG got changed to 32,
applications would observe extra padding.
In other parts of the uapi headers we seem to have a mix of those
expecting either 32 or 64 on x32 applications, so we can't easily revert
the path that broke these two structures.
Instead, this patch decouples x32 from the other architectures and moves
it back into arch specific headers, partially reverting the even older
commit 73a2d096fd ("x86: remove all now-duplicate header files").
It's not clear whether this ever made any difference, since at least
glibc carries its own (correct) copy of both of these header files,
so possibly no application has ever observed the definitions here.
Based on a suggestion from H.J. Lu, I tried out the tool from
https://github.com/hjl-tools/linux-header to find other such
bugs, which pointed out the same bug in statfs(), which also has
a separate (correct) copy in glibc.
Fixes: f4b4aae182 ("x86/headers/uapi: Fix __BITS_PER_LONG value for x32 builds")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: "H . J . Lu" <hjl.tools@gmail.com>
Cc: Jeffrey Walton <noloader@gmail.com>
Cc: stable@vger.kernel.org
Cc: "H. Peter Anvin" <hpa@zytor.com>
Link: https://lkml.kernel.org/r/20180424212013.3967461-1-arnd@arndb.de
Xen PV domains cannot shut down and start a crash kernel. Instead,
the crashing kernel makes a SCHEDOP_shutdown hypercall with the
reason code SHUTDOWN_crash, cf. xen_crash_shutdown() machine op in
arch/x86/xen/enlighten_pv.c.
A crash kernel reservation is merely a waste of RAM in this case. It
may also confuse users of kexec_load(2) and/or kexec_file_load(2).
When flags include KEXEC_ON_CRASH or KEXEC_FILE_ON_CRASH,
respectively, these syscalls return success, which is technically
correct, but the crash kexec image will never be actually used.
Signed-off-by: Petr Tesarik <ptesarik@suse.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Juergen Gross <jgross@suse.com>
Cc: Tom Lendacky <thomas.lendacky@amd.com>
Cc: Dou Liyang <douly.fnst@cn.fujitsu.com>
Cc: Mikulas Patocka <mpatocka@redhat.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: xen-devel@lists.xenproject.org
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: Jean Delvare <jdelvare@suse.de>
Link: https://lkml.kernel.org/r/20180425120835.23cef60c@ezekiel.suse.cz
- Add workqueue forward declaration (for new work, but a nice clean up)
- seftest fixes for the new histogram code
- Print output fix for hwlat tracer
- Fix missing system call events - due to change in x86 syscall naming
- Fix kprobe address being used by perf being hashed
-----BEGIN PGP SIGNATURE-----
iIoEABYIADIWIQRRSw7ePDh/lE+zeZMp5XQQmuv6qgUCWuIMShQccm9zdGVkdEBn
b29kbWlzLm9yZwAKCRAp5XQQmuv6qkrdAQDRrgIGcm4pRGrvPiGhp4FeQKUx3woM
LY10qMYo3St7zwEAn5oor/e/7KQaQSdKQ7QkL690QU2bTO6FXz4VwE1OcgM=
=OHJk
-----END PGP SIGNATURE-----
Merge tag 'trace-v4.17-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace
Pull tracing fixes from Steven Rostedt:
- Add workqueue forward declaration (for new work, but a nice clean up)
- seftest fixes for the new histogram code
- Print output fix for hwlat tracer
- Fix missing system call events - due to change in x86 syscall naming
- Fix kprobe address being used by perf being hashed
* tag 'trace-v4.17-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
tracing: Fix missing tab for hwlat_detector print format
selftests: ftrace: Add a testcase for multiple actions on trigger
selftests: ftrace: Fix trigger extended error testcase
kprobes: Fix random address output of blacklist file
tracing: Fix kernel crash while using empty filter with perf
tracing/x86: Update syscall trace events to handle new prefixed syscall func names
tracing: Add missing forward declaration
Make kernel print the correct number of TLB entries on Intel Xeon Phi 7210
(and others)
Before:
[ 0.320005] Last level dTLB entries: 4KB 0, 2MB 0, 4MB 0, 1GB 0
After:
[ 0.320005] Last level dTLB entries: 4KB 256, 2MB 128, 4MB 128, 1GB 16
The entries do exist in the official Intel SMD but the type column there is
incorrect (states "Cache" where it should read "TLB"), but the entries for
the values 0x6B, 0x6C and 0x6D are correctly described as 'Data TLB'.
Signed-off-by: Jacek Tomaka <jacek.tomaka@poczta.fm>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/20180423161425.24366-1-jacekt@dugeo.com
Recent AMD systems support using MWAIT for C1 state. However, MWAIT will
not allow deeper cstates than C1 on current systems.
play_dead() expects to use the deepest state available. The deepest state
available on AMD systems is reached through SystemIO or HALT. If MWAIT is
available, it is preferred over the other methods, so the CPU never reaches
the deepest possible state.
Don't try to use MWAIT to play_dead() on AMD systems. Instead, use CPUIDLE
to enter the deepest state advertised by firmware. If CPUIDLE is not
available then fallback to HALT.
Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Borislav Petkov <bp@suse.de>
Cc: stable@vger.kernel.org
Cc: Yazen Ghannam <Yazen.Ghannam@amd.com>
Link: https://lkml.kernel.org/r/20180403140228.58540-1-Yazen.Ghannam@amd.com
Commits 9b46a051e4 ("x86/mm: Initialize vmemmap_base at boot-time") and
a7412546d8 ("x86/mm: Adjust vmalloc base and size at boot-time") lost the
type information for __VMALLOC_BASE_L4, __VMALLOC_BASE_L5,
__VMEMMAP_BASE_L4 and __VMEMMAP_BASE_L5 constants.
Declare them explicitly unsigned long again.
Fixes: 9b46a051e4 ("x86/mm: Initialize vmemmap_base at boot-time")
Fixes: a7412546d8 ("x86/mm: Adjust vmalloc base and size at boot-time")
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Link: https://lkml.kernel.org/r/nycvar.YFH.7.76.1804121437350.28129@cbobk.fhfr.pm
Now, Linux uses matrix allocator for vector assignment, the original
assignment code which used VECTOR_OFFSET_START has been removed.
So remove the stale macro as well.
Fixes: commit 69cde0004a ("x86/vector: Use matrix allocator for vector assignment")
Signed-off-by: Dou Liyang <douly.fnst@cn.fujitsu.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: David Rientjes <rientjes@google.com>
Cc: hpa@zytor.com
Link: https://lkml.kernel.org/r/20180425020553.17210-1-douly.fnst@cn.fujitsu.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
cldemote is a new instruction in future x86 processors. It hints
to hardware that a specified cache line should be moved ("demoted")
from the cache(s) closest to the processor core to a level more
distant from the processor core. This instruction is faster than
snooping to make the cache line available for other cores.
cldemote instruction is indicated by the presence of the CPUID
feature flag CLDEMOTE (CPUID.(EAX=0x7, ECX=0):ECX[bit25]).
More details on cldemote instruction can be found in the latest
Intel Architecture Instruction Set Extensions and Future Features
Programming Reference.
Signed-off-by: Fenghua Yu <fenghua.yu@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: "Ravi V Shankar" <ravi.v.shankar@intel.com>
Cc: "H. Peter Anvin" <hpa@linux.intel.com>
Cc: "Ashok Raj" <ashok.raj@intel.com>
Link: https://lkml.kernel.org/r/1524508162-192587-1-git-send-email-fenghua.yu@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Daniel Borkmann says:
====================
pull-request: bpf 2018-04-25
The following pull-request contains BPF updates for your *net* tree.
The main changes are:
1) Fix to clear the percpu metadata_dst that could otherwise carry
stale ip_tunnel_info, from William.
2) Fix that reduces the number of passes in x64 JIT with regards to
dead code sanitation to avoid risk of prog rejection, from Gianluca.
3) Several fixes of sockmap programs, besides others, fixing a double
page_put() in error path, missing refcount hold for pinned sockmap,
adding required -target bpf for clang in sample Makefile, from John.
4) Fix to disable preemption in __BPF_PROG_RUN_ARRAY() paths, from Roman.
5) Fix tools/bpf/ Makefile with regards to a lex/yacc build error
seen on older gcc-5, from John.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>