Add support for decoding CXL Protocol Error Section as defined in UEFI 2.10
Section N.2.13.

Do the section decoding in a new cper_cxl.c file. This new file will be
used in the future for more CXL CPERs decode support. Add this to the
existing UEFI_CPER config.
Signed-off-by: Smita Koralahalli <Smita.KoralahalliChannabasappa@amd.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>

commit f4dc7fff ("efi: libstub: unify initrd loading between
architectures") merge the first and the second parameters into a
struct without updating the kernel-doc. Let's fix it.
Signed-off-by: Jialin Zhang <zhangjialin11@huawei.com>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>

The EFI runtime map code is only wired up on x86, which is the only
architecture that has a need for it in its implementation of kexec.

So let's move this code under arch/x86 and drop all references to it
from generic code. To ensure that the efi_runtime_map_init() is invoked
at the appropriate time use a 'sync' subsys_initcall() that will be
called right after the EFI initcall made from generic code where the
original invocation of efi_runtime_map_init() resided.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Reviewed-by: Dave Young <dyoung@redhat.com>

The current Kconfig logic for CONFIG_EFI_RUNTIME_MAPS does not convey
that without it, a kexec kernel is not able to boot in EFI mode at all.
So clarify this, and make the option only configurable via the menu
system if CONFIG_EXPERT is set.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Reviewed-by: Dave Young <dyoung@redhat.com>

By default, the efi-pstore backend hardcode the UEFI variable size
as 1024 bytes. The historical reasons for that were discussed by
Ard in threads [0][1]:

"there is some cargo cult from prehistoric EFI times going
on here, it seems. Or maybe just misinterpretation of the maximum
size for the variable *name* vs the variable itself.".

"OVMF has
OvmfPkg/OvmfPkgX64.dsc:
gEfiMdeModulePkgTokenSpaceGuid.PcdMaxVariableSize|0x2000
OvmfPkg/OvmfPkgX64.dsc:
gEfiMdeModulePkgTokenSpaceGuid.PcdMaxVariableSize|0x8400

where the first one is without secure boot and the second with secure
boot. Interestingly, the default is

gEfiMdeModulePkgTokenSpaceGuid.PcdMaxVariableSize|0x400

so this is probably where this 1k number comes from."

With that, and since there is not such a limit in the UEFI spec, we
have the confidence to hereby add a module parameter to enable advanced
users to change the UEFI record size for efi-pstore data collection,
this way allowing a much easier reading of the collected log, which
wouldn't be scattered anymore among many small files.

Through empirical analysis we observed that extreme low values (like 8
bytes) could eventually cause writing issues, so given that and the OVMF
default discussed, we limited the minimum value to 1024 bytes, which also
is still the default.

[0] https://lore.kernel.org/lkml/CAMj1kXF4UyRMh2Y_KakeNBHvkHhTtavASTAxXinDO1rhPe_wYg@mail.gmail.com/
[1] https://lore.kernel.org/lkml/CAMj1kXFy-2KddGu+dgebAdU9v2sindxVoiHLWuVhqYw+R=kqng@mail.gmail.com/Signed-off-by: Guilherme G. Piccoli <gpiccoli@igalia.com>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>

Currently, the EFI_PARAVIRT flag is only used by Xen dom0 boot on x86,
even though other architectures also support pseudo-EFI boot, where the
core kernel is invoked directly and provided with a set of data tables
that resemble the ones constructed by the EFI stub, which never actually
runs in that case.

Let's fix this inconsistency, and always set this flag when booting dom0
via the EFI boot path. Note that Xen on x86 does not provide the EFI
memory map in this case, whereas other architectures do, so move the
associated EFI_PARAVIRT check into the x86 platform code.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>

The EFI memory map is a description of the memory layout as provided by
the firmware, and only x86 manipulates it in various different ways for
its own memory bookkeeping. So let's move the memmap routines that are
only used by x86 into the x86 arch tree.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>

The EFI fake memmap support is specific to x86, which manipulates the
EFI memory map in various different ways after receiving it from the EFI
stub. On other architectures, we have managed to push back on this, and
the EFI memory map is kept pristine.

So let's move the fake memmap code into the x86 arch tree, where it
arguably belongs.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>

The initrd= command line loader can be useful for development, but it
was limited to loading files from the same file system as the loaded
kernel (and it didn't work on x86 mixed mode).

As both issues have been fixed, and the initrd= can now be used with
files residing on any simple file system exposed by the EFI firmware,
let's permit it to be enabled on RISC-V and LoongArch, which did not
support it up to this point.

Note that LoadFile2 remains the preferred option, as it is much simpler
to use and implement, but generic loaders (including the UEFI shell) may
not implement this so there, initrd= can now be used as well (if enabled
in the build)
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>

Now that we have support for calling protocols that need additional
marshalling for mixed mode, wire up the initrd command line loader.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>

Rework the EFI stub macro wrappers around protocol method calls and
other indirect calls in order to allow return types other than
efi_status_t. This means the widening should be conditional on whether
or not the return type is efi_status_t, and should be omitted otherwise.

Also, switch to _Generic() to implement the type based compile time
conditionals, which is more concise, and distinguishes between
efi_status_t and u64 properly.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>

Currently, the initrd= command line option to the EFI stub only supports
loading files that reside on the same volume as the loaded image, which
is not workable for loaders like GRUB that don't even implement the
volume abstraction (EFI_SIMPLE_FILE_SYSTEM_PROTOCOL), and load the
kernel from an anonymous buffer in memory. For this reason, another
method was devised that relies on the LoadFile2 protocol.

However, the command line loader is rather useful when using the UEFI
shell or other generic loaders that have no awareness of Linux specific
protocols so let's make it a bit more flexible, by permitting textual
device paths to be provided to initrd= as well, provided that they refer
to a file hosted on a EFI_SIMPLE_FILE_SYSTEM_PROTOCOL volume. E.g.,

  initrd=PciRoot(0x0)/Pci(0x3,0x0)/HD(1,MBR,0xBE1AFDFA,0x3F,0xFBFC1)/rootfs.cpio.gz
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>

The EFI spec is not very clear about which permissions are being given
when allocating pages of a certain type. However, it is quite obvious
that EFI_LOADER_CODE is more likely to permit execution than
EFI_LOADER_DATA, which becomes relevant once we permit booting the
kernel proper with the firmware's 1:1 mapping still active.

Ostensibly, recent systems such as the Surface Pro X grant executable
permissions to EFI_LOADER_CODE regions but not EFI_LOADER_DATA regions.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>

Pull MIPS fixes from Thomas Bogendoerfer:

 - fix jump label branch range check

 - check kmalloc failures in Loongson64 kexec

 - fix builds with clang-14

 - fix char/int handling in pic32

* tag 'mips-fixes_6.1_1' of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux:
  MIPS: pic32: treat port as signed integer
  MIPS: jump_label: Fix compat branch range check
  mips: alchemy: gpio: Include the right header
  MIPS: Loongson64: Add WARN_ON on kexec related kmalloc failed
  MIPS: fix duplicate definitions for exported symbols
  mips: boot/compressed: use __NO_FORTIFY

Pull EFI fixes from Ard Biesheuvel:

 - Force the use of SetVirtualAddressMap() on Ampera Altra arm64
   machines, which crash in SetTime() if no virtual remapping is used

   This is the first time we've added an SMBIOS based quirk on arm64,
   but fortunately, we can just call a EFI protocol to grab the type #1
   SMBIOS record when running in the stub, so we don't need all the
   machinery we have in the kernel proper to parse SMBIOS data.

 - Drop a spurious warning on misaligned runtime regions when using 16k
   or 64k pages on arm64

* tag 'efi-fixes-for-v6.1-3' of git://git.kernel.org/pub/scm/linux/kernel/git/efi/efi:
  arm64: efi: Fix handling of misaligned runtime regions and drop warning
  arm64: efi: Force the use of SetVirtualAddressMap() on Altra machines

Pull SCSI fixes from James Bottomley:
 "Three small fixes, all in drivers.

  The sas one is in an unlikely error leg, the debug one is to make it
  more standards conformant and the ibmvfc one is to fix a user visible
  bug where a failover could lose all paths to the device"

* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
  scsi: scsi_debug: Make the READ CAPACITY response compliant with ZBC
  scsi: scsi_transport_sas: Fix error handling in sas_phy_add()
  scsi: ibmvfc: Avoid path failures during live migration

Pull additional sound fix from Takashi Iwai:
 "A regression fix for the latest memalloc helper change"

* tag 'sound-fix-6.1-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
  ALSA: memalloc: Try dma_alloc_noncontiguous() at first

The latest fix for the non-contiguous memalloc helper changed the
allocation method for a non-IOMMU system to use only the fallback
allocator.  This should have worked, but it caused a problem sometimes
when too many non-contiguous pages are allocated that can't be treated
by HD-audio controller.

As a quirk workaround, go back to the original strategy: use
dma_alloc_noncontiguous() at first, and apply the fallback only when
it fails, but only for non-IOMMU case.

We'll need a better fix in the fallback code as well, but this
workaround should paper over most cases.

Fixes: 9736a325 ("ALSA: memalloc: Don't fall back for SG-buffer with IOMMU")
Reported-by: Linus Torvalds <torvalds@linux-foundation.org>
Link: https://lore.kernel.org/r/CAHk-=wgSH5ubdvt76gNwa004ooZAEJL_1Q-Fyw5M2FDdqL==dg@mail.gmail.com
Link: https://lore.kernel.org/r/20221112084718.3305-1-tiwai@suse.deSigned-off-by: Takashi Iwai <tiwai@suse.de>

Pull ata fixes from Damien Le Moal:
 "Several libata generic code fixes for rc5:

   - Add missing translation of the SYNCHRONIZE CACHE 16 scsi command as
     this command is mandatory for host-managed ZBC drives.

     The lack of support for it in libata-scsi was causing issues with
     some passthrough applications using ZBC drives (from Shin'ichiro).

   - Fix the error path of libata-transport host, port, link and device
     attributes initialization (from Yingliang).

   - Prevent issuing new commands to a drive that is in the NCQ error
     state and undergoing recovery (From Niklas).

     This bug went unnoticed for a long time as commands issued to a
     drive in error state are aborted immediately and retried by the
     scsi layer, hiding the useless abort-and-retry sequence"

* tag 'ata-6.1-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/dlemoal/libata:
  ata: libata-core: do not issue non-internal commands once EH is pending
  ata: libata-transport: fix error handling in ata_tdev_add()
  ata: libata-transport: fix error handling in ata_tlink_add()
  ata: libata-transport: fix error handling in ata_tport_add()
  ata: libata-transport: fix double ata_host_put() in ata_tport_add()
  ata: libata-scsi: fix SYNCHRONIZE CACHE (16) command failure

Pull misc hotfixes from Andrew Morton:
 "22 hotfixes.

  Eight are cc:stable and the remainder address issues which were
  introduced post-6.0 or which aren't considered serious enough to
  justify a -stable backport"

* tag 'mm-hotfixes-stable-2022-11-11' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm: (22 commits)
  docs: kmsan: fix formatting of "Example report"
  mm/damon/dbgfs: check if rm_contexts input is for a real context
  maple_tree: don't set a new maximum on the node when not reusing nodes
  maple_tree: fix depth tracking in maple_state
  arch/x86/mm/hugetlbpage.c: pud_huge() returns 0 when using 2-level paging
  fs: fix leaked psi pressure state
  nilfs2: fix use-after-free bug of ns_writer on remount
  x86/traps: avoid KMSAN bugs originating from handle_bug()
  kmsan: make sure PREEMPT_RT is off
  Kconfig.debug: ensure early check for KMSAN in CONFIG_KMSAN_WARN
  x86/uaccess: instrument copy_from_user_nmi()
  kmsan: core: kmsan_in_runtime() should return true in NMI context
  mm: hugetlb_vmemmap: include missing linux/moduleparam.h
  mm/shmem: use page_mapping() to detect page cache for uffd continue
  mm/memremap.c: map FS_DAX device memory as decrypted
  Partly revert "mm/thp: carry over dirty bit when thp splits on pmd"
  nilfs2: fix deadlock in nilfs_count_free_blocks()
  mm/mmap: fix memory leak in mmap_region()
  hugetlbfs: don't delete error page from pagecache
  maple_tree: reorganize testing to restore module testing
  ...

Pull arm64 fixes from Catalin Marinas:

 - Another fix for rodata=full. Since rodata= is not a simple boolean on
   arm64 (accepting 'full' as well), it got inadvertently broken by
   changes in the core code. If rodata=on is the default and rodata=off
   is passed on the kernel command line, rodata_full is never disabled

 - Fix gcc compiler warning of shifting 0xc0 into bits 31:24 without an
   explicit conversion to u32 (triggered by the AMPERE1 MIDR definition)

 - Include asm/ptrace.h in asm/syscall_wrapper.h to fix an incomplete
   struct pt_regs type causing the BPF verifier to refuse to load a
   tracing program which accesses pt_regs

* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
  arm64/syscall: Include asm/ptrace.h in syscall_wrapper header.
  arm64: Fix bit-shifting UB in the MIDR_CPU_MODEL() macro
  arm64: fix rodata=full again

While the ATA specification states that a device should return command
aborted for all commands queued after the device has entered error state,
since ATA only keeps the sense data for the latest command (in non-NCQ
case), we really don't want to send block layer commands to the device
after it has entered error state. (Only ATA EH commands should be sent,
to read the sense data etc.)

Currently, scsi_queue_rq() will check if scsi_host_in_recovery()
(state is SHOST_RECOVERY), and if so, it will _not_ issue a command via:
scsi_dispatch_cmd() -> host->hostt->queuecommand() (ata_scsi_queuecmd())
-> __ata_scsi_queuecmd() -> ata_scsi_translate() -> ata_qc_issue()

Before commit e494f6a7 ("[SCSI] improved eh timeout handler"),
when receiving a TFES error IRQ, the call chain looked like this:
ahci_error_intr() -> ata_port_abort() -> ata_do_link_abort() ->
ata_qc_complete() -> ata_qc_schedule_eh() -> blk_abort_request() ->
blk_rq_timed_out() -> q->rq_timed_out_fn() (scsi_times_out()) ->
scsi_eh_scmd_add() -> scsi_host_set_state(shost, SHOST_RECOVERY)

Which meant that as soon as an error IRQ was serviced, SHOST_RECOVERY
would be set.

However, after commit e494f6a7 ("[SCSI] improved eh timeout handler"),
scsi_times_out() will instead call scsi_abort_command() which will queue
delayed work, and the worker function scmd_eh_abort_handler() will call
scsi_eh_scmd_add(), which calls scsi_host_set_state(shost, SHOST_RECOVERY).

So now, after the TFES error IRQ has been serviced, we need to wait for
the SCSI workqueue to run its work before SHOST_RECOVERY gets set.

It is worth noting that, even before commit e494f6a7 ("[SCSI] improved
eh timeout handler"), we could receive an error IRQ from the time when
scsi_queue_rq() checks scsi_host_in_recovery(), to the time when
ata_scsi_queuecmd() is actually called.

In order to handle both the delayed setting of SHOST_RECOVERY and the
window where we can receive an error IRQ, add a check against
ATA_PFLAG_EH_PENDING (which gets set when servicing the error IRQ),
inside ata_scsi_queuecmd() itself, while holding the ap->lock.
(Since the ap->lock is held while servicing IRQs.)

Fixes: e494f6a7 ("[SCSI] improved eh timeout handler")
Signed-off-by: Niklas Cassel <niklas.cassel@wdc.com>
Tested-by: John Garry <john.g.garry@oracle.com>
Signed-off-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>

Pull block fixes from Jens Axboe:

 - NVMe pull request via Christoph:
        - Quiet user passthrough command errors (Keith Busch)
        - Fix memory leak in nvmet_subsys_attr_model_store_locked
        - Fix a memory leak in nvmet-auth (Sagi Grimberg)

 - Fix a potential NULL point deref in bfq (Yu)

 - Allocate command/response buffers separately for DMA for sed-opal,
   rather than rely on embedded alignment (Serge)

* tag 'block-6.1-2022-11-11' of git://git.kernel.dk/linux:
  nvmet: fix a memory leak
  nvmet: fix memory leak in nvmet_subsys_attr_model_store_locked
  nvme: quiet user passthrough command errors
  block: sed-opal: kmalloc the cmd/resp buffers
  block, bfq: fix null pointer dereference in bfq_bio_bfqg()

Pull io_uring fixes from Jens Axboe:
 "Nothing major, just a few minor tweaks:

   - Tweak for the TCP zero-copy io_uring self test (Pavel)

   - Rather than use our internal cached value of number of CQ events
     available, use what the user can see (Dylan)

   - Fix a typo in a comment, added in this release (me)

   - Don't allow wrapping while adding provided buffers (me)

   - Fix a double poll race, and add a lockdep assertion for it too
     (Pavel)"

* tag 'io_uring-6.1-2022-11-11' of git://git.kernel.dk/linux:
  io_uring/poll: lockdep annote io_poll_req_insert_locked
  io_uring/poll: fix double poll req->flags races
  io_uring: check for rollover of buffer ID when providing buffers
  io_uring: calculate CQEs from the user visible value
  io_uring: fix typo in io_uring.h comment
  selftests/net: don't tests batched TCP io_uring zc

Pull s390 fixes from Alexander Gordeev:

 - fix memcpy warning about field-spanning write in zcrypt driver

 - minor updates to defconfigs

 - remove CONFIG_DEBUG_INFO_BTF from all defconfigs and add btf.config
   addon config file. It significantly decreases compile time and allows
   quickly enabling that option into the current kernel config

 - add kasan.config addon config file which allows to easily enable
   KASAN into the current kernel config

 - binutils commit 906f69cf65da ("IBM zSystems: Issue error for *DBL
   relocs on misaligned symbols") caused several link errors. Always
   build relocatable kernel to avoid this problem

 - raise the minimum clang version to 15.0.0 to avoid silent generation
   of a corrupted code

* tag 's390-6.1-4' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:
  scripts/min-tool-version.sh: raise minimum clang version to 15.0.0 for s390
  s390: always build relocatable kernel
  s390/configs: add kasan.config addon config file
  s390/configs: move CONFIG_DEBUG_INFO_BTF into btf.config addon config
  s390: update defconfigs
  s390/zcrypt: fix warning about field-spanning write

Pull kernel hardening fix from Kees Cook:

 - Fix !SMP placement of '.data..decrypted' section (Nathan Chancellor)

* tag 'hardening-v6.1-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux:
  vmlinux.lds.h: Fix placement of '.data..decrypted' section

Pull nfsd fixes from Chuck Lever:

 - Fix an export leak

 - Fix a potential tracepoint crash

* tag 'nfsd-6.1-4' of git://git.kernel.org/pub/scm/linux/kernel/git/cel/linux:
  nfsd: put the export reference in nfsd4_verify_deleg_dentry
  nfsd: fix use-after-free in nfsd_file_do_acquire tracepoint

Pull UDF fix from Jan Kara:
 "Fix a possible memory corruption with UDF"

* tag 'fixes_for_v6.1-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs:
  udf: Fix a slab-out-of-bounds write bug in udf_find_entry()

Merge tag 'perf-tools-fixes-for-v6.1-2-2022-11-10' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux

Pull perf tools fixes from Arnaldo Carvalho de Melo:

 - Fix 'perf stat' crash with --per-node --metric-only in CSV mode, due
   to the AGGR_NODE slot in the 'aggr_header_csv' array not being set.

 - Fix printing prefix in CSV output of 'perf stat' metrics in interval
   mode (-I), where an extra separator was being added to the start of
   some lines.

 - Fix skipping branch stack sampling 'perf test' entry, that was using
   both --branch-any and --branch-filter, which can't be used together.

* tag 'perf-tools-fixes-for-v6.1-2-2022-11-10' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux:
  perf tools: Add the include/perf/ directory to .gitignore
  perf test: Fix skipping branch stack sampling test
  perf stat: Fix printing os->prefix in CSV metrics output
  perf stat: Fix crash with --per-node --metric-only in CSV mode

Pull RISC-V fixes from Palmer Dabbelt:

 - A fix to add the missing PWM LEDs into the SiFive HiFive Unleashed
   device tree.

 - A fix to fully clear a task's registers on creation, as they end up
   in userspace and thus leak kernel memory.

 - A pair of VDSO-related build fixes that manifest on recent LLVM-based
   toolchains.

 - A fix to our early init to ensure the DT is adequately processed
   before reserved memory nodes are processed.

* tag 'riscv-for-linus-6.1-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux:
  RISC-V: vdso: Do not add missing symbols to version section in linker script
  riscv: fix reserved memory setup
  riscv: vdso: fix build with llvm
  riscv: process: fix kernel info leakage
  riscv: dts: sifive unleashed: Add PWM controlled LEDs

Pull kvm
 "This is a pretty large diffstat for this time of the release. The main
  culprit is a reorganization of the AMD assembly trampoline, allowing
  percpu variables to be accessed early.

  This is needed for the return stack depth tracking retbleed mitigation
  that will be in 6.2, but it also makes it possible to tighten the IBRS
  restore on vmexit. The latter change is a long tail of the
  spectrev2/retbleed patches (the corresponding Intel change was simpler
  and went in already last June), which is why I am including it right
  now instead of sharing a topic branch with tip.

  Being assembly and being rich in comments makes the line count balloon
  a bit, but I am pretty confident in the change (famous last words)
  because the reorganization actually makes everything simpler and more
  understandable than before. It has also had external review and has
  been tested on the aforementioned 6.2 changes, which explode quite
  brutally without the fix.

  Apart from this, things are pretty normal.

  s390:

   - PCI fix

   - PV clock fix

  x86:

   - Fix clash between PMU MSRs and other MSRs

   - Prepare SVM assembly trampoline for 6.2 retbleed mitigation and
     for...

   - ... tightening IBRS restore on vmexit, moving it before the first
     RET or indirect branch

   - Fix log level for VMSA dump

   - Block all page faults during kvm_zap_gfn_range()

  Tools:

   - kvm_stat: fix incorrect detection of debugfs

   - kvm_stat: update vmexit definitions"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
  KVM: x86/mmu: Block all page faults during kvm_zap_gfn_range()
  KVM: x86/pmu: Limit the maximum number of supported AMD GP counters
  KVM: x86/pmu: Limit the maximum number of supported Intel GP counters
  KVM: x86/pmu: Do not speculatively query Intel GP PMCs that don't exist yet
  KVM: SVM: Only dump VMSA to klog at KERN_DEBUG level
  tools/kvm_stat: update exit reasons for vmx/svm/aarch64/userspace
  tools/kvm_stat: fix incorrect detection of debugfs
  x86, KVM: remove unnecessary argument to x86_virt_spec_ctrl and callers
  KVM: SVM: move MSR_IA32_SPEC_CTRL save/restore to assembly
  KVM: SVM: restore host save area from assembly
  KVM: SVM: move guest vmsave/vmload back to assembly
  KVM: SVM: do not allocate struct svm_cpu_data dynamically
  KVM: SVM: remove dead field from struct svm_cpu_data
  KVM: SVM: remove unused field from struct vcpu_svm
  KVM: SVM: retrieve VMCB from assembly
  KVM: SVM: adjust register allocation for __svm_vcpu_run()
  KVM: SVM: replace regs argument of __svm_vcpu_run() with vcpu_svm
  KVM: x86: use a separate asm-offsets.c file
  KVM: s390: pci: Fix allocation size of aift kzdev elements
  KVM: s390: pv: don't allow userspace to set the clock under PV

Merge tag 'hyperv-fixes-signed-20221110' of git://git.kernel.org/pub/scm/linux/kernel/git/hyperv/linux

Pull hyperv fixes from Wei Liu:

 - Fix TSC MSR write for root partition (Anirudh Rayabharam)

 - Fix definition of vector in pci-hyperv driver (Dexuan Cui)

 - A few other misc patches

* tag 'hyperv-fixes-signed-20221110' of git://git.kernel.org/pub/scm/linux/kernel/git/hyperv/linux:
  PCI: hv: Fix the definition of vector in hv_compose_msi_msg()
  MAINTAINERS: remove sthemmin
  x86/hyperv: fix invalid writes to MSRs during root partition kexec
  clocksource/drivers/hyperv: add data structure for reference TSC MSR
  Drivers: hv: fix repeated words in comments
  x86/hyperv: Remove BUG_ON() for kmap_local_page()

Pull dmaengine fixes from Vinod Koul:
 "Misc minor driver fixes and a big pile of at_hdmac driver fixes. More
  work on this driver is done and sitting in next:

   - Pile of at_hdmac driver rework which fixes many long standing
     issues for this driver.

   - couple of stm32 driver fixes for clearing structure and race fix

   - idxd fixes for RO device state and batch size

   - ti driver mem leak fix

   - apple fix for grabbing channels in xlate

   - resource leak fix in mv xor"

* tag 'dmaengine-fix-6.1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine: (24 commits)
  dmaengine: at_hdmac: Check return code of dma_async_device_register
  dmaengine: at_hdmac: Fix impossible condition
  dmaengine: at_hdmac: Don't allow CPU to reorder channel enable
  dmaengine: at_hdmac: Fix completion of unissued descriptor in case of errors
  dmaengine: at_hdmac: Fix descriptor handling when issuing it to hardware
  dmaengine: at_hdmac: Fix concurrency over the active list
  dmaengine: at_hdmac: Free the memset buf without holding the chan lock
  dmaengine: at_hdmac: Fix concurrency over descriptor
  dmaengine: at_hdmac: Fix concurrency problems by removing atc_complete_all()
  dmaengine: at_hdmac: Protect atchan->status with the channel lock
  dmaengine: at_hdmac: Do not call the complete callback on device_terminate_all
  dmaengine: at_hdmac: Fix premature completion of desc in issue_pending
  dmaengine: at_hdmac: Start transfer for cyclic channels in issue_pending
  dmaengine: at_hdmac: Don't start transactions at tx_submit level
  dmaengine: at_hdmac: Fix at_lli struct definition
  dmaengine: stm32-dma: fix potential race between pause and resume
  dmaengine: ti: k3-udma-glue: fix memory leak when register device fail
  dmaengine: mv_xor_v2: Fix a resource leak in mv_xor_v2_remove()
  dmaengine: apple-admac: Fix grabbing of channels in of_xlate
  dmaengine: idxd: fix RO device state error after been disabled/reset
  ...

Pull spi fixes from Mark Brown:
 "A relatively large batch of fixes here but all device specific, plus
  an update to MAINTAINERS.

  The summary print change to the STM32 driver is fixing an issue where
  the driver could easily end up spamming the logs with something that
  should be a debug message"

* tag 'spi-fix-v6.1-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi:
  spi: amd: Fix SPI_SPD7 value
  spi: stm32: fix stm32_spi_prepare_mbr() that halves spi clk for every run
  spi: meson-spicc: fix do_div build error on non-arm64
  spi: intel: Use correct mask for flash and protected regions
  spi: mediatek: Fix package division error
  spi: tegra210-quad: Don't initialise DMA if not supported
  MAINTAINERS: Update HiSilicon SFC Driver maintainer
  spi: meson-spicc: move wait completion in driver to take bursts delay in account
  spi: stm32: Print summary 'callbacks suppressed' message

Pull MMC fixes from Ulf Hansson:

 - Provide helper for resetting both SDHCI and CQHCI

 - Fix reset for CQHCI (am654, brcmstb, esdhc-imx, of-arasan, tegra)

 - Fixup support for MMC_CAP_8_BIT_DATA (esdhc-imx)

* tag 'mmc-v6.1-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc:
  mmc: sdhci-esdhc-imx: use the correct host caps for MMC_CAP_8_BIT_DATA
  mmc: sdhci_am654: Fix SDHCI_RESET_ALL for CQHCI
  mmc: sdhci-tegra: Fix SDHCI_RESET_ALL for CQHCI
  mms: sdhci-esdhc-imx: Fix SDHCI_RESET_ALL for CQHCI
  mmc: sdhci-brcmstb: Fix SDHCI_RESET_ALL for CQHCI
  mmc: sdhci-of-arasan: Fix SDHCI_RESET_ALL for CQHCI
  mmc: cqhci: Provide helper for resetting both SDHCI and CQHCI

Pull HID fixes from Jiri Kosina:

 - fix for memory leak (on error path) in Hyper-V driver (Yang
   Yingliang)

 - regression fix for handling 3rd barrel switch emulation in Wacom
   driver (Jason Gerecke)

* tag 'for-linus-2022111101' of git://git.kernel.org/pub/scm/linux/kernel/git/hid/hid:
  HID: wacom: Fix logic used for 3rd barrel switch emulation
  HID: hyperv: fix possible memory leak in mousevsc_probe()
  HID: asus: Remove unused variable in asus_report_tool_width()

Add a lockdep annotation in io_poll_req_insert_locked().
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/8115d8e702733754d0aea119e9b5bb63d1eb8b24.1668184658.git.asml.silence@gmail.comSigned-off-by: Jens Axboe <axboe@kernel.dk>

io_poll_double_prepare()            | io_poll_wake()
                                    | poll->head = NULL
smp_load(&poll->head); /* NULL */   |
flags = req->flags;                 |
                                    | req->flags &= ~SINGLE_POLL;
req->flags = flags | DOUBLE_POLL    |

The idea behind io_poll_double_prepare() is to serialise with the
first poll entry by taking the wq lock. However, it's not safe to assume
that io_poll_wake() is not running when we can't grab the lock and so we
may race modifying req->flags.

Skip double poll setup if that happens. It's ok because the first poll
entry will only be removed when it's definitely completing, e.g.
pollfree or oneshot with a valid mask.

Fixes: 49f1c68e ("io_uring: optimise submission side poll_refs")
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/b7fab2d502f6121a7d7b199fe4d914a43ca9cdfd.1668184658.git.asml.silence@gmail.comSigned-off-by: Jens Axboe <axboe@kernel.dk>