The kernel will currently disable all SGX support if the hardware does
not support launch control. Make it more permissive to allow SGX
virtualization on systems without Launch Control support. This will
allow KVM to expose SGX to guests that have less-strict requirements on
the availability of flexible launch control.
Improve error message to distinguish between three cases. There are two
cases where SGX support is completely disabled:
1) SGX has been disabled completely by the BIOS
2) SGX LC is locked by the BIOS. Bare-metal support is disabled because
of LC unavailability. SGX virtualization is unavailable (because of
Kconfig).
One where it is partially available:
3) SGX LC is locked by the BIOS. Bare-metal support is disabled because
of LC unavailability. SGX virtualization is supported.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Co-developed-by: Kai Huang <kai.huang@intel.com>
Signed-off-by: Kai Huang <kai.huang@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: Jarkko Sakkinen <jarkko@kernel.org>
Acked-by: Dave Hansen <dave.hansen@intel.com>
Link: https://lkml.kernel.org/r/b3329777076509b3b601550da288c8f3c406a865.1616136308.git.kai.huang@intel.com
Add a misc device /dev/sgx_vepc to allow userspace to allocate "raw"
Enclave Page Cache (EPC) without an associated enclave. The intended
and only known use case for raw EPC allocation is to expose EPC to a
KVM guest, hence the 'vepc' moniker, virt.{c,h} files and X86_SGX_KVM
Kconfig.
The SGX driver uses the misc device /dev/sgx_enclave to support
userspace in creating an enclave. Each file descriptor returned from
opening /dev/sgx_enclave represents an enclave. Unlike the SGX driver,
KVM doesn't control how the guest uses the EPC, therefore EPC allocated
to a KVM guest is not associated with an enclave, and /dev/sgx_enclave
is not suitable for allocating EPC for a KVM guest.
Having separate device nodes for the SGX driver and KVM virtual EPC also
allows separate permission control for running host SGX enclaves and KVM
SGX guests.
To use /dev/sgx_vepc to allocate a virtual EPC instance with particular
size, the hypervisor opens /dev/sgx_vepc, and uses mmap() with the
intended size to get an address range of virtual EPC. Then it may use
the address range to create one KVM memory slot as virtual EPC for
a guest.
Implement the "raw" EPC allocation in the x86 core-SGX subsystem via
/dev/sgx_vepc rather than in KVM. Doing so has two major advantages:
- Does not require changes to KVM's uAPI, e.g. EPC gets handled as
just another memory backend for guests.
- EPC management is wholly contained in the SGX subsystem, e.g. SGX
does not have to export any symbols, changes to reclaim flows don't
need to be routed through KVM, SGX's dirty laundry doesn't have to
get aired out for the world to see, and so on and so forth.
The virtual EPC pages allocated to guests are currently not reclaimable.
Reclaiming an EPC page used by enclave requires a special reclaim
mechanism separate from normal page reclaim, and that mechanism is not
supported for virutal EPC pages. Due to the complications of handling
reclaim conflicts between guest and host, reclaiming virtual EPC pages
is significantly more complex than basic support for SGX virtualization.
[ bp:
- Massage commit message and comments
- use cpu_feature_enabled()
- vertically align struct members init
- massage Virtual EPC clarification text
- move Kconfig prompt to Virtualization ]
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Co-developed-by: Kai Huang <kai.huang@intel.com>
Signed-off-by: Kai Huang <kai.huang@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: Dave Hansen <dave.hansen@intel.com>
Acked-by: Jarkko Sakkinen <jarkko@kernel.org>
Link: https://lkml.kernel.org/r/0c38ced8c8e5a69872db4d6a1c0dabd01e07cad7.1616136308.git.kai.huang@intel.com
SGX driver can accurately track how enclave pages are used. This
enables SECS to be specifically targeted and EREMOVE'd only after all
child pages have been EREMOVE'd. This ensures that SGX driver will
never encounter SGX_CHILD_PRESENT in normal operation.
Virtual EPC is different. The host does not track how EPC pages are
used by the guest, so it cannot guarantee EREMOVE success. It might,
for instance, encounter a SECS with a non-zero child count.
Add a definition of SGX_CHILD_PRESENT. It will be used exclusively by
the SGX virtualization driver to handle recoverable EREMOVE errors when
saniziting EPC pages after they are freed.
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Kai Huang <kai.huang@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: Dave Hansen <dave.hansen@intel.com>
Acked-by: Jarkko Sakkinen <jarkko@kernel.org>
Link: https://lkml.kernel.org/r/050b198e882afde7e6eba8e6a0d4da39161dbb5a.1616136308.git.kai.huang@intel.com
EREMOVE takes a page and removes any association between that page and
an enclave. It must be run on a page before it can be added into another
enclave. Currently, EREMOVE is run as part of pages being freed into the
SGX page allocator. It is not expected to fail, as it would indicate a
use-after-free of EPC pages. Rather than add the page back to the pool
of available EPC pages, the kernel intentionally leaks the page to avoid
additional errors in the future.
However, KVM does not track how guest pages are used, which means that
SGX virtualization use of EREMOVE might fail. Specifically, it is
legitimate that EREMOVE returns SGX_CHILD_PRESENT for EPC assigned to
KVM guest, because KVM/kernel doesn't track SECS pages.
To allow SGX/KVM to introduce a more permissive EREMOVE helper and
to let the SGX virtualization code use the allocator directly, break
out the EREMOVE call from the SGX page allocator. Rename the original
sgx_free_epc_page() to sgx_encl_free_epc_page(), indicating that
it is used to free an EPC page assigned to a host enclave. Replace
sgx_free_epc_page() with sgx_encl_free_epc_page() in all call sites so
there's no functional change.
At the same time, improve the error message when EREMOVE fails, and
add documentation to explain to the user what that failure means and
to suggest to the user what to do when this bug happens in the case it
happens.
[ bp: Massage commit message, fix typos and sanitize text, simplify. ]
Signed-off-by: Kai Huang <kai.huang@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Reviewed-by: Jarkko Sakkinen <jarkko@kernel.org>
Link: https://lkml.kernel.org/r/20210325093057.122834-1-kai.huang@intel.com
Add SGX1 and SGX2 feature flags, via CPUID.0x12.0x0.EAX, as scattered
features, since adding a new leaf for only two bits would be wasteful.
As part of virtualizing SGX, KVM will expose the SGX CPUID leafs to its
guest, and to do so correctly needs to query hardware and kernel support
for SGX1 and SGX2.
Suppress both SGX1 and SGX2 from /proc/cpuinfo. SGX1 basically means
SGX, and for SGX2 there is no concrete use case of using it in
/proc/cpuinfo.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Kai Huang <kai.huang@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: Dave Hansen <dave.hansen@intel.com>
Acked-by: Jarkko Sakkinen <jarkko@kernel.org>
Link: https://lkml.kernel.org/r/d787827dbfca6b3210ac3e432e3ac1202727e786.1616136308.git.kai.huang@intel.com
kmap() is inefficient and is being replaced by kmap_local_page(), if
possible. There is no readily apparent reason why initp_page needs to be
allocated and kmap'ed() except that 'sigstruct' needs to be page-aligned
and 'token' 512 byte-aligned.
Rather than change it to kmap_local_page(), use kmalloc() instead
because kmalloc() can give this alignment when allocating PAGE_SIZE
bytes.
Remove the alloc_page()/kmap() and replace with kmalloc(PAGE_SIZE, ...)
to get a page aligned kernel address.
In addition, add a comment to document the alignment requirements so that
others don't attempt to 'fix' this again.
[ bp: Massage commit message. ]
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20210324182246.2484875-1-ira.weiny@intel.com
The SGX device file (/dev/sgx_enclave) is unusual in that it requires
execute permissions. It has to be both "chmod +x" *and* be on a
filesystem without 'noexec'.
In the future, udev and systemd should get updates to set up systems
automatically. But, for now, nobody's systems do this automatically,
and everybody gets error messages like this when running ./test_sgx:
0x0000000000000000 0x0000000000002000 0x03
0x0000000000002000 0x0000000000001000 0x05
0x0000000000003000 0x0000000000003000 0x03
mmap() failed, errno=1.
That isn't very user friendly, even for forgetful kernel developers.
Further, the test case is rather haphazard about its use of fprintf()
versus perror().
Improve the error messages. Use perror() where possible. Lastly,
do some sanity checks on opening and mmap()ing the device file so
that we can get a decent error message out to the user.
Now, if your user doesn't have permission, you'll get the following:
$ ls -l /dev/sgx_enclave
crw------- 1 root root 10, 126 Mar 18 11:29 /dev/sgx_enclave
$ ./test_sgx
Unable to open /dev/sgx_enclave: Permission denied
If you then 'chown dave:dave /dev/sgx_enclave' (or whatever), but
you leave execute permissions off, you'll get:
$ ls -l /dev/sgx_enclave
crw------- 1 dave dave 10, 126 Mar 18 11:29 /dev/sgx_enclave
$ ./test_sgx
no execute permissions on device file
If you fix that with "chmod ug+x /dev/sgx" but you leave /dev as
noexec, you'll get this:
$ mount | grep "/dev .*noexec"
udev on /dev type devtmpfs (rw,nosuid,noexec,...)
$ ./test_sgx
ERROR: mmap for exec: Operation not permitted
mmap() succeeded for PROT_READ, but failed for PROT_EXEC
check that user has execute permissions on /dev/sgx_enclave and
that /dev does not have noexec set: 'mount | grep "/dev .*noexec"'
That can be fixed with:
mount -o remount,noexec /devESC
Hopefully, the combination of better error messages and the search
engines indexing this message will help people fix their systems
until we do this properly.
[ bp: Improve error messages more. ]
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Borislav Petkov <bp@suse.de>
Reviewed-by: Jarkko Sakkinen <jarkko@kernel.org>
Link: https://lore.kernel.org/r/20210318194301.11D9A984@viggo.jf.intel.com
Background
==========
SGX enclave memory is enumerated by the processor in contiguous physical
ranges called Enclave Page Cache (EPC) sections. Currently, there is a
free list per section, but allocations simply target the lowest-numbered
sections. This is functional, but has no NUMA awareness.
Fortunately, EPC sections are covered by entries in the ACPI SRAT table.
These entries allow each EPC section to be associated with a NUMA node,
just like normal RAM.
Solution
========
Implement a NUMA-aware enclave page allocator. Mirror the buddy allocator
and maintain a list of enclave pages for each NUMA node. Attempt to
allocate enclave memory first from local nodes, then fall back to other
nodes.
Note that the fallback is not as sophisticated as the buddy allocator
and is itself not aware of NUMA distances. When a node's free list is
empty, it searches for the next-highest node with enclave pages (and
will wrap if necessary). This could be improved in the future.
Other
=====
NUMA_KEEP_MEMINFO dependency is required for phys_to_target_node().
[ Kai Huang: Do not return NULL from __sgx_alloc_epc_page() because
callers do not expect that and that leads to a NULL ptr deref. ]
[ dhansen: Fix an uninitialized 'nid' variable in
__sgx_alloc_epc_page() as
Reported-by: kernel test robot <lkp@intel.com>
to avoid any potential allocations from the wrong NUMA node or even
premature allocation failures. ]
Signed-off-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Kai Huang <kai.huang@intel.com>
Signed-off-by: Dave Hansen <dave.hansen@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: Dave Hansen <dave.hansen@linux.intel.com>
Link: https://lore.kernel.org/lkml/158188326978.894464.217282995221175417.stgit@dwillia2-desk3.amr.corp.intel.com/
Link: https://lkml.kernel.org/r/20210319040602.178558-1-kai.huang@intel.com
Link: https://lkml.kernel.org/r/20210318214933.29341-1-dave.hansen@intel.com
Link: https://lkml.kernel.org/r/20210317235332.362001-2-jarkko.sakkinen@intel.com
During normal runtime, the "ksgxd" daemon behaves like a version of
kswapd just for SGX. But, before it starts acting like kswapd, its first
job is to initialize enclave memory.
Currently, the SGX boot code places each enclave page on a
epc_section->init_laundry_list. Once it starts up, the ksgxd code walks
over that list and populates the actual SGX page allocator.
However, the per-section structures are going away to make way for the
SGX NUMA allocator. There's also little need to have a per-section
structure; the enclave pages are all treated identically, and they can
be placed on the correct allocator list from metadata stored in the
enclave page (struct sgx_epc_page) itself.
Modify sgx_sanitize_section() to take a single page list instead of
taking a section and deriving the list from there.
Signed-off-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Borislav Petkov <bp@suse.de>
Acked-by: Dave Hansen <dave.hansen@linux.intel.com>
Link: https://lkml.kernel.org/r/20210317235332.362001-1-jarkko.sakkinen@intel.com
Doing a
prctl(PR_SET_MM, PR_SET_MM_AUXV, addr, 1);
will copy 1 byte from userspace to (quite big) on-stack array
and then stash everything to mm->saved_auxv.
AT_NULL terminator will be inserted at the very end.
/proc/*/auxv handler will find that AT_NULL terminator
and copy original stack contents to userspace.
This devious scheme requires CAP_SYS_RESOURCE.
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull irq fixes from Thomas Gleixner:
"A set of irqchip updates:
- Make the GENERIC_IRQ_MULTI_HANDLER configuration correct
- Add a missing DT compatible string for the Ingenic driver
- Remove the pointless debugfs_file pointer from struct irqdomain"
* tag 'irq-urgent-2021-03-14' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
irqchip/ingenic: Add support for the JZ4760
dt-bindings/irq: Add compatible string for the JZ4760B
irqchip: Do not blindly select CONFIG_GENERIC_IRQ_MULTI_HANDLER
ARM: ep93xx: Select GENERIC_IRQ_MULTI_HANDLER directly
irqdomain: Remove debugfs_file from struct irq_domain
Pull timer fix from Thomas Gleixner:
"A single fix in for hrtimers to prevent an interrupt storm caused by
the lack of reevaluation of the timers which expire in softirq context
under certain circumstances, e.g. when the clock was set"
* tag 'timers-urgent-2021-03-14' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
hrtimer: Update softirq_expires_next correctly after __hrtimer_get_next_event()
Pull scheduler fixes from Thomas Gleixner:
"A set of scheduler updates:
- Prevent a NULL pointer dereference in the migration_stop_cpu()
mechanims
- Prevent self concurrency of affine_move_task()
- Small fixes and cleanups related to task migration/affinity setting
- Ensure that sync_runqueues_membarrier_state() is invoked on the
current CPU when it is in the cpu mask"
* tag 'sched-urgent-2021-03-14' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
sched/membarrier: fix missing local execution of ipi_sync_rq_state()
sched: Simplify set_affinity_pending refcounts
sched: Fix affine_move_task() self-concurrency
sched: Optimize migration_cpu_stop()
sched: Collate affine_move_task() stoppers
sched: Simplify migration_cpu_stop()
sched: Fix migration_cpu_stop() requeueing
Pull objtool fix from Thomas Gleixner:
"A single objtool fix to handle the PUSHF/POPF validation correctly for
the paravirt changes which modified arch_local_irq_restore not to use
popf"
* tag 'objtool-urgent-2021-03-14' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
objtool,x86: Fix uaccess PUSHF/POPF validation
Pull locking fixes from Thomas Gleixner:
"A couple of locking fixes:
- A fix for the static_call mechanism so it handles unaligned
addresses correctly.
- Make u64_stats_init() a macro so every instance gets a seperate
lockdep key.
- Make seqcount_latch_init() a macro as well to preserve the static
variable which is used for the lockdep key"
* tag 'locking-urgent-2021-03-14' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
seqlock,lockdep: Fix seqcount_latch_init()
u64_stats,lockdep: Fix u64_stats_init() vs lockdep
static_call: Fix the module key fixup
Pull perf fixes from Borislav Petkov:
- Make sure PMU internal buffers are flushed for per-CPU events too and
properly handle PID/TID for large PEBS.
- Handle the case properly when there's no PMU and therefore return an
empty list of perf MSRs for VMX to switch instead of reading random
garbage from the stack.
* tag 'perf_urgent_for_v5.12-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/perf: Use RET0 as default for guest_get_msrs to handle "no PMU" case
perf/x86/intel: Set PERF_ATTACH_SCHED_CB for large PEBS and LBR
perf/core: Flush PMU internal buffers for per-CPU events
Pull EFI fix from Ard Biesheuvel via Borislav Petkov:
"Fix an oversight in the handling of EFI_RT_PROPERTIES_TABLE, which was
added v5.10, but failed to take the SetVirtualAddressMap() RT service
into account"
* tag 'efi-urgent-for-v5.12-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
efi: stub: omit SetVirtualAddressMap() if marked unsupported in RT_PROP table
Pull x86 fixes from Borislav Petkov:
- A couple of SEV-ES fixes and robustifications: verify usermode stack
pointer in NMI is not coming from the syscall gap, correctly track
IRQ states in the #VC handler and access user insn bytes atomically
in same handler as latter cannot sleep.
- Balance 32-bit fast syscall exit path to do the proper work on exit
and thus not confuse audit and ptrace frameworks.
- Two fixes for the ORC unwinder going "off the rails" into KASAN
redzones and when ORC data is missing.
* tag 'x86_urgent_for_v5.12_rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/sev-es: Use __copy_from_user_inatomic()
x86/sev-es: Correctly track IRQ states in runtime #VC handler
x86/sev-es: Check regs->sp is trusted before adjusting #VC IST stack
x86/sev-es: Introduce ip_within_syscall_gap() helper
x86/entry: Fix entry/exit mismatch on failed fast 32-bit syscalls
x86/unwind/orc: Silence warnings caused by missing ORC data
x86/unwind/orc: Disable KASAN checking in the ORC unwinder, part 2
Pull powerpc fixes from Michael Ellerman:
"Some more powerpc fixes for 5.12:
- Fix wrong instruction encoding for lis in ppc_function_entry(),
which could potentially lead to missed kprobes.
- Fix SET_FULL_REGS on 32-bit and 64e, which prevented ptrace of
non-volatile GPRs immediately after exec.
- Clean up a missed SRR specifier in the recent interrupt rework.
- Don't treat unrecoverable_exception() as an interrupt handler, it's
called from other handlers so shouldn't do the interrupt entry/exit
accounting itself.
- Fix build errors caused by missing declarations for
[en/dis]able_kernel_vsx().
Thanks to Christophe Leroy, Daniel Axtens, Geert Uytterhoeven, Jiri
Olsa, Naveen N. Rao, and Nicholas Piggin"
* tag 'powerpc-5.12-3' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
powerpc/traps: unrecoverable_exception() is not an interrupt handler
powerpc: Fix missing declaration of [en/dis]able_kernel_vsx()
powerpc/64s/exception: Clean up a missed SRR specifier
powerpc: Fix inverted SET_FULL_REGS bitop
powerpc/64s: Use symbolic macros for function entry encoding
powerpc/64s: Fix instruction encoding for lis in ppc_function_entry()
Pull KVM fixes from Paolo Bonzini:
"More fixes for ARM and x86"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
KVM: LAPIC: Advancing the timer expiration on guest initiated write
KVM: x86/mmu: Skip !MMU-present SPTEs when removing SP in exclusive mode
KVM: kvmclock: Fix vCPUs > 64 can't be online/hotpluged
kvm: x86: annotate RCU pointers
KVM: arm64: Fix exclusive limit for IPA size
KVM: arm64: Reject VM creation when the default IPA size is unsupported
KVM: arm64: Ensure I-cache isolation between vcpus of a same VM
KVM: arm64: Don't use cbz/adr with external symbols
KVM: arm64: Fix range alignment when walking page tables
KVM: arm64: Workaround firmware wrongly advertising GICv2-on-v3 compatibility
KVM: arm64: Rename __vgic_v3_get_ich_vtr_el2() to __vgic_v3_get_gic_config()
KVM: arm64: Don't access PMSELR_EL0/PMUSERENR_EL0 when no PMU is available
KVM: arm64: Turn kvm_arm_support_pmu_v3() into a static key
KVM: arm64: Fix nVHE hyp panic host context restore
KVM: arm64: Avoid corrupting vCPU context register in guest exit
KVM: arm64: nvhe: Save the SPE context early
kvm: x86: use NULL instead of using plain integer as pointer
KVM: SVM: Connect 'npt' module param to KVM's internal 'npt_enabled'
KVM: x86: Ensure deadline timer has truly expired before posting its IRQ
Sean Christopherson
Kai Huang
Sean Christopherson
Ira Weiny
Tianjia Zhang
Dave Hansen
Jarkko Sakkinen
Linus Torvalds
Alexey Dobriyan