KVM x86 fixes for 7.1-rcN

 - Include the kernel's linux/mman.h in KVM selftests to ensure MADV_COLLAPSE
   is defined, as older libc versions may not provide it.

 - Include execinfo.h if and only if KVM selftests are building against glibc,
   and provide a test_dump_stack() for non-glibc builds.

 - Fudge around an RCU splat in the emegerncy reboot code that is technically
   a legitimate flaw, but in practice is a non-issue and fixing the flaw, e.g.
   by adding locking, would incur meaningful risk, i.e. do more harm than good.

 - Rate-limit global clock updates once again (but without delayed work), as
   KVM was subtly relying on the old rate-limiting for NPT correction to guard
   against "update storms" when running without a master clock on systems with
   overcommitted CPUs.

 - Fix a brown paper bag goof where KVM checked if ERAPS is "dirty" instead of
   marking it dirty when emulating INVPCID.

 - Flush the TLB when transitioning from xAVIC => x2AVIC to ensure the CPU TLB
   doesn't contain AVIC-tagged entries for the APIC base GPA.

<bits/endian.h> is a glibc-internal header that explicitly states it
should never be included directly:

  #error "Never use <bits/endian.h> directly; include <endian.h> instead."

Replace it with the correct public header <endian.h> which works on
all C libraries including musl. Building KVM selftests with musl-gcc
fails with:

  lib/elf.c:10:10: fatal error: bits/endian.h: No such file or directory

Fixes: 6089ae0bd5e1 ("kvm: selftests: add sync_regs_test")
Signed-off-by: Hisam Mehboob <hisamshar@gmail.com>
Message-ID: <20260409164020.1575176-4-hisamshar@gmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

The backtrace() function and execinfo.h are GNU extensions available
in glibc but not in non-glibc C libraries such as musl. Building KVM
selftests with musl-gcc fails with:

  lib/assert.c:9:10: fatal error: execinfo.h: No such file or directory

Fix this by guarding the inclusion of execinfo.h and the stack dumping
logic under #ifdef __GLIBC__. For non-glibc builds, provide a local
stub for test_dump_stack().

Suggested-by: Aqib Faruqui <aqibaf@amazon.com>
Suggested-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Hisam Mehboob <hisamshar@gmail.com>
Link: https://patch.msgid.link/20260409153846.1502656-2-hisamshar@gmail.com
Signed-off-by: Sean Christopherson <seanjc@google.com>

Include both linux/mman.h (the kernel provided version) and sys/mman.h (the
libc provided version) throughout KVM selftests, by way of kvm_syscalls.h
(which should have been including sys/mman.h anyways).  Pulling in the
kernel's version fixes compilation errors with the guest_memfd test on
older versions of libc due to a recent commit adding MADV_COLLAPSE testing.

  In file included from include/kvm_util.h:8,
                   from guest_memfd_test.c:21:
  guest_memfd_test.c: In function ‘test_collapse’:
  guest_memfd_test.c:219:47: error: ‘MADV_COLLAPSE’ undeclared (first use in this function); did you mean ‘MADV_COLD’?
      219 |         TEST_ASSERT_EQ(madvise(mem, pmd_size, MADV_COLLAPSE), -1);
          |                                               ^~~~~~~~~~~~~
    include/test_util.h:62:16: note: in definition of macro ‘TEST_ASSERT_EQ’
       62 |         typeof(a) __a = (a);                                            \
          |                ^
    guest_memfd_test.c:219:47: note: each undeclared identifier is reported only once for each function it appears in
      219 |         TEST_ASSERT_EQ(madvise(mem, pmd_size, MADV_COLLAPSE), -1);
          |                                               ^~~~~~~~~~~~~
    include/test_util.h:62:16: note: in definition of macro ‘TEST_ASSERT_EQ’
       62 |         typeof(a) __a = (a);                                            \
          |                ^

Route the includes through kvm_syscalls.h to try and avoid a future game
of whack-a-mole, i.e. so that future expansion of test coverage doesn't run
into the same problem.

To discourage use of sys/mman.h, opportunistically include the kernel's
version of mman.h in test_util.h as it only needs MAP_SHARED, i.e. only
needs the full set of kernel defs, not the libc syscall wrappers.

Fixes: 9830209b4ae8 ("KVM: selftests: Test MADV_COLLAPSE on guest_memfd")
Reported-by: Rick Edgecombe <rick.p.edgecombe@intel.com>
Closes: https://lore.kernel.org/all/20260427204313.50741-1-rick.p.edgecombe@intel.com
Link: https://patch.msgid.link/20260428012503.1213654-1-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>

Replace all variations of "paddr" variables in KVM selftests with "gpa",
with the exception of the ELF structures, as those fields are not specific
to guest virtual addresses, to complete the conversion from vm_paddr_t to
gpa_t.

No functional change intended.

Link: https://patch.msgid.link/20260420212004.3938325-20-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>

In x86's nested TDP APIs, use the appropriate gpa_t typedef and rename
variables from nested_paddr to l2_gpa to match KVM x86's nomenclature.

No functional change intended.

Link: https://patch.msgid.link/20260420212004.3938325-19-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>

Use gpa_t instead of u64 for obvious declarations of GPA variables.

No functional change intended.

Link: https://patch.msgid.link/20260420212004.3938325-18-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>

Replace all variations of "vaddr" variables in KVM selftests with "gva",
with the exception of the ELF structures, as those fields are not specific
to guest virtual addresses, to complete the conversion from vm_vaddr_t to
gva_t.

Opportunistically use gva_t instead of u64 for relevant variables, and
fixup indentation as appropriate.

No functional change intended.

Link: https://patch.msgid.link/20260420212004.3938325-17-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>

Now that KVM selftests use gva_t instead of vm_vaddr_t, rename the helper
for populating the initial GVA bitmap to drop the defunct terminology and
use "vm" for the scope.

Opportunistically fixup the declaration of the API, which has been broken
since day 1.  The flaw went unnoticed because the sole caller is defined
after the weak version, i.e. can see the prototype without a previous
declaration.

No functional change intended.

Fixes: e8b9a055fa04 ("KVM: arm64: selftests: Align VA space allocator with TTBR0")
Link: https://patch.msgid.link/20260420212004.3938325-14-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>

Now that KVM selftests use gva_t instead of vm_vaddr_t, rename the API
for finding an unused range of virtual memory to drop the defunct
terminology and use "vm" for the scope.

Opportunistically clean up the function comment to drop superfluous
and redundant information.

No functional change intended.

Link: https://patch.msgid.link/20260420212004.3938325-13-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>