summaryrefslogtreecommitdiff
path: root/mm/huge_memory.c
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2020-01-31 09:30:41 -0800
committerLinus Torvalds <torvalds@linux-foundation.org>2020-01-31 09:30:41 -0800
commite813e65038389b66d2f8dd87588694caf8dc2923 (patch)
tree4595d8ebaf672b79b412bd663a13907fd785478d /mm/huge_memory.c
parentccaaaf6fe5a5e1fffca5cca0f3fc4ec84d7ae752 (diff)
parent4cbc418a44d5067133271bb6eeac2382f2bf94f7 (diff)
Merge tag 'kvm-5.6-1' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Pull KVM updates from Paolo Bonzini: "This is the first batch of KVM changes. ARM: - cleanups and corner case fixes. PPC: - Bugfixes x86: - Support for mapping DAX areas with large nested page table entries. - Cleanups and bugfixes here too. A particularly important one is a fix for FPU load when the thread has TIF_NEED_FPU_LOAD. There is also a race condition which could be used in guest userspace to exploit the guest kernel, for which the embargo expired today. - Fast path for IPI delivery vmexits, shaving about 200 clock cycles from IPI latency. - Protect against "Spectre-v1/L1TF" (bring data in the cache via speculative out of bound accesses, use L1TF on the sibling hyperthread to read it), which unfortunately is an even bigger whack-a-mole game than SpectreV1. Sean continues his mission to rewrite KVM. In addition to a sizable number of x86 patches, this time he contributed a pretty large refactoring of vCPU creation that affects all architectures but should not have any visible effect. s390 will come next week together with some more x86 patches" * tag 'kvm-5.6-1' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (204 commits) x86/KVM: Clean up host's steal time structure x86/KVM: Make sure KVM_VCPU_FLUSH_TLB flag is not missed x86/kvm: Cache gfn to pfn translation x86/kvm: Introduce kvm_(un)map_gfn() x86/kvm: Be careful not to clear KVM_VCPU_FLUSH_TLB bit KVM: PPC: Book3S PR: Fix -Werror=return-type build failure KVM: PPC: Book3S HV: Release lock on page-out failure path KVM: arm64: Treat emulated TVAL TimerValue as a signed 32-bit integer KVM: arm64: pmu: Only handle supported event counters KVM: arm64: pmu: Fix chained SW_INCR counters KVM: arm64: pmu: Don't mark a counter as chained if the odd one is disabled KVM: arm64: pmu: Don't increment SW_INCR if PMCR.E is unset KVM: x86: Use a typedef for fastop functions KVM: X86: Add 'else' to unify fastop and execute call path KVM: x86: inline memslot_valid_for_gpte KVM: x86/mmu: Use huge pages for DAX-backed files KVM: x86/mmu: Remove lpage_is_disallowed() check from set_spte() KVM: x86/mmu: Fold max_mapping_level() into kvm_mmu_hugepage_adjust() KVM: x86/mmu: Zap any compound page when collapsing sptes KVM: x86/mmu: Remove obsolete gfn restoration in FNAME(fetch) ...
Diffstat (limited to 'mm/huge_memory.c')
-rw-r--r--mm/huge_memory.c11
1 files changed, 11 insertions, 0 deletions
diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index a88093213674..f39689a29128 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -527,6 +527,17 @@ void prep_transhuge_page(struct page *page)
set_compound_page_dtor(page, TRANSHUGE_PAGE_DTOR);
}
+bool is_transparent_hugepage(struct page *page)
+{
+ if (!PageCompound(page))
+ return 0;
+
+ page = compound_head(page);
+ return is_huge_zero_page(page) ||
+ page[1].compound_dtor == TRANSHUGE_PAGE_DTOR;
+}
+EXPORT_SYMBOL_GPL(is_transparent_hugepage);
+
static unsigned long __thp_get_unmapped_area(struct file *filp,
unsigned long addr, unsigned long len,
loff_t off, unsigned long flags, unsigned long size)