free_ldt_pgtables() uses the MMU gather API for batching TLB flushes
over the call to free_pgd_range(). However, tlb_gather_mmu() expects
to operate on user addresses and so passing LDT_{BASE,END}_ADDR will
confuse the range setting logic in __tlb_adjust_range(), causing the
gather to identify a range starting at TASK_SIZE. Such a large range
will be converted into a 'fullmm' flush by the low-level invalidation
code, so change the caller to invoke tlb_gather_mmu_fullmm() directly.

Cc: Thomas Gleixner <redacted>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Yu Zhao <redacted>
Signed-off-by: Will Deacon <will@kernel.org>
---
 arch/x86/kernel/ldt.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/arch/x86/kernel/ldt.c b/arch/x86/kernel/ldt.c
index 7ad9834e0d95..aa15132228da 100644
--- a/arch/x86/kernel/ldt.c
+++ b/arch/x86/kernel/ldt.c

@@ -398,7 +398,13 @@ static void free_ldt_pgtables(struct mm_struct *mm)
 	if (!boot_cpu_has(X86_FEATURE_PTI))
 		return;
 
-	tlb_gather_mmu(&tlb, mm);
+	/*
+	 * Although free_pgd_range() is intended for freeing user
+	 * page-tables, it also works out for kernel mappings on x86.
+	 * We use tlb_gather_mmu_fullmm() to avoid confusing the
+	 * range-tracking logic in __tlb_adjust_range().
+	 */
+	tlb_gather_mmu_fullmm(&tlb, mm);
 	free_pgd_range(&tlb, start, end, start, end);
 	tlb_finish_mmu(&tlb);
 #endif

-- 
2.30.0.365.g02bc693789-goog