KVM: arm64: nv: Fix incorrect VNCR invalidation range calculation

The code for invalidating VNCR entries in both kvm_invalidate_vncr_ipa()
and invalidate_vncr_va() incorrectly uses a bitwise AND with `(size - 1)`
instead of `~(size - 1)` to align the start address. This results
in masking the address bits instead of aligning them down to the start
of the block.

This bug may cause stale VNCR TLB entries to remain valid even after a
TLBI or MMU notifier, leading to incorrect memory translation and
unexpected guest behavior.

Credit to Team 0xB6 in bob14: DongHa Lee, Gyujeong Jin, Daehyeon Ko,
Geonha Lee, Hyungyu Oh, and Jaewon Yang.

Reviewed-by: Marc Zyngier <[email protected]>
Signed-off-by: Dongha Lee <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Oliver Upton <[email protected]>

Author: GAP-dev

arch/arm64/kvm/nested.c

@@ -847,7 +847,7 @@ static void kvm_invalidate_vncr_ipa(struct kvm *kvm, u64 start, u64 end)
 
 		ipa_size = ttl_to_size(pgshift_level_to_ttl(vt->wi.pgshift,
 							    vt->wr.level));
-		ipa_start = vt->wr.pa & (ipa_size - 1);
+		ipa_start = vt->wr.pa & ~(ipa_size - 1);
 		ipa_end = ipa_start + ipa_size;
 
 		if (ipa_end <= start || ipa_start >= end)
@@ -887,7 +887,7 @@ static void invalidate_vncr_va(struct kvm *kvm,
 
 		va_size = ttl_to_size(pgshift_level_to_ttl(vt->wi.pgshift,
 							   vt->wr.level));
-		va_start = vt->gva & (va_size - 1);
+		va_start = vt->gva & ~(va_size - 1);
 		va_end = va_start + va_size;
 
 		switch (scope->type) {