> -----Original Message----- > From: Oliver Upton [mailto:oliver.upton@xxxxxxxxx] > Sent: 15 September 2023 23:06 > To: Shameerali Kolothum Thodi <shameerali.kolothum.thodi@xxxxxxxxxx> > Cc: kvmarm@xxxxxxxxxxxxxxx; kvm@xxxxxxxxxxxxxxx; > linux-arm-kernel@xxxxxxxxxxxxxxxxxxx; maz@xxxxxxxxxx; will@xxxxxxxxxx; > catalin.marinas@xxxxxxx; james.morse@xxxxxxx; > suzuki.poulose@xxxxxxx; yuzenghui <yuzenghui@xxxxxxxxxx>; zhukeqian > <zhukeqian1@xxxxxxxxxx>; Jonathan Cameron > <jonathan.cameron@xxxxxxxxxx>; Linuxarm <linuxarm@xxxxxxxxxx> > Subject: Re: [RFC PATCH v2 2/8] KVM: arm64: Add > KVM_PGTABLE_WALK_HW_DBM for HW DBM support > > Hi Shameer, > > On Fri, Aug 25, 2023 at 10:35:22AM +0100, Shameer Kolothum wrote: > > KVM_PGTABLE_WALK_HW_DBM - Indicates page table walk is for HW > DBM > > related updates. > > > > No functional changes here. Only apply any HW DBM bit updates to last > > level only. These will be used by a future commit where we will add > > support for HW DBM. > > > > Signed-off-by: Shameer Kolothum > <shameerali.kolothum.thodi@xxxxxxxxxx> > > --- > > arch/arm64/include/asm/kvm_pgtable.h | 3 +++ > > arch/arm64/kvm/hyp/pgtable.c | 10 ++++++++++ > > 2 files changed, 13 insertions(+) > > > > diff --git a/arch/arm64/include/asm/kvm_pgtable.h > b/arch/arm64/include/asm/kvm_pgtable.h > > index d3e354bb8351..3f96bdd2086f 100644 > > --- a/arch/arm64/include/asm/kvm_pgtable.h > > +++ b/arch/arm64/include/asm/kvm_pgtable.h > > @@ -219,6 +219,8 @@ typedef bool (*kvm_pgtable_force_pte_cb_t)(u64 > addr, u64 end, > > * @KVM_PGTABLE_WALK_SKIP_CMO: Visit and update table > entries > > * without Cache maintenance > > * operations required. > > + * @KVM_PGTABLE_WALK_HW_DBM: Indicates that the attribute > update is > > + * HW DBM related. > > */ > > enum kvm_pgtable_walk_flags { > > KVM_PGTABLE_WALK_LEAF = BIT(0), > > @@ -228,6 +230,7 @@ enum kvm_pgtable_walk_flags { > > KVM_PGTABLE_WALK_HANDLE_FAULT = BIT(4), > > KVM_PGTABLE_WALK_SKIP_BBM_TLBI = BIT(5), > > KVM_PGTABLE_WALK_SKIP_CMO = BIT(6), > > + KVM_PGTABLE_WALK_HW_DBM = BIT(7), > > }; > > Rather than making this DBM specific, call it > KVM_PGTABLE_WALK_FORCE_PTE > and get rid of stage2_map_data::force_pte. Then it becomes immediately > obvious what this flag implies. Ok. Will change accordingly. Thanks, Shameer
