|
From: Zhang, X. <xia...@in...> - 2008-02-01 04:01:35
|
Akio Takebe wrote:
> Hi, Xiantao
>=20
>> +void thash_vhpt_insert(VCPU *v, u64 pte, u64 itir, u64 va, int
>> type) +{ + u64 phy_pte, psr;
>> + ia64_rr mrr;
>> +
>> + mrr.val =3D ia64_get_rr(va);
>> + phy_pte =3D translate_phy_pte(&pte, itir, va);
>> +
>> + if (itir_ps(itir) >=3D mrr.ps) {
>> + vhpt_insert(phy_pte, itir, va, pte);
>> + } else {
>> + phy_pte &=3D ~PAGE_FLAGS_RV_MASK;
>> + psr =3D ia64_clear_ic();
>> + ia64_itc(type, va, phy_pte, itir_ps(itir));
>> + ia64_set_psr(psr);
>> + ia64_srlz_i();
>> + }
>> +}
> You add ia64_srlz_i() into ia64_set_psr() with [02]patch.
> So is this a redundancy if the patch is applied?
Yes, we need to remove it. Once the second patch is picked up. Thanks
Xiantao
|