I found the problem...
The real VHPT insertion is done based on the machine PTE returned from
translate_domain_pte, which does the appropriate offset calculations.
However, the insertion into the one-entry TLB uses the original PTE, but
the page size has been reset to PAGE_SIZE [1]. Thus the entry in the
one-entry TLB incorrectly maps the PAGE_SIZE sub-page which was faulted
on to the PAGE_SIZE sub-page at the bottom of the superpage.
I think it makes most sense to simply use the original itir when
inserting into the single-entry TLB, as per attached patch. I've moved
the vcpu_set_tr_entry calls up a level into vcpu_itc_d and vcpu_itc_i;
the third caller previously used the 4 flag to specify "don't do that".
Matt
[1] In fact, this is enforced twice, once in translate_domain_pte and
again in vcpu_itc_no_srlz.
diff -r 0114b372dfae xen/arch/ia64/xen/faults.c
--- a/xen/arch/ia64/xen/faults.c Wed Nov 22 10:13:31 2006 -0700
+++ b/xen/arch/ia64/xen/faults.c Thu Nov 30 14:51:33 2006 +1100
@@ -215,7 +215,7 @@ void ia64_do_page_fault(unsigned long ad
unsigned long m_pteval;
m_pteval = translate_domain_pte(pteval, address, itir,
&logps, &entry);
- vcpu_itc_no_srlz(current, (is_data ? 2 : 1) | 4,
+ vcpu_itc_no_srlz(current, is_data ? 2 : 1,
address, m_pteval, pteval, logps, &entry);
if ((fault == IA64_USE_TLB && !current->arch.dtlb.pte.p) ||
p2m_entry_retry(&entry)) {
diff -r 0114b372dfae xen/arch/ia64/xen/vcpu.c
--- a/xen/arch/ia64/xen/vcpu.c Wed Nov 22 10:13:31 2006 -0700
+++ b/xen/arch/ia64/xen/vcpu.c Thu Nov 30 14:54:11 2006 +1100
@@ -2181,14 +2181,6 @@ vcpu_itc_no_srlz(VCPU * vcpu, u64 IorD,
else
vhpt_insert(vaddr, pte, PAGE_SHIFT << 2);
#endif
- if (IorD & 0x4) /* don't place in 1-entry TLB */
- return;
- if (IorD & 0x1) {
- vcpu_set_tr_entry(&PSCBX(vcpu, itlb), mp_pte, ps << 2, vaddr);
- }
- if (IorD & 0x2) {
- vcpu_set_tr_entry(&PSCBX(vcpu, dtlb), mp_pte, ps << 2, vaddr);
- }
}
IA64FAULT vcpu_itc_d(VCPU * vcpu, u64 pte, u64 itir, u64 ifa)
@@ -2215,6 +2207,7 @@ IA64FAULT vcpu_itc_d(VCPU * vcpu, u64 pt
vcpu_flush_tlb_vhpt_range(ifa, logps);
goto again;
}
+ vcpu_set_tr_entry(&PSCBX(vcpu, dtlb), pte, itir, ifa);
return IA64_NO_FAULT;
}
@@ -2241,6 +2234,7 @@ IA64FAULT vcpu_itc_i(VCPU * vcpu, u64 pt
vcpu_flush_tlb_vhpt_range(ifa, logps);
goto again;
}
+ vcpu_set_tr_entry(&PSCBX(vcpu, itlb), pte, itir, ifa);
return IA64_NO_FAULT;
}
_______________________________________________
Xen-ia64-devel mailing list
[email protected]
http://lists.xensource.com/xen-ia64-devel
