search for: mod_l1_entry

...> add the missing vram pages in the guest. Maybe this hypercall is failing > > in your case? > > > Hi, > > I continue to invesgate this bug and find hypercall_mmu_update in > qemu_remap_bucket(xc_map_foreign_bulk) is failing: > > do_mmu_update > ->mod_l1_entry > -> if ( !p2m_is_ram(p2mt) || unlikely(mfn == INVALID_MFN) ) > return -EINVAL; > > mfn==INVALID_MFN, because : > > mod_l1_entry > ->gfn_to_mfn(p2m_get_hostp2m(pg_dom), l1e_get_pfn(nl1e), &p2mt)); > -&...

...he guest. Maybe this hypercall is failing >> > in your case? >> >> >> Hi, >> >>    I continue to invesgate this bug and find hypercall_mmu_update in >> qemu_remap_bucket(xc_map_foreign_bulk) is failing: >> >> do_mmu_update >>       ->mod_l1_entry >>              ->  if ( !p2m_is_ram(p2mt) || unlikely(mfn == INVALID_MFN) ) >>                          return -EINVAL; >> >>    mfn==INVALID_MFN, because : >> >> mod_l1_entry >>       ->gfn_to_mfn(p2m_get_hostp2m(pg_dom), l1e_get_pfn(nl1e), &p2m...

...devices via IOMMU. - x86/cpuidle: add Westmere-EX support to hw residencies reading logic Hardware support. - x86/cpuidle: deny access to the I/O port used for EM_SYSIO - xen: Fix schedule()'s grabbing of the schedule lock - SVM: fix performance decrease with asid assignment - x86/mm: fix mod_l1_entry() return value when encountering r/o MMIO page - x86/PCI: fix guest_io_read() when pci_cfg_ok() denies access - passthrough: fix xsm-related oversight Bastian -- A father doesn't destroy his children. -- Lt. Carolyn Palamas, "Who Mourns for Adonais?", stardate 3468.1.

While working on a database scaling problem using a SAP database test suite we discovered that these enterprise level large shared memory databases are very heavy users of mprotect(), to the extent that the performance overhead in current Xenolinux impacts scaling beyond a few cpus quite badly. A single cpu run under Xen has a nominal impact, but scaling out to 8 cpus results in a performance of

...unsigned long val, + l1_pgentry_t val, unsigned long flags) { struct exec_domain *ed = current; @@ -2050,8 +2054,7 @@ int do_update_va_mapping(unsigned long v * the case of updating L2 entries. */ - if ( unlikely(!mod_l1_entry(&linear_pg_table[l1_linear_offset(va)], - mk_l1_pgentry(val))) ) + if ( unlikely(!mod_l1_entry(&linear_pg_table[l1_linear_offset(va)],val)) ) rc = -EINVAL; if ( unlikely(shadow_mode_enabled(d)) ) @@ -2075,7 +2078,7 @@ int do_update_va_mappin...