search for: p2m_domain

From: Zhang Xiantao <xiantao.zhang@intel.com> With virtual EPT support, L1 hyerpvisor can use EPT hardware for L2 guest''s memory virtualization. In this way, L2 guest''s performance can be improved sharply. According to our testing, some benchmarks can show > 5x performance gain. Signed-off-by: Zhang Xiantao <xiantao.zhang@intel.com> Zhang Xiantao (11):

...* 256 concurrent domains. */ +DECLARE_BITMAP(vmid_mask, MAX_VMID); + +void p2m_vmid_allocator_init(void) +{ + /* VMID 0 is reserved */ + set_bit(0, vmid_mask); +} + +/* p2m_alloc|free_vmid must both be called with the p2m->lock */ +static int p2m_alloc_vmid(struct domain *d) +{ + struct p2m_domain *p2m = &d->arch.p2m; + + int nr = find_first_zero_bit(vmid_mask, MAX_VMID); + + ASSERT(nr > 0); /* VMID is reserved */ + + if ( nr == MAX_VMID ) + { + printk("p2m.c: dom%d: VMID pool exhausted\n", d->domain_id); + return -EBUSY; + } + + set_bit...

Populate-on-demand: Check pages being returned by the balloon driver This patch series is the second result of my work last summer on decreasing fragmentation of superpages in a guests'' p2m when using populate-on-demand. This patch series is against 4.1; I''m posting it to get feedback on the viability of getting a ported version of this patch into 4.2. As with the previous

This patch series tidies up a few bits ofthe nested p2m code. The main thing it does is reorganize the locking so that most of the changes to nested p2m tables happen only under the p2m lock, and the nestedp2m lock is only needed to reassign p2m tables to new cr3 values. Changes since v1: - a few minor fixes - more sensible flushing policy in p2m_get_nestedp2m() - smoke-tested this time!

...ES : 0; > } > > +/* When populating a new superpage, look at recently populated superpages > + * hoping that they''ve been zeroed. This will snap up zeroed pages as > soon as > + * the guest OS is done with them. */ > +static void > +p2m_pod_check_last_super(struct p2m_domain *p2m, unsigned long > gfn_aligned) > +{ > + unsigned long check_gfn; > + > + ASSERT(p2m->pod.last_populated_index < POD_HISTORY_MAX); > + > + check_gfn = p2m->pod.last_populated[p2m->pod.last_populated_index]; > + > + p2m->pod.last_populated[p2m...

This patch series restores the use of superpages when restoring or migrating a VM, while retaining efficient batching of 4k pages when superpages are not appropriate or available. Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com> _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel

...@gmail.com> --- xen/arch/x86/mm/p2m.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/xen/arch/x86/mm/p2m.c b/xen/arch/x86/mm/p2m.c index de1dd82..ad1f676 100644 --- a/xen/arch/x86/mm/p2m.c +++ b/xen/arch/x86/mm/p2m.c @@ -490,15 +490,13 @@ void p2m_teardown(struct p2m_domain *p2m) void p2m_final_teardown(struct domain *d) { - /* Iterate over all p2m tables per domain */ - struct p2m_domain *p2m = p2m_get_hostp2m(d); - if ( p2m ) - p2m_teardown_hostp2m(d); - /* We must teardown unconditionally because * we initialise them unconditionally....

-- Advanced Micro Devices GmbH Sitz: Dornach, Gemeinde Aschheim, Landkreis München Registergericht München, HRB Nr. 43632 WEEE-Reg-Nr: DE 12919551 Geschäftsführer: Alberto Bozzo, Andrew Bowd _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel

There has been some progress, but still no joy. Definitely not intended for inclusion at this point. Tim, Wei, I added a Xen command line toggle to disable IOMMU and P2M table sharing. Tim, I verified that changes to p2m-pt.c don''t break shadow mode (64bit hypervisor and Win 7 guest). Hongkaixing, I incorporated your suggestion in patch 2, so I should add your Signed-off-by eventually.

ging_prep ensures that an mfn is backing the paged-out gfn, and transitions to the next state in the paging state machine for this page. Foreign mappings of the gfn will now succeed. This is the key idea, as it allows the pager to now map the gfn and fill in its contents. Unfortunately, it also allows any other foreign mapper to map the gfn and read its contents. This is particularly dangerous

Hi, V6: The only change from V5 is in patch #6: - changed comment to reflect autoxlate - removed a redundant ASSERT - reworked logic a bit so that get_page_from_gfn() is called with NULL for p2m type as before. arm has ASSERT wanting it to be NULL. Tim: patch 4 needs your approval. Daniel: patch 5 needs your approval. These patches implement PVH dom0. Patches 1 and 2

George has release acked all of these. Otherwise mostly minor updates this time around. Summary: A == acked, M == modified A xen: arm64: Add 8250 earlyprintk support A xen: arm64: Add Basic Platform support for APM X-Gene Storm. A xen: arm64: Add APM implementor id to processor implementers. M xen: arm: add a quirk to handle platforms with unusual GIC layout A xen: arm: allow platform

The following two patches are about how to speed up in xenpaging when page in pages. On suse11-64 with 4G memory,if we page out 2G pages,it will cost about 15.5 seconds, but take 2088 seconds to finish paging in.If page-in costs too much time,it will cause unmesurable problems when vm or dom0 access the paged_out page,such as BSOD,crash. What鈥檚 more,the dom0 is always in high I/O pressure.

Another bundle of issues from Coverity triage. The first one is in x86/mm, and looks scarier than it is. The others are all in xen/drivers and AFAICT are pretty minor. Cheers, Tim.

This patch series includes a series of bugs related to p2m, ept, and PoD code which were found as part of our XenServer product testing. Each of these fixes actual bugs, and the 3.4-based version of the patch has been tested thoroughly. (There may be bugs in porting the patches, but most of them are simple enough as to make it unlikely.) Each patch is conceptually independent, so they can each

The following two patches allow the parallel use of memsharing, xenpaging and xen-access by using an independent ring buffer for each feature. Please review. v2: - update mem_event_check_ring arguments, check domain rather than domain_id - check ring_full first because its value was just evaluated - check if ring buffer is initialized before calling mem_access_domctl/mem_paging_domctl

I''m afraid this series is rather a grab bag and it is distressingly large at this stage. With this series I can boot an Xgene board until it fails to find its SATA controller. This is a dom0 issue for which patches are pending from APM (/me nudges Anup). As well as the APM specific platform stuff there are also some generic improvements which were either necessary or useful during this

...entry, pte); return 0; @@ -408,6 +414,206 @@ unsigned long gmfn_to_mfn(struct domain *d, unsigned long gpfn) return p >> PAGE_SHIFT; } +/* Change types across all p2m entries in a domain */ +void p2m_change_entry_type_global(struct domain *d, enum mg nt) +{ + struct p2m_domain *p2m = &d->arch.p2m; + paddr_t ram_base; + int i1, i2, i3; + int first_index, second_index, third_index; + lpae_t *first = __map_domain_page(p2m->first_level); + lpae_t pte, *second = NULL, *third = NULL; + + get_gma_start_end(d, &ram_base, NULL); + + fir...

From: Yang Zhang <yang.z.zhang@Intel.com> The following patches fix the issue that fail to boot L2 guest on APIC-v available machine. The main problem is that with APIC-v, virtual interrupt inject L1 is totally through APIC-v. But if virtual interrupt is arrived when L2 is running, L1 will detect interrupt through vmexit with reason external interrupt. If this happens, we should update

These two patches were originally part of the XSM series that I have posted, and remain prerequisites for that series. However, they are independent of the XSM changes and are a useful simplification regardless of the use of XSM. The Acked-bys on these patches were provided before rebasing them over the copyback changes in 26268:1b72138bddda, which had minor conflicts that I resolved. [PATCH