similar to: pte_offset_map + lazy mmu

This patch is a significant performance improvement for the m2p_override: about 6% using the gntdev device. Each m2p_add/remove_override call issues a MULTI_grant_table_op and a __flush_tlb_single if kmap_op != NULL. Batching all the calls together is a great performance benefit because it means issuing one hypercall total rather than two hypercall per page. If paravirt_lazy_mode is set

Implement lazy MMU update hooks which are SMP safe for both direct and shadow page tables. The idea is that PTE updates and page invalidations while in lazy mode can be batched into a single hypercall. We use this in VMI for shadow page table synchronization, and it is a win. It also can be used by PPC and for direct page tables on Xen. For SMP, the enter / leave must happen under protection

(This is a continuation of "[PATCH v9] xen/grant-table: Avoid m2p_override during mapping") The grant mapping API does m2p_override unnecessarily: only gntdev needs it, for blkback and future netback patches it just cause a lock contention, as those pages never go to userspace. Therefore this series does the following: - the bulk of the original function (everything after the mapping

(This is a continuation of "[PATCH v9] xen/grant-table: Avoid m2p_override during mapping") The grant mapping API does m2p_override unnecessarily: only gntdev needs it, for blkback and future netback patches it just cause a lock contention, as those pages never go to userspace. Therefore this series does the following: - the bulk of the original function (everything after the mapping

[ REPOST: Apologies to anyone who has seen this before. It didn't make it onto any of the lists it should have. -J ] Hi Andrew, This series of patches lays the basic ground work for the paravirtualized kernel patches coming later on. I think this lot is ready for the rough-and-tumble world of the -mm tree. For the most part, these patches do nothing or very little. The patches should

[ REPOST: Apologies to anyone who has seen this before. It didn't make it onto any of the lists it should have. -J ] Hi Andrew, This series of patches lays the basic ground work for the paravirtualized kernel patches coming later on. I think this lot is ready for the rough-and-tumble world of the -mm tree. For the most part, these patches do nothing or very little. The patches should

At OLS I gave a talk on some of the Xen scalability inhibitors, and one of these was writable page tables. We went over why the feature does not scale, but just as important, we found that the uniprocessor case does not provide any advantage either. These tests were done on x86_64, so I wanted to run the 1-way test on 32 bit to show the same problem. So, I have run with writable PTs and

Create a new PTE function which combines clearing a kernel PTE with the subsequent flush. This allows the two to be easily combined into a single hypercall or paravirt-op. More subtly, reverse the order of the flush for kmap_atomic. Instead of flushing on establishing a mapping, flush on clearing a mapping. This eliminates the possibility of leaving stale kmap entries which may still have

Currently, the set_lazy_mode pv_op is overloaded with 5 functions: 1. enter lazy cpu mode 2. leave lazy cpu mode 3. enter lazy mmu mode 4. leave lazy mmu mode 5. flush pending batched operations This complicates each paravirt backend, since it needs to deal with all the possible state transitions, handling flushing, etc. In particular, flushing is quite distinct from the other 4 functions,

Currently, the set_lazy_mode pv_op is overloaded with 5 functions: 1. enter lazy cpu mode 2. leave lazy cpu mode 3. enter lazy mmu mode 4. leave lazy mmu mode 5. flush pending batched operations This complicates each paravirt backend, since it needs to deal with all the possible state transitions, handling flushing, etc. In particular, flushing is quite distinct from the other 4 functions,

From: Matt Wilson <msw@amazon.com> Commit f62805f1 introduced a bug where lazy MMU mode isn''t exited if a m2p_add/remove_override call fails. Cc: Stefano Stabellini <stefano.stabellini@eu.citrix.com> Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com> Cc: David Vrabel <david.vrabel@citrix.com> Cc: Anthony

[ I think this is a straight repost this patch, which addresses all the previous comments. I'd like to submit this for .24 as the basis for a unified paravirt_ops. Any objections? ] Currently, the set_lazy_mode pv_op is overloaded with 5 functions: 1. enter lazy cpu mode 2. leave lazy cpu mode 3. enter lazy mmu mode 4. leave lazy mmu mode 5. flush pending batched operations This

[ I think this is a straight repost this patch, which addresses all the previous comments. I'd like to submit this for .24 as the basis for a unified paravirt_ops. Any objections? ] Currently, the set_lazy_mode pv_op is overloaded with 5 functions: 1. enter lazy cpu mode 2. leave lazy cpu mode 3. enter lazy mmu mode 4. leave lazy mmu mode 5. flush pending batched operations This

Today paravirtualization is a all-or-nothing game: either a kernel is compiled with no paravirtualization support at all, or it is supporting paravirtualized environments like Xen pv-guests or lguest additionally to some paravirtualized tuning for KVM, Hyperv, VMWare or Xen HVM-guests. As support of pv-guests requires quite intrusive pv-hooks (e.g. all access functions to page table entries,

Today paravirtualization is a all-or-nothing game: either a kernel is compiled with no paravirtualization support at all, or it is supporting paravirtualized environments like Xen pv-guests or lguest additionally to some paravirtualized tuning for KVM, Hyperv, VMWare or Xen HVM-guests. As support of pv-guests requires quite intrusive pv-hooks (e.g. all access functions to page table entries,

MMU code movement. Unfortunately, this one is a little bit more complicated than the rest. We have to override the default accessors that directly write to page table entries. Because of the 2/3-level PAE split in Linux, this turned out to be really ugly at first, but by allowing the sub-arch layer to override the definitions and keeping the native definitions in place, the code becomes much

MMU code movement. Unfortunately, this one is a little bit more complicated than the rest. We have to override the default accessors that directly write to page table entries. Because of the 2/3-level PAE split in Linux, this turned out to be really ugly at first, but by allowing the sub-arch layer to override the definitions and keeping the native definitions in place, the code becomes much

Hi Andi, This series of patches updates paravirt_ops in various ways. Some of the changes are plain cleanups and improvements, and some add some interfaces necessary for Xen. The brief overview: add-MAINTAINERS.patch - obvious remove-CONFIG_DEBUG_PARAVIRT.patch - no longer needed paravirt-nop.patch - mark nop operations consistently paravirt-pte-accessors.patch - operations to pack/unpack

Hi Andi, This series of patches updates paravirt_ops in various ways. Some of the changes are plain cleanups and improvements, and some add some interfaces necessary for Xen. The brief overview: add-MAINTAINERS.patch - obvious remove-CONFIG_DEBUG_PARAVIRT.patch - no longer needed paravirt-nop.patch - mark nop operations consistently paravirt-pte-accessors.patch - operations to pack/unpack

The VMI ROM has a mode where hypercalls can be queued and batched. This turns out to be a significant win during context switch, but must be done at a specific point before side effects to CPU state are visible to subsequent instructions. This is similar to the MMU batching hooks already provided. The same hooks could be used by the Xen backend to implement a context switch multicall. To