search for: ptep_modify_prot_commit

...tion emulator is inherently expensive. And, 2: because ptep_get_and_clear has atomic-fetch-and-update semantics, it's impossible to implement in a way which can be batched to amortize the cost of trapping into the hypervisor. This series adds the ptep_modify_prot_start() and ptep_modify_prot_commit() operations, which change this sequence to: ptent = ptep_modify_prot_start(mm, addr, pte); ptent = pte_modify(ptent, newprot); /* ... */ ptep_modify_prot_commit(mm, addr, pte, ptent); Which looks very familiar. And, indeed, when compiled without CONFIG_PARAVIRT (or on a non-x86 architecture...

...tion emulator is inherently expensive. And, 2: because ptep_get_and_clear has atomic-fetch-and-update semantics, it's impossible to implement in a way which can be batched to amortize the cost of trapping into the hypervisor. This series adds the ptep_modify_prot_start() and ptep_modify_prot_commit() operations, which change this sequence to: ptent = ptep_modify_prot_start(mm, addr, pte); ptent = pte_modify(ptent, newprot); /* ... */ ptep_modify_prot_commit(mm, addr, pte, ptent); Which looks very familiar. And, indeed, when compiled without CONFIG_PARAVIRT (or on a non-x86 architecture...

...tion emulator is inherently expensive. And, 2: because ptep_get_and_clear has atomic-fetch-and-update semantics, it's impossible to implement in a way which can be batched to amortize the cost of trapping into the hypervisor. This series adds the ptep_modify_prot_start() and ptep_modify_prot_commit() operations, which change this sequence to: ptent = ptep_modify_prot_start(mm, addr, pte); ptent = pte_modify(ptent, newprot); /* ... */ ptep_modify_prot_commit(mm, addr, pte, ptent); Which looks very familiar. And, indeed, when compiled without CONFIG_PARAVIRT (or on a non-x86 architecture...

...v_cpu_ops.io_delay(); + pv_ops.pv_cpu_ops.io_delay(); #ifdef REALLY_SLOW_IO - pv_cpu_ops.io_delay(); - pv_cpu_ops.io_delay(); - pv_cpu_ops.io_delay(); + pv_ops.pv_cpu_ops.io_delay(); + pv_ops.pv_cpu_ops.io_delay(); + pv_ops.pv_cpu_ops.io_delay(); #endif } @@ -432,7 +432,7 @@ static inline void ptep_modify_prot_commit(struct mm_struct *mm, unsigned long a { if (sizeof(pteval_t) > sizeof(long)) /* 5 arg words */ - pv_mmu_ops.ptep_modify_prot_commit(mm, addr, ptep, pte); + pv_ops.pv_mmu_ops.ptep_modify_prot_commit(mm, addr, ptep, pte); else PVOP_VCALL4(pv_mmu_ops.ptep_modify_prot_commit, mm...

...f(pgdval_t) > sizeof(long)) - ret = PVOP_CALLEE2(pgdval_t, mmu.pgd_val, - pgd.pgd, (u64)pgd.pgd >> 32); - else - ret = PVOP_CALLEE1(pgdval_t, mmu.pgd_val, pgd.pgd); + ret = PVOP_CALLEE1(pgdval_t, mmu.pgd_val, pgd.pgd); return ret; } @@ -438,51 +420,32 @@ static inline void ptep_modify_prot_commit(struct vm_area_struct *vma, unsigned pte_t *ptep, pte_t old_pte, pte_t pte) { - if (sizeof(pteval_t) > sizeof(long)) - /* 5 arg words */ - pv_ops.mmu.ptep_modify_prot_commit(vma, addr, ptep, pte); - else - PVOP_VCALL4(mmu.ptep_modify_prot_commit, - vma, addr, ptep, pte.pte);...

...(pgdval_t, mmu.pgd_val, - pgd.pgd, (u64)pgd.pgd >> 32); - else - ret = PVOP_CALLEE1(pgdval_t, mmu.pgd_val, pgd.pgd); - - return ret; + return PVOP_CALLEE1(pgdval_t, mmu.pgd_val, pgd.pgd); } #define __HAVE_ARCH_PTEP_MODIFY_PROT_TRANSACTION @@ -438,78 +404,40 @@ static inline void ptep_modify_prot_commit(struct vm_area_struct *vma, unsigned pte_t *ptep, pte_t old_pte, pte_t pte) { - if (sizeof(pteval_t) > sizeof(long)) - /* 5 arg words */ - pv_ops.mmu.ptep_modify_prot_commit(vma, addr, ptep, pte); - else - PVOP_VCALL4(mmu.ptep_modify_prot_commit, - vma, addr, ptep, pte.pte);...

The long term plan has been to replace Xen PV guests by PVH. The first victim of that plan are now 32-bit PV guests, as those are used only rather seldom these days. Xen on x86 requires 64-bit support and with Grub2 now supporting PVH officially since version 2.04 there is no need to keep 32-bit PV guest support alive in the Linux kernel. Additionally Meltdown mitigation is not available in the

The long term plan has been to replace Xen PV guests by PVH. The first victim of that plan are now 32-bit PV guests, as those are used only rather seldom these days. Xen on x86 requires 64-bit support and with Grub2 now supporting PVH officially since version 2.04 there is no need to keep 32-bit PV guest support alive in the Linux kernel. Additionally Meltdown mitigation is not available in the

The long term plan has been to replace Xen PV guests by PVH. The first victim of that plan are now 32-bit PV guests, as those are used only rather seldom these days. Xen on x86 requires 64-bit support and with Grub2 now supporting PVH officially since version 2.04 there is no need to keep 32-bit PV guest support alive in the Linux kernel. Additionally Meltdown mitigation is not available in the

The long term plan has been to replace Xen PV guests by PVH. The first victim of that plan are now 32-bit PV guests, as those are used only rather seldom these days. Xen on x86 requires 64-bit support and with Grub2 now supporting PVH officially since version 2.04 there is no need to keep 32-bit PV guest support alive in the Linux kernel. Additionally Meltdown mitigation is not available in the

The long term plan has been to replace Xen PV guests by PVH. The first victim of that plan are now 32-bit PV guests, as those are used only rather seldom these days. Xen on x86 requires 64-bit support and with Grub2 now supporting PVH officially since version 2.04 there is no need to keep 32-bit PV guest support alive in the Linux kernel. Additionally Meltdown mitigation is not available in the

A lot of cleanup after removal of 32-bit Xen PV guest support in paravirt code. Changes in V4: - dropped patches 1-3, as already committed - addressed comments to V3 - added new patches 5+6 Changes in V3: - addressed comments to V2 - split patch 1 into 2 patches - new patches 3 and 7 Changes in V2: - rebase to 5.8 kernel - addressed comments to V1 - new patches 3 and 4 Juergen Gross (6):

This series removes some no longer needed stuff from paravirt infrastructure and puts large quantities of paravirt ops under a new config option PARAVIRT_XXL which is selected by XEN_PV only. A pvops kernel without XEN_PV being configured is about 2.5% smaller with this series applied. tip commit 5800dc5c19f34e6e03b5adab1282535cb102fafd ("x86/paravirt: Fix spectre-v2 mitigations for

...or_init, }; static const struct pv_time_ops xen_time_ops __initdata = { @@ -1103,7 +1111,7 @@ .set_pte = NULL, /* see xen_pagetable_setup_* */ .set_pte_at = xen_set_pte_at, - .set_pmd = xen_set_pmd, + .set_pmd = xen_set_pmd_hyper, .ptep_modify_prot_start = __ptep_modify_prot_start, .ptep_modify_prot_commit = __ptep_modify_prot_commit, @@ -1117,7 +1125,7 @@ .set_pte_atomic = xen_set_pte_atomic, .set_pte_present = xen_set_pte_at, - .set_pud = xen_set_pud, + .set_pud = xen_set_pud_hyper, .pte_clear = xen_pte_clear, .pmd_clear = xen_pmd_clear, =================================================...

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,

This series removes some no longer needed stuff from paravirt infrastructure and puts large quantities of paravirt ops under a new config option PARAVIRT_XXL which is selected by XEN_PV only. A pvops kernel without XEN_PV being configured is about 2.5% smaller with this series applied. tip commit 5800dc5c19f34e6e03b5adab1282535cb102fafd ("x86/paravirt: Fix spectre-v2 mitigations for

...9 +444,7 @@ struct pv_mmu_ops pv_mmu_ops = { .set_pmd = native_set_pmd, .set_pmd_at = native_set_pmd_at, .pte_update = paravirt_nop, - .pte_update_defer = paravirt_nop, .pmd_update = paravirt_nop, - .pmd_update_defer = paravirt_nop, .ptep_modify_prot_start = __ptep_modify_prot_start, .ptep_modify_prot_commit = __ptep_modify_prot_commit, diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c index a0d09f6..a1900d4 100644 --- a/arch/x86/lguest/boot.c +++ b/arch/x86/lguest/boot.c @@ -1472,7 +1472,6 @@ __init void lguest_init(void) pv_mmu_ops.lazy_mode.leave = lguest_leave_lazy_mmu_mode; pv_mmu_o...

...9 +444,7 @@ struct pv_mmu_ops pv_mmu_ops = { .set_pmd = native_set_pmd, .set_pmd_at = native_set_pmd_at, .pte_update = paravirt_nop, - .pte_update_defer = paravirt_nop, .pmd_update = paravirt_nop, - .pmd_update_defer = paravirt_nop, .ptep_modify_prot_start = __ptep_modify_prot_start, .ptep_modify_prot_commit = __ptep_modify_prot_commit, diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c index a0d09f6..a1900d4 100644 --- a/arch/x86/lguest/boot.c +++ b/arch/x86/lguest/boot.c @@ -1472,7 +1472,6 @@ __init void lguest_init(void) pv_mmu_ops.lazy_mode.leave = lguest_leave_lazy_mmu_mode; pv_mmu_o...

...9 +444,6 @@ struct pv_mmu_ops pv_mmu_ops = { .set_pmd = native_set_pmd, .set_pmd_at = native_set_pmd_at, .pte_update = paravirt_nop, - .pte_update_defer = paravirt_nop, - .pmd_update = paravirt_nop, - .pmd_update_defer = paravirt_nop, .ptep_modify_prot_start = __ptep_modify_prot_start, .ptep_modify_prot_commit = __ptep_modify_prot_commit, diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c index a0d09f6..a1900d4 100644 --- a/arch/x86/lguest/boot.c +++ b/arch/x86/lguest/boot.c @@ -1472,7 +1472,6 @@ __init void lguest_init(void) pv_mmu_ops.lazy_mode.leave = lguest_leave_lazy_mmu_mode; pv_mmu_o...