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 paravirt guests") is a prerequisite for this series. The last 4 patches of this series require my Xen cleanup series https://lore.kernel.org/lkml/20180717120113.12756-1-jgross at suse.com/ which hides more Xen PV-only code behind CONFIG_XEN_PV. Juergen Gross (10): x86/paravirt: make paravirt_patch_call() and paravirt_patch_jmp() static x86/paravirt: remove clobbers parameter from paravirt patch functions x86/paravirt: remove clobbers from struct paravirt_patch_site x86/paravirt: use a single ops structure x86/paravirt: remove unused paravirt bits x86/paravirt: introduce new config option PARAVIRT_XXL x86/paravirt: move items in pv_info under PARAVIRT_XXL umbrella x86/paravirt: move the Xen-only pv_cpu_ops under the PARAVIRT_XXL umbrella x86/paravirt: move the Xen-only pv_irq_ops under the PARAVIRT_XXL umbrella x86/paravirt: move the Xen-only pv_mmu_ops under the PARAVIRT_XXL umbrella arch/x86/Kconfig | 3 + arch/x86/hyperv/mmu.c | 4 +- arch/x86/include/asm/debugreg.h | 2 +- arch/x86/include/asm/desc.h | 4 +- arch/x86/include/asm/fixmap.h | 2 +- arch/x86/include/asm/irqflags.h | 56 +++--- arch/x86/include/asm/mmu_context.h | 4 +- arch/x86/include/asm/msr.h | 4 +- arch/x86/include/asm/paravirt.h | 183 +++++++++--------- arch/x86/include/asm/paravirt_types.h | 65 +++---- arch/x86/include/asm/pgalloc.h | 2 +- arch/x86/include/asm/pgtable-3level_types.h | 2 +- arch/x86/include/asm/pgtable.h | 7 +- arch/x86/include/asm/processor.h | 4 +- arch/x86/include/asm/ptrace.h | 3 +- arch/x86/include/asm/segment.h | 2 +- arch/x86/include/asm/special_insns.h | 4 +- arch/x86/kernel/alternative.c | 2 +- arch/x86/kernel/asm-offsets.c | 15 +- arch/x86/kernel/asm-offsets_64.c | 9 +- arch/x86/kernel/cpu/common.c | 4 +- arch/x86/kernel/cpu/vmware.c | 4 +- arch/x86/kernel/head_64.S | 2 +- arch/x86/kernel/kvm.c | 18 +- arch/x86/kernel/kvmclock.c | 4 +- arch/x86/kernel/paravirt-spinlocks.c | 15 +- arch/x86/kernel/paravirt.c | 290 ++++++++++++++-------------- arch/x86/kernel/paravirt_patch_32.c | 9 +- arch/x86/kernel/paravirt_patch_64.c | 11 +- arch/x86/kernel/tsc.c | 2 +- arch/x86/kernel/vsmp_64.c | 17 +- arch/x86/xen/Kconfig | 1 + arch/x86/xen/enlighten_pv.c | 32 +-- arch/x86/xen/irq.c | 2 +- arch/x86/xen/mmu_hvm.c | 2 +- arch/x86/xen/mmu_pv.c | 28 +-- arch/x86/xen/spinlock.c | 12 +- arch/x86/xen/time.c | 4 +- drivers/xen/time.c | 2 +- 39 files changed, 430 insertions(+), 406 deletions(-) -- 2.13.7
Juergen Gross
2018-Aug-10 11:52 UTC
[PATCH 01/10] x86/paravirt: make paravirt_patch_call() and paravirt_patch_jmp() static
paravirt_patch_call() and paravirt_patch_jmp() are used in paravirt.c only. Convert them to static. Signed-off-by: Juergen Gross <jgross at suse.com> --- arch/x86/include/asm/paravirt_types.h | 6 ------ arch/x86/kernel/paravirt.c | 12 ++++++------ 2 files changed, 6 insertions(+), 12 deletions(-) diff --git a/arch/x86/include/asm/paravirt_types.h b/arch/x86/include/asm/paravirt_types.h index 180bc0bff0fb..036b2f88f105 100644 --- a/arch/x86/include/asm/paravirt_types.h +++ b/arch/x86/include/asm/paravirt_types.h @@ -370,12 +370,6 @@ extern struct pv_lock_ops pv_lock_ops; unsigned paravirt_patch_ident_32(void *insnbuf, unsigned len); unsigned paravirt_patch_ident_64(void *insnbuf, unsigned len); -unsigned paravirt_patch_call(void *insnbuf, - const void *target, u16 tgt_clobbers, - unsigned long addr, u16 site_clobbers, - unsigned len); -unsigned paravirt_patch_jmp(void *insnbuf, const void *target, - unsigned long addr, unsigned len); unsigned paravirt_patch_default(u8 type, u16 clobbers, void *insnbuf, unsigned long addr, unsigned len); diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c index 930c88341e4e..ce560b916b1f 100644 --- a/arch/x86/kernel/paravirt.c +++ b/arch/x86/kernel/paravirt.c @@ -80,10 +80,10 @@ struct branch { u32 delta; } __attribute__((packed)); -unsigned paravirt_patch_call(void *insnbuf, - const void *target, u16 tgt_clobbers, - unsigned long addr, u16 site_clobbers, - unsigned len) +static unsigned paravirt_patch_call(void *insnbuf, + const void *target, u16 tgt_clobbers, + unsigned long addr, u16 site_clobbers, + unsigned len) { struct branch *b = insnbuf; unsigned long delta = (unsigned long)target - (addr+5); @@ -102,8 +102,8 @@ unsigned paravirt_patch_call(void *insnbuf, return 5; } -unsigned paravirt_patch_jmp(void *insnbuf, const void *target, - unsigned long addr, unsigned len) +static unsigned paravirt_patch_jmp(void *insnbuf, const void *target, + unsigned long addr, unsigned len) { struct branch *b = insnbuf; unsigned long delta = (unsigned long)target - (addr+5); -- 2.13.7
Juergen Gross
2018-Aug-10 11:52 UTC
[PATCH 02/10] x86/paravirt: remove clobbers parameter from paravirt patch functions
The clobbers parameter from paravirt_patch_default() et al isn't used any longer. Remove it. Signed-off-by: Juergen Gross <jgross at suse.com> --- arch/x86/include/asm/paravirt_types.h | 7 +++---- arch/x86/kernel/alternative.c | 2 +- arch/x86/kernel/paravirt.c | 14 +++++--------- arch/x86/kernel/paravirt_patch_32.c | 5 ++--- arch/x86/kernel/paravirt_patch_64.c | 5 ++--- arch/x86/kernel/vsmp_64.c | 6 +++--- 6 files changed, 16 insertions(+), 23 deletions(-) diff --git a/arch/x86/include/asm/paravirt_types.h b/arch/x86/include/asm/paravirt_types.h index 036b2f88f105..f6e24e78633b 100644 --- a/arch/x86/include/asm/paravirt_types.h +++ b/arch/x86/include/asm/paravirt_types.h @@ -84,7 +84,7 @@ struct pv_init_ops { * the number of bytes of code generated, as we nop pad the * rest in generic code. */ - unsigned (*patch)(u8 type, u16 clobber, void *insnbuf, + unsigned (*patch)(u8 type, void *insnbuf, unsigned long addr, unsigned len); } __no_randomize_layout; @@ -370,14 +370,13 @@ extern struct pv_lock_ops pv_lock_ops; unsigned paravirt_patch_ident_32(void *insnbuf, unsigned len); unsigned paravirt_patch_ident_64(void *insnbuf, unsigned len); -unsigned paravirt_patch_default(u8 type, u16 clobbers, void *insnbuf, +unsigned paravirt_patch_default(u8 type, void *insnbuf, unsigned long addr, unsigned len); unsigned paravirt_patch_insns(void *insnbuf, unsigned len, const char *start, const char *end); -unsigned native_patch(u8 type, u16 clobbers, void *ibuf, - unsigned long addr, unsigned len); +unsigned native_patch(u8 type, void *ibuf, unsigned long addr, unsigned len); int paravirt_disable_iospace(void); diff --git a/arch/x86/kernel/alternative.c b/arch/x86/kernel/alternative.c index a481763a3776..9729cee11149 100644 --- a/arch/x86/kernel/alternative.c +++ b/arch/x86/kernel/alternative.c @@ -594,7 +594,7 @@ void __init_or_module apply_paravirt(struct paravirt_patch_site *start, BUG_ON(p->len > MAX_PATCH_LEN); /* prep the buffer with the original instructions */ memcpy(insnbuf, p->instr, p->len); - used = pv_init_ops.patch(p->instrtype, p->clobbers, insnbuf, + used = pv_init_ops.patch(p->instrtype, insnbuf, (unsigned long)p->instr, p->len); BUG_ON(used > p->len); diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c index ce560b916b1f..f0c462fe2808 100644 --- a/arch/x86/kernel/paravirt.c +++ b/arch/x86/kernel/paravirt.c @@ -80,10 +80,8 @@ struct branch { u32 delta; } __attribute__((packed)); -static unsigned paravirt_patch_call(void *insnbuf, - const void *target, u16 tgt_clobbers, - unsigned long addr, u16 site_clobbers, - unsigned len) +static unsigned paravirt_patch_call(void *insnbuf, const void *target, + unsigned long addr, unsigned len) { struct branch *b = insnbuf; unsigned long delta = (unsigned long)target - (addr+5); @@ -148,7 +146,7 @@ static void *get_call_destination(u8 type) return *((void **)&tmpl + type); } -unsigned paravirt_patch_default(u8 type, u16 clobbers, void *insnbuf, +unsigned paravirt_patch_default(u8 type, void *insnbuf, unsigned long addr, unsigned len) { void *opfunc = get_call_destination(type); @@ -171,10 +169,8 @@ unsigned paravirt_patch_default(u8 type, u16 clobbers, void *insnbuf, /* If operation requires a jmp, then jmp */ ret = paravirt_patch_jmp(insnbuf, opfunc, addr, len); else - /* Otherwise call the function; assume target could - clobber any caller-save reg */ - ret = paravirt_patch_call(insnbuf, opfunc, CLBR_ANY, - addr, clobbers, len); + /* Otherwise call the function. */ + ret = paravirt_patch_call(insnbuf, opfunc, addr, len); return ret; } diff --git a/arch/x86/kernel/paravirt_patch_32.c b/arch/x86/kernel/paravirt_patch_32.c index 758e69d72ebf..e5c3a438149e 100644 --- a/arch/x86/kernel/paravirt_patch_32.c +++ b/arch/x86/kernel/paravirt_patch_32.c @@ -30,8 +30,7 @@ unsigned paravirt_patch_ident_64(void *insnbuf, unsigned len) extern bool pv_is_native_spin_unlock(void); extern bool pv_is_native_vcpu_is_preempted(void); -unsigned native_patch(u8 type, u16 clobbers, void *ibuf, - unsigned long addr, unsigned len) +unsigned native_patch(u8 type, void *ibuf, unsigned long addr, unsigned len) { const unsigned char *start, *end; unsigned ret; @@ -70,7 +69,7 @@ unsigned native_patch(u8 type, u16 clobbers, void *ibuf, default: patch_default: __maybe_unused - ret = paravirt_patch_default(type, clobbers, ibuf, addr, len); + ret = paravirt_patch_default(type, ibuf, addr, len); break; patch_site: diff --git a/arch/x86/kernel/paravirt_patch_64.c b/arch/x86/kernel/paravirt_patch_64.c index 9edadabf04f6..893ef87eb268 100644 --- a/arch/x86/kernel/paravirt_patch_64.c +++ b/arch/x86/kernel/paravirt_patch_64.c @@ -38,8 +38,7 @@ unsigned paravirt_patch_ident_64(void *insnbuf, unsigned len) extern bool pv_is_native_spin_unlock(void); extern bool pv_is_native_vcpu_is_preempted(void); -unsigned native_patch(u8 type, u16 clobbers, void *ibuf, - unsigned long addr, unsigned len) +unsigned native_patch(u8 type, void *ibuf, unsigned long addr, unsigned len) { const unsigned char *start, *end; unsigned ret; @@ -80,7 +79,7 @@ unsigned native_patch(u8 type, u16 clobbers, void *ibuf, default: patch_default: __maybe_unused - ret = paravirt_patch_default(type, clobbers, ibuf, addr, len); + ret = paravirt_patch_default(type, ibuf, addr, len); break; patch_site: diff --git a/arch/x86/kernel/vsmp_64.c b/arch/x86/kernel/vsmp_64.c index 44685fb2a192..f194e5e1e95c 100644 --- a/arch/x86/kernel/vsmp_64.c +++ b/arch/x86/kernel/vsmp_64.c @@ -69,7 +69,7 @@ asmlinkage __visible void vsmp_irq_enable(void) } PV_CALLEE_SAVE_REGS_THUNK(vsmp_irq_enable); -static unsigned __init vsmp_patch(u8 type, u16 clobbers, void *ibuf, +static unsigned __init vsmp_patch(u8 type, void *ibuf, unsigned long addr, unsigned len) { switch (type) { @@ -77,9 +77,9 @@ static unsigned __init vsmp_patch(u8 type, u16 clobbers, void *ibuf, case PARAVIRT_PATCH(pv_irq_ops.irq_disable): case PARAVIRT_PATCH(pv_irq_ops.save_fl): case PARAVIRT_PATCH(pv_irq_ops.restore_fl): - return paravirt_patch_default(type, clobbers, ibuf, addr, len); + return paravirt_patch_default(type, ibuf, addr, len); default: - return native_patch(type, clobbers, ibuf, addr, len); + return native_patch(type, ibuf, addr, len); } } -- 2.13.7
Juergen Gross
2018-Aug-10 11:52 UTC
[PATCH 03/10] x86/paravirt: remove clobbers from struct paravirt_patch_site
There is no need any longer to store the clobbers in struct paravirt_patch_site. Remove clobbers from the struct and from the related macros. While at it fix some lines longer than 80 characters. Signed-off-by: Juergen Gross <jgross at suse.com> --- arch/x86/include/asm/paravirt.h | 33 +++++++++++++++------------------ arch/x86/include/asm/paravirt_types.h | 1 - 2 files changed, 15 insertions(+), 19 deletions(-) diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h index d49bbf4bb5c8..76b4b5c056f3 100644 --- a/arch/x86/include/asm/paravirt.h +++ b/arch/x86/include/asm/paravirt.h @@ -822,7 +822,7 @@ extern void default_banner(void); #else /* __ASSEMBLY__ */ -#define _PVSITE(ptype, clobbers, ops, word, algn) \ +#define _PVSITE(ptype, ops, word, algn) \ 771:; \ ops; \ 772:; \ @@ -831,7 +831,6 @@ extern void default_banner(void); word 771b; \ .byte ptype; \ .byte 772b-771b; \ - .short clobbers; \ .popsection @@ -864,7 +863,7 @@ extern void default_banner(void); COND_POP(set, CLBR_RAX, rax) #define PARA_PATCH(struct, off) ((PARAVIRT_PATCH_##struct + (off)) / 8) -#define PARA_SITE(ptype, clobbers, ops) _PVSITE(ptype, clobbers, ops, .quad, 8) +#define PARA_SITE(ptype, ops) _PVSITE(ptype, ops, .quad, 8) #define PARA_INDIRECT(addr) *addr(%rip) #else #define PV_SAVE_REGS(set) \ @@ -879,26 +878,26 @@ extern void default_banner(void); COND_POP(set, CLBR_EAX, eax) #define PARA_PATCH(struct, off) ((PARAVIRT_PATCH_##struct + (off)) / 4) -#define PARA_SITE(ptype, clobbers, ops) _PVSITE(ptype, clobbers, ops, .long, 4) +#define PARA_SITE(ptype, ops) _PVSITE(ptype, ops, .long, 4) #define PARA_INDIRECT(addr) *%cs:addr #endif #define INTERRUPT_RETURN \ - PARA_SITE(PARA_PATCH(pv_cpu_ops, PV_CPU_iret), CLBR_NONE, \ - ANNOTATE_RETPOLINE_SAFE; \ + PARA_SITE(PARA_PATCH(pv_cpu_ops, PV_CPU_iret), \ + ANNOTATE_RETPOLINE_SAFE; \ jmp PARA_INDIRECT(pv_cpu_ops+PV_CPU_iret);) #define DISABLE_INTERRUPTS(clobbers) \ - PARA_SITE(PARA_PATCH(pv_irq_ops, PV_IRQ_irq_disable), clobbers, \ + PARA_SITE(PARA_PATCH(pv_irq_ops, PV_IRQ_irq_disable), \ PV_SAVE_REGS(clobbers | CLBR_CALLEE_SAVE); \ - ANNOTATE_RETPOLINE_SAFE; \ + ANNOTATE_RETPOLINE_SAFE; \ call PARA_INDIRECT(pv_irq_ops+PV_IRQ_irq_disable); \ PV_RESTORE_REGS(clobbers | CLBR_CALLEE_SAVE);) #define ENABLE_INTERRUPTS(clobbers) \ - PARA_SITE(PARA_PATCH(pv_irq_ops, PV_IRQ_irq_enable), clobbers, \ + PARA_SITE(PARA_PATCH(pv_irq_ops, PV_IRQ_irq_enable), \ PV_SAVE_REGS(clobbers | CLBR_CALLEE_SAVE); \ - ANNOTATE_RETPOLINE_SAFE; \ + ANNOTATE_RETPOLINE_SAFE; \ call PARA_INDIRECT(pv_irq_ops+PV_IRQ_irq_enable); \ PV_RESTORE_REGS(clobbers | CLBR_CALLEE_SAVE);) @@ -916,8 +915,7 @@ extern void default_banner(void); * inlined, or the swapgs instruction must be trapped and emulated. */ #define SWAPGS_UNSAFE_STACK \ - PARA_SITE(PARA_PATCH(pv_cpu_ops, PV_CPU_swapgs), CLBR_NONE, \ - swapgs) + PARA_SITE(PARA_PATCH(pv_cpu_ops, PV_CPU_swapgs), swapgs) /* * Note: swapgs is very special, and in practise is either going to be @@ -926,8 +924,8 @@ extern void default_banner(void); * it. */ #define SWAPGS \ - PARA_SITE(PARA_PATCH(pv_cpu_ops, PV_CPU_swapgs), CLBR_NONE, \ - ANNOTATE_RETPOLINE_SAFE; \ + PARA_SITE(PARA_PATCH(pv_cpu_ops, PV_CPU_swapgs), \ + ANNOTATE_RETPOLINE_SAFE; \ call PARA_INDIRECT(pv_cpu_ops+PV_CPU_swapgs); \ ) @@ -937,15 +935,14 @@ extern void default_banner(void); #define USERGS_SYSRET64 \ PARA_SITE(PARA_PATCH(pv_cpu_ops, PV_CPU_usergs_sysret64), \ - CLBR_NONE, \ - ANNOTATE_RETPOLINE_SAFE; \ + ANNOTATE_RETPOLINE_SAFE; \ jmp PARA_INDIRECT(pv_cpu_ops+PV_CPU_usergs_sysret64);) #ifdef CONFIG_DEBUG_ENTRY #define SAVE_FLAGS(clobbers) \ - PARA_SITE(PARA_PATCH(pv_irq_ops, PV_IRQ_save_fl), clobbers, \ + PARA_SITE(PARA_PATCH(pv_irq_ops, PV_IRQ_save_fl), \ PV_SAVE_REGS(clobbers | CLBR_CALLEE_SAVE); \ - ANNOTATE_RETPOLINE_SAFE; \ + ANNOTATE_RETPOLINE_SAFE; \ call PARA_INDIRECT(pv_irq_ops+PV_IRQ_save_fl); \ PV_RESTORE_REGS(clobbers | CLBR_CALLEE_SAVE);) #endif diff --git a/arch/x86/include/asm/paravirt_types.h b/arch/x86/include/asm/paravirt_types.h index f6e24e78633b..b900088cd244 100644 --- a/arch/x86/include/asm/paravirt_types.h +++ b/arch/x86/include/asm/paravirt_types.h @@ -678,7 +678,6 @@ struct paravirt_patch_site { u8 *instr; /* original instructions */ u8 instrtype; /* type of this instruction */ u8 len; /* length of original instruction */ - u16 clobbers; /* what registers you may clobber */ }; extern struct paravirt_patch_site __parainstructions[], -- 2.13.7
Juergen Gross
2018-Aug-10 11:52 UTC
[PATCH 04/10] x86/paravirt: use a single ops structure
Instead of using six globally visible paravirt ops structures combine them in a single structure, keeping the original structures as sub-structures. This avoids the need to assemble struct paravirt_patch_template at runtime on the stack each time apply_paravirt() is being called (i.e. when loading a module). Signed-off-by: Juergen Gross <jgross at suse.com> --- arch/x86/hyperv/mmu.c | 4 +- arch/x86/include/asm/paravirt.h | 51 ++++--- arch/x86/include/asm/paravirt_types.h | 13 +- arch/x86/kernel/alternative.c | 2 +- arch/x86/kernel/asm-offsets.c | 14 +- arch/x86/kernel/asm-offsets_64.c | 7 +- arch/x86/kernel/cpu/common.c | 2 +- arch/x86/kernel/cpu/vmware.c | 4 +- arch/x86/kernel/kvm.c | 18 ++- arch/x86/kernel/kvmclock.c | 4 +- arch/x86/kernel/paravirt-spinlocks.c | 15 +- arch/x86/kernel/paravirt.c | 278 +++++++++++++++++----------------- arch/x86/kernel/tsc.c | 2 +- arch/x86/kernel/vsmp_64.c | 11 +- arch/x86/xen/enlighten_pv.c | 32 ++-- arch/x86/xen/irq.c | 2 +- arch/x86/xen/mmu_hvm.c | 2 +- arch/x86/xen/mmu_pv.c | 28 ++-- arch/x86/xen/spinlock.c | 12 +- arch/x86/xen/time.c | 4 +- drivers/xen/time.c | 2 +- 21 files changed, 249 insertions(+), 258 deletions(-) diff --git a/arch/x86/hyperv/mmu.c b/arch/x86/hyperv/mmu.c index de27615c51ea..b34768cfb204 100644 --- a/arch/x86/hyperv/mmu.c +++ b/arch/x86/hyperv/mmu.c @@ -228,9 +228,9 @@ void hyperv_setup_mmu_ops(void) if (!(ms_hyperv.hints & HV_X64_EX_PROCESSOR_MASKS_RECOMMENDED)) { pr_info("Using hypercall for remote TLB flush\n"); - pv_mmu_ops.flush_tlb_others = hyperv_flush_tlb_others; + pv_ops.pv_mmu_ops.flush_tlb_others = hyperv_flush_tlb_others; } else { pr_info("Using ext hypercall for remote TLB flush\n"); - pv_mmu_ops.flush_tlb_others = hyperv_flush_tlb_others_ex; + pv_ops.pv_mmu_ops.flush_tlb_others = hyperv_flush_tlb_others_ex; } } diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h index 76b4b5c056f3..1b86bb319393 100644 --- a/arch/x86/include/asm/paravirt.h +++ b/arch/x86/include/asm/paravirt.h @@ -265,11 +265,11 @@ static inline void set_iopl_mask(unsigned mask) /* The paravirtualized I/O functions */ static inline void slow_down_io(void) { - pv_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, addr, ptep, pte.pte); @@ -453,7 +453,7 @@ static inline void set_pte_at(struct mm_struct *mm, unsigned long addr, { if (sizeof(pteval_t) > sizeof(long)) /* 5 arg words */ - pv_mmu_ops.set_pte_at(mm, addr, ptep, pte); + pv_ops.pv_mmu_ops.set_pte_at(mm, addr, ptep, pte); else PVOP_VCALL4(pv_mmu_ops.set_pte_at, mm, addr, ptep, pte.pte); } @@ -663,7 +663,7 @@ static inline void arch_flush_lazy_mmu_mode(void) static inline void __set_fixmap(unsigned /* enum fixed_addresses */ idx, phys_addr_t phys, pgprot_t flags) { - pv_mmu_ops.set_fixmap(idx, phys, flags); + pv_ops.pv_mmu_ops.set_fixmap(idx, phys, flags); } #if defined(CONFIG_SMP) && defined(CONFIG_PARAVIRT_SPINLOCKS) @@ -694,6 +694,9 @@ static __always_inline bool pv_vcpu_is_preempted(long cpu) return PVOP_CALLEE1(bool, pv_lock_ops.vcpu_is_preempted, cpu); } +void __raw_callee_save___native_queued_spin_unlock(struct qspinlock *lock); +bool __raw_callee_save___native_vcpu_is_preempted(long cpu); + #endif /* SMP && PARAVIRT_SPINLOCKS */ #ifdef CONFIG_X86_32 @@ -862,7 +865,7 @@ extern void default_banner(void); COND_POP(set, CLBR_RCX, rcx); \ COND_POP(set, CLBR_RAX, rax) -#define PARA_PATCH(struct, off) ((PARAVIRT_PATCH_##struct + (off)) / 8) +#define PARA_PATCH(off) ((off) / 8) #define PARA_SITE(ptype, ops) _PVSITE(ptype, ops, .quad, 8) #define PARA_INDIRECT(addr) *addr(%rip) #else @@ -877,35 +880,35 @@ extern void default_banner(void); COND_POP(set, CLBR_EDI, edi); \ COND_POP(set, CLBR_EAX, eax) -#define PARA_PATCH(struct, off) ((PARAVIRT_PATCH_##struct + (off)) / 4) +#define PARA_PATCH(off) ((off) / 4) #define PARA_SITE(ptype, ops) _PVSITE(ptype, ops, .long, 4) #define PARA_INDIRECT(addr) *%cs:addr #endif #define INTERRUPT_RETURN \ - PARA_SITE(PARA_PATCH(pv_cpu_ops, PV_CPU_iret), \ + PARA_SITE(PARA_PATCH(PV_CPU_iret), \ ANNOTATE_RETPOLINE_SAFE; \ - jmp PARA_INDIRECT(pv_cpu_ops+PV_CPU_iret);) + jmp PARA_INDIRECT(pv_ops+PV_CPU_iret);) #define DISABLE_INTERRUPTS(clobbers) \ - PARA_SITE(PARA_PATCH(pv_irq_ops, PV_IRQ_irq_disable), \ + PARA_SITE(PARA_PATCH(PV_IRQ_irq_disable), \ PV_SAVE_REGS(clobbers | CLBR_CALLEE_SAVE); \ ANNOTATE_RETPOLINE_SAFE; \ - call PARA_INDIRECT(pv_irq_ops+PV_IRQ_irq_disable); \ + call PARA_INDIRECT(pv_ops+PV_IRQ_irq_disable); \ PV_RESTORE_REGS(clobbers | CLBR_CALLEE_SAVE);) #define ENABLE_INTERRUPTS(clobbers) \ - PARA_SITE(PARA_PATCH(pv_irq_ops, PV_IRQ_irq_enable), \ + PARA_SITE(PARA_PATCH(PV_IRQ_irq_enable), \ PV_SAVE_REGS(clobbers | CLBR_CALLEE_SAVE); \ ANNOTATE_RETPOLINE_SAFE; \ - call PARA_INDIRECT(pv_irq_ops+PV_IRQ_irq_enable); \ + call PARA_INDIRECT(pv_ops+PV_IRQ_irq_enable); \ PV_RESTORE_REGS(clobbers | CLBR_CALLEE_SAVE);) #ifdef CONFIG_X86_32 #define GET_CR0_INTO_EAX \ push %ecx; push %edx; \ ANNOTATE_RETPOLINE_SAFE; \ - call PARA_INDIRECT(pv_cpu_ops+PV_CPU_read_cr0); \ + call PARA_INDIRECT(pv_ops+PV_CPU_read_cr0); \ pop %edx; pop %ecx #else /* !CONFIG_X86_32 */ @@ -915,7 +918,7 @@ extern void default_banner(void); * inlined, or the swapgs instruction must be trapped and emulated. */ #define SWAPGS_UNSAFE_STACK \ - PARA_SITE(PARA_PATCH(pv_cpu_ops, PV_CPU_swapgs), swapgs) + PARA_SITE(PARA_PATCH(PV_CPU_swapgs), swapgs) /* * Note: swapgs is very special, and in practise is either going to be @@ -924,26 +927,26 @@ extern void default_banner(void); * it. */ #define SWAPGS \ - PARA_SITE(PARA_PATCH(pv_cpu_ops, PV_CPU_swapgs), \ + PARA_SITE(PARA_PATCH(PV_CPU_swapgs), \ ANNOTATE_RETPOLINE_SAFE; \ - call PARA_INDIRECT(pv_cpu_ops+PV_CPU_swapgs); \ + call PARA_INDIRECT(pv_ops+PV_CPU_swapgs); \ ) #define GET_CR2_INTO_RAX \ ANNOTATE_RETPOLINE_SAFE; \ - call PARA_INDIRECT(pv_mmu_ops+PV_MMU_read_cr2); + call PARA_INDIRECT(pv_ops+PV_MMU_read_cr2); #define USERGS_SYSRET64 \ - PARA_SITE(PARA_PATCH(pv_cpu_ops, PV_CPU_usergs_sysret64), \ + PARA_SITE(PARA_PATCH(PV_CPU_usergs_sysret64), \ ANNOTATE_RETPOLINE_SAFE; \ - jmp PARA_INDIRECT(pv_cpu_ops+PV_CPU_usergs_sysret64);) + jmp PARA_INDIRECT(pv_ops+PV_CPU_usergs_sysret64);) #ifdef CONFIG_DEBUG_ENTRY #define SAVE_FLAGS(clobbers) \ - PARA_SITE(PARA_PATCH(pv_irq_ops, PV_IRQ_save_fl), \ + PARA_SITE(PARA_PATCH(PV_IRQ_save_fl), \ PV_SAVE_REGS(clobbers | CLBR_CALLEE_SAVE); \ ANNOTATE_RETPOLINE_SAFE; \ - call PARA_INDIRECT(pv_irq_ops+PV_IRQ_save_fl); \ + call PARA_INDIRECT(pv_ops+PV_IRQ_save_fl); \ PV_RESTORE_REGS(clobbers | CLBR_CALLEE_SAVE);) #endif diff --git a/arch/x86/include/asm/paravirt_types.h b/arch/x86/include/asm/paravirt_types.h index b900088cd244..ed024e90b863 100644 --- a/arch/x86/include/asm/paravirt_types.h +++ b/arch/x86/include/asm/paravirt_types.h @@ -327,19 +327,14 @@ struct paravirt_patch_template { } __no_randomize_layout; extern struct pv_info pv_info; -extern struct pv_init_ops pv_init_ops; -extern struct pv_time_ops pv_time_ops; -extern struct pv_cpu_ops pv_cpu_ops; -extern struct pv_irq_ops pv_irq_ops; -extern struct pv_mmu_ops pv_mmu_ops; -extern struct pv_lock_ops pv_lock_ops; +extern struct paravirt_patch_template pv_ops; #define PARAVIRT_PATCH(x) \ (offsetof(struct paravirt_patch_template, x) / sizeof(void *)) #define paravirt_type(op) \ [paravirt_typenum] "i" (PARAVIRT_PATCH(op)), \ - [paravirt_opptr] "i" (&(op)) + [paravirt_opptr] "i" (&(pv_ops.op)) #define paravirt_clobber(clobber) \ [paravirt_clobber] "i" (clobber) @@ -500,9 +495,9 @@ int paravirt_disable_iospace(void); #endif /* CONFIG_X86_32 */ #ifdef CONFIG_PARAVIRT_DEBUG -#define PVOP_TEST_NULL(op) BUG_ON(op == NULL) +#define PVOP_TEST_NULL(op) BUG_ON(pv_ops.op == NULL) #else -#define PVOP_TEST_NULL(op) ((void)op) +#define PVOP_TEST_NULL(op) ((void)pv_ops.op) #endif #define PVOP_RETMASK(rettype) \ diff --git a/arch/x86/kernel/alternative.c b/arch/x86/kernel/alternative.c index 9729cee11149..7cfeda749382 100644 --- a/arch/x86/kernel/alternative.c +++ b/arch/x86/kernel/alternative.c @@ -594,7 +594,7 @@ void __init_or_module apply_paravirt(struct paravirt_patch_site *start, BUG_ON(p->len > MAX_PATCH_LEN); /* prep the buffer with the original instructions */ memcpy(insnbuf, p->instr, p->len); - used = pv_init_ops.patch(p->instrtype, insnbuf, + used = pv_ops.pv_init_ops.patch(p->instrtype, insnbuf, (unsigned long)p->instr, p->len); BUG_ON(used > p->len); diff --git a/arch/x86/kernel/asm-offsets.c b/arch/x86/kernel/asm-offsets.c index dcb008c320fe..bec9fc3498f8 100644 --- a/arch/x86/kernel/asm-offsets.c +++ b/arch/x86/kernel/asm-offsets.c @@ -66,13 +66,13 @@ void common(void) { #ifdef CONFIG_PARAVIRT BLANK(); - OFFSET(PARAVIRT_PATCH_pv_cpu_ops, paravirt_patch_template, pv_cpu_ops); - OFFSET(PARAVIRT_PATCH_pv_irq_ops, paravirt_patch_template, pv_irq_ops); - OFFSET(PV_IRQ_irq_disable, pv_irq_ops, irq_disable); - OFFSET(PV_IRQ_irq_enable, pv_irq_ops, irq_enable); - OFFSET(PV_CPU_iret, pv_cpu_ops, iret); - OFFSET(PV_CPU_read_cr0, pv_cpu_ops, read_cr0); - OFFSET(PV_MMU_read_cr2, pv_mmu_ops, read_cr2); + OFFSET(PV_IRQ_irq_disable, paravirt_patch_template, + pv_irq_ops.irq_disable); + OFFSET(PV_IRQ_irq_enable, paravirt_patch_template, + pv_irq_ops.irq_enable); + OFFSET(PV_CPU_iret, paravirt_patch_template, pv_cpu_ops.iret); + OFFSET(PV_CPU_read_cr0, paravirt_patch_template, pv_cpu_ops.read_cr0); + OFFSET(PV_MMU_read_cr2, paravirt_patch_template, pv_mmu_ops.read_cr2); #endif #ifdef CONFIG_XEN diff --git a/arch/x86/kernel/asm-offsets_64.c b/arch/x86/kernel/asm-offsets_64.c index b2dcd161f514..2add567c1b2a 100644 --- a/arch/x86/kernel/asm-offsets_64.c +++ b/arch/x86/kernel/asm-offsets_64.c @@ -21,10 +21,11 @@ static char syscalls_ia32[] = { int main(void) { #ifdef CONFIG_PARAVIRT - OFFSET(PV_CPU_usergs_sysret64, pv_cpu_ops, usergs_sysret64); - OFFSET(PV_CPU_swapgs, pv_cpu_ops, swapgs); + OFFSET(PV_CPU_usergs_sysret64, paravirt_patch_template, + pv_cpu_ops.usergs_sysret64); + OFFSET(PV_CPU_swapgs, paravirt_patch_template, pv_cpu_ops.swapgs); #ifdef CONFIG_DEBUG_ENTRY - OFFSET(PV_IRQ_save_fl, pv_irq_ops, save_fl); + OFFSET(PV_IRQ_save_fl, paravirt_patch_template, pv_irq_ops.save_fl); #endif BLANK(); #endif diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index eb4cb3efd20e..3893df059174 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -1225,7 +1225,7 @@ static void generic_identify(struct cpuinfo_x86 *c) # ifdef CONFIG_PARAVIRT do { extern void native_iret(void); - if (pv_cpu_ops.iret == native_iret) + if (pv_ops.pv_cpu_ops.iret == native_iret) set_cpu_bug(c, X86_BUG_ESPFIX); } while (0); # else diff --git a/arch/x86/kernel/cpu/vmware.c b/arch/x86/kernel/cpu/vmware.c index 8e005329648b..f85302a308f2 100644 --- a/arch/x86/kernel/cpu/vmware.c +++ b/arch/x86/kernel/cpu/vmware.c @@ -97,14 +97,14 @@ static void __init vmware_sched_clock_setup(void) d->cyc2ns_offset = mul_u64_u32_shr(tsc_now, d->cyc2ns_mul, d->cyc2ns_shift); - pv_time_ops.sched_clock = vmware_sched_clock; + pv_ops.pv_time_ops.sched_clock = vmware_sched_clock; pr_info("using sched offset of %llu ns\n", d->cyc2ns_offset); } static void __init vmware_paravirt_ops_setup(void) { pv_info.name = "VMware hypervisor"; - pv_cpu_ops.io_delay = paravirt_nop; + pv_ops.pv_cpu_ops.io_delay = paravirt_nop; if (vmware_tsc_khz && vmw_sched_clock) vmware_sched_clock_setup(); diff --git a/arch/x86/kernel/kvm.c b/arch/x86/kernel/kvm.c index 5b2300b818af..610da165aa26 100644 --- a/arch/x86/kernel/kvm.c +++ b/arch/x86/kernel/kvm.c @@ -292,7 +292,7 @@ static void __init paravirt_ops_setup(void) pv_info.name = "KVM"; if (kvm_para_has_feature(KVM_FEATURE_NOP_IO_DELAY)) - pv_cpu_ops.io_delay = kvm_io_delay; + pv_ops.pv_cpu_ops.io_delay = kvm_io_delay; #ifdef CONFIG_X86_IO_APIC no_timer_check = 1; @@ -549,13 +549,13 @@ static void __init kvm_guest_init(void) if (kvm_para_has_feature(KVM_FEATURE_STEAL_TIME)) { has_steal_clock = 1; - pv_time_ops.steal_clock = kvm_steal_clock; + pv_ops.pv_time_ops.steal_clock = kvm_steal_clock; } if (kvm_para_has_feature(KVM_FEATURE_PV_TLB_FLUSH) && !kvm_para_has_hint(KVM_HINTS_REALTIME) && kvm_para_has_feature(KVM_FEATURE_STEAL_TIME)) - pv_mmu_ops.flush_tlb_others = kvm_flush_tlb_others; + pv_ops.pv_mmu_ops.flush_tlb_others = kvm_flush_tlb_others; if (kvm_para_has_feature(KVM_FEATURE_PV_EOI)) apic_set_eoi_write(kvm_guest_apic_eoi_write); @@ -749,13 +749,15 @@ void __init kvm_spinlock_init(void) return; __pv_init_lock_hash(); - pv_lock_ops.queued_spin_lock_slowpath = __pv_queued_spin_lock_slowpath; - pv_lock_ops.queued_spin_unlock = PV_CALLEE_SAVE(__pv_queued_spin_unlock); - pv_lock_ops.wait = kvm_wait; - pv_lock_ops.kick = kvm_kick_cpu; + pv_ops.pv_lock_ops.queued_spin_lock_slowpath + __pv_queued_spin_lock_slowpath; + pv_ops.pv_lock_ops.queued_spin_unlock + PV_CALLEE_SAVE(__pv_queued_spin_unlock); + pv_ops.pv_lock_ops.wait = kvm_wait; + pv_ops.pv_lock_ops.kick = kvm_kick_cpu; if (kvm_para_has_feature(KVM_FEATURE_STEAL_TIME)) { - pv_lock_ops.vcpu_is_preempted + pv_ops.pv_lock_ops.vcpu_is_preempted PV_CALLEE_SAVE(__kvm_vcpu_is_preempted); } } diff --git a/arch/x86/kernel/kvmclock.c b/arch/x86/kernel/kvmclock.c index 3b8e7c13c614..10a6071b209a 100644 --- a/arch/x86/kernel/kvmclock.c +++ b/arch/x86/kernel/kvmclock.c @@ -104,13 +104,13 @@ static u64 kvm_sched_clock_read(void) static inline void kvm_sched_clock_init(bool stable) { if (!stable) { - pv_time_ops.sched_clock = kvm_clock_read; + pv_ops.pv_time_ops.sched_clock = kvm_clock_read; clear_sched_clock_stable(); return; } kvm_sched_clock_offset = kvm_clock_read(); - pv_time_ops.sched_clock = kvm_sched_clock_read; + pv_ops.pv_time_ops.sched_clock = kvm_sched_clock_read; printk(KERN_INFO "kvm-clock: using sched offset of %llu cycles\n", kvm_sched_clock_offset); diff --git a/arch/x86/kernel/paravirt-spinlocks.c b/arch/x86/kernel/paravirt-spinlocks.c index 71f2d1125ec0..9569481cadb3 100644 --- a/arch/x86/kernel/paravirt-spinlocks.c +++ b/arch/x86/kernel/paravirt-spinlocks.c @@ -17,7 +17,7 @@ PV_CALLEE_SAVE_REGS_THUNK(__native_queued_spin_unlock); bool pv_is_native_spin_unlock(void) { - return pv_lock_ops.queued_spin_unlock.func =+ return pv_ops.pv_lock_ops.queued_spin_unlock.func = __raw_callee_save___native_queued_spin_unlock; } @@ -29,17 +29,6 @@ PV_CALLEE_SAVE_REGS_THUNK(__native_vcpu_is_preempted); bool pv_is_native_vcpu_is_preempted(void) { - return pv_lock_ops.vcpu_is_preempted.func =+ return pv_ops.pv_lock_ops.vcpu_is_preempted.func = __raw_callee_save___native_vcpu_is_preempted; } - -struct pv_lock_ops pv_lock_ops = { -#ifdef CONFIG_SMP - .queued_spin_lock_slowpath = native_queued_spin_lock_slowpath, - .queued_spin_unlock = PV_CALLEE_SAVE(__native_queued_spin_unlock), - .wait = paravirt_nop, - .kick = paravirt_nop, - .vcpu_is_preempted = PV_CALLEE_SAVE(__native_vcpu_is_preempted), -#endif /* SMP */ -}; -EXPORT_SYMBOL(pv_lock_ops); diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c index f0c462fe2808..40ec68135f7a 100644 --- a/arch/x86/kernel/paravirt.c +++ b/arch/x86/kernel/paravirt.c @@ -127,29 +127,14 @@ void __init native_pv_lock_init(void) static_branch_disable(&virt_spin_lock_key); } -/* - * Neat trick to map patch type back to the call within the - * corresponding structure. - */ -static void *get_call_destination(u8 type) -{ - struct paravirt_patch_template tmpl = { - .pv_init_ops = pv_init_ops, - .pv_time_ops = pv_time_ops, - .pv_cpu_ops = pv_cpu_ops, - .pv_irq_ops = pv_irq_ops, - .pv_mmu_ops = pv_mmu_ops, -#ifdef CONFIG_PARAVIRT_SPINLOCKS - .pv_lock_ops = pv_lock_ops, -#endif - }; - return *((void **)&tmpl + type); -} - unsigned paravirt_patch_default(u8 type, void *insnbuf, unsigned long addr, unsigned len) { - void *opfunc = get_call_destination(type); + /* + * Neat trick to map patch type back to the call within the + * corresponding structure. + */ + void *opfunc = *((void **)&pv_ops + type); unsigned ret; if (opfunc == NULL) @@ -315,77 +300,6 @@ struct pv_info pv_info = { #endif }; -struct pv_init_ops pv_init_ops = { - .patch = native_patch, -}; - -struct pv_time_ops pv_time_ops = { - .sched_clock = native_sched_clock, - .steal_clock = native_steal_clock, -}; - -__visible struct pv_irq_ops pv_irq_ops = { - .save_fl = __PV_IS_CALLEE_SAVE(native_save_fl), - .restore_fl = __PV_IS_CALLEE_SAVE(native_restore_fl), - .irq_disable = __PV_IS_CALLEE_SAVE(native_irq_disable), - .irq_enable = __PV_IS_CALLEE_SAVE(native_irq_enable), - .safe_halt = native_safe_halt, - .halt = native_halt, -}; - -__visible struct pv_cpu_ops pv_cpu_ops = { - .cpuid = native_cpuid, - .get_debugreg = native_get_debugreg, - .set_debugreg = native_set_debugreg, - .read_cr0 = native_read_cr0, - .write_cr0 = native_write_cr0, - .write_cr4 = native_write_cr4, -#ifdef CONFIG_X86_64 - .read_cr8 = native_read_cr8, - .write_cr8 = native_write_cr8, -#endif - .wbinvd = native_wbinvd, - .read_msr = native_read_msr, - .write_msr = native_write_msr, - .read_msr_safe = native_read_msr_safe, - .write_msr_safe = native_write_msr_safe, - .read_pmc = native_read_pmc, - .load_tr_desc = native_load_tr_desc, - .set_ldt = native_set_ldt, - .load_gdt = native_load_gdt, - .load_idt = native_load_idt, - .store_tr = native_store_tr, - .load_tls = native_load_tls, -#ifdef CONFIG_X86_64 - .load_gs_index = native_load_gs_index, -#endif - .write_ldt_entry = native_write_ldt_entry, - .write_gdt_entry = native_write_gdt_entry, - .write_idt_entry = native_write_idt_entry, - - .alloc_ldt = paravirt_nop, - .free_ldt = paravirt_nop, - - .load_sp0 = native_load_sp0, - -#ifdef CONFIG_X86_64 - .usergs_sysret64 = native_usergs_sysret64, -#endif - .iret = native_iret, - .swapgs = native_swapgs, - - .set_iopl_mask = native_set_iopl_mask, - .io_delay = native_io_delay, - - .start_context_switch = paravirt_nop, - .end_context_switch = paravirt_nop, -}; - -/* At this point, native_get/set_debugreg has real function entries */ -NOKPROBE_SYMBOL(native_get_debugreg); -NOKPROBE_SYMBOL(native_set_debugreg); -NOKPROBE_SYMBOL(native_load_idt); - #if defined(CONFIG_X86_32) && !defined(CONFIG_X86_PAE) /* 32-bit pagetable entries */ #define PTE_IDENT __PV_IS_CALLEE_SAVE(_paravirt_ident_32) @@ -394,84 +308,162 @@ NOKPROBE_SYMBOL(native_load_idt); #define PTE_IDENT __PV_IS_CALLEE_SAVE(_paravirt_ident_64) #endif -struct pv_mmu_ops pv_mmu_ops __ro_after_init = { - - .read_cr2 = native_read_cr2, - .write_cr2 = native_write_cr2, - .read_cr3 = __native_read_cr3, - .write_cr3 = native_write_cr3, - - .flush_tlb_user = native_flush_tlb, - .flush_tlb_kernel = native_flush_tlb_global, - .flush_tlb_one_user = native_flush_tlb_one_user, - .flush_tlb_others = native_flush_tlb_others, - - .pgd_alloc = __paravirt_pgd_alloc, - .pgd_free = paravirt_nop, +struct paravirt_patch_template pv_ops = { + /* Init ops. */ + .pv_init_ops.patch = native_patch, + + /* Time ops. */ + .pv_time_ops.sched_clock = native_sched_clock, + .pv_time_ops.steal_clock = native_steal_clock, + + /* Cpu ops. */ + .pv_cpu_ops.cpuid = native_cpuid, + .pv_cpu_ops.get_debugreg = native_get_debugreg, + .pv_cpu_ops.set_debugreg = native_set_debugreg, + .pv_cpu_ops.read_cr0 = native_read_cr0, + .pv_cpu_ops.write_cr0 = native_write_cr0, + .pv_cpu_ops.write_cr4 = native_write_cr4, +#ifdef CONFIG_X86_64 + .pv_cpu_ops.read_cr8 = native_read_cr8, + .pv_cpu_ops.write_cr8 = native_write_cr8, +#endif + .pv_cpu_ops.wbinvd = native_wbinvd, + .pv_cpu_ops.read_msr = native_read_msr, + .pv_cpu_ops.write_msr = native_write_msr, + .pv_cpu_ops.read_msr_safe = native_read_msr_safe, + .pv_cpu_ops.write_msr_safe = native_write_msr_safe, + .pv_cpu_ops.read_pmc = native_read_pmc, + .pv_cpu_ops.load_tr_desc = native_load_tr_desc, + .pv_cpu_ops.set_ldt = native_set_ldt, + .pv_cpu_ops.load_gdt = native_load_gdt, + .pv_cpu_ops.load_idt = native_load_idt, + .pv_cpu_ops.store_tr = native_store_tr, + .pv_cpu_ops.load_tls = native_load_tls, +#ifdef CONFIG_X86_64 + .pv_cpu_ops.load_gs_index = native_load_gs_index, +#endif + .pv_cpu_ops.write_ldt_entry = native_write_ldt_entry, + .pv_cpu_ops.write_gdt_entry = native_write_gdt_entry, + .pv_cpu_ops.write_idt_entry = native_write_idt_entry, - .alloc_pte = paravirt_nop, - .alloc_pmd = paravirt_nop, - .alloc_pud = paravirt_nop, - .alloc_p4d = paravirt_nop, - .release_pte = paravirt_nop, - .release_pmd = paravirt_nop, - .release_pud = paravirt_nop, - .release_p4d = paravirt_nop, + .pv_cpu_ops.alloc_ldt = paravirt_nop, + .pv_cpu_ops.free_ldt = paravirt_nop, - .set_pte = native_set_pte, - .set_pte_at = native_set_pte_at, - .set_pmd = native_set_pmd, + .pv_cpu_ops.load_sp0 = native_load_sp0, - .ptep_modify_prot_start = __ptep_modify_prot_start, - .ptep_modify_prot_commit = __ptep_modify_prot_commit, +#ifdef CONFIG_X86_64 + .pv_cpu_ops.usergs_sysret64 = native_usergs_sysret64, +#endif + .pv_cpu_ops.iret = native_iret, + .pv_cpu_ops.swapgs = native_swapgs, + + .pv_cpu_ops.set_iopl_mask = native_set_iopl_mask, + .pv_cpu_ops.io_delay = native_io_delay, + + .pv_cpu_ops.start_context_switch = paravirt_nop, + .pv_cpu_ops.end_context_switch = paravirt_nop, + + /* Irq ops. */ + .pv_irq_ops.save_fl = __PV_IS_CALLEE_SAVE(native_save_fl), + .pv_irq_ops.restore_fl = __PV_IS_CALLEE_SAVE(native_restore_fl), + .pv_irq_ops.irq_disable = __PV_IS_CALLEE_SAVE(native_irq_disable), + .pv_irq_ops.irq_enable = __PV_IS_CALLEE_SAVE(native_irq_enable), + .pv_irq_ops.safe_halt = native_safe_halt, + .pv_irq_ops.halt = native_halt, + + /* Mmu ops. */ + .pv_mmu_ops.read_cr2 = native_read_cr2, + .pv_mmu_ops.write_cr2 = native_write_cr2, + .pv_mmu_ops.read_cr3 = __native_read_cr3, + .pv_mmu_ops.write_cr3 = native_write_cr3, + + .pv_mmu_ops.flush_tlb_user = native_flush_tlb, + .pv_mmu_ops.flush_tlb_kernel = native_flush_tlb_global, + .pv_mmu_ops.flush_tlb_one_user = native_flush_tlb_one_user, + .pv_mmu_ops.flush_tlb_others = native_flush_tlb_others, + + .pv_mmu_ops.pgd_alloc = __paravirt_pgd_alloc, + .pv_mmu_ops.pgd_free = paravirt_nop, + + .pv_mmu_ops.alloc_pte = paravirt_nop, + .pv_mmu_ops.alloc_pmd = paravirt_nop, + .pv_mmu_ops.alloc_pud = paravirt_nop, + .pv_mmu_ops.alloc_p4d = paravirt_nop, + .pv_mmu_ops.release_pte = paravirt_nop, + .pv_mmu_ops.release_pmd = paravirt_nop, + .pv_mmu_ops.release_pud = paravirt_nop, + .pv_mmu_ops.release_p4d = paravirt_nop, + + .pv_mmu_ops.set_pte = native_set_pte, + .pv_mmu_ops.set_pte_at = native_set_pte_at, + .pv_mmu_ops.set_pmd = native_set_pmd, + + .pv_mmu_ops.ptep_modify_prot_start = __ptep_modify_prot_start, + .pv_mmu_ops.ptep_modify_prot_commit = __ptep_modify_prot_commit, #if CONFIG_PGTABLE_LEVELS >= 3 #ifdef CONFIG_X86_PAE - .set_pte_atomic = native_set_pte_atomic, - .pte_clear = native_pte_clear, - .pmd_clear = native_pmd_clear, + .pv_mmu_ops.set_pte_atomic = native_set_pte_atomic, + .pv_mmu_ops.pte_clear = native_pte_clear, + .pv_mmu_ops.pmd_clear = native_pmd_clear, #endif - .set_pud = native_set_pud, + .pv_mmu_ops.set_pud = native_set_pud, - .pmd_val = PTE_IDENT, - .make_pmd = PTE_IDENT, + .pv_mmu_ops.pmd_val = PTE_IDENT, + .pv_mmu_ops.make_pmd = PTE_IDENT, #if CONFIG_PGTABLE_LEVELS >= 4 - .pud_val = PTE_IDENT, - .make_pud = PTE_IDENT, + .pv_mmu_ops.pud_val = PTE_IDENT, + .pv_mmu_ops.make_pud = PTE_IDENT, - .set_p4d = native_set_p4d, + .pv_mmu_ops.set_p4d = native_set_p4d, #if CONFIG_PGTABLE_LEVELS >= 5 - .p4d_val = PTE_IDENT, - .make_p4d = PTE_IDENT, + .pv_mmu_ops.p4d_val = PTE_IDENT, + .pv_mmu_ops.make_p4d = PTE_IDENT, - .set_pgd = native_set_pgd, + .pv_mmu_ops.set_pgd = native_set_pgd, #endif /* CONFIG_PGTABLE_LEVELS >= 5 */ #endif /* CONFIG_PGTABLE_LEVELS >= 4 */ #endif /* CONFIG_PGTABLE_LEVELS >= 3 */ - .pte_val = PTE_IDENT, - .pgd_val = PTE_IDENT, + .pv_mmu_ops.pte_val = PTE_IDENT, + .pv_mmu_ops.pgd_val = PTE_IDENT, - .make_pte = PTE_IDENT, - .make_pgd = PTE_IDENT, + .pv_mmu_ops.make_pte = PTE_IDENT, + .pv_mmu_ops.make_pgd = PTE_IDENT, - .dup_mmap = paravirt_nop, - .exit_mmap = paravirt_nop, - .activate_mm = paravirt_nop, + .pv_mmu_ops.dup_mmap = paravirt_nop, + .pv_mmu_ops.exit_mmap = paravirt_nop, + .pv_mmu_ops.activate_mm = paravirt_nop, - .lazy_mode = { + .pv_mmu_ops.lazy_mode = { .enter = paravirt_nop, .leave = paravirt_nop, .flush = paravirt_nop, }, - .set_fixmap = native_set_fixmap, + .pv_mmu_ops.set_fixmap = native_set_fixmap, + +#if defined(CONFIG_PARAVIRT_SPINLOCKS) + /* Lock ops. */ +#ifdef CONFIG_SMP + .pv_lock_ops.queued_spin_lock_slowpath + native_queued_spin_lock_slowpath, + .pv_lock_ops.queued_spin_unlock + PV_CALLEE_SAVE(__native_queued_spin_unlock), + .pv_lock_ops.wait = paravirt_nop, + .pv_lock_ops.kick = paravirt_nop, + .pv_lock_ops.vcpu_is_preempted + PV_CALLEE_SAVE(__native_vcpu_is_preempted), +#endif /* SMP */ +#endif }; -EXPORT_SYMBOL_GPL(pv_time_ops); -EXPORT_SYMBOL (pv_cpu_ops); -EXPORT_SYMBOL (pv_mmu_ops); +/* At this point, native_get/set_debugreg has real function entries */ +NOKPROBE_SYMBOL(native_get_debugreg); +NOKPROBE_SYMBOL(native_set_debugreg); +NOKPROBE_SYMBOL(native_load_idt); + +EXPORT_SYMBOL_GPL(pv_ops); EXPORT_SYMBOL_GPL(pv_info); -EXPORT_SYMBOL (pv_irq_ops); diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c index 74392d9d51e0..42b936dd5846 100644 --- a/arch/x86/kernel/tsc.c +++ b/arch/x86/kernel/tsc.c @@ -230,7 +230,7 @@ unsigned long long sched_clock(void) bool using_native_sched_clock(void) { - return pv_time_ops.sched_clock == native_sched_clock; + return pv_ops.pv_time_ops.sched_clock == native_sched_clock; } #else unsigned long long diff --git a/arch/x86/kernel/vsmp_64.c b/arch/x86/kernel/vsmp_64.c index f194e5e1e95c..85425584f9a7 100644 --- a/arch/x86/kernel/vsmp_64.c +++ b/arch/x86/kernel/vsmp_64.c @@ -111,11 +111,12 @@ static void __init set_vsmp_pv_ops(void) if (cap & ctl & (1 << 4)) { /* Setup irq ops and turn on vSMP IRQ fastpath handling */ - pv_irq_ops.irq_disable = PV_CALLEE_SAVE(vsmp_irq_disable); - pv_irq_ops.irq_enable = PV_CALLEE_SAVE(vsmp_irq_enable); - pv_irq_ops.save_fl = PV_CALLEE_SAVE(vsmp_save_fl); - pv_irq_ops.restore_fl = PV_CALLEE_SAVE(vsmp_restore_fl); - pv_init_ops.patch = vsmp_patch; + pv_ops.pv_irq_ops.irq_disable + PV_CALLEE_SAVE(vsmp_irq_disable); + pv_ops.pv_irq_ops.irq_enable = PV_CALLEE_SAVE(vsmp_irq_enable); + pv_ops.pv_irq_ops.save_fl = PV_CALLEE_SAVE(vsmp_save_fl); + pv_ops.pv_irq_ops.restore_fl = PV_CALLEE_SAVE(vsmp_restore_fl); + pv_ops.pv_init_ops.patch = vsmp_patch; ctl &= ~(1 << 4); } writel(ctl, address + 4); diff --git a/arch/x86/xen/enlighten_pv.c b/arch/x86/xen/enlighten_pv.c index 439a94bf89ad..dbb3a3b24cf8 100644 --- a/arch/x86/xen/enlighten_pv.c +++ b/arch/x86/xen/enlighten_pv.c @@ -998,11 +998,15 @@ void __ref xen_setup_vcpu_info_placement(void) * percpu area for all cpus, so make use of it. */ if (xen_have_vcpu_info_placement) { - pv_irq_ops.save_fl = __PV_IS_CALLEE_SAVE(xen_save_fl_direct); - pv_irq_ops.restore_fl = __PV_IS_CALLEE_SAVE(xen_restore_fl_direct); - pv_irq_ops.irq_disable = __PV_IS_CALLEE_SAVE(xen_irq_disable_direct); - pv_irq_ops.irq_enable = __PV_IS_CALLEE_SAVE(xen_irq_enable_direct); - pv_mmu_ops.read_cr2 = xen_read_cr2_direct; + pv_ops.pv_irq_ops.save_fl + __PV_IS_CALLEE_SAVE(xen_save_fl_direct); + pv_ops.pv_irq_ops.restore_fl + __PV_IS_CALLEE_SAVE(xen_restore_fl_direct); + pv_ops.pv_irq_ops.irq_disable + __PV_IS_CALLEE_SAVE(xen_irq_disable_direct); + pv_ops.pv_irq_ops.irq_enable + __PV_IS_CALLEE_SAVE(xen_irq_enable_direct); + pv_ops.pv_mmu_ops.read_cr2 = xen_read_cr2_direct; } } @@ -1177,14 +1181,14 @@ static void __init xen_boot_params_init_edd(void) */ static void xen_setup_gdt(int cpu) { - pv_cpu_ops.write_gdt_entry = xen_write_gdt_entry_boot; - pv_cpu_ops.load_gdt = xen_load_gdt_boot; + pv_ops.pv_cpu_ops.write_gdt_entry = xen_write_gdt_entry_boot; + pv_ops.pv_cpu_ops.load_gdt = xen_load_gdt_boot; setup_stack_canary_segment(0); switch_to_new_gdt(0); - pv_cpu_ops.write_gdt_entry = xen_write_gdt_entry; - pv_cpu_ops.load_gdt = xen_load_gdt; + pv_ops.pv_cpu_ops.write_gdt_entry = xen_write_gdt_entry; + pv_ops.pv_cpu_ops.load_gdt = xen_load_gdt; } static void __init xen_dom0_set_legacy_features(void) @@ -1209,8 +1213,8 @@ asmlinkage __visible void __init xen_start_kernel(void) /* Install Xen paravirt ops */ pv_info = xen_info; - pv_init_ops.patch = paravirt_patch_default; - pv_cpu_ops = xen_cpu_ops; + pv_ops.pv_init_ops.patch = paravirt_patch_default; + pv_ops.pv_cpu_ops = xen_cpu_ops; xen_init_irq_ops(); /* @@ -1274,8 +1278,10 @@ asmlinkage __visible void __init xen_start_kernel(void) #endif if (xen_feature(XENFEAT_mmu_pt_update_preserve_ad)) { - pv_mmu_ops.ptep_modify_prot_start = xen_ptep_modify_prot_start; - pv_mmu_ops.ptep_modify_prot_commit = xen_ptep_modify_prot_commit; + pv_ops.pv_mmu_ops.ptep_modify_prot_start + xen_ptep_modify_prot_start; + pv_ops.pv_mmu_ops.ptep_modify_prot_commit + xen_ptep_modify_prot_commit; } machine_ops = xen_machine_ops; diff --git a/arch/x86/xen/irq.c b/arch/x86/xen/irq.c index 7515a19fd324..2df69ffc33d6 100644 --- a/arch/x86/xen/irq.c +++ b/arch/x86/xen/irq.c @@ -128,6 +128,6 @@ static const struct pv_irq_ops xen_irq_ops __initconst = { void __init xen_init_irq_ops(void) { - pv_irq_ops = xen_irq_ops; + pv_ops.pv_irq_ops = xen_irq_ops; x86_init.irqs.intr_init = xen_init_IRQ; } diff --git a/arch/x86/xen/mmu_hvm.c b/arch/x86/xen/mmu_hvm.c index dd2ad82eee80..5ef3ba3e748f 100644 --- a/arch/x86/xen/mmu_hvm.c +++ b/arch/x86/xen/mmu_hvm.c @@ -73,7 +73,7 @@ static int is_pagetable_dying_supported(void) void __init xen_hvm_init_mmu_ops(void) { if (is_pagetable_dying_supported()) - pv_mmu_ops.exit_mmap = xen_hvm_exit_mmap; + pv_ops.pv_mmu_ops.exit_mmap = xen_hvm_exit_mmap; #ifdef CONFIG_PROC_VMCORE WARN_ON(register_oldmem_pfn_is_ram(&xen_oldmem_pfn_is_ram)); #endif diff --git a/arch/x86/xen/mmu_pv.c b/arch/x86/xen/mmu_pv.c index b7ec689320c7..9f46c67787bc 100644 --- a/arch/x86/xen/mmu_pv.c +++ b/arch/x86/xen/mmu_pv.c @@ -2215,7 +2215,7 @@ static void __init xen_write_cr3_init(unsigned long cr3) set_page_prot(initial_page_table, PAGE_KERNEL); set_page_prot(initial_kernel_pmd, PAGE_KERNEL); - pv_mmu_ops.write_cr3 = &xen_write_cr3; + pv_ops.pv_mmu_ops.write_cr3 = &xen_write_cr3; } /* @@ -2364,27 +2364,27 @@ static void xen_set_fixmap(unsigned idx, phys_addr_t phys, pgprot_t prot) static void __init xen_post_allocator_init(void) { - pv_mmu_ops.set_pte = xen_set_pte; - pv_mmu_ops.set_pmd = xen_set_pmd; - pv_mmu_ops.set_pud = xen_set_pud; + pv_ops.pv_mmu_ops.set_pte = xen_set_pte; + pv_ops.pv_mmu_ops.set_pmd = xen_set_pmd; + pv_ops.pv_mmu_ops.set_pud = xen_set_pud; #ifdef CONFIG_X86_64 - pv_mmu_ops.set_p4d = xen_set_p4d; + pv_ops.pv_mmu_ops.set_p4d = xen_set_p4d; #endif /* This will work as long as patching hasn't happened yet (which it hasn't) */ - pv_mmu_ops.alloc_pte = xen_alloc_pte; - pv_mmu_ops.alloc_pmd = xen_alloc_pmd; - pv_mmu_ops.release_pte = xen_release_pte; - pv_mmu_ops.release_pmd = xen_release_pmd; + pv_ops.pv_mmu_ops.alloc_pte = xen_alloc_pte; + pv_ops.pv_mmu_ops.alloc_pmd = xen_alloc_pmd; + pv_ops.pv_mmu_ops.release_pte = xen_release_pte; + pv_ops.pv_mmu_ops.release_pmd = xen_release_pmd; #ifdef CONFIG_X86_64 - pv_mmu_ops.alloc_pud = xen_alloc_pud; - pv_mmu_ops.release_pud = xen_release_pud; + pv_ops.pv_mmu_ops.alloc_pud = xen_alloc_pud; + pv_ops.pv_mmu_ops.release_pud = xen_release_pud; #endif - pv_mmu_ops.make_pte = PV_CALLEE_SAVE(xen_make_pte); + pv_ops.pv_mmu_ops.make_pte = PV_CALLEE_SAVE(xen_make_pte); #ifdef CONFIG_X86_64 - pv_mmu_ops.write_cr3 = &xen_write_cr3; + pv_ops.pv_mmu_ops.write_cr3 = &xen_write_cr3; #endif } @@ -2471,7 +2471,7 @@ void __init xen_init_mmu_ops(void) x86_init.paging.pagetable_init = xen_pagetable_init; x86_init.hyper.init_after_bootmem = xen_after_bootmem; - pv_mmu_ops = xen_mmu_ops; + pv_ops.pv_mmu_ops = xen_mmu_ops; memset(dummy_mapping, 0xff, PAGE_SIZE); } diff --git a/arch/x86/xen/spinlock.c b/arch/x86/xen/spinlock.c index cd97a62394e7..53b213af8c26 100644 --- a/arch/x86/xen/spinlock.c +++ b/arch/x86/xen/spinlock.c @@ -137,11 +137,13 @@ void __init xen_init_spinlocks(void) printk(KERN_DEBUG "xen: PV spinlocks enabled\n"); __pv_init_lock_hash(); - pv_lock_ops.queued_spin_lock_slowpath = __pv_queued_spin_lock_slowpath; - pv_lock_ops.queued_spin_unlock = PV_CALLEE_SAVE(__pv_queued_spin_unlock); - pv_lock_ops.wait = xen_qlock_wait; - pv_lock_ops.kick = xen_qlock_kick; - pv_lock_ops.vcpu_is_preempted = PV_CALLEE_SAVE(xen_vcpu_stolen); + pv_ops.pv_lock_ops.queued_spin_lock_slowpath + __pv_queued_spin_lock_slowpath; + pv_ops.pv_lock_ops.queued_spin_unlock + PV_CALLEE_SAVE(__pv_queued_spin_unlock); + pv_ops.pv_lock_ops.wait = xen_qlock_wait; + pv_ops.pv_lock_ops.kick = xen_qlock_kick; + pv_ops.pv_lock_ops.vcpu_is_preempted = PV_CALLEE_SAVE(xen_vcpu_stolen); } static __init int xen_parse_nopvspin(char *arg) diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c index e0f1bcf01d63..66311c3e90b0 100644 --- a/arch/x86/xen/time.c +++ b/arch/x86/xen/time.c @@ -505,7 +505,7 @@ static void __init xen_time_init(void) void __ref xen_init_time_ops(void) { - pv_time_ops = xen_time_ops; + pv_ops.pv_time_ops = xen_time_ops; x86_init.timers.timer_init = xen_time_init; x86_init.timers.setup_percpu_clockev = x86_init_noop; @@ -547,7 +547,7 @@ void __init xen_hvm_init_time_ops(void) return; } - pv_time_ops = xen_time_ops; + pv_ops.pv_time_ops = xen_time_ops; x86_init.timers.setup_percpu_clockev = xen_time_init; x86_cpuinit.setup_percpu_clockev = xen_hvm_setup_cpu_clockevents; diff --git a/drivers/xen/time.c b/drivers/xen/time.c index 3e741cd1409c..994fb1ae64b3 100644 --- a/drivers/xen/time.c +++ b/drivers/xen/time.c @@ -175,7 +175,7 @@ void __init xen_time_setup_guest(void) xen_runstate_remote = !HYPERVISOR_vm_assist(VMASST_CMD_enable, VMASST_TYPE_runstate_update_flag); - pv_time_ops.steal_clock = xen_steal_clock; + pv_ops.pv_time_ops.steal_clock = xen_steal_clock; static_key_slow_inc(¶virt_steal_enabled); if (xen_runstate_remote) -- 2.13.7
Juergen Gross
2018-Aug-10 11:52 UTC
[PATCH 05/10] x86/paravirt: remove unused paravirt bits
The macros ENABLE_INTERRUPTS_SYSEXIT, GET_CR0_INTO_EAX and PARAVIRT_ADJUST_EXCEPTION_FRAME are used nowhere. Remove their definitions. Signed-off-by: Juergen Gross <jgross at suse.com> --- arch/x86/include/asm/irqflags.h | 4 ---- arch/x86/include/asm/paravirt.h | 9 +-------- arch/x86/kernel/asm-offsets.c | 1 - 3 files changed, 1 insertion(+), 13 deletions(-) diff --git a/arch/x86/include/asm/irqflags.h b/arch/x86/include/asm/irqflags.h index c4fc17220df9..b7a790d03229 100644 --- a/arch/x86/include/asm/irqflags.h +++ b/arch/x86/include/asm/irqflags.h @@ -132,8 +132,6 @@ static inline notrace unsigned long arch_local_irq_save(void) */ #define SWAPGS_UNSAFE_STACK swapgs -#define PARAVIRT_ADJUST_EXCEPTION_FRAME /* */ - #define INTERRUPT_RETURN jmp native_iret #define USERGS_SYSRET64 \ swapgs; \ @@ -147,8 +145,6 @@ static inline notrace unsigned long arch_local_irq_save(void) #endif #else #define INTERRUPT_RETURN iret -#define ENABLE_INTERRUPTS_SYSEXIT sti; sysexit -#define GET_CR0_INTO_EAX movl %cr0, %eax #endif diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h index 1b86bb319393..436d270e622b 100644 --- a/arch/x86/include/asm/paravirt.h +++ b/arch/x86/include/asm/paravirt.h @@ -904,14 +904,7 @@ extern void default_banner(void); call PARA_INDIRECT(pv_ops+PV_IRQ_irq_enable); \ PV_RESTORE_REGS(clobbers | CLBR_CALLEE_SAVE);) -#ifdef CONFIG_X86_32 -#define GET_CR0_INTO_EAX \ - push %ecx; push %edx; \ - ANNOTATE_RETPOLINE_SAFE; \ - call PARA_INDIRECT(pv_ops+PV_CPU_read_cr0); \ - pop %edx; pop %ecx -#else /* !CONFIG_X86_32 */ - +#ifdef CONFIG_X86_64 /* * If swapgs is used while the userspace stack is still current, * there's no way to call a pvop. The PV replacement *must* be diff --git a/arch/x86/kernel/asm-offsets.c b/arch/x86/kernel/asm-offsets.c index bec9fc3498f8..d0f0348209cb 100644 --- a/arch/x86/kernel/asm-offsets.c +++ b/arch/x86/kernel/asm-offsets.c @@ -71,7 +71,6 @@ void common(void) { OFFSET(PV_IRQ_irq_enable, paravirt_patch_template, pv_irq_ops.irq_enable); OFFSET(PV_CPU_iret, paravirt_patch_template, pv_cpu_ops.iret); - OFFSET(PV_CPU_read_cr0, paravirt_patch_template, pv_cpu_ops.read_cr0); OFFSET(PV_MMU_read_cr2, paravirt_patch_template, pv_mmu_ops.read_cr2); #endif -- 2.13.7
Juergen Gross
2018-Aug-10 11:52 UTC
[PATCH 06/10] x86/paravirt: introduce new config option PARAVIRT_XXL
A large amount of paravirt ops is used by Xen PV guests only. Add a new config option PARAVIRT_XXL which is selected by XEN_PV. Later we can put the Xen PV only paravirt ops under the PARACVIRT_XXL umbrella. Signed-off-by: Juergen Gross <jgross at suse.com> --- arch/x86/Kconfig | 3 +++ arch/x86/xen/Kconfig | 1 + 2 files changed, 4 insertions(+) diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 887d3a7bb646..3c967b803c21 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -754,6 +754,9 @@ config PARAVIRT over full virtualization. However, when run without a hypervisor the kernel is theoretically slower and slightly larger. +config PARAVIRT_XXL + bool + config PARAVIRT_DEBUG bool "paravirt-ops debugging" depends on PARAVIRT && DEBUG_KERNEL diff --git a/arch/x86/xen/Kconfig b/arch/x86/xen/Kconfig index c1f98f32c45f..dd92d7bd3613 100644 --- a/arch/x86/xen/Kconfig +++ b/arch/x86/xen/Kconfig @@ -18,6 +18,7 @@ config XEN_PV bool "Xen PV guest support" default y depends on XEN + select PARAVIRT_XXL select XEN_HAVE_PVMMU select XEN_HAVE_VPMU help -- 2.13.7
Juergen Gross
2018-Aug-10 11:52 UTC
[PATCH 07/10] x86/paravirt: move items in pv_info under PARAVIRT_XXL umbrella
All items but name in pv_info are needed by Xen PV only. Define them with CONFIG_PARAVIRT_XXL set only. Signed-off-by: Juergen Gross <jgross at suse.com> --- arch/x86/include/asm/paravirt.h | 2 ++ arch/x86/include/asm/paravirt_types.h | 2 ++ arch/x86/include/asm/pgtable-3level_types.h | 2 +- arch/x86/include/asm/ptrace.h | 3 ++- arch/x86/include/asm/segment.h | 2 +- arch/x86/kernel/paravirt.c | 2 ++ 6 files changed, 10 insertions(+), 3 deletions(-) diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h index 436d270e622b..afc0469979f7 100644 --- a/arch/x86/include/asm/paravirt.h +++ b/arch/x86/include/asm/paravirt.h @@ -104,7 +104,9 @@ static inline void wbinvd(void) PVOP_VCALL0(pv_cpu_ops.wbinvd); } +#ifdef CONFIG_PARAVIRT_XXL #define get_kernel_rpl() (pv_info.kernel_rpl) +#endif static inline u64 paravirt_read_msr(unsigned msr) { diff --git a/arch/x86/include/asm/paravirt_types.h b/arch/x86/include/asm/paravirt_types.h index ed024e90b863..f1bdc4c9ff4c 100644 --- a/arch/x86/include/asm/paravirt_types.h +++ b/arch/x86/include/asm/paravirt_types.h @@ -65,12 +65,14 @@ struct paravirt_callee_save { /* general info */ struct pv_info { +#ifdef CONFIG_PARAVIRT_XXL unsigned int kernel_rpl; int shared_kernel_pmd; #ifdef CONFIG_X86_64 u16 extra_user_64bit_cs; /* __USER_CS if none */ #endif +#endif const char *name; }; diff --git a/arch/x86/include/asm/pgtable-3level_types.h b/arch/x86/include/asm/pgtable-3level_types.h index 6a59a6d0cc50..1aa68ca1907c 100644 --- a/arch/x86/include/asm/pgtable-3level_types.h +++ b/arch/x86/include/asm/pgtable-3level_types.h @@ -20,7 +20,7 @@ typedef union { } pte_t; #endif /* !__ASSEMBLY__ */ -#ifdef CONFIG_PARAVIRT +#ifdef CONFIG_PARAVIRT_XXL #define SHARED_KERNEL_PMD (pv_info.shared_kernel_pmd) #else #define SHARED_KERNEL_PMD 1 diff --git a/arch/x86/include/asm/ptrace.h b/arch/x86/include/asm/ptrace.h index 6de1fd3d0097..c9ac6ff5f7d2 100644 --- a/arch/x86/include/asm/ptrace.h +++ b/arch/x86/include/asm/ptrace.h @@ -144,7 +144,8 @@ static inline int v8086_mode(struct pt_regs *regs) static inline bool user_64bit_mode(struct pt_regs *regs) { #ifdef CONFIG_X86_64 -#ifndef CONFIG_PARAVIRT +/* Early boot code has CONFIG_PARAVIRT undefined! */ +#if !defined(CONFIG_PARAVIRT) || !defined(CONFIG_PARAVIRT_XXL) /* * On non-paravirt systems, this is the only long mode CPL 3 * selector. We do not allow long mode selectors in the LDT. diff --git a/arch/x86/include/asm/segment.h b/arch/x86/include/asm/segment.h index e293c122d0d5..0ffbe9519e68 100644 --- a/arch/x86/include/asm/segment.h +++ b/arch/x86/include/asm/segment.h @@ -211,7 +211,7 @@ #endif -#ifndef CONFIG_PARAVIRT +#ifndef CONFIG_PARAVIRT_XXL # define get_kernel_rpl() 0 #endif diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c index 40ec68135f7a..168901f4dc09 100644 --- a/arch/x86/kernel/paravirt.c +++ b/arch/x86/kernel/paravirt.c @@ -292,12 +292,14 @@ enum paravirt_lazy_mode paravirt_get_lazy_mode(void) struct pv_info pv_info = { .name = "bare hardware", +#ifdef CONFIG_PARAVIRT_XXL .kernel_rpl = 0, .shared_kernel_pmd = 1, /* Only used when CONFIG_X86_PAE is set */ #ifdef CONFIG_X86_64 .extra_user_64bit_cs = __USER_CS, #endif +#endif }; #if defined(CONFIG_X86_32) && !defined(CONFIG_X86_PAE) -- 2.13.7
Juergen Gross
2018-Aug-10 11:52 UTC
[PATCH 08/10] x86/paravirt: move the Xen-only pv_cpu_ops under the PARAVIRT_XXL umbrella
Most of the paravirt ops defined in pv_cpu_ops are for Xen PV guests only. Define them only if CONFIG_PARAVIRT_XXL is set. Signed-off-by: Juergen Gross <jgross at suse.com> --- arch/x86/include/asm/debugreg.h | 2 +- arch/x86/include/asm/desc.h | 4 ++-- arch/x86/include/asm/irqflags.h | 16 +++++++++++----- arch/x86/include/asm/msr.h | 4 ++-- arch/x86/include/asm/paravirt.h | 15 +++++++++++++-- arch/x86/include/asm/paravirt_types.h | 5 ++++- arch/x86/include/asm/pgtable.h | 6 ++++-- arch/x86/include/asm/processor.h | 4 ++-- arch/x86/include/asm/special_insns.h | 9 +++++++-- arch/x86/kernel/asm-offsets.c | 2 ++ arch/x86/kernel/asm-offsets_64.c | 2 ++ arch/x86/kernel/cpu/common.c | 2 +- arch/x86/kernel/head_64.S | 2 ++ arch/x86/kernel/paravirt.c | 13 ++++++++++++- arch/x86/kernel/paravirt_patch_32.c | 4 ++++ arch/x86/kernel/paravirt_patch_64.c | 6 +++++- 16 files changed, 74 insertions(+), 22 deletions(-) diff --git a/arch/x86/include/asm/debugreg.h b/arch/x86/include/asm/debugreg.h index 4505ac2735ad..9e5ca30738e5 100644 --- a/arch/x86/include/asm/debugreg.h +++ b/arch/x86/include/asm/debugreg.h @@ -8,7 +8,7 @@ DECLARE_PER_CPU(unsigned long, cpu_dr7); -#ifndef CONFIG_PARAVIRT +#ifndef CONFIG_PARAVIRT_XXL /* * These special macros can be used to get or set a debugging register */ diff --git a/arch/x86/include/asm/desc.h b/arch/x86/include/asm/desc.h index 13c5ee878a47..68a99d2a5f33 100644 --- a/arch/x86/include/asm/desc.h +++ b/arch/x86/include/asm/desc.h @@ -108,7 +108,7 @@ static inline int desc_empty(const void *ptr) return !(desc[0] | desc[1]); } -#ifdef CONFIG_PARAVIRT +#ifdef CONFIG_PARAVIRT_XXL #include <asm/paravirt.h> #else #define load_TR_desc() native_load_tr_desc() @@ -134,7 +134,7 @@ static inline void paravirt_alloc_ldt(struct desc_struct *ldt, unsigned entries) static inline void paravirt_free_ldt(struct desc_struct *ldt, unsigned entries) { } -#endif /* CONFIG_PARAVIRT */ +#endif /* CONFIG_PARAVIRT_XXL */ #define store_ldt(ldt) asm("sldt %0" : "=m"(ldt)) diff --git a/arch/x86/include/asm/irqflags.h b/arch/x86/include/asm/irqflags.h index b7a790d03229..03bb451e4e6b 100644 --- a/arch/x86/include/asm/irqflags.h +++ b/arch/x86/include/asm/irqflags.h @@ -120,6 +120,16 @@ static inline notrace unsigned long arch_local_irq_save(void) #define DISABLE_INTERRUPTS(x) cli #ifdef CONFIG_X86_64 +#ifdef CONFIG_DEBUG_ENTRY +#define SAVE_FLAGS(x) pushfq; popq %rax +#endif +#endif +#endif /* __ASSEMBLY__ */ +#endif /* CONFIG_PARAVIRT */ + +#ifndef CONFIG_PARAVIRT_XXL +#ifdef __ASSEMBLY__ +#ifdef CONFIG_X86_64 #define SWAPGS swapgs /* * Currently paravirt can't handle swapgs nicely when we @@ -140,16 +150,12 @@ static inline notrace unsigned long arch_local_irq_save(void) swapgs; \ sysretl -#ifdef CONFIG_DEBUG_ENTRY -#define SAVE_FLAGS(x) pushfq; popq %rax -#endif #else #define INTERRUPT_RETURN iret #endif - #endif /* __ASSEMBLY__ */ -#endif /* CONFIG_PARAVIRT */ +#endif /* CONFIG_PARAVIRT_XXL */ #ifndef __ASSEMBLY__ static inline int arch_irqs_disabled_flags(unsigned long flags) diff --git a/arch/x86/include/asm/msr.h b/arch/x86/include/asm/msr.h index 04addd6e0a4a..91e4cf189914 100644 --- a/arch/x86/include/asm/msr.h +++ b/arch/x86/include/asm/msr.h @@ -242,7 +242,7 @@ static inline unsigned long long native_read_pmc(int counter) return EAX_EDX_VAL(val, low, high); } -#ifdef CONFIG_PARAVIRT +#ifdef CONFIG_PARAVIRT_XXL #include <asm/paravirt.h> #else #include <linux/errno.h> @@ -305,7 +305,7 @@ do { \ #define rdpmcl(counter, val) ((val) = native_read_pmc(counter)) -#endif /* !CONFIG_PARAVIRT */ +#endif /* !CONFIG_PARAVIRT_XXL */ /* * 64-bit version of wrmsr_safe(): diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h index afc0469979f7..bc9a72a767c8 100644 --- a/arch/x86/include/asm/paravirt.h +++ b/arch/x86/include/asm/paravirt.h @@ -17,6 +17,7 @@ #include <linux/cpumask.h> #include <asm/frame.h> +#ifdef CONFIG_PARAVIRT_XXL static inline void load_sp0(unsigned long sp0) { PVOP_VCALL1(pv_cpu_ops.load_sp0, sp0); @@ -51,6 +52,7 @@ static inline void write_cr0(unsigned long x) { PVOP_VCALL1(pv_cpu_ops.write_cr0, x); } +#endif static inline unsigned long read_cr2(void) { @@ -72,6 +74,7 @@ static inline void write_cr3(unsigned long x) PVOP_VCALL1(pv_mmu_ops.write_cr3, x); } +#ifdef CONFIG_PARAVIRT_XXL static inline void __write_cr4(unsigned long x) { PVOP_VCALL1(pv_cpu_ops.write_cr4, x); @@ -88,6 +91,7 @@ static inline void write_cr8(unsigned long x) PVOP_VCALL1(pv_cpu_ops.write_cr8, x); } #endif +#endif static inline void arch_safe_halt(void) { @@ -99,14 +103,13 @@ static inline void halt(void) PVOP_VCALL0(pv_irq_ops.halt); } +#ifdef CONFIG_PARAVIRT_XXL static inline void wbinvd(void) { PVOP_VCALL0(pv_cpu_ops.wbinvd); } -#ifdef CONFIG_PARAVIRT_XXL #define get_kernel_rpl() (pv_info.kernel_rpl) -#endif static inline u64 paravirt_read_msr(unsigned msr) { @@ -171,6 +174,7 @@ static inline int rdmsrl_safe(unsigned msr, unsigned long long *p) *p = paravirt_read_msr_safe(msr, &err); return err; } +#endif static inline unsigned long long paravirt_sched_clock(void) { @@ -186,6 +190,7 @@ static inline u64 paravirt_steal_clock(int cpu) return PVOP_CALL1(u64, pv_time_ops.steal_clock, cpu); } +#ifdef CONFIG_PARAVIRT_XXL static inline unsigned long long paravirt_read_pmc(int counter) { return PVOP_CALL1(u64, pv_cpu_ops.read_pmc, counter); @@ -230,6 +235,7 @@ static inline unsigned long paravirt_store_tr(void) { return PVOP_CALL0(unsigned long, pv_cpu_ops.store_tr); } + #define store_tr(tr) ((tr) = paravirt_store_tr()) static inline void load_TLS(struct thread_struct *t, unsigned cpu) { @@ -263,6 +269,7 @@ static inline void set_iopl_mask(unsigned mask) { PVOP_VCALL1(pv_cpu_ops.set_iopl_mask, mask); } +#endif /* The paravirtualized I/O functions */ static inline void slow_down_io(void) @@ -635,6 +642,7 @@ static inline void pmd_clear(pmd_t *pmdp) } #endif /* CONFIG_X86_PAE */ +#ifdef CONFIG_PARAVIRT_XXL #define __HAVE_ARCH_START_CONTEXT_SWITCH static inline void arch_start_context_switch(struct task_struct *prev) { @@ -645,6 +653,7 @@ static inline void arch_end_context_switch(struct task_struct *next) { PVOP_VCALL1(pv_cpu_ops.end_context_switch, next); } +#endif #define __HAVE_ARCH_ENTER_LAZY_MMU_MODE static inline void arch_enter_lazy_mmu_mode(void) @@ -887,10 +896,12 @@ extern void default_banner(void); #define PARA_INDIRECT(addr) *%cs:addr #endif +#ifdef CONFIG_PARAVIRT_XXL #define INTERRUPT_RETURN \ PARA_SITE(PARA_PATCH(PV_CPU_iret), \ ANNOTATE_RETPOLINE_SAFE; \ jmp PARA_INDIRECT(pv_ops+PV_CPU_iret);) +#endif #define DISABLE_INTERRUPTS(clobbers) \ PARA_SITE(PARA_PATCH(PV_IRQ_irq_disable), \ diff --git a/arch/x86/include/asm/paravirt_types.h b/arch/x86/include/asm/paravirt_types.h index f1bdc4c9ff4c..be356aacc82c 100644 --- a/arch/x86/include/asm/paravirt_types.h +++ b/arch/x86/include/asm/paravirt_types.h @@ -105,6 +105,9 @@ struct pv_time_ops { struct pv_cpu_ops { /* hooks for various privileged instructions */ + void (*io_delay)(void); + +#ifdef CONFIG_PARAVIRT_XXL unsigned long (*get_debugreg)(int regno); void (*set_debugreg)(int regno, unsigned long value); @@ -142,7 +145,6 @@ struct pv_cpu_ops { void (*set_iopl_mask)(unsigned mask); void (*wbinvd)(void); - void (*io_delay)(void); /* cpuid emulation, mostly so that caps bits can be disabled */ void (*cpuid)(unsigned int *eax, unsigned int *ebx, @@ -177,6 +179,7 @@ struct pv_cpu_ops { void (*start_context_switch)(struct task_struct *prev); void (*end_context_switch)(struct task_struct *next); +#endif } __no_randomize_layout; struct pv_irq_ops { diff --git a/arch/x86/include/asm/pgtable.h b/arch/x86/include/asm/pgtable.h index 5715647fc4fe..9ea291fe7107 100644 --- a/arch/x86/include/asm/pgtable.h +++ b/arch/x86/include/asm/pgtable.h @@ -108,10 +108,12 @@ extern pmdval_t early_pmd_flags; #define pte_val(x) native_pte_val(x) #define __pte(x) native_make_pte(x) -#define arch_end_context_switch(prev) do {} while(0) - #endif /* CONFIG_PARAVIRT */ +#ifndef CONFIG_PARAVIRT_XXL +#define arch_end_context_switch(prev) do {} while(0) +#endif /* CONFIG_PARAVIRT_XXL */ + /* * The following only work if pte_present() is true. * Undefined behaviour if not.. diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h index cfd29ee8c3da..7a8fa57218c2 100644 --- a/arch/x86/include/asm/processor.h +++ b/arch/x86/include/asm/processor.h @@ -571,7 +571,7 @@ static inline bool on_thread_stack(void) current_stack_pointer) < THREAD_SIZE; } -#ifdef CONFIG_PARAVIRT +#ifdef CONFIG_PARAVIRT_XXL #include <asm/paravirt.h> #else #define __cpuid native_cpuid @@ -582,7 +582,7 @@ static inline void load_sp0(unsigned long sp0) } #define set_iopl_mask native_set_iopl_mask -#endif /* CONFIG_PARAVIRT */ +#endif /* CONFIG_PARAVIRT_XXL */ /* Free all resources held by a thread. */ extern void release_thread(struct task_struct *); diff --git a/arch/x86/include/asm/special_insns.h b/arch/x86/include/asm/special_insns.h index 317fc59b512c..2aa6ce4bf159 100644 --- a/arch/x86/include/asm/special_insns.h +++ b/arch/x86/include/asm/special_insns.h @@ -143,8 +143,9 @@ static inline unsigned long __read_cr4(void) #ifdef CONFIG_PARAVIRT #include <asm/paravirt.h> -#else +#endif +#ifndef CONFIG_PARAVIRT_XXL static inline unsigned long read_cr0(void) { return native_read_cr0(); @@ -154,7 +155,9 @@ static inline void write_cr0(unsigned long x) { native_write_cr0(x); } +#endif +#ifndef CONFIG_PARAVIRT static inline unsigned long read_cr2(void) { return native_read_cr2(); @@ -178,7 +181,9 @@ static inline void write_cr3(unsigned long x) { native_write_cr3(x); } +#endif +#ifndef CONFIG_PARAVIRT_XXL static inline void __write_cr4(unsigned long x) { native_write_cr4(x); @@ -208,7 +213,7 @@ static inline void load_gs_index(unsigned selector) #endif -#endif/* CONFIG_PARAVIRT */ +#endif/* CONFIG_PARAVIRT_XXL */ static inline void clflush(volatile void *__p) { diff --git a/arch/x86/kernel/asm-offsets.c b/arch/x86/kernel/asm-offsets.c index d0f0348209cb..429db3c8a0cc 100644 --- a/arch/x86/kernel/asm-offsets.c +++ b/arch/x86/kernel/asm-offsets.c @@ -70,7 +70,9 @@ void common(void) { pv_irq_ops.irq_disable); OFFSET(PV_IRQ_irq_enable, paravirt_patch_template, pv_irq_ops.irq_enable); +#ifdef CONFIG_PARAVIRT_XXL OFFSET(PV_CPU_iret, paravirt_patch_template, pv_cpu_ops.iret); +#endif OFFSET(PV_MMU_read_cr2, paravirt_patch_template, pv_mmu_ops.read_cr2); #endif diff --git a/arch/x86/kernel/asm-offsets_64.c b/arch/x86/kernel/asm-offsets_64.c index 2add567c1b2a..c98d2935af94 100644 --- a/arch/x86/kernel/asm-offsets_64.c +++ b/arch/x86/kernel/asm-offsets_64.c @@ -21,9 +21,11 @@ static char syscalls_ia32[] = { int main(void) { #ifdef CONFIG_PARAVIRT +#ifdef CONFIG_PARAVIRT_XXL OFFSET(PV_CPU_usergs_sysret64, paravirt_patch_template, pv_cpu_ops.usergs_sysret64); OFFSET(PV_CPU_swapgs, paravirt_patch_template, pv_cpu_ops.swapgs); +#endif #ifdef CONFIG_DEBUG_ENTRY OFFSET(PV_IRQ_save_fl, paravirt_patch_template, pv_irq_ops.save_fl); #endif diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index 3893df059174..c97f4e0ebad6 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -1222,7 +1222,7 @@ static void generic_identify(struct cpuinfo_x86 *c) * ESPFIX issue, we can change this. */ #ifdef CONFIG_X86_32 -# ifdef CONFIG_PARAVIRT +# ifdef CONFIG_PARAVIRT_XXL do { extern void native_iret(void); if (pv_ops.pv_cpu_ops.iret == native_iret) diff --git a/arch/x86/kernel/head_64.S b/arch/x86/kernel/head_64.S index 8344dd2f310a..e11b96b2dc6b 100644 --- a/arch/x86/kernel/head_64.S +++ b/arch/x86/kernel/head_64.S @@ -31,6 +31,8 @@ #define GET_CR2_INTO(reg) GET_CR2_INTO_RAX ; movq %rax, reg #else #define GET_CR2_INTO(reg) movq %cr2, reg +#endif +#ifndef CONFIG_PARAVIRT_XXL #define INTERRUPT_RETURN iretq #endif diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c index 168901f4dc09..437be9454cab 100644 --- a/arch/x86/kernel/paravirt.c +++ b/arch/x86/kernel/paravirt.c @@ -100,6 +100,7 @@ static unsigned paravirt_patch_call(void *insnbuf, const void *target, return 5; } +#ifdef CONFIG_PARAVIRT_XXL static unsigned paravirt_patch_jmp(void *insnbuf, const void *target, unsigned long addr, unsigned len) { @@ -118,6 +119,7 @@ static unsigned paravirt_patch_jmp(void *insnbuf, const void *target, return 5; } +#endif DEFINE_STATIC_KEY_TRUE(virt_spin_lock_key); @@ -149,10 +151,12 @@ unsigned paravirt_patch_default(u8 type, void *insnbuf, else if (opfunc == _paravirt_ident_64) ret = paravirt_patch_ident_64(insnbuf, len); +#ifdef CONFIG_PARAVIRT_XXL else if (type == PARAVIRT_PATCH(pv_cpu_ops.iret) || type == PARAVIRT_PATCH(pv_cpu_ops.usergs_sysret64)) /* If operation requires a jmp, then jmp */ ret = paravirt_patch_jmp(insnbuf, opfunc, addr, len); +#endif else /* Otherwise call the function. */ ret = paravirt_patch_call(insnbuf, opfunc, addr, len); @@ -261,6 +265,7 @@ void paravirt_flush_lazy_mmu(void) preempt_enable(); } +#ifdef CONFIG_PARAVIRT_XXL void paravirt_start_context_switch(struct task_struct *prev) { BUG_ON(preemptible()); @@ -281,6 +286,7 @@ void paravirt_end_context_switch(struct task_struct *next) if (test_and_clear_ti_thread_flag(task_thread_info(next), TIF_LAZY_MMU_UPDATES)) arch_enter_lazy_mmu_mode(); } +#endif enum paravirt_lazy_mode paravirt_get_lazy_mode(void) { @@ -319,6 +325,9 @@ struct paravirt_patch_template pv_ops = { .pv_time_ops.steal_clock = native_steal_clock, /* Cpu ops. */ + .pv_cpu_ops.io_delay = native_io_delay, + +#ifdef CONFIG_PARAVIRT_XXL .pv_cpu_ops.cpuid = native_cpuid, .pv_cpu_ops.get_debugreg = native_get_debugreg, .pv_cpu_ops.set_debugreg = native_set_debugreg, @@ -360,10 +369,10 @@ struct paravirt_patch_template pv_ops = { .pv_cpu_ops.swapgs = native_swapgs, .pv_cpu_ops.set_iopl_mask = native_set_iopl_mask, - .pv_cpu_ops.io_delay = native_io_delay, .pv_cpu_ops.start_context_switch = paravirt_nop, .pv_cpu_ops.end_context_switch = paravirt_nop, +#endif /* Irq ops. */ .pv_irq_ops.save_fl = __PV_IS_CALLEE_SAVE(native_save_fl), @@ -462,10 +471,12 @@ struct paravirt_patch_template pv_ops = { #endif }; +#ifdef CONFIG_PARAVIRT_XXL /* At this point, native_get/set_debugreg has real function entries */ NOKPROBE_SYMBOL(native_get_debugreg); NOKPROBE_SYMBOL(native_set_debugreg); NOKPROBE_SYMBOL(native_load_idt); +#endif EXPORT_SYMBOL_GPL(pv_ops); EXPORT_SYMBOL_GPL(pv_info); diff --git a/arch/x86/kernel/paravirt_patch_32.c b/arch/x86/kernel/paravirt_patch_32.c index e5c3a438149e..704ed7718062 100644 --- a/arch/x86/kernel/paravirt_patch_32.c +++ b/arch/x86/kernel/paravirt_patch_32.c @@ -5,7 +5,9 @@ DEF_NATIVE(pv_irq_ops, irq_disable, "cli"); DEF_NATIVE(pv_irq_ops, irq_enable, "sti"); DEF_NATIVE(pv_irq_ops, restore_fl, "push %eax; popf"); DEF_NATIVE(pv_irq_ops, save_fl, "pushf; pop %eax"); +#ifdef CONFIG_PARAVIRT_XXL DEF_NATIVE(pv_cpu_ops, iret, "iret"); +#endif DEF_NATIVE(pv_mmu_ops, read_cr2, "mov %cr2, %eax"); DEF_NATIVE(pv_mmu_ops, write_cr3, "mov %eax, %cr3"); DEF_NATIVE(pv_mmu_ops, read_cr3, "mov %cr3, %eax"); @@ -45,7 +47,9 @@ unsigned native_patch(u8 type, void *ibuf, unsigned long addr, unsigned len) PATCH_SITE(pv_irq_ops, irq_enable); PATCH_SITE(pv_irq_ops, restore_fl); PATCH_SITE(pv_irq_ops, save_fl); +#ifdef CONFIG_PARAVIRT_XXL PATCH_SITE(pv_cpu_ops, iret); +#endif PATCH_SITE(pv_mmu_ops, read_cr2); PATCH_SITE(pv_mmu_ops, read_cr3); PATCH_SITE(pv_mmu_ops, write_cr3); diff --git a/arch/x86/kernel/paravirt_patch_64.c b/arch/x86/kernel/paravirt_patch_64.c index 893ef87eb268..131a7eb01a13 100644 --- a/arch/x86/kernel/paravirt_patch_64.c +++ b/arch/x86/kernel/paravirt_patch_64.c @@ -10,10 +10,12 @@ DEF_NATIVE(pv_irq_ops, save_fl, "pushfq; popq %rax"); DEF_NATIVE(pv_mmu_ops, read_cr2, "movq %cr2, %rax"); DEF_NATIVE(pv_mmu_ops, read_cr3, "movq %cr3, %rax"); DEF_NATIVE(pv_mmu_ops, write_cr3, "movq %rdi, %cr3"); +#ifdef CONFIG_PARAVIRT_XXL DEF_NATIVE(pv_cpu_ops, wbinvd, "wbinvd"); DEF_NATIVE(pv_cpu_ops, usergs_sysret64, "swapgs; sysretq"); DEF_NATIVE(pv_cpu_ops, swapgs, "swapgs"); +#endif DEF_NATIVE(, mov32, "mov %edi, %eax"); DEF_NATIVE(, mov64, "mov %rdi, %rax"); @@ -53,12 +55,14 @@ unsigned native_patch(u8 type, void *ibuf, unsigned long addr, unsigned len) PATCH_SITE(pv_irq_ops, save_fl); PATCH_SITE(pv_irq_ops, irq_enable); PATCH_SITE(pv_irq_ops, irq_disable); +#ifdef CONFIG_PARAVIRT_XXL PATCH_SITE(pv_cpu_ops, usergs_sysret64); PATCH_SITE(pv_cpu_ops, swapgs); + PATCH_SITE(pv_cpu_ops, wbinvd); +#endif PATCH_SITE(pv_mmu_ops, read_cr2); PATCH_SITE(pv_mmu_ops, read_cr3); PATCH_SITE(pv_mmu_ops, write_cr3); - PATCH_SITE(pv_cpu_ops, wbinvd); #if defined(CONFIG_PARAVIRT_SPINLOCKS) case PARAVIRT_PATCH(pv_lock_ops.queued_spin_unlock): if (pv_is_native_spin_unlock()) { -- 2.13.7
Juergen Gross
2018-Aug-10 11:52 UTC
[PATCH 09/10] x86/paravirt: move the Xen-only pv_irq_ops under the PARAVIRT_XXL umbrella
Some of the paravirt ops defined in pv_irq_ops are for Xen PV guests only. Define them only if CONFIG_PARAVIRT_XXL is set. Signed-off-by: Juergen Gross <jgross at suse.com> --- arch/x86/include/asm/irqflags.h | 38 ++++++++++++++++++----------------- arch/x86/include/asm/paravirt.h | 2 -- arch/x86/include/asm/paravirt_types.h | 2 ++ arch/x86/kernel/paravirt.c | 2 ++ 4 files changed, 24 insertions(+), 20 deletions(-) diff --git a/arch/x86/include/asm/irqflags.h b/arch/x86/include/asm/irqflags.h index 03bb451e4e6b..205e43e55144 100644 --- a/arch/x86/include/asm/irqflags.h +++ b/arch/x86/include/asm/irqflags.h @@ -88,24 +88,6 @@ static inline notrace void arch_local_irq_enable(void) } /* - * Used in the idle loop; sti takes one instruction cycle - * to complete: - */ -static inline __cpuidle void arch_safe_halt(void) -{ - native_safe_halt(); -} - -/* - * Used when interrupts are already enabled or to - * shutdown the processor: - */ -static inline __cpuidle void halt(void) -{ - native_halt(); -} - -/* * For spinlocks, etc: */ static inline notrace unsigned long arch_local_irq_save(void) @@ -154,6 +136,26 @@ static inline notrace unsigned long arch_local_irq_save(void) #define INTERRUPT_RETURN iret #endif +#else + +/* + * Used in the idle loop; sti takes one instruction cycle + * to complete: + */ +static inline __cpuidle void arch_safe_halt(void) +{ + native_safe_halt(); +} + +/* + * Used when interrupts are already enabled or to + * shutdown the processor: + */ +static inline __cpuidle void halt(void) +{ + native_halt(); +} + #endif /* __ASSEMBLY__ */ #endif /* CONFIG_PARAVIRT_XXL */ diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h index bc9a72a767c8..220c13d7e846 100644 --- a/arch/x86/include/asm/paravirt.h +++ b/arch/x86/include/asm/paravirt.h @@ -91,7 +91,6 @@ static inline void write_cr8(unsigned long x) PVOP_VCALL1(pv_cpu_ops.write_cr8, x); } #endif -#endif static inline void arch_safe_halt(void) { @@ -103,7 +102,6 @@ static inline void halt(void) PVOP_VCALL0(pv_irq_ops.halt); } -#ifdef CONFIG_PARAVIRT_XXL static inline void wbinvd(void) { PVOP_VCALL0(pv_cpu_ops.wbinvd); diff --git a/arch/x86/include/asm/paravirt_types.h b/arch/x86/include/asm/paravirt_types.h index be356aacc82c..938ac2bece81 100644 --- a/arch/x86/include/asm/paravirt_types.h +++ b/arch/x86/include/asm/paravirt_types.h @@ -197,8 +197,10 @@ struct pv_irq_ops { struct paravirt_callee_save irq_disable; struct paravirt_callee_save irq_enable; +#ifdef CONFIG_PARAVIRT_XXL void (*safe_halt)(void); void (*halt)(void); +#endif } __no_randomize_layout; diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c index 437be9454cab..19bfb3d2083f 100644 --- a/arch/x86/kernel/paravirt.c +++ b/arch/x86/kernel/paravirt.c @@ -379,8 +379,10 @@ struct paravirt_patch_template pv_ops = { .pv_irq_ops.restore_fl = __PV_IS_CALLEE_SAVE(native_restore_fl), .pv_irq_ops.irq_disable = __PV_IS_CALLEE_SAVE(native_irq_disable), .pv_irq_ops.irq_enable = __PV_IS_CALLEE_SAVE(native_irq_enable), +#ifdef CONFIG_PARAVIRT_XXL .pv_irq_ops.safe_halt = native_safe_halt, .pv_irq_ops.halt = native_halt, +#endif /* Mmu ops. */ .pv_mmu_ops.read_cr2 = native_read_cr2, -- 2.13.7
Juergen Gross
2018-Aug-10 11:52 UTC
[PATCH 10/10] x86/paravirt: move the Xen-only pv_mmu_ops under the PARAVIRT_XXL umbrella
Most of the paravirt ops defined in pv_mmu_ops are for Xen PV guests only. Define them only if CONFIG_PARAVIRT_XXL is set. Signed-off-by: Juergen Gross <jgross at suse.com> --- arch/x86/include/asm/fixmap.h | 2 +- arch/x86/include/asm/mmu_context.h | 4 +- arch/x86/include/asm/paravirt.h | 115 +++++++++++++++++----------------- arch/x86/include/asm/paravirt_types.h | 29 ++++----- arch/x86/include/asm/pgalloc.h | 2 +- arch/x86/include/asm/pgtable.h | 7 +-- arch/x86/include/asm/special_insns.h | 11 +--- arch/x86/kernel/asm-offsets.c | 2 +- arch/x86/kernel/head_64.S | 4 +- arch/x86/kernel/paravirt.c | 15 +++-- arch/x86/kernel/paravirt_patch_32.c | 4 +- arch/x86/kernel/paravirt_patch_64.c | 4 +- 12 files changed, 97 insertions(+), 102 deletions(-) diff --git a/arch/x86/include/asm/fixmap.h b/arch/x86/include/asm/fixmap.h index e203169931c7..ac80e7eadc3a 100644 --- a/arch/x86/include/asm/fixmap.h +++ b/arch/x86/include/asm/fixmap.h @@ -152,7 +152,7 @@ void __native_set_fixmap(enum fixed_addresses idx, pte_t pte); void native_set_fixmap(enum fixed_addresses idx, phys_addr_t phys, pgprot_t flags); -#ifndef CONFIG_PARAVIRT +#ifndef CONFIG_PARAVIRT_XXL static inline void __set_fixmap(enum fixed_addresses idx, phys_addr_t phys, pgprot_t flags) { diff --git a/arch/x86/include/asm/mmu_context.h b/arch/x86/include/asm/mmu_context.h index bbc796eb0a3b..ffae17a8db36 100644 --- a/arch/x86/include/asm/mmu_context.h +++ b/arch/x86/include/asm/mmu_context.h @@ -16,12 +16,12 @@ extern atomic64_t last_mm_ctx_id; -#ifndef CONFIG_PARAVIRT +#ifndef CONFIG_PARAVIRT_XXL static inline void paravirt_activate_mm(struct mm_struct *prev, struct mm_struct *next) { } -#endif /* !CONFIG_PARAVIRT */ +#endif /* !CONFIG_PARAVIRT_XXL */ #ifdef CONFIG_PERF_EVENTS diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h index 220c13d7e846..520c85b74c74 100644 --- a/arch/x86/include/asm/paravirt.h +++ b/arch/x86/include/asm/paravirt.h @@ -17,6 +17,57 @@ #include <linux/cpumask.h> #include <asm/frame.h> +static inline unsigned long long paravirt_sched_clock(void) +{ + return PVOP_CALL0(unsigned long long, pv_time_ops.sched_clock); +} + +struct static_key; +extern struct static_key paravirt_steal_enabled; +extern struct static_key paravirt_steal_rq_enabled; + +static inline u64 paravirt_steal_clock(int cpu) +{ + return PVOP_CALL1(u64, pv_time_ops.steal_clock, cpu); +} + +/* The paravirtualized I/O functions */ +static inline void slow_down_io(void) +{ + pv_ops.pv_cpu_ops.io_delay(); +#ifdef REALLY_SLOW_IO + pv_ops.pv_cpu_ops.io_delay(); + pv_ops.pv_cpu_ops.io_delay(); + pv_ops.pv_cpu_ops.io_delay(); +#endif +} + +static inline void __flush_tlb(void) +{ + PVOP_VCALL0(pv_mmu_ops.flush_tlb_user); +} + +static inline void __flush_tlb_global(void) +{ + PVOP_VCALL0(pv_mmu_ops.flush_tlb_kernel); +} + +static inline void __flush_tlb_one_user(unsigned long addr) +{ + PVOP_VCALL1(pv_mmu_ops.flush_tlb_one_user, addr); +} + +static inline void flush_tlb_others(const struct cpumask *cpumask, + const struct flush_tlb_info *info) +{ + PVOP_VCALL2(pv_mmu_ops.flush_tlb_others, cpumask, info); +} + +static inline void paravirt_arch_exit_mmap(struct mm_struct *mm) +{ + PVOP_VCALL1(pv_mmu_ops.exit_mmap, mm); +} + #ifdef CONFIG_PARAVIRT_XXL static inline void load_sp0(unsigned long sp0) { @@ -52,7 +103,6 @@ static inline void write_cr0(unsigned long x) { PVOP_VCALL1(pv_cpu_ops.write_cr0, x); } -#endif static inline unsigned long read_cr2(void) { @@ -74,7 +124,6 @@ static inline void write_cr3(unsigned long x) PVOP_VCALL1(pv_mmu_ops.write_cr3, x); } -#ifdef CONFIG_PARAVIRT_XXL static inline void __write_cr4(unsigned long x) { PVOP_VCALL1(pv_cpu_ops.write_cr4, x); @@ -172,23 +221,7 @@ static inline int rdmsrl_safe(unsigned msr, unsigned long long *p) *p = paravirt_read_msr_safe(msr, &err); return err; } -#endif -static inline unsigned long long paravirt_sched_clock(void) -{ - return PVOP_CALL0(unsigned long long, pv_time_ops.sched_clock); -} - -struct static_key; -extern struct static_key paravirt_steal_enabled; -extern struct static_key paravirt_steal_rq_enabled; - -static inline u64 paravirt_steal_clock(int cpu) -{ - return PVOP_CALL1(u64, pv_time_ops.steal_clock, cpu); -} - -#ifdef CONFIG_PARAVIRT_XXL static inline unsigned long long paravirt_read_pmc(int counter) { return PVOP_CALL1(u64, pv_cpu_ops.read_pmc, counter); @@ -267,18 +300,6 @@ static inline void set_iopl_mask(unsigned mask) { PVOP_VCALL1(pv_cpu_ops.set_iopl_mask, mask); } -#endif - -/* The paravirtualized I/O functions */ -static inline void slow_down_io(void) -{ - pv_ops.pv_cpu_ops.io_delay(); -#ifdef REALLY_SLOW_IO - pv_ops.pv_cpu_ops.io_delay(); - pv_ops.pv_cpu_ops.io_delay(); - pv_ops.pv_cpu_ops.io_delay(); -#endif -} static inline void paravirt_activate_mm(struct mm_struct *prev, struct mm_struct *next) @@ -292,30 +313,6 @@ static inline void paravirt_arch_dup_mmap(struct mm_struct *oldmm, PVOP_VCALL2(pv_mmu_ops.dup_mmap, oldmm, mm); } -static inline void paravirt_arch_exit_mmap(struct mm_struct *mm) -{ - PVOP_VCALL1(pv_mmu_ops.exit_mmap, mm); -} - -static inline void __flush_tlb(void) -{ - PVOP_VCALL0(pv_mmu_ops.flush_tlb_user); -} -static inline void __flush_tlb_global(void) -{ - PVOP_VCALL0(pv_mmu_ops.flush_tlb_kernel); -} -static inline void __flush_tlb_one_user(unsigned long addr) -{ - PVOP_VCALL1(pv_mmu_ops.flush_tlb_one_user, addr); -} - -static inline void flush_tlb_others(const struct cpumask *cpumask, - const struct flush_tlb_info *info) -{ - PVOP_VCALL2(pv_mmu_ops.flush_tlb_others, cpumask, info); -} - static inline int paravirt_pgd_alloc(struct mm_struct *mm) { return PVOP_CALL1(int, pv_mmu_ops.pgd_alloc, mm); @@ -640,7 +637,6 @@ static inline void pmd_clear(pmd_t *pmdp) } #endif /* CONFIG_X86_PAE */ -#ifdef CONFIG_PARAVIRT_XXL #define __HAVE_ARCH_START_CONTEXT_SWITCH static inline void arch_start_context_switch(struct task_struct *prev) { @@ -651,7 +647,6 @@ static inline void arch_end_context_switch(struct task_struct *next) { PVOP_VCALL1(pv_cpu_ops.end_context_switch, next); } -#endif #define __HAVE_ARCH_ENTER_LAZY_MMU_MODE static inline void arch_enter_lazy_mmu_mode(void) @@ -674,6 +669,7 @@ static inline void __set_fixmap(unsigned /* enum fixed_addresses */ idx, { pv_ops.pv_mmu_ops.set_fixmap(idx, phys, flags); } +#endif #if defined(CONFIG_SMP) && defined(CONFIG_PARAVIRT_SPINLOCKS) @@ -959,15 +955,20 @@ extern void default_banner(void); #endif /* __ASSEMBLY__ */ #else /* CONFIG_PARAVIRT */ # define default_banner x86_init_noop +#endif /* !CONFIG_PARAVIRT */ + #ifndef __ASSEMBLY__ +#ifndef CONFIG_PARAVIRT_XXL static inline void paravirt_arch_dup_mmap(struct mm_struct *oldmm, struct mm_struct *mm) { } +#endif +#ifndef CONFIG_PARAVIRT static inline void paravirt_arch_exit_mmap(struct mm_struct *mm) { } +#endif #endif /* __ASSEMBLY__ */ -#endif /* !CONFIG_PARAVIRT */ #endif /* _ASM_X86_PARAVIRT_H */ diff --git a/arch/x86/include/asm/paravirt_types.h b/arch/x86/include/asm/paravirt_types.h index 938ac2bece81..a9b72dae18b2 100644 --- a/arch/x86/include/asm/paravirt_types.h +++ b/arch/x86/include/asm/paravirt_types.h @@ -90,13 +90,14 @@ struct pv_init_ops { unsigned long addr, unsigned len); } __no_randomize_layout; - +#ifdef CONFIG_PARAVIRT_XXL struct pv_lazy_ops { /* Set deferred update mode, used for batching operations. */ void (*enter)(void); void (*leave)(void); void (*flush)(void); } __no_randomize_layout; +#endif struct pv_time_ops { unsigned long long (*sched_clock)(void); @@ -205,29 +206,28 @@ struct pv_irq_ops { } __no_randomize_layout; struct pv_mmu_ops { + /* TLB operations */ + void (*flush_tlb_user)(void); + void (*flush_tlb_kernel)(void); + void (*flush_tlb_one_user)(unsigned long addr); + void (*flush_tlb_others)(const struct cpumask *cpus, + const struct flush_tlb_info *info); + + /* Hook for intercepting the destruction of an mm_struct. */ + void (*exit_mmap)(struct mm_struct *mm); + +#ifdef CONFIG_PARAVIRT_XXL unsigned long (*read_cr2)(void); void (*write_cr2)(unsigned long); unsigned long (*read_cr3)(void); void (*write_cr3)(unsigned long); - /* - * Hooks for intercepting the creation/use/destruction of an - * mm_struct. - */ + /* Hooks for intercepting the creation/use of an mm_struct. */ void (*activate_mm)(struct mm_struct *prev, struct mm_struct *next); void (*dup_mmap)(struct mm_struct *oldmm, struct mm_struct *mm); - void (*exit_mmap)(struct mm_struct *mm); - - - /* TLB operations */ - void (*flush_tlb_user)(void); - void (*flush_tlb_kernel)(void); - void (*flush_tlb_one_user)(unsigned long addr); - void (*flush_tlb_others)(const struct cpumask *cpus, - const struct flush_tlb_info *info); /* Hooks for allocating and freeing a pagetable top-level */ int (*pgd_alloc)(struct mm_struct *mm); @@ -302,6 +302,7 @@ struct pv_mmu_ops { an mfn. We can tell which is which from the index. */ void (*set_fixmap)(unsigned /* enum fixed_addresses */ idx, phys_addr_t phys, pgprot_t flags); +#endif } __no_randomize_layout; struct arch_spinlock; diff --git a/arch/x86/include/asm/pgalloc.h b/arch/x86/include/asm/pgalloc.h index fbd578daa66e..ec7f43327033 100644 --- a/arch/x86/include/asm/pgalloc.h +++ b/arch/x86/include/asm/pgalloc.h @@ -8,7 +8,7 @@ static inline int __paravirt_pgd_alloc(struct mm_struct *mm) { return 0; } -#ifdef CONFIG_PARAVIRT +#ifdef CONFIG_PARAVIRT_XXL #include <asm/paravirt.h> #else #define paravirt_pgd_alloc(mm) __paravirt_pgd_alloc(mm) diff --git a/arch/x86/include/asm/pgtable.h b/arch/x86/include/asm/pgtable.h index 9ea291fe7107..b9abc525ece3 100644 --- a/arch/x86/include/asm/pgtable.h +++ b/arch/x86/include/asm/pgtable.h @@ -52,9 +52,9 @@ extern struct mm_struct *pgd_page_get_mm(struct page *page); extern pmdval_t early_pmd_flags; -#ifdef CONFIG_PARAVIRT +#ifdef CONFIG_PARAVIRT_XXL #include <asm/paravirt.h> -#else /* !CONFIG_PARAVIRT */ +#else /* !CONFIG_PARAVIRT_XXL */ #define set_pte(ptep, pte) native_set_pte(ptep, pte) #define set_pte_at(mm, addr, ptep, pte) native_set_pte_at(mm, addr, ptep, pte) @@ -108,9 +108,6 @@ extern pmdval_t early_pmd_flags; #define pte_val(x) native_pte_val(x) #define __pte(x) native_make_pte(x) -#endif /* CONFIG_PARAVIRT */ - -#ifndef CONFIG_PARAVIRT_XXL #define arch_end_context_switch(prev) do {} while(0) #endif /* CONFIG_PARAVIRT_XXL */ diff --git a/arch/x86/include/asm/special_insns.h b/arch/x86/include/asm/special_insns.h index 2aa6ce4bf159..43c029cdc3fe 100644 --- a/arch/x86/include/asm/special_insns.h +++ b/arch/x86/include/asm/special_insns.h @@ -141,11 +141,10 @@ static inline unsigned long __read_cr4(void) return native_read_cr4(); } -#ifdef CONFIG_PARAVIRT +#ifdef CONFIG_PARAVIRT_XXL #include <asm/paravirt.h> -#endif +#else -#ifndef CONFIG_PARAVIRT_XXL static inline unsigned long read_cr0(void) { return native_read_cr0(); @@ -155,9 +154,7 @@ static inline void write_cr0(unsigned long x) { native_write_cr0(x); } -#endif -#ifndef CONFIG_PARAVIRT static inline unsigned long read_cr2(void) { return native_read_cr2(); @@ -181,9 +178,7 @@ static inline void write_cr3(unsigned long x) { native_write_cr3(x); } -#endif -#ifndef CONFIG_PARAVIRT_XXL static inline void __write_cr4(unsigned long x) { native_write_cr4(x); @@ -213,7 +208,7 @@ static inline void load_gs_index(unsigned selector) #endif -#endif/* CONFIG_PARAVIRT_XXL */ +#endif /* CONFIG_PARAVIRT_XXL */ static inline void clflush(volatile void *__p) { diff --git a/arch/x86/kernel/asm-offsets.c b/arch/x86/kernel/asm-offsets.c index 429db3c8a0cc..1871bb388bb0 100644 --- a/arch/x86/kernel/asm-offsets.c +++ b/arch/x86/kernel/asm-offsets.c @@ -72,9 +72,9 @@ void common(void) { pv_irq_ops.irq_enable); #ifdef CONFIG_PARAVIRT_XXL OFFSET(PV_CPU_iret, paravirt_patch_template, pv_cpu_ops.iret); -#endif OFFSET(PV_MMU_read_cr2, paravirt_patch_template, pv_mmu_ops.read_cr2); #endif +#endif #ifdef CONFIG_XEN BLANK(); diff --git a/arch/x86/kernel/head_64.S b/arch/x86/kernel/head_64.S index e11b96b2dc6b..981fd802830f 100644 --- a/arch/x86/kernel/head_64.S +++ b/arch/x86/kernel/head_64.S @@ -25,14 +25,12 @@ #include <asm/export.h> #include <asm/nospec-branch.h> -#ifdef CONFIG_PARAVIRT +#ifdef CONFIG_PARAVIRT_XXL #include <asm/asm-offsets.h> #include <asm/paravirt.h> #define GET_CR2_INTO(reg) GET_CR2_INTO_RAX ; movq %rax, reg #else #define GET_CR2_INTO(reg) movq %cr2, reg -#endif -#ifndef CONFIG_PARAVIRT_XXL #define INTERRUPT_RETURN iretq #endif diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c index 19bfb3d2083f..36c60fbf61fd 100644 --- a/arch/x86/kernel/paravirt.c +++ b/arch/x86/kernel/paravirt.c @@ -385,16 +385,19 @@ struct paravirt_patch_template pv_ops = { #endif /* Mmu ops. */ - .pv_mmu_ops.read_cr2 = native_read_cr2, - .pv_mmu_ops.write_cr2 = native_write_cr2, - .pv_mmu_ops.read_cr3 = __native_read_cr3, - .pv_mmu_ops.write_cr3 = native_write_cr3, - .pv_mmu_ops.flush_tlb_user = native_flush_tlb, .pv_mmu_ops.flush_tlb_kernel = native_flush_tlb_global, .pv_mmu_ops.flush_tlb_one_user = native_flush_tlb_one_user, .pv_mmu_ops.flush_tlb_others = native_flush_tlb_others, + .pv_mmu_ops.exit_mmap = paravirt_nop, + +#ifdef CONFIG_PARAVIRT_XXL + .pv_mmu_ops.read_cr2 = native_read_cr2, + .pv_mmu_ops.write_cr2 = native_write_cr2, + .pv_mmu_ops.read_cr3 = __native_read_cr3, + .pv_mmu_ops.write_cr3 = native_write_cr3, + .pv_mmu_ops.pgd_alloc = __paravirt_pgd_alloc, .pv_mmu_ops.pgd_free = paravirt_nop, @@ -447,7 +450,6 @@ struct paravirt_patch_template pv_ops = { .pv_mmu_ops.make_pgd = PTE_IDENT, .pv_mmu_ops.dup_mmap = paravirt_nop, - .pv_mmu_ops.exit_mmap = paravirt_nop, .pv_mmu_ops.activate_mm = paravirt_nop, .pv_mmu_ops.lazy_mode = { @@ -457,6 +459,7 @@ struct paravirt_patch_template pv_ops = { }, .pv_mmu_ops.set_fixmap = native_set_fixmap, +#endif #if defined(CONFIG_PARAVIRT_SPINLOCKS) /* Lock ops. */ diff --git a/arch/x86/kernel/paravirt_patch_32.c b/arch/x86/kernel/paravirt_patch_32.c index 704ed7718062..a4e66a912121 100644 --- a/arch/x86/kernel/paravirt_patch_32.c +++ b/arch/x86/kernel/paravirt_patch_32.c @@ -7,10 +7,10 @@ DEF_NATIVE(pv_irq_ops, restore_fl, "push %eax; popf"); DEF_NATIVE(pv_irq_ops, save_fl, "pushf; pop %eax"); #ifdef CONFIG_PARAVIRT_XXL DEF_NATIVE(pv_cpu_ops, iret, "iret"); -#endif DEF_NATIVE(pv_mmu_ops, read_cr2, "mov %cr2, %eax"); DEF_NATIVE(pv_mmu_ops, write_cr3, "mov %eax, %cr3"); DEF_NATIVE(pv_mmu_ops, read_cr3, "mov %cr3, %eax"); +#endif #if defined(CONFIG_PARAVIRT_SPINLOCKS) DEF_NATIVE(pv_lock_ops, queued_spin_unlock, "movb $0, (%eax)"); @@ -49,10 +49,10 @@ unsigned native_patch(u8 type, void *ibuf, unsigned long addr, unsigned len) PATCH_SITE(pv_irq_ops, save_fl); #ifdef CONFIG_PARAVIRT_XXL PATCH_SITE(pv_cpu_ops, iret); -#endif PATCH_SITE(pv_mmu_ops, read_cr2); PATCH_SITE(pv_mmu_ops, read_cr3); PATCH_SITE(pv_mmu_ops, write_cr3); +#endif #if defined(CONFIG_PARAVIRT_SPINLOCKS) case PARAVIRT_PATCH(pv_lock_ops.queued_spin_unlock): if (pv_is_native_spin_unlock()) { diff --git a/arch/x86/kernel/paravirt_patch_64.c b/arch/x86/kernel/paravirt_patch_64.c index 131a7eb01a13..abc0c94b8caa 100644 --- a/arch/x86/kernel/paravirt_patch_64.c +++ b/arch/x86/kernel/paravirt_patch_64.c @@ -7,10 +7,10 @@ DEF_NATIVE(pv_irq_ops, irq_disable, "cli"); DEF_NATIVE(pv_irq_ops, irq_enable, "sti"); DEF_NATIVE(pv_irq_ops, restore_fl, "pushq %rdi; popfq"); DEF_NATIVE(pv_irq_ops, save_fl, "pushfq; popq %rax"); +#ifdef CONFIG_PARAVIRT_XXL DEF_NATIVE(pv_mmu_ops, read_cr2, "movq %cr2, %rax"); DEF_NATIVE(pv_mmu_ops, read_cr3, "movq %cr3, %rax"); DEF_NATIVE(pv_mmu_ops, write_cr3, "movq %rdi, %cr3"); -#ifdef CONFIG_PARAVIRT_XXL DEF_NATIVE(pv_cpu_ops, wbinvd, "wbinvd"); DEF_NATIVE(pv_cpu_ops, usergs_sysret64, "swapgs; sysretq"); @@ -59,10 +59,10 @@ unsigned native_patch(u8 type, void *ibuf, unsigned long addr, unsigned len) PATCH_SITE(pv_cpu_ops, usergs_sysret64); PATCH_SITE(pv_cpu_ops, swapgs); PATCH_SITE(pv_cpu_ops, wbinvd); -#endif PATCH_SITE(pv_mmu_ops, read_cr2); PATCH_SITE(pv_mmu_ops, read_cr3); PATCH_SITE(pv_mmu_ops, write_cr3); +#endif #if defined(CONFIG_PARAVIRT_SPINLOCKS) case PARAVIRT_PATCH(pv_lock_ops.queued_spin_unlock): if (pv_is_native_spin_unlock()) { -- 2.13.7
Jan Beulich
2018-Aug-10 12:06 UTC
[Xen-devel] [PATCH 04/10] x86/paravirt: use a single ops structure
>>> On 10.08.18 at 13:52, <jgross at suse.com> wrote: > --- a/arch/x86/hyperv/mmu.c > +++ b/arch/x86/hyperv/mmu.c > @@ -228,9 +228,9 @@ void hyperv_setup_mmu_ops(void) > > if (!(ms_hyperv.hints & HV_X64_EX_PROCESSOR_MASKS_RECOMMENDED)) { > pr_info("Using hypercall for remote TLB flush\n"); > - pv_mmu_ops.flush_tlb_others = hyperv_flush_tlb_others; > + pv_ops.pv_mmu_ops.flush_tlb_others = hyperv_flush_tlb_others;Taking just this as example, why not pv_ops.mmu.flush_tlb_others = hyperv_flush_tlb_others; ? Both pv_ and _ops are redundant on the field names. Jan
Juergen Gross
2018-Aug-10 12:30 UTC
[Xen-devel] [PATCH 04/10] x86/paravirt: use a single ops structure
On 10/08/18 14:06, Jan Beulich wrote:>>>> On 10.08.18 at 13:52, <jgross at suse.com> wrote: >> --- a/arch/x86/hyperv/mmu.c >> +++ b/arch/x86/hyperv/mmu.c >> @@ -228,9 +228,9 @@ void hyperv_setup_mmu_ops(void) >> >> if (!(ms_hyperv.hints & HV_X64_EX_PROCESSOR_MASKS_RECOMMENDED)) { >> pr_info("Using hypercall for remote TLB flush\n"); >> - pv_mmu_ops.flush_tlb_others = hyperv_flush_tlb_others; >> + pv_ops.pv_mmu_ops.flush_tlb_others = hyperv_flush_tlb_others; > > Taking just this as example, why not > > pv_ops.mmu.flush_tlb_others = hyperv_flush_tlb_others; > > ? Both pv_ and _ops are redundant on the field names.Good idea. Juergen
Boris Ostrovsky
2018-Aug-10 14:22 UTC
[PATCH 06/10] x86/paravirt: introduce new config option PARAVIRT_XXL
On 08/10/2018 07:52 AM, Juergen Gross wrote:> A large amount of paravirt ops is used by Xen PV guests only. Add a new > config option PARAVIRT_XXL which is selected by XEN_PV. Later we can > put the Xen PV only paravirt ops under the PARACVIRT_XXL umbrella.What does "XXL" stand for? My immediate thought was "extra extra large" but I suspect it's something else. -boris