search for: restore_all

This set of patches converts the way frames gets created from using PUSHes/POPs to using MOVes, thus allowing (in certain cases) to avoid saving/restoring part of the register set. While the place where the (small) win from this comes from varies between CPUs, the net effect is a 1 to 2% reduction on a combined interruption entry and exit when the full state save can be avoided. 1: use MOV

...hen I try to create a domU BUG: warning at kernel/lockdep.c:1814/trace_hardirqs_on() (Not tainted) [<c040566e>] show_trace_log_lvl+0x58/0x177 [<c0405c73>] show_trace+0xd/0x10 [<c0405cb1>] dump_stack+0x19/0x1b [<c043646e>] trace_hardirqs_on+0xa4/0x120 [<c0404e67>] restore_all+0x37/0x3a DWARF2 unwinder stuck at restore_all+0x37/0x3a Leftover inexact backtrace: Inexact backtrace: [<c0405c73>] show_trace+0xd/0x10 [<c0405cb1>] dump_stack+0x19/0x1b [<c043646e>] trace_hardirqs_on+0xa4/0x120 [<c0404e67>] restore_all+0x37/0x3a any help ? ---...

...hen I try to create a domU BUG: warning at kernel/lockdep.c:1814/trace_hardirqs_on() (Not tainted) [<c040566e>] show_trace_log_lvl+0x58/0x177 [<c0405c73>] show_trace+0xd/0x10 [<c0405cb1>] dump_stack+0x19/0x1b [<c043646e>] trace_hardirqs_on+0xa4/0x120 [<c0404e67>] restore_all+0x37/0x3a DWARF2 unwinder stuck at restore_all+0x37/0x3a Leftover inexact backtrace: Inexact backtrace: [<c0405c73>] show_trace+0xd/0x10 [<c0405cb1>] dump_stack+0x19/0x1b [<c043646e>] trace_hardirqs_on+0xa4/0x120 [<c0404e67>] restore_all+0x37/0x3a any help ? ---...

...MASK | 3), %eax - jz resume_kernel + andl $(VM_MASK | SEGMENT_RPL_MASK), %eax + cmpl $USER_RPL, %eax + jb resume_kernel # not returning to v8086 or userspace ENTRY(resume_userspace) cli # make sure we don't miss an interrupt # setting need_resched or sigpending @@ -367,8 +368,8 @@ restore_all: # See comments in process.c:copy_thread() for details. movb OLDSS(%esp), %ah movb CS(%esp), %al - andl $(VM_MASK | (4 << 8) | 3), %eax - cmpl $((4 << 8) | 3), %eax + andl $(VM_MASK | (4 << 8) | SEGMENT_RPL_MASK), %eax + cmpl $((4 << 8) | USER_RPL), %eax CFI_REMEMBER...

...MASK | 3), %eax - jz resume_kernel + andl $(VM_MASK | SEGMENT_RPL_MASK), %eax + cmpl $USER_RPL, %eax + jb resume_kernel # not returning to v8086 or userspace ENTRY(resume_userspace) cli # make sure we don't miss an interrupt # setting need_resched or sigpending @@ -367,8 +368,8 @@ restore_all: # See comments in process.c:copy_thread() for details. movb OLDSS(%esp), %ah movb CS(%esp), %al - andl $(VM_MASK | (4 << 8) | 3), %eax - cmpl $((4 << 8) | 3), %eax + andl $(VM_MASK | (4 << 8) | SEGMENT_RPL_MASK), %eax + cmpl $((4 << 8) | USER_RPL), %eax CFI_REMEMBER...

...-work/arch/i386/kernel/entry.S =================================================================== --- linux-2.6.14-zach-work.orig/arch/i386/kernel/entry.S 2005-10-27 17:02:08.000000000 -0700 +++ linux-2.6.14-zach-work/arch/i386/kernel/entry.S 2005-11-04 18:22:07.000000000 -0800 @@ -250,8 +250,8 @@ restore_all: # See comments in process.c:copy_thread() for details. movb OLDSS(%esp), %ah movb CS(%esp), %al - andl $(VM_MASK | (4 << 8) | 3), %eax - cmpl $((4 << 8) | 3), %eax + andl $(VM_MASK | (LDT_SEGMENT << 8) | 3), %eax + cmpl $((LDT_SEGMENT << 8) | 3), %eax je ldt_ss #...

..., forget to join lists. Regards, Jinsong -----Original Message----- From: Liu, Jinsong Sent: 2007年5月16日 8:23 To: ''Keir Fraser''; Tian, Kevin; xen-devel@lists.xensource.com Subject: RE: [Xen-devel] trap/interrupt gate for hypercall In native linux, syscall (gate type 15) and its RESTORE_ALL code provides weak fixup mechanism when reload segment register, by sending 0 selector to ds/es/gs, or kill the process when iret meets problem. Native linux can pay the price with this way since it doesn''t matter when a ring3 process meet problem (and the possibility of doing so is very l...

...-work/arch/i386/kernel/entry.S =================================================================== --- linux-2.6.14-zach-work.orig/arch/i386/kernel/entry.S 2005-10-27 17:02:08.000000000 -0700 +++ linux-2.6.14-zach-work/arch/i386/kernel/entry.S 2005-11-04 18:22:07.000000000 -0800 @@ -250,8 +250,8 @@ restore_all: # See comments in process.c:copy_thread() for details. movb OLDSS(%esp), %ah movb CS(%esp), %al - andl $(VM_MASK | (4 << 8) | 3), %eax - cmpl $((4 << 8) | 3), %eax + andl $(VM_MASK | (LDT_SEGMENT << 8) | 3), %eax + cmpl $((LDT_SEGMENT << 8) | 3), %eax je ldt_ss #...

...stack+0x48/0x60 [<c056f3d4>] warn_alloc_failed+0xd4/0x110 [<c0571ade>] __alloc_pages_nodemask+0x81e/0xc30 [<c04e4b65>] ? clockevents_program_event+0x45/0x150 [<c05b81c8>] new_slab+0x258/0x3a0 [<c05b9490>] __slab_alloc.constprop.55+0x5f0/0x790 [<c0b34636>] ? restore_all+0xf/0xf [<c0755622>] ? radix_tree_node_alloc+0x22/0x90 [<c04a9728>] ? __lock_is_held+0x48/0x70 [<c05bac55>] kmem_cache_alloc+0x295/0x3c0 [<c0755622>] ? radix_tree_node_alloc+0x22/0x90 [<c0755622>] radix_tree_node_alloc+0x22/0x90 [<c0755e7c>] __radix_tree_cr...

...FLAGS and CS movb CS(%esp), %al andl $(VM_MASK | SEGMENT_RPL_MASK), %eax - cmpl $SEGMENT_RPL_MASK, %eax + cmpl $USER_RPL, %eax jb resume_kernel # not returning to v8086 or userspace ENTRY(resume_userspace) DISABLE_INTERRUPTS # make sure we don't miss an interrupt @@ -374,8 +374,8 @@ restore_all: # See comments in process.c:copy_thread() for details. movb OLDSS(%esp), %ah movb CS(%esp), %al - andl $(VM_MASK | (4 << 8) | 3), %eax - cmpl $((4 << 8) | 3), %eax + andl $(VM_MASK | (SEGMENT_TI_MASK << 8) | SEGMENT_RPL_MASK), %eax + cmpl $((SEGMENT_LDT << 8) | USER_R...

...FLAGS and CS movb CS(%esp), %al andl $(VM_MASK | SEGMENT_RPL_MASK), %eax - cmpl $SEGMENT_RPL_MASK, %eax + cmpl $USER_RPL, %eax jb resume_kernel # not returning to v8086 or userspace ENTRY(resume_userspace) DISABLE_INTERRUPTS # make sure we don't miss an interrupt @@ -374,8 +374,8 @@ restore_all: # See comments in process.c:copy_thread() for details. movb OLDSS(%esp), %ah movb CS(%esp), %al - andl $(VM_MASK | (4 << 8) | 3), %eax - cmpl $((4 << 8) | 3), %eax + andl $(VM_MASK | (SEGMENT_TI_MASK << 8) | SEGMENT_RPL_MASK), %eax + cmpl $((SEGMENT_LDT << 8) | USER_R...

...eax,EAX(%esp) # store the return value syscall_exit: - cli # make sure we don't miss an interrupt + CLI # make sure we don't miss an interrupt # setting need_resched or sigpending # between sampling and the iret #ifdef CONFIG_TRAP_BAD_SYSCALL_EXITS @@ -274,10 +273,10 @@ restore_all: restore_nocheck: RESTORE_REGS addl $4, %esp -1: iret +1: IRET .section .fixup,"ax" iret_exc: - sti + STI pushl $0 # no error code pushl $do_iret_error jmp error_code @@ -299,14 +298,14 @@ ldt_ss: * CPUs, which we can try to work around to make * dosemu and wine happy...

...eax,EAX(%esp) # store the return value syscall_exit: - cli # make sure we don't miss an interrupt + CLI # make sure we don't miss an interrupt # setting need_resched or sigpending # between sampling and the iret #ifdef CONFIG_TRAP_BAD_SYSCALL_EXITS @@ -274,10 +273,10 @@ restore_all: restore_nocheck: RESTORE_REGS addl $4, %esp -1: iret +1: IRET .section .fixup,"ax" iret_exc: - sti + STI pushl $0 # no error code pushl $do_iret_error jmp error_code @@ -299,14 +298,14 @@ ldt_ss: * CPUs, which we can try to work around to make * dosemu and wine happy...

The self_nmi() code cause''s an NMI to be triggered by sending an APIC message to the local processor. However, NMIs are blocked by the VMEXIT, until the next iret or VMENTER. Volume 3 Chapter 27 Section 1 of the Intel SDM states: An NMI causes subsequent NMIs to be blocked, but only after the VM exit completes. As a result, as soon as the VMENTER happens, an immediate VMEXIT happens

...) movl %eax,EAX(%esp) # store the return value syscall_exit: - cli # make sure we don't miss an interrupt + CLI # make sure we don't miss an interrupt # setting need_resched or sigpending # between sampling and the iret movl TI_flags(%ebp), %ecx @@ -256,14 +256,14 @@ restore_all: restore_nocheck: RESTORE_REGS addl $4, %esp -1: iret -.section .fixup,"ax" +1: IRET +.pushsection .fixup,"ax" iret_exc: - sti + STI pushl $0 # no error code pushl $do_iret_error jmp error_code -.previous +.popsection .section __ex_table,"a" .align 4...

...) movl %eax,EAX(%esp) # store the return value syscall_exit: - cli # make sure we don't miss an interrupt + CLI # make sure we don't miss an interrupt # setting need_resched or sigpending # between sampling and the iret movl TI_flags(%ebp), %ecx @@ -256,14 +256,14 @@ restore_all: restore_nocheck: RESTORE_REGS addl $4, %esp -1: iret -.section .fixup,"ax" +1: IRET +.pushsection .fixup,"ax" iret_exc: - sti + STI pushl $0 # no error code pushl $do_iret_error jmp error_code -.previous +.popsection .section __ex_table,"a" .align 4...

...t;, \ - "jmp .Lsyscall_32_done", X86_FEATURE_XENPV + #define JMP_IF_IRET testl %eax, %eax; jz .Lsyscall_32_done + ALTERNATIVE(JMP_IF_IRET, jmp .Lsyscall_32_done, X86_FEATURE_XENPV) /* Opportunistic SYSEXIT */ TRACE_IRQS_ON /* User mode traces as IRQs on. */ @@ -536,7 +536,7 @@ restore_all: TRACE_IRQS_IRET .Lrestore_all_notrace: #ifdef CONFIG_X86_ESPFIX32 - ALTERNATIVE "jmp .Lrestore_nocheck", "", X86_BUG_ESPFIX + ALTERNATIVE(jmp .Lrestore_nocheck, , X86_BUG_ESPFIX) movl PT_EFLAGS(%esp), %eax # mix EFLAGS, SS and CS /* @@ -692,9 +692,9 @@ ENTRY(simd_cop...

...t addq $EXCEPTION_STKSZ, per_cpu__init_tss + TSS_ist + (\ist - 1) * 8(%rbp) .endif - cli + DISABLE_INTERRUPTS(CLBR_NONE) .if \irqtrace TRACE_IRQS_OFF .endif @@ -767,10 +784,10 @@ paranoid_swapgs\trace: .if \trace TRACE_IRQS_IRETQ 0 .endif - swapgs + SWAPGS paranoid_restore\trace: RESTORE_ALL 8 - iretq + INTERRUPT_RETURN paranoid_userspace\trace: GET_THREAD_INFO(%rcx) movl threadinfo_flags(%rcx),%ebx @@ -785,11 +802,11 @@ paranoid_userspace\trace: .if \trace TRACE_IRQS_ON .endif - sti + ENABLE_INTERRUPTS(CLBR_NONE) xorl %esi,%esi /* arg2: oldset */ movq %rsp,%rdi /*...

...t addq $EXCEPTION_STKSZ, per_cpu__init_tss + TSS_ist + (\ist - 1) * 8(%rbp) .endif - cli + DISABLE_INTERRUPTS(CLBR_NONE) .if \irqtrace TRACE_IRQS_OFF .endif @@ -767,10 +784,10 @@ paranoid_swapgs\trace: .if \trace TRACE_IRQS_IRETQ 0 .endif - swapgs + SWAPGS paranoid_restore\trace: RESTORE_ALL 8 - iretq + INTERRUPT_RETURN paranoid_userspace\trace: GET_THREAD_INFO(%rcx) movl threadinfo_flags(%rcx),%ebx @@ -785,11 +802,11 @@ paranoid_userspace\trace: .if \trace TRACE_IRQS_ON .endif - sti + ENABLE_INTERRUPTS(CLBR_NONE) xorl %esi,%esi /* arg2: oldset */ movq %rsp,%rdi /*...

...@@ -738,7 +733,7 @@ END(spurious_interrupt) .if \ist addq $EXCEPTION_STKSZ, per_cpu__init_tss + TSS_ist + (\ist - 1) * 8(%rbp) .endif - cli + DISABLE_INTERRUPTS(CLBR_NONE) .if \irqtrace TRACE_IRQS_OFF .endif @@ -770,7 +765,7 @@ paranoid_swapgs\trace: swapgs paranoid_restore\trace: RESTORE_ALL 8 - iretq + INTERRUPT_RETURN paranoid_userspace\trace: GET_THREAD_INFO(%rcx) movl threadinfo_flags(%rcx),%ebx @@ -785,11 +780,11 @@ paranoid_userspace\trace: .if \trace TRACE_IRQS_ON .endif - sti + ENABLE_INTERRUPTS(CLBR_NONE) xorl %esi,%esi /* arg2: oldset */ movq %rsp,%rdi /*...