X-Git-Url: http://git.onelab.eu/?a=blobdiff_plain;f=arch%2Fx86_64%2Fkernel%2Fvsyscall.c;h=0bc61763d23fa2a5f1124439f63485300b4a5d6f;hb=refs%2Fheads%2Fvserver;hp=2e5734425949648ea92bd8d7fd9e65e6f99a8ed5;hpb=f7f1b0f1e2fbadeab12d24236000e778aa9b1ead;p=linux-2.6.git diff --git a/arch/x86_64/kernel/vsyscall.c b/arch/x86_64/kernel/vsyscall.c index 2e5734425..0bc61763d 100644 --- a/arch/x86_64/kernel/vsyscall.c +++ b/arch/x86_64/kernel/vsyscall.c @@ -26,6 +26,10 @@ #include #include #include +#include +#include +#include +#include #include #include @@ -33,16 +37,20 @@ #include #include #include +#include +#include +#include #define __vsyscall(nr) __attribute__ ((unused,__section__(".vsyscall_" #nr))) -#define force_inline __attribute__((always_inline)) inline +#define __syscall_clobber "r11","rcx","memory" int __sysctl_vsyscall __section_sysctl_vsyscall = 1; seqlock_t __xtime_lock __section_xtime_lock = SEQLOCK_UNLOCKED; +int __vgetcpu_mode __section_vgetcpu_mode; #include -static force_inline void timeval_normalize(struct timeval * tv) +static __always_inline void timeval_normalize(struct timeval * tv) { time_t __sec; @@ -53,7 +61,7 @@ static force_inline void timeval_normalize(struct timeval * tv) } } -static force_inline void do_vgettimeofday(struct timeval * tv) +static __always_inline void do_vgettimeofday(struct timeval * tv) { long sequence, t; unsigned long sec, usec; @@ -62,19 +70,18 @@ static force_inline void do_vgettimeofday(struct timeval * tv) sequence = read_seqbegin(&__xtime_lock); sec = __xtime.tv_sec; - usec = (__xtime.tv_nsec / 1000) + - (__jiffies - __wall_jiffies) * (1000000 / HZ); + usec = __xtime.tv_nsec / 1000; if (__vxtime.mode != VXTIME_HPET) { - sync_core(); - rdtscll(t); + t = get_cycles_sync(); if (t < __vxtime.last_tsc) t = __vxtime.last_tsc; usec += ((t - __vxtime.last_tsc) * __vxtime.tsc_quot) >> 32; /* See comment in x86_64 do_gettimeofday. */ } else { - usec += ((readl((void *)fix_to_virt(VSYSCALL_HPET) + 0xf0) - + usec += ((readl((void __iomem *) + fix_to_virt(VSYSCALL_HPET) + 0xf0) - __vxtime.last) * __vxtime.quot) >> 32; } } while (read_seqretry(&__xtime_lock, sequence)); @@ -84,12 +91,12 @@ static force_inline void do_vgettimeofday(struct timeval * tv) } /* RED-PEN may want to readd seq locking, but then the variable should be write-once. */ -static force_inline void do_get_tz(struct timezone * tz) +static __always_inline void do_get_tz(struct timezone * tz) { *tz = __sys_tz; } -static force_inline int gettimeofday(struct timeval *tv, struct timezone *tz) +static __always_inline int gettimeofday(struct timeval *tv, struct timezone *tz) { int ret; asm volatile("vsysc2: syscall" @@ -98,7 +105,7 @@ static force_inline int gettimeofday(struct timeval *tv, struct timezone *tz) return ret; } -static force_inline long time_syscall(long *t) +static __always_inline long time_syscall(long *t) { long secs; asm volatile("vsysc1: syscall" @@ -107,9 +114,9 @@ static force_inline long time_syscall(long *t) return secs; } -static int __vsyscall(0) vgettimeofday(struct timeval * tv, struct timezone * tz) +int __vsyscall(0) vgettimeofday(struct timeval * tv, struct timezone * tz) { - if (unlikely(!__sysctl_vsyscall)) + if (!__sysctl_vsyscall) return gettimeofday(tv,tz); if (tv) do_vgettimeofday(tv); @@ -120,21 +127,58 @@ static int __vsyscall(0) vgettimeofday(struct timeval * tv, struct timezone * tz /* This will break when the xtime seconds get inaccurate, but that is * unlikely */ -static time_t __vsyscall(1) vtime(time_t *t) +time_t __vsyscall(1) vtime(time_t *t) { - if (unlikely(!__sysctl_vsyscall)) + if (!__sysctl_vsyscall) return time_syscall(t); else if (t) *t = __xtime.tv_sec; return __xtime.tv_sec; } -static long __vsyscall(2) venosys_0(void) +/* Fast way to get current CPU and node. + This helps to do per node and per CPU caches in user space. + The result is not guaranteed without CPU affinity, but usually + works out because the scheduler tries to keep a thread on the same + CPU. + + tcache must point to a two element sized long array. + All arguments can be NULL. */ +long __vsyscall(2) +vgetcpu(unsigned *cpu, unsigned *node, struct getcpu_cache *tcache) { - return -ENOSYS; + unsigned int dummy, p; + unsigned long j = 0; + + /* Fast cache - only recompute value once per jiffies and avoid + relatively costly rdtscp/cpuid otherwise. + This works because the scheduler usually keeps the process + on the same CPU and this syscall doesn't guarantee its + results anyways. + We do this here because otherwise user space would do it on + its own in a likely inferior way (no access to jiffies). + If you don't like it pass NULL. */ + if (tcache && tcache->blob[0] == (j = __jiffies)) { + p = tcache->blob[1]; + } else if (__vgetcpu_mode == VGETCPU_RDTSCP) { + /* Load per CPU data from RDTSCP */ + rdtscp(dummy, dummy, p); + } else { + /* Load per CPU data from GDT */ + asm("lsl %1,%0" : "=r" (p) : "r" (__PER_CPU_SEG)); + } + if (tcache) { + tcache->blob[0] = j; + tcache->blob[1] = p; + } + if (cpu) + *cpu = p & 0xfff; + if (node) + *node = p >> 12; + return 0; } -static long __vsyscall(3) venosys_1(void) +long __vsyscall(3) venosys_1(void) { return -ENOSYS; } @@ -151,7 +195,8 @@ static int vsyscall_sysctl_change(ctl_table *ctl, int write, struct file * filp, void __user *buffer, size_t *lenp, loff_t *ppos) { extern u16 vsysc1, vsysc2; - u16 *map1, *map2; + u16 __iomem *map1; + u16 __iomem *map2; int ret = proc_dointvec(ctl, write, filp, buffer, lenp, ppos); if (!write) return ret; @@ -166,11 +211,11 @@ static int vsyscall_sysctl_change(ctl_table *ctl, int write, struct file * filp, goto out; } if (!sysctl_vsyscall) { - *map1 = SYSCALL; - *map2 = SYSCALL; + writew(SYSCALL, map1); + writew(SYSCALL, map2); } else { - *map1 = NOP2; - *map2 = NOP2; + writew(NOP2, map1); + writew(NOP2, map2); } iounmap(map2); out: @@ -180,8 +225,7 @@ out: static int vsyscall_sysctl_nostrat(ctl_table *t, int __user *name, int nlen, void __user *oldval, size_t __user *oldlenp, - void __user *newval, size_t newlen, - void **context) + void __user *newval, size_t newlen) { return -ENOSYS; } @@ -202,23 +246,83 @@ static ctl_table kernel_root_table2[] = { #endif +#ifndef CONFIG_XEN +/* Assume __initcall executes before all user space. Hopefully kmod + doesn't violate that. We'll find out if it does. */ +static void __cpuinit vsyscall_set_cpu(int cpu) +{ + unsigned long *d; + unsigned long node = 0; +#ifdef CONFIG_NUMA + node = cpu_to_node[cpu]; +#endif + if (cpu_has(&cpu_data[cpu], X86_FEATURE_RDTSCP)) + write_rdtscp_aux((node << 12) | cpu); + + /* Store cpu number in limit so that it can be loaded quickly + in user space in vgetcpu. + 12 bits for the CPU and 8 bits for the node. */ + d = (unsigned long *)(cpu_gdt(cpu) + GDT_ENTRY_PER_CPU); + *d = 0x0f40000000000ULL; + *d |= cpu; + *d |= (node & 0xf) << 12; + *d |= (node >> 4) << 48; +} + +static void __cpuinit cpu_vsyscall_init(void *arg) +{ + /* preemption should be already off */ + vsyscall_set_cpu(raw_smp_processor_id()); +} + +static int __cpuinit +cpu_vsyscall_notifier(struct notifier_block *n, unsigned long action, void *arg) +{ + long cpu = (long)arg; + if (action == CPU_ONLINE) + smp_call_function_single(cpu, cpu_vsyscall_init, NULL, 0, 1); + return NOTIFY_DONE; +} +#endif + static void __init map_vsyscall(void) { extern char __vsyscall_0; unsigned long physaddr_page0 = __pa_symbol(&__vsyscall_0); + /* Note that VSYSCALL_MAPPED_PAGES must agree with the code below. */ __set_fixmap(VSYSCALL_FIRST_PAGE, physaddr_page0, PAGE_KERNEL_VSYSCALL); } +#ifdef CONFIG_XEN +static void __init map_vsyscall_user(void) +{ + extern void __set_fixmap_user(enum fixed_addresses, unsigned long, pgprot_t); + extern char __vsyscall_0; + unsigned long physaddr_page0 = __pa_symbol(&__vsyscall_0); + + __set_fixmap_user(VSYSCALL_FIRST_PAGE, physaddr_page0, PAGE_KERNEL_VSYSCALL); +} +#endif + static int __init vsyscall_init(void) { BUG_ON(((unsigned long) &vgettimeofday != VSYSCALL_ADDR(__NR_vgettimeofday))); BUG_ON((unsigned long) &vtime != VSYSCALL_ADDR(__NR_vtime)); BUG_ON((VSYSCALL_ADDR(0) != __fix_to_virt(VSYSCALL_FIRST_PAGE))); + BUG_ON((unsigned long) &vgetcpu != VSYSCALL_ADDR(__NR_vgetcpu)); map_vsyscall(); +#ifdef CONFIG_XEN + map_vsyscall_user(); + sysctl_vsyscall = 0; /* disable vgettimeofay() */ +#endif #ifdef CONFIG_SYSCTL register_sysctl_table(kernel_root_table2, 0); +#endif +#ifndef CONFIG_XEN + on_each_cpu(cpu_vsyscall_init, NULL, 0, 1); + hotcpu_notifier(cpu_vsyscall_notifier, 0); #endif return 0; }