From 4421011120b2304e5c248ae4165a2704588aedf1 Mon Sep 17 00:00:00 2001 From: Roland McGrath Date: Wed, 30 Jan 2008 13:31:50 +0100 Subject: x86: x86 i387 user_regset This revamps the i387 code to be shared across 32-bit, 64-bit, and 32-on-64. It does so by consolidating the code in one place based on the user_regset accessor interfaces. This switches 32-bit to using the i387_64.h header and 64-bit to using the i387.c that was previously i387_32.c, but that's what took the least cleanup in each file. Here i387.h is stubbed to always include i387_64.h rather than renaming the file, to keep this diff smaller and easier to read. Signed-off-by: Roland McGrath Signed-off-by: Ingo Molnar Signed-off-by: Thomas Gleixner --- arch/x86/kernel/i387.c | 481 +++++++++++++++++++++++++------------------------ 1 file changed, 248 insertions(+), 233 deletions(-) (limited to 'arch/x86/kernel/i387.c') diff --git a/arch/x86/kernel/i387.c b/arch/x86/kernel/i387.c index bebe03463461..f7f7568dd7bc 100644 --- a/arch/x86/kernel/i387.c +++ b/arch/x86/kernel/i387.c @@ -8,6 +8,7 @@ #include #include +#include #include #include #include @@ -16,13 +17,29 @@ #include #include +#ifdef CONFIG_X86_64 + +#include +#include + +#else + +#define save_i387_ia32 save_i387 +#define restore_i387_ia32 restore_i387 + +#define _fpstate_ia32 _fpstate +#define user_i387_ia32_struct user_i387_struct +#define user32_fxsr_struct user_fxsr_struct + +#endif + #ifdef CONFIG_MATH_EMULATION #define HAVE_HWFP (boot_cpu_data.hard_math) #else #define HAVE_HWFP 1 #endif -static unsigned long mxcsr_feature_mask __read_mostly = 0xffffffff; +unsigned int mxcsr_feature_mask __read_mostly = 0xffffffffu; void mxcsr_feature_mask_init(void) { @@ -40,6 +57,30 @@ void mxcsr_feature_mask_init(void) stts(); } +#ifdef CONFIG_X86_64 +/* + * Called at bootup to set up the initial FPU state that is later cloned + * into all processes. + */ +void __cpuinit fpu_init(void) +{ + unsigned long oldcr0 = read_cr0(); + extern void __bad_fxsave_alignment(void); + + if (offsetof(struct task_struct, thread.i387.fxsave) & 15) + __bad_fxsave_alignment(); + set_in_cr4(X86_CR4_OSFXSR); + set_in_cr4(X86_CR4_OSXMMEXCPT); + + write_cr0(oldcr0 & ~((1UL<<3)|(1UL<<2))); /* clear TS and EM */ + + mxcsr_feature_mask_init(); + /* clean state in init */ + current_thread_info()->status = 0; + clear_used_math(); +} +#endif /* CONFIG_X86_64 */ + /* * The _current_ task is using the FPU for the first time * so initialize it and set the mxcsr to its default @@ -48,12 +89,18 @@ void mxcsr_feature_mask_init(void) */ void init_fpu(struct task_struct *tsk) { + if (tsk_used_math(tsk)) { + if (tsk == current) + unlazy_fpu(tsk); + return; + } + if (cpu_has_fxsr) { memset(&tsk->thread.i387.fxsave, 0, sizeof(struct i387_fxsave_struct)); tsk->thread.i387.fxsave.cwd = 0x37f; if (cpu_has_xmm) - tsk->thread.i387.fxsave.mxcsr = 0x1f80; + tsk->thread.i387.fxsave.mxcsr = MXCSR_DEFAULT; } else { memset(&tsk->thread.i387.fsave, 0, sizeof(struct i387_fsave_struct)); @@ -62,27 +109,59 @@ void init_fpu(struct task_struct *tsk) tsk->thread.i387.fsave.twd = 0xffffffffu; tsk->thread.i387.fsave.fos = 0xffff0000u; } - /* only the device not available exception - * or ptrace can call init_fpu */ + /* + * Only the device not available exception or ptrace can call init_fpu. + */ set_stopped_child_used_math(tsk); } -/* - * FPU lazy state save handling. - */ +int fpregs_active(struct task_struct *target, const struct user_regset *regset) +{ + return tsk_used_math(target) ? regset->n : 0; +} -void kernel_fpu_begin(void) +int xfpregs_active(struct task_struct *target, const struct user_regset *regset) { - struct thread_info *thread = current_thread_info(); + return (cpu_has_fxsr && tsk_used_math(target)) ? regset->n : 0; +} - preempt_disable(); - if (thread->status & TS_USEDFPU) { - __save_init_fpu(thread->task); - return; - } - clts(); +int xfpregs_get(struct task_struct *target, const struct user_regset *regset, + unsigned int pos, unsigned int count, + void *kbuf, void __user *ubuf) +{ + if (!cpu_has_fxsr) + return -ENODEV; + + unlazy_fpu(target); + + return user_regset_copyout(&pos, &count, &kbuf, &ubuf, + &target->thread.i387.fxsave, 0, -1); } -EXPORT_SYMBOL_GPL(kernel_fpu_begin); + +int xfpregs_set(struct task_struct *target, const struct user_regset *regset, + unsigned int pos, unsigned int count, + const void *kbuf, const void __user *ubuf) +{ + int ret; + + if (!cpu_has_fxsr) + return -ENODEV; + + unlazy_fpu(target); + set_stopped_child_used_math(target); + + ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf, + &target->thread.i387.fxsave, 0, -1); + + /* + * mxcsr reserved bits must be masked to zero for security reasons. + */ + target->thread.i387.fxsave.mxcsr &= mxcsr_feature_mask; + + return ret; +} + +#if defined CONFIG_X86_32 || defined CONFIG_IA32_EMULATION /* * FPU tag word conversions. @@ -94,210 +173,187 @@ static inline unsigned short twd_i387_to_fxsr(unsigned short twd) /* Transform each pair of bits into 01 (valid) or 00 (empty) */ tmp = ~twd; - tmp = (tmp | (tmp >> 1)) & 0x5555; /* 0V0V0V0V0V0V0V0V */ + tmp = (tmp | (tmp>>1)) & 0x5555; /* 0V0V0V0V0V0V0V0V */ /* and move the valid bits to the lower byte. */ tmp = (tmp | (tmp >> 1)) & 0x3333; /* 00VV00VV00VV00VV */ tmp = (tmp | (tmp >> 2)) & 0x0f0f; /* 0000VVVV0000VVVV */ tmp = (tmp | (tmp >> 4)) & 0x00ff; /* 00000000VVVVVVVV */ - return tmp; } -static inline unsigned long twd_fxsr_to_i387(struct i387_fxsave_struct *fxsave) -{ - struct _fpxreg *st = NULL; - unsigned long tos = (fxsave->swd >> 11) & 7; - unsigned long twd = (unsigned long) fxsave->twd; - unsigned long tag; - unsigned long ret = 0xffff0000u; - int i; - #define FPREG_ADDR(f, n) ((void *)&(f)->st_space + (n) * 16); +#define FP_EXP_TAG_VALID 0 +#define FP_EXP_TAG_ZERO 1 +#define FP_EXP_TAG_SPECIAL 2 +#define FP_EXP_TAG_EMPTY 3 + +static inline u32 twd_fxsr_to_i387(struct i387_fxsave_struct *fxsave) +{ + struct _fpxreg *st; + u32 tos = (fxsave->swd >> 11) & 7; + u32 twd = (unsigned long) fxsave->twd; + u32 tag; + u32 ret = 0xffff0000u; + int i; - for (i = 0; i < 8; i++) { + for (i = 0; i < 8; i++, twd >>= 1) { if (twd & 0x1) { st = FPREG_ADDR(fxsave, (i - tos) & 7); switch (st->exponent & 0x7fff) { case 0x7fff: - tag = 2; /* Special */ + tag = FP_EXP_TAG_SPECIAL; break; case 0x0000: if (!st->significand[0] && !st->significand[1] && !st->significand[2] && - !st->significand[3]) { - tag = 1; /* Zero */ - } else { - tag = 2; /* Special */ - } + !st->significand[3]) + tag = FP_EXP_TAG_ZERO; + else + tag = FP_EXP_TAG_SPECIAL; break; default: - if (st->significand[3] & 0x8000) { - tag = 0; /* Valid */ - } else { - tag = 2; /* Special */ - } + if (st->significand[3] & 0x8000) + tag = FP_EXP_TAG_VALID; + else + tag = FP_EXP_TAG_SPECIAL; break; } } else { - tag = 3; /* Empty */ + tag = FP_EXP_TAG_EMPTY; } - ret |= (tag << (2 * i)); - twd = twd >> 1; + ret |= tag << (2 * i); } return ret; } /* - * FPU state interaction. + * FXSR floating point environment conversions. */ -unsigned short get_fpu_cwd(struct task_struct *tsk) +static void convert_from_fxsr(struct user_i387_ia32_struct *env, + struct task_struct *tsk) { - if (cpu_has_fxsr) { - return tsk->thread.i387.fxsave.cwd; - } else { - return (unsigned short)tsk->thread.i387.fsave.cwd; - } -} + struct i387_fxsave_struct *fxsave = &tsk->thread.i387.fxsave; + struct _fpreg *to = (struct _fpreg *) &env->st_space[0]; + struct _fpxreg *from = (struct _fpxreg *) &fxsave->st_space[0]; + int i; -unsigned short get_fpu_swd(struct task_struct *tsk) -{ - if (cpu_has_fxsr) { - return tsk->thread.i387.fxsave.swd; + env->cwd = fxsave->cwd | 0xffff0000u; + env->swd = fxsave->swd | 0xffff0000u; + env->twd = twd_fxsr_to_i387(fxsave); + +#ifdef CONFIG_X86_64 + env->fip = fxsave->rip; + env->foo = fxsave->rdp; + if (tsk == current) { + /* + * should be actually ds/cs at fpu exception time, but + * that information is not available in 64bit mode. + */ + asm("mov %%ds,%0" : "=r" (env->fos)); + asm("mov %%cs,%0" : "=r" (env->fcs)); } else { - return (unsigned short)tsk->thread.i387.fsave.swd; + struct pt_regs *regs = task_pt_regs(tsk); + env->fos = 0xffff0000 | tsk->thread.ds; + env->fcs = regs->cs; } -} - -#if 0 -unsigned short get_fpu_twd(struct task_struct *tsk) -{ - if (cpu_has_fxsr) { - return tsk->thread.i387.fxsave.twd; - } else { - return (unsigned short)tsk->thread.i387.fsave.twd; - } -} -#endif /* 0 */ +#else + env->fip = fxsave->fip; + env->fcs = fxsave->fcs; + env->foo = fxsave->foo; + env->fos = fxsave->fos; +#endif -unsigned short get_fpu_mxcsr(struct task_struct *tsk) -{ - if (cpu_has_xmm) { - return tsk->thread.i387.fxsave.mxcsr; - } else { - return 0x1f80; - } + for (i = 0; i < 8; ++i) + memcpy(&to[i], &from[i], sizeof(to[0])); } -#if 0 +static void convert_to_fxsr(struct task_struct *tsk, + const struct user_i387_ia32_struct *env) -void set_fpu_cwd(struct task_struct *tsk, unsigned short cwd) { - if (cpu_has_fxsr) { - tsk->thread.i387.fxsave.cwd = cwd; - } else { - tsk->thread.i387.fsave.cwd = ((long)cwd | 0xffff0000u); - } -} + struct i387_fxsave_struct *fxsave = &tsk->thread.i387.fxsave; + struct _fpreg *from = (struct _fpreg *) &env->st_space[0]; + struct _fpxreg *to = (struct _fpxreg *) &fxsave->st_space[0]; + int i; -void set_fpu_swd(struct task_struct *tsk, unsigned short swd) -{ - if (cpu_has_fxsr) { - tsk->thread.i387.fxsave.swd = swd; - } else { - tsk->thread.i387.fsave.swd = ((long)swd | 0xffff0000u); - } -} + fxsave->cwd = env->cwd; + fxsave->swd = env->swd; + fxsave->twd = twd_i387_to_fxsr(env->twd); + fxsave->fop = (u16) ((u32) env->fcs >> 16); +#ifdef CONFIG_X86_64 + fxsave->rip = env->fip; + fxsave->rdp = env->foo; + /* cs and ds ignored */ +#else + fxsave->fip = env->fip; + fxsave->fcs = (env->fcs & 0xffff); + fxsave->foo = env->foo; + fxsave->fos = env->fos; +#endif -void set_fpu_twd(struct task_struct *tsk, unsigned short twd) -{ - if (cpu_has_fxsr) { - tsk->thread.i387.fxsave.twd = twd_i387_to_fxsr(twd); - } else { - tsk->thread.i387.fsave.twd = ((long)twd | 0xffff0000u); - } + for (i = 0; i < 8; ++i) + memcpy(&to[i], &from[i], sizeof(from[0])); } -#endif /* 0 */ - -/* - * FXSR floating point environment conversions. - */ - -static int convert_fxsr_to_user(struct _fpstate __user *buf, - struct i387_fxsave_struct *fxsave) +int fpregs_get(struct task_struct *target, const struct user_regset *regset, + unsigned int pos, unsigned int count, + void *kbuf, void __user *ubuf) { - unsigned long env[7]; - struct _fpreg __user *to; - struct _fpxreg *from; - int i; + struct user_i387_ia32_struct env; - env[0] = (unsigned long)fxsave->cwd | 0xffff0000ul; - env[1] = (unsigned long)fxsave->swd | 0xffff0000ul; - env[2] = twd_fxsr_to_i387(fxsave); - env[3] = fxsave->fip; - env[4] = fxsave->fcs | ((unsigned long)fxsave->fop << 16); - env[5] = fxsave->foo; - env[6] = fxsave->fos; + if (!HAVE_HWFP) + return fpregs_soft_get(target, regset, pos, count, kbuf, ubuf); - if (__copy_to_user(buf, env, 7 * sizeof(unsigned long))) - return 1; + unlazy_fpu(target); - to = &buf->_st[0]; - from = (struct _fpxreg *) &fxsave->st_space[0]; - for (i = 0; i < 8; i++, to++, from++) { - unsigned long __user *t = (unsigned long __user *)to; - unsigned long *f = (unsigned long *)from; + if (!cpu_has_fxsr) + return user_regset_copyout(&pos, &count, &kbuf, &ubuf, + &target->thread.i387.fsave, 0, -1); - if (__put_user(*f, t) || - __put_user(*(f + 1), t + 1) || - __put_user(from->exponent, &to->exponent)) - return 1; + if (kbuf && pos == 0 && count == sizeof(env)) { + convert_from_fxsr(kbuf, target); + return 0; } - return 0; + + convert_from_fxsr(&env, target); + return user_regset_copyout(&pos, &count, &kbuf, &ubuf, &env, 0, -1); } -static int convert_fxsr_from_user(struct i387_fxsave_struct *fxsave, - struct _fpstate __user *buf) +int fpregs_set(struct task_struct *target, const struct user_regset *regset, + unsigned int pos, unsigned int count, + const void *kbuf, const void __user *ubuf) { - unsigned long env[7]; - struct _fpxreg *to; - struct _fpreg __user *from; - int i; + struct user_i387_ia32_struct env; + int ret; - if (__copy_from_user(env, buf, 7 * sizeof(long))) - return 1; + if (!HAVE_HWFP) + return fpregs_soft_set(target, regset, pos, count, kbuf, ubuf); - fxsave->cwd = (unsigned short)(env[0] & 0xffff); - fxsave->swd = (unsigned short)(env[1] & 0xffff); - fxsave->twd = twd_i387_to_fxsr((unsigned short)(env[2] & 0xffff)); - fxsave->fip = env[3]; - fxsave->fop = (unsigned short)((env[4] & 0xffff0000ul) >> 16); - fxsave->fcs = (env[4] & 0xffff); - fxsave->foo = env[5]; - fxsave->fos = env[6]; - - to = (struct _fpxreg *) &fxsave->st_space[0]; - from = &buf->_st[0]; - for (i = 0; i < 8; i++, to++, from++) { - unsigned long *t = (unsigned long *)to; - unsigned long __user *f = (unsigned long __user *)from; - - if (__get_user(*t, f) || - __get_user(*(t + 1), f + 1) || - __get_user(to->exponent, &from->exponent)) - return 1; - } - return 0; + unlazy_fpu(target); + set_stopped_child_used_math(target); + + if (!cpu_has_fxsr) + return user_regset_copyin(&pos, &count, &kbuf, &ubuf, + &target->thread.i387.fsave, 0, -1); + + if (pos > 0 || count < sizeof(env)) + convert_from_fxsr(&env, target); + + ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf, &env, 0, -1); + if (!ret) + convert_to_fxsr(target, &env); + + return ret; } /* * Signal frame handlers. */ -static inline int save_i387_fsave(struct _fpstate __user *buf) +static inline int save_i387_fsave(struct _fpstate_ia32 __user *buf) { struct task_struct *tsk = current; @@ -309,14 +365,16 @@ static inline int save_i387_fsave(struct _fpstate __user *buf) return 1; } -static int save_i387_fxsave(struct _fpstate __user *buf) +static int save_i387_fxsave(struct _fpstate_ia32 __user *buf) { struct task_struct *tsk = current; + struct user_i387_ia32_struct env; int err = 0; unlazy_fpu(tsk); - if (convert_fxsr_to_user(buf, &tsk->thread.i387.fxsave)) + convert_from_fxsr(&env, tsk); + if (__copy_to_user(buf, &env, sizeof(env))) return -1; err |= __put_user(tsk->thread.i387.fxsave.swd, &buf->status); @@ -330,7 +388,7 @@ static int save_i387_fxsave(struct _fpstate __user *buf) return 1; } -int save_i387(struct _fpstate __user *buf) +int save_i387_ia32(struct _fpstate_ia32 __user *buf) { if (!used_math()) return 0; @@ -347,11 +405,13 @@ int save_i387(struct _fpstate __user *buf) return save_i387_fsave(buf); } } else { - return save_i387_soft(¤t->thread.i387.soft, buf); + return fpregs_soft_get(current, NULL, + 0, sizeof(struct user_i387_ia32_struct), + NULL, buf) ? -1 : 1; } } -static inline int restore_i387_fsave(struct _fpstate __user *buf) +static inline int restore_i387_fsave(struct _fpstate_ia32 __user *buf) { struct task_struct *tsk = current; clear_fpu(tsk); @@ -359,19 +419,23 @@ static inline int restore_i387_fsave(struct _fpstate __user *buf) sizeof(struct i387_fsave_struct)); } -static int restore_i387_fxsave(struct _fpstate __user *buf) +static int restore_i387_fxsave(struct _fpstate_ia32 __user *buf) { int err; struct task_struct *tsk = current; + struct user_i387_ia32_struct env; clear_fpu(tsk); err = __copy_from_user(&tsk->thread.i387.fxsave, &buf->_fxsr_env[0], sizeof(struct i387_fxsave_struct)); /* mxcsr reserved bits must be masked to zero for security reasons */ tsk->thread.i387.fxsave.mxcsr &= mxcsr_feature_mask; - return err ? 1 : convert_fxsr_from_user(&tsk->thread.i387.fxsave, buf); + if (err || __copy_from_user(&env, buf, sizeof(env))) + return 1; + convert_to_fxsr(tsk, &env); + return 0; } -int restore_i387(struct _fpstate __user *buf) +int restore_i387_ia32(struct _fpstate_ia32 __user *buf) { int err; @@ -382,101 +446,52 @@ int restore_i387(struct _fpstate __user *buf) err = restore_i387_fsave(buf); } } else { - err = restore_i387_soft(¤t->thread.i387.soft, buf); + err = fpregs_soft_set(current, NULL, + 0, sizeof(struct user_i387_ia32_struct), + NULL, buf) != 0; } set_used_math(); return err; } -/* - * ptrace request handlers. - */ +#endif /* CONFIG_X86_32 || CONFIG_IA32_EMULATION */ -static inline int get_fpregs_fsave(struct user_i387_struct __user *buf, - struct task_struct *tsk) -{ - return __copy_to_user(buf, &tsk->thread.i387.fsave, - sizeof(struct user_i387_struct)); -} - -static inline int get_fpregs_fxsave(struct user_i387_struct __user *buf, - struct task_struct *tsk) -{ - return convert_fxsr_to_user((struct _fpstate __user *)buf, - &tsk->thread.i387.fxsave); -} +#ifdef CONFIG_X86_64 int get_fpregs(struct user_i387_struct __user *buf, struct task_struct *tsk) { - if (HAVE_HWFP) { - if (cpu_has_fxsr) { - return get_fpregs_fxsave(buf, tsk); - } else { - return get_fpregs_fsave(buf, tsk); - } - } else { - return save_i387_soft(&tsk->thread.i387.soft, - (struct _fpstate __user *)buf); - } + return xfpregs_get(tsk, NULL, 0, sizeof(*buf), NULL, buf); } -static inline int set_fpregs_fsave(struct task_struct *tsk, - struct user_i387_struct __user *buf) +int set_fpregs(struct task_struct *tsk, struct user_i387_struct __user *buf) { - return __copy_from_user(&tsk->thread.i387.fsave, buf, - sizeof(struct user_i387_struct)); + return xfpregs_set(tsk, NULL, 0, sizeof(*buf), NULL, buf); } -static inline int set_fpregs_fxsave(struct task_struct *tsk, - struct user_i387_struct __user *buf) +#else + +int get_fpregs(struct user_i387_struct __user *buf, struct task_struct *tsk) { - return convert_fxsr_from_user(&tsk->thread.i387.fxsave, - (struct _fpstate __user *)buf); + return fpregs_get(tsk, NULL, 0, sizeof(*buf), NULL, buf); } int set_fpregs(struct task_struct *tsk, struct user_i387_struct __user *buf) { - if (HAVE_HWFP) { - if (cpu_has_fxsr) { - return set_fpregs_fxsave(tsk, buf); - } else { - return set_fpregs_fsave(tsk, buf); - } - } else { - return restore_i387_soft(&tsk->thread.i387.soft, - (struct _fpstate __user *)buf); - } + return fpregs_set(tsk, NULL, 0, sizeof(*buf), NULL, buf); } int get_fpxregs(struct user_fxsr_struct __user *buf, struct task_struct *tsk) { - if (cpu_has_fxsr) { - if (__copy_to_user(buf, &tsk->thread.i387.fxsave, - sizeof(struct user_fxsr_struct))) - return -EFAULT; - return 0; - } else { - return -EIO; - } + return xfpregs_get(tsk, NULL, 0, sizeof(*buf), NULL, buf); } int set_fpxregs(struct task_struct *tsk, struct user_fxsr_struct __user *buf) { - int ret = 0; - - if (cpu_has_fxsr) { - if (__copy_from_user(&tsk->thread.i387.fxsave, buf, - sizeof(struct user_fxsr_struct))) - ret = -EFAULT; - /* mxcsr reserved bits must be masked to zero - * for security reasons */ - tsk->thread.i387.fxsave.mxcsr &= mxcsr_feature_mask; - } else { - ret = -EIO; - } - return ret; + return xfpregs_get(tsk, NULL, 0, sizeof(*buf), NULL, buf); } +#endif + /* * FPU state for core dumps. */ @@ -538,7 +553,7 @@ int dump_task_fpu(struct task_struct *tsk, struct user_i387_struct *fpu) } int dump_task_extended_fpu(struct task_struct *tsk, - struct user_fxsr_struct *fpu) + struct user32_fxsr_struct *fpu) { int fpvalid = tsk_used_math(tsk) && cpu_has_fxsr; -- cgit v1.2.3