Skip to content

L
linux
Commit 72a671ce authored by Suresh Siddha's avatar Suresh Siddha Committed by H. Peter Anvin
1 Browse files
Options

x86, fpu: Unify signal handling code paths for x86 and x86_64 kernels 

Currently for x86 and x86_32 binaries, fpstate in the user sigframe is copied
to/from the fpstate in the task struct.

And in the case of signal delivery for x86_64 binaries, if the fpstate is live
in the CPU registers, then the live state is copied directly to the user
sigframe. Otherwise  fpstate in the task struct is copied to the user sigframe.
During restore, fpstate in the user sigframe is restored directly to the live
CPU registers.

Historically, different code paths led to different bugs. For example,
x86_64 code path was not preemption safe till recently. Also there is lot
of code duplication for support of new features like xsave etc.

Unify signal handling code paths for x86 and x86_64 kernels.

New strategy is as follows:

Signal delivery: Both for 32/64-bit frames, align the core math frame area to
64bytes as needed by xsave (this where the main fpu/extended state gets copied
to and excludes the legacy compatibility fsave header for the 32-bit [f]xsave
frames). If the state is live, copy the register state directly to the user
frame. If not live, copy the state in the thread struct to the user frame. And
for 32-bit [f]xsave frames, construct the fsave header separately before
the actual [f]xsave area.

Signal return: As the 32-bit frames with [f]xstate has an additional
'fsave' header, copy everything back from the user sigframe to the
fpstate in the task structure and reconstruct the fxstate from the 'fsave'
header (Also user passed pointers may not be correctly aligned for
any attempt to directly restore any partial state). At the next fpstate usage,
everything will be restored to the live CPU registers.
For all the 64-bit frames and the 32-bit fsave frame, restore the state from
the user sigframe directly to the live CPU registers. 64-bit signals always
restored the math frame directly, so we can expect the math frame pointer
to be correctly aligned. For 32-bit fsave frames, there are no alignment
requirements, so we can restore the state directly.

"lat_sig catch" microbenchmark numbers (for x86, x86_64, x86_32 binaries) are
with in the noise range with this change.
Signed-off-by: default avatarSuresh Siddha <suresh.b.siddha@intel.com>
Link: http://lkml.kernel.org/r/1343171129-2747-4-git-send-email-suresh.b.siddha@intel.com
[ Merged in compilation fix ]
Link: http://lkml.kernel.org/r/1344544736.8326.17.camel@sbsiddha-desk.sc.intel.comSigned-off-by: default avatarH. Peter Anvin <hpa@linux.intel.com>
parent 0ca5bd0d
Expand all Hide whitespace changes
Inline Side-by-side
Showing with 348 additions and 484 deletions
arch/x86/ia32/ia32_signal.c
View file @ 72a671ce
......@@ -251,7 +251,7 @@ static int ia32_restore_sigcontext(struct pt_regs *regs,
get_user_ex(tmp, &sc->fpstate);
buf = compat_ptr(tmp);
err |= restore_i387_xstate_ia32(buf);
err |= restore_xstate_sig(buf, 1);
get_user_ex(*pax, &sc->ax);
} get_user_catch(err);
......@@ -382,9 +382,12 @@ static void __user *get_sigframe(struct k_sigaction *ka, struct pt_regs *regs,
sp = (unsigned long) ka->sa.sa_restorer;
if (used_math()) {
sp = sp - sig_xstate_ia32_size;
unsigned long fx_aligned, math_size;
sp = alloc_mathframe(sp, 1, &fx_aligned, &math_size);
*fpstate = (struct _fpstate_ia32 __user *) sp;
if (save_i387_xstate_ia32(*fpstate) < 0)
if (save_xstate_sig(*fpstate, (void __user *)fx_aligned,
math_size) < 0)
return (void __user *) -1L;
}
......
arch/x86/include/asm/fpu-internal.h
View file @ 72a671ce
......@@ -22,11 +22,30 @@
#include <asm/uaccess.h>
#include <asm/xsave.h>
extern unsigned int sig_xstate_size;
#ifdef CONFIG_X86_64
# include <asm/sigcontext32.h>
# include <asm/user32.h>
int ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
compat_sigset_t *set, struct pt_regs *regs);
int ia32_setup_frame(int sig, struct k_sigaction *ka,
compat_sigset_t *set, struct pt_regs *regs);
#else
# define user_i387_ia32_struct user_i387_struct
# define user32_fxsr_struct user_fxsr_struct
# define ia32_setup_frame __setup_frame
# define ia32_setup_rt_frame __setup_rt_frame
#endif
extern unsigned int mxcsr_feature_mask;
extern void fpu_init(void);
DECLARE_PER_CPU(struct task_struct *, fpu_owner_task);
extern void convert_from_fxsr(struct user_i387_ia32_struct *env,
struct task_struct *tsk);
extern void convert_to_fxsr(struct task_struct *tsk,
const struct user_i387_ia32_struct *env);
extern user_regset_active_fn fpregs_active, xfpregs_active;
extern user_regset_get_fn fpregs_get, xfpregs_get, fpregs_soft_get,
xstateregs_get;
......@@ -39,19 +58,11 @@ extern user_regset_set_fn fpregs_set, xfpregs_set, fpregs_soft_set,
*/
#define xstateregs_active fpregs_active
extern struct _fpx_sw_bytes fx_sw_reserved;
#ifdef CONFIG_IA32_EMULATION
extern unsigned int sig_xstate_ia32_size;
extern struct _fpx_sw_bytes fx_sw_reserved_ia32;
struct _fpstate_ia32;
struct _xstate_ia32;
extern int save_i387_xstate_ia32(void __user *buf);
extern int restore_i387_xstate_ia32(void __user *buf);
#endif
#ifdef CONFIG_MATH_EMULATION
# define HAVE_HWFP (boot_cpu_data.hard_math)
extern void finit_soft_fpu(struct i387_soft_struct *soft);
#else
# define HAVE_HWFP 1
static inline void finit_soft_fpu(struct i387_soft_struct *soft) {}
#endif
......@@ -119,17 +130,6 @@ static inline int fsave_user(struct i387_fsave_struct __user *fx)
static inline int fxsave_user(struct i387_fxsave_struct __user *fx)
{
int err;
/*
* Clear the bytes not touched by the fxsave and reserved
* for the SW usage.
*/
err = __clear_user(&fx->sw_reserved,
sizeof(struct _fpx_sw_bytes));
if (unlikely(err))
return -EFAULT;
if (config_enabled(CONFIG_X86_32))
return check_insn(fxsave %[fx], [fx] "=m" (*fx), "m" (*fx));
else if (config_enabled(CONFIG_AS_FXSAVEQ))
......@@ -189,19 +189,6 @@ static inline void fpu_fxsave(struct fpu *fpu)
: [fx] "R" (&fpu->state->fxsave));
}
}
#ifdef CONFIG_X86_64
int ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
compat_sigset_t *set, struct pt_regs *regs);
int ia32_setup_frame(int sig, struct k_sigaction *ka,
compat_sigset_t *set, struct pt_regs *regs);
#else /* CONFIG_X86_32 */
#define ia32_setup_frame __setup_frame
#define ia32_setup_rt_frame __setup_rt_frame
#endif /* CONFIG_X86_64 */
/*
* These must be called with preempt disabled. Returns
......@@ -392,10 +379,28 @@ static inline void switch_fpu_finish(struct task_struct *new, fpu_switch_t fpu)
/*
* Signal frame handlers...
*/
extern int save_i387_xstate(void __user *buf);
extern int restore_i387_xstate(void __user *buf);
extern int save_xstate_sig(void __user *buf, void __user *fx, int size);
extern int __restore_xstate_sig(void __user *buf, void __user *fx, int size);
static inline int xstate_sigframe_size(void)
{
return use_xsave() ? xstate_size + FP_XSTATE_MAGIC2_SIZE : xstate_size;
}
static inline int restore_xstate_sig(void __user *buf, int ia32_frame)
{
void __user *buf_fx = buf;
int size = xstate_sigframe_size();
if (ia32_frame && use_fxsr()) {
buf_fx = buf + sizeof(struct i387_fsave_struct);
size += sizeof(struct i387_fsave_struct);
}
return __restore_xstate_sig(buf, buf_fx, size);
}
static inline void __clear_fpu(struct task_struct *tsk)
static inline void __drop_fpu(struct task_struct *tsk)
{
if (__thread_has_fpu(tsk)) {
/* Ignore delayed exceptions from user space */
......@@ -443,11 +448,21 @@ static inline void save_init_fpu(struct task_struct *tsk)
preempt_enable();
}
static inline void clear_fpu(struct task_struct *tsk)
static inline void stop_fpu_preload(struct task_struct *tsk)
{
tsk->fpu_counter = 0;
}
static inline void drop_fpu(struct task_struct *tsk)
{
/*
* Forget coprocessor state..
*/
stop_fpu_preload(tsk);
preempt_disable();
__clear_fpu(tsk);
__drop_fpu(tsk);
preempt_enable();
clear_used_math();
}
/*
......@@ -511,4 +526,20 @@ static inline void fpu_copy(struct fpu *dst, struct fpu *src)
extern void fpu_finit(struct fpu *fpu);
static inline unsigned long
alloc_mathframe(unsigned long sp, int ia32_frame, unsigned long *buf_fx,
unsigned long *size)
{
unsigned long frame_size = xstate_sigframe_size();
*buf_fx = sp = round_down(sp - frame_size, 64);
if (ia32_frame && use_fxsr()) {
frame_size += sizeof(struct i387_fsave_struct);
sp -= sizeof(struct i387_fsave_struct);
}
*size = frame_size;
return sp;
}
#endif
arch/x86/include/asm/xsave.h
View file @ 72a671ce
......@@ -38,9 +38,6 @@ extern u64 xstate_fx_sw_bytes[USER_XSTATE_FX_SW_WORDS];
extern void xsave_init(void);
extern void update_regset_xstate_info(unsigned int size, u64 xstate_mask);
extern int init_fpu(struct task_struct *child);
extern int check_for_xstate(struct i387_fxsave_struct __user *buf,
void __user *fpstate,
struct _fpx_sw_bytes *sw);
static inline int fpu_xrstor_checking(struct xsave_struct *fx)
{
......@@ -68,8 +65,7 @@ static inline int xsave_user(struct xsave_struct __user *buf)
* Clear the xsave header first, so that reserved fields are
* initialized to zero.
*/
err = __clear_user(&buf->xsave_hdr,
sizeof(struct xsave_hdr_struct));
err = __clear_user(&buf->xsave_hdr, sizeof(buf->xsave_hdr));
if (unlikely(err))
return -EFAULT;
......
arch/x86/kernel/i387.c
View file @ 72a671ce
......@@ -19,20 +19,6 @@
#include <asm/fpu-internal.h>
#include <asm/user.h>
#ifdef CONFIG_X86_64
# include <asm/sigcontext32.h>
# include <asm/user32.h>
#else
# define save_i387_xstate_ia32 save_i387_xstate
# define restore_i387_xstate_ia32 restore_i387_xstate
# define _fpstate_ia32 _fpstate
# define _xstate_ia32 _xstate
# define sig_xstate_ia32_size sig_xstate_size
# define fx_sw_reserved_ia32 fx_sw_reserved
# define user_i387_ia32_struct user_i387_struct
# define user32_fxsr_struct user_fxsr_struct
#endif
/*
* Were we in an interrupt that interrupted kernel mode?
*
......@@ -113,16 +99,9 @@ void unlazy_fpu(struct task_struct *tsk)
}
EXPORT_SYMBOL(unlazy_fpu);
#ifdef CONFIG_MATH_EMULATION
# define HAVE_HWFP (boot_cpu_data.hard_math)
#else
# define HAVE_HWFP 1
#endif
static unsigned int mxcsr_feature_mask __read_mostly = 0xffffffffu;
unsigned int mxcsr_feature_mask __read_mostly = 0xffffffffu;
unsigned int xstate_size;
EXPORT_SYMBOL_GPL(xstate_size);
unsigned int sig_xstate_ia32_size = sizeof(struct _fpstate_ia32);
static struct i387_fxsave_struct fx_scratch __cpuinitdata;
static void __cpuinit mxcsr_feature_mask_init(void)
......@@ -454,7 +433,7 @@ static inline u32 twd_fxsr_to_i387(struct i387_fxsave_struct *fxsave)
* FXSR floating point environment conversions.
*/
static void
void
convert_from_fxsr(struct user_i387_ia32_struct *env, struct task_struct *tsk)
{
struct i387_fxsave_struct *fxsave = &tsk->thread.fpu.state->fxsave;
......@@ -491,8 +470,8 @@ convert_from_fxsr(struct user_i387_ia32_struct *env, struct task_struct *tsk)
memcpy(&to[i], &from[i], sizeof(to[0]));
}
static void convert_to_fxsr(struct task_struct *tsk,
const struct user_i387_ia32_struct *env)
void convert_to_fxsr(struct task_struct *tsk,
const struct user_i387_ia32_struct *env)
{
struct i387_fxsave_struct *fxsave = &tsk->thread.fpu.state->fxsave;
......@@ -588,223 +567,6 @@ int fpregs_set(struct task_struct *target, const struct user_regset *regset,
return ret;
}
/*
* Signal frame handlers.
*/
static inline int save_i387_fsave(struct _fpstate_ia32 __user *buf)
{
struct task_struct *tsk = current;
struct i387_fsave_struct *fp = &tsk->thread.fpu.state->fsave;
fp->status = fp->swd;
if (__copy_to_user(buf, fp, sizeof(struct i387_fsave_struct)))
return -1;
return 1;
}
static int save_i387_fxsave(struct _fpstate_ia32 __user *buf)
{
struct task_struct *tsk = current;
struct i387_fxsave_struct *fx = &tsk->thread.fpu.state->fxsave;
struct user_i387_ia32_struct env;
int err = 0;
convert_from_fxsr(&env, tsk);
if (__copy_to_user(buf, &env, sizeof(env)))
return -1;
err |= __put_user(fx->swd, &buf->status);
err |= __put_user(X86_FXSR_MAGIC, &buf->magic);
if (err)
return -1;
if (__copy_to_user(&buf->_fxsr_env[0], fx, xstate_size))
return -1;
return 1;
}
static int save_i387_xsave(void __user *buf)
{
struct task_struct *tsk = current;
struct _fpstate_ia32 __user *fx = buf;
int err = 0;
sanitize_i387_state(tsk);
/*
* For legacy compatible, we always set FP/SSE bits in the bit
* vector while saving the state to the user context.
* This will enable us capturing any changes(during sigreturn) to
* the FP/SSE bits by the legacy applications which don't touch
* xstate_bv in the xsave header.
*
* xsave aware applications can change the xstate_bv in the xsave
* header as well as change any contents in the memory layout.
* xrestore as part of sigreturn will capture all the changes.
*/
tsk->thread.fpu.state->xsave.xsave_hdr.xstate_bv |= XSTATE_FPSSE;
if (save_i387_fxsave(fx) < 0)
return -1;
err = __copy_to_user(&fx->sw_reserved, &fx_sw_reserved_ia32,
sizeof(struct _fpx_sw_bytes));
err |= __put_user(FP_XSTATE_MAGIC2,
(__u32 __user *) (buf + sig_xstate_ia32_size
- FP_XSTATE_MAGIC2_SIZE));
if (err)
return -1;
return 1;
}
int save_i387_xstate_ia32(void __user *buf)
{
struct _fpstate_ia32 __user *fp = (struct _fpstate_ia32 __user *) buf;
struct task_struct *tsk = current;
if (!used_math())
return 0;
if (!access_ok(VERIFY_WRITE, buf, sig_xstate_ia32_size))
return -EACCES;
/*
* This will cause a "finit" to be triggered by the next
* attempted FPU operation by the 'current' process.
*/
clear_used_math();
if (!HAVE_HWFP) {
return fpregs_soft_get(current, NULL,
0, sizeof(struct user_i387_ia32_struct),
NULL, fp) ? -1 : 1;
}
unlazy_fpu(tsk);
if (cpu_has_xsave)
return save_i387_xsave(fp);
if (cpu_has_fxsr)
return save_i387_fxsave(fp);
else
return save_i387_fsave(fp);
}
static inline int restore_i387_fsave(struct _fpstate_ia32 __user *buf)
{
struct task_struct *tsk = current;
return __copy_from_user(&tsk->thread.fpu.state->fsave, buf,
sizeof(struct i387_fsave_struct));
}
static int restore_i387_fxsave(struct _fpstate_ia32 __user *buf,
unsigned int size)
{
struct task_struct *tsk = current;
struct user_i387_ia32_struct env;
int err;
err = __copy_from_user(&tsk->thread.fpu.state->fxsave, &buf->_fxsr_env[0],
size);
/* mxcsr reserved bits must be masked to zero for security reasons */
tsk->thread.fpu.state->fxsave.mxcsr &= mxcsr_feature_mask;
if (err || __copy_from_user(&env, buf, sizeof(env)))
return 1;
convert_to_fxsr(tsk, &env);
return 0;
}
static int restore_i387_xsave(void __user *buf)
{
struct _fpx_sw_bytes fx_sw_user;
struct _fpstate_ia32 __user *fx_user =
((struct _fpstate_ia32 __user *) buf);
struct i387_fxsave_struct __user *fx =
(struct i387_fxsave_struct __user *) &fx_user->_fxsr_env[0];
struct xsave_hdr_struct *xsave_hdr =
&current->thread.fpu.state->xsave.xsave_hdr;
u64 mask;
int err;
if (check_for_xstate(fx, buf, &fx_sw_user))
goto fx_only;
mask = fx_sw_user.xstate_bv;
err = restore_i387_fxsave(buf, fx_sw_user.xstate_size);
xsave_hdr->xstate_bv &= pcntxt_mask;
/*
* These bits must be zero.
*/
xsave_hdr->reserved1[0] = xsave_hdr->reserved1[1] = 0;
/*
* Init the state that is not present in the memory layout
* and enabled by the OS.
*/
mask = ~(pcntxt_mask & ~mask);
xsave_hdr->xstate_bv &= mask;
return err;
fx_only:
/*
* Couldn't find the extended state information in the memory
* layout. Restore the FP/SSE and init the other extended state
* enabled by the OS.
*/
xsave_hdr->xstate_bv = XSTATE_FPSSE;
return restore_i387_fxsave(buf, sizeof(struct i387_fxsave_struct));
}
int restore_i387_xstate_ia32(void __user *buf)
{
int err;
struct task_struct *tsk = current;
struct _fpstate_ia32 __user *fp = (struct _fpstate_ia32 __user *) buf;
if (HAVE_HWFP)
clear_fpu(tsk);
if (!buf) {
if (used_math()) {
clear_fpu(tsk);
clear_used_math();
}
return 0;
} else
if (!access_ok(VERIFY_READ, buf, sig_xstate_ia32_size))
return -EACCES;
if (!used_math()) {
err = init_fpu(tsk);
if (err)
return err;
}
if (HAVE_HWFP) {
if (cpu_has_xsave)
err = restore_i387_xsave(buf);
else if (cpu_has_fxsr)
err = restore_i387_fxsave(fp, sizeof(struct
i387_fxsave_struct));
else
err = restore_i387_fsave(fp);
} else {
err = fpregs_soft_set(current, NULL,
0, sizeof(struct user_i387_ia32_struct),
NULL, fp) != 0;
}
set_used_math();
return err;
}
/*
* FPU state for core dumps.
* This is only used for a.out dumps now.
......
arch/x86/kernel/process.c
View file @ 72a671ce
......@@ -97,16 +97,6 @@ void arch_task_cache_init(void)
SLAB_PANIC | SLAB_NOTRACK, NULL);
}
static inline void drop_fpu(struct task_struct *tsk)
{
/*
* Forget coprocessor state..
*/
tsk->fpu_counter = 0;
clear_fpu(tsk);
clear_used_math();
}
/*
* Free current thread data structures etc..
*/
......
arch/x86/kernel/ptrace.c
View file @ 72a671ce
......@@ -1332,9 +1332,6 @@ static const struct user_regset_view user_x86_64_view = {
#define genregs32_get genregs_get
#define genregs32_set genregs_set
#define user_i387_ia32_struct user_i387_struct
#define user32_fxsr_struct user_fxsr_struct
#endif /* CONFIG_X86_64 */
#if defined CONFIG_X86_32 || defined CONFIG_IA32_EMULATION
......
arch/x86/kernel/signal.c
View file @ 72a671ce
......@@ -114,7 +114,7 @@ int restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc,
regs->orig_ax = -1; /* disable syscall checks */
get_user_ex(buf, &sc->fpstate);
err |= restore_i387_xstate(buf);
err |= restore_xstate_sig(buf, config_enabled(CONFIG_X86_32));
get_user_ex(*pax, &sc->ax);
} get_user_catch(err);
......@@ -206,7 +206,9 @@ get_sigframe(struct k_sigaction *ka, struct pt_regs *regs, size_t frame_size,
void __user **fpstate)
{
/* Default to using normal stack */
unsigned long math_size = 0;
unsigned long sp = regs->sp;
unsigned long buf_fx = 0;
int onsigstack = on_sig_stack(sp);
/* redzone */
......@@ -228,10 +230,8 @@ get_sigframe(struct k_sigaction *ka, struct pt_regs *regs, size_t frame_size,
}
if (used_math()) {
sp -= sig_xstate_size;
#ifdef CONFIG_X86_64
sp = round_down(sp, 64);
#endif /* CONFIG_X86_64 */
sp = alloc_mathframe(sp, config_enabled(CONFIG_X86_32),
&buf_fx, &math_size);
*fpstate = (void __user *)sp;
}
......@@ -244,8 +244,9 @@ get_sigframe(struct k_sigaction *ka, struct pt_regs *regs, size_t frame_size,
if (onsigstack && !likely(on_sig_stack(sp)))
return (void __user *)-1L;
/* save i387 state */
if (used_math() && save_i387_xstate(*fpstate) < 0)
/* save i387 and extended state */
if (used_math() &&
save_xstate_sig(*fpstate, (void __user *)buf_fx, math_size) < 0)
return (void __user *)-1L;
return (void __user *)sp;
......
arch/x86/kernel/xsave.c
View file @ 72a671ce
This diff is collapsed.
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment
GitLab Nexedi Edition | About GitLab | About Nexedi | 沪ICP备2021021310号-2 | 沪ICP备2021021310号-7