[PATCH 0/3] x86/entry: 32-bit C exit conversion and code deletion

[PATCH 0/3] x86/entry: 32-bit C exit conversion and code deletion

[Andy Lutomirski]

[Resend, this time to lkml, too.  Sigh, I'm bad at this email thing.]

As promised, here's the 32-bit code.  Lightly tested.  Brian, thanks for
cleaning up the vm86 mess.

I can imagine some objections to patch 1.  I wonder how many users
actually care about 32-bit audit performance.  It ought to be relatively
straightforward to come up with a clean opportunistic sysexit
implementation for 32-bit kernels, which would fix the performance
regression (mostly, at least) and would speed up lots of other
workloads.

Are these okay as is, or is opportunistic sysexit a prerequisite?

Applies to tip/x86/asm.

P.S. Denys, I think we should do opportunistic sysretl on 64-bit
kernels.  That would be much nicer on top of your sysexit cleanup
series.

Andy Lutomirski (3):
  x86/entry/32: Remove 32-bit syscall audit optimizations
  x86/entry/32: Migrate to C exit path
  x86/entry: Remove do_notify_resume, syscall_trace_leave, and their TIF
    masks

 arch/x86/entry/common.c            |  57 --------------------
 arch/x86/entry/entry_32.S          | 108 +++++--------------------------------
 arch/x86/include/asm/ptrace.h      |   1 -
 arch/x86/include/asm/signal.h      |   1 -
 arch/x86/include/asm/thread_info.h |  16 ------
 5 files changed, 12 insertions(+), 171 deletions(-)

-- 
2.4.3

[PATCH 1/3] x86/entry/32: Remove 32-bit syscall audit optimizations

[Andy Lutomirski]

The asm audit optimizations are ugly and obfuscate the code too
much.  Remove them.

This will regress performance if syscall auditing is enabled on
32-bit kernels and sysenter is in use.  If this becomes a problem,
interested parties are encouraged to implement the equivalent of the
64-bit opportunistic sysret optimization.

Alternatively, a case could be made that, on 32-bit kernels, a less
messy asm audit optimization could be done.  32-bit kernels don't have
the complicated partial register saving tricks that 64-bit kernels
have, so the sysenter post-syscall path could just call the audit
hooks directly.  Any reimplementation of this ought to demonstrate
that it only calls the audit hook once per syscall, though, which does
not currently appear to be true.  Someone would have to make the case
that doing so would be better than implementing opportunistic sysexit,
though.

Signed-off-by: Andy Lutomirski <luto@kernel.org>
---
 arch/x86/entry/entry_32.S | 48 ++---------------------------------------------
 1 file changed, 2 insertions(+), 46 deletions(-)

diff --git a/arch/x86/entry/entry_32.S b/arch/x86/entry/entry_32.S
index f940e24acaf0..a3c307ad5ac4 100644
--- a/arch/x86/entry/entry_32.S
+++ b/arch/x86/entry/entry_32.S
@@ -45,16 +45,6 @@
 #include <asm/asm.h>
 #include <asm/smap.h>
 
-/* Avoid __ASSEMBLER__'ifying <linux/audit.h> just for this.  */
-#include <linux/elf-em.h>
-#define AUDIT_ARCH_I386		(EM_386|__AUDIT_ARCH_LE)
-#define __AUDIT_ARCH_LE		0x40000000
-
-#ifndef CONFIG_AUDITSYSCALL
-# define sysenter_audit		syscall_trace_entry
-# define sysexit_audit		syscall_exit_work
-#endif
-
 	.section .entry.text, "ax"
 
 /*
@@ -339,7 +329,7 @@ sysenter_past_esp:
 	GET_THREAD_INFO(%ebp)
 
 	testl	$_TIF_WORK_SYSCALL_ENTRY, TI_flags(%ebp)
-	jnz	sysenter_audit
+	jnz	syscall_trace_entry
 sysenter_do_call:
 	cmpl	$(NR_syscalls), %eax
 	jae	sysenter_badsys
@@ -351,7 +341,7 @@ sysenter_after_call:
 	TRACE_IRQS_OFF
 	movl	TI_flags(%ebp), %ecx
 	testl	$_TIF_ALLWORK_MASK, %ecx
-	jnz	sysexit_audit
+	jnz	syscall_exit_work
 sysenter_exit:
 /* if something modifies registers it must also disable sysexit */
 	movl	PT_EIP(%esp), %edx
@@ -362,40 +352,6 @@ sysenter_exit:
 	PTGS_TO_GS
 	ENABLE_INTERRUPTS_SYSEXIT
 
-#ifdef CONFIG_AUDITSYSCALL
-sysenter_audit:
-	testl	$(_TIF_WORK_SYSCALL_ENTRY & ~_TIF_SYSCALL_AUDIT), TI_flags(%ebp)
-	jnz	syscall_trace_entry
-	/* movl	PT_EAX(%esp), %eax already set, syscall number: 1st arg to audit */
-	movl	PT_EBX(%esp), %edx		/* ebx/a0: 2nd arg to audit */
-	/* movl	PT_ECX(%esp), %ecx already set, a1: 3nd arg to audit */
-	pushl	PT_ESI(%esp)			/* a3: 5th arg */
-	pushl	PT_EDX+4(%esp)			/* a2: 4th arg */
-	call	__audit_syscall_entry
-	popl	%ecx				/* get that remapped edx off the stack */
-	popl	%ecx				/* get that remapped esi off the stack */
-	movl	PT_EAX(%esp), %eax		/* reload syscall number */
-	jmp	sysenter_do_call
-
-sysexit_audit:
-	testl	$(_TIF_ALLWORK_MASK & ~_TIF_SYSCALL_AUDIT), %ecx
-	jnz	syscall_exit_work
-	TRACE_IRQS_ON
-	ENABLE_INTERRUPTS(CLBR_ANY)
-	movl	%eax, %edx			/* second arg, syscall return value */
-	cmpl	$-MAX_ERRNO, %eax		/* is it an error ? */
-	setbe %al				/* 1 if so, 0 if not */
-	movzbl %al, %eax			/* zero-extend that */
-	call	__audit_syscall_exit
-	DISABLE_INTERRUPTS(CLBR_ANY)
-	TRACE_IRQS_OFF
-	movl	TI_flags(%ebp), %ecx
-	testl	$(_TIF_ALLWORK_MASK & ~_TIF_SYSCALL_AUDIT), %ecx
-	jnz	syscall_exit_work
-	movl	PT_EAX(%esp), %eax		/* reload syscall return value */
-	jmp	sysenter_exit
-#endif
-
 .pushsection .fixup, "ax"
 2:	movl	$0, PT_FS(%esp)
 	jmp	1b
-- 
2.4.3

[PATCH 2/3] x86/entry/32: Migrate to C exit path

[Andy Lutomirski]

This removes the hybrid asm-and-C implementation of exit work.

Signed-off-by: Andy Lutomirski <luto@kernel.org>
---
 arch/x86/entry/entry_32.S | 62 +++++++++--------------------------------------
 1 file changed, 11 insertions(+), 51 deletions(-)

diff --git a/arch/x86/entry/entry_32.S b/arch/x86/entry/entry_32.S
index a3c307ad5ac4..b2909bf8cf70 100644
--- a/arch/x86/entry/entry_32.S
+++ b/arch/x86/entry/entry_32.S
@@ -256,14 +256,10 @@ ret_from_intr:
 
 ENTRY(resume_userspace)
 	LOCKDEP_SYS_EXIT
-	DISABLE_INTERRUPTS(CLBR_ANY)		# make sure we don't miss an interrupt
-						# setting need_resched or sigpending
-						# between sampling and the iret
+	DISABLE_INTERRUPTS(CLBR_ANY)
 	TRACE_IRQS_OFF
-	movl	TI_flags(%ebp), %ecx
-	andl	$_TIF_WORK_MASK, %ecx		# is there any work to be done on
-						# int/exception return?
-	jne	work_pending
+	movl	%esp, %eax
+	call	prepare_exit_to_usermode
 	jmp	restore_all
 END(ret_from_exception)
 
@@ -341,7 +337,7 @@ sysenter_after_call:
 	TRACE_IRQS_OFF
 	movl	TI_flags(%ebp), %ecx
 	testl	$_TIF_ALLWORK_MASK, %ecx
-	jnz	syscall_exit_work
+	jnz	syscall_exit_work_irqs_off
 sysenter_exit:
 /* if something modifies registers it must also disable sysexit */
 	movl	PT_EIP(%esp), %edx
@@ -377,13 +373,7 @@ syscall_after_call:
 	movl	%eax, PT_EAX(%esp)		# store the return value
 syscall_exit:
 	LOCKDEP_SYS_EXIT
-	DISABLE_INTERRUPTS(CLBR_ANY)		# make sure we don't miss an interrupt
-						# setting need_resched or sigpending
-						# between sampling and the iret
-	TRACE_IRQS_OFF
-	movl	TI_flags(%ebp), %ecx
-	testl	$_TIF_ALLWORK_MASK, %ecx	# current->work
-	jnz	syscall_exit_work
+	jmp	syscall_exit_work
 
 restore_all:
 	TRACE_IRQS_IRET
@@ -460,35 +450,6 @@ ldt_ss:
 #endif
 ENDPROC(entry_INT80_32)
 
-	# perform work that needs to be done immediately before resumption
-	ALIGN
-work_pending:
-	testb	$_TIF_NEED_RESCHED, %cl
-	jz	work_notifysig
-work_resched:
-	call	schedule
-	LOCKDEP_SYS_EXIT
-	DISABLE_INTERRUPTS(CLBR_ANY)		# make sure we don't miss an interrupt
-						# setting need_resched or sigpending
-						# between sampling and the iret
-	TRACE_IRQS_OFF
-	movl	TI_flags(%ebp), %ecx
-	andl	$_TIF_WORK_MASK, %ecx		# is there any work to be done other
-						# than syscall tracing?
-	jz	restore_all
-	testb	$_TIF_NEED_RESCHED, %cl
-	jnz	work_resched
-
-work_notifysig:					# deal with pending signals and
-						# notify-resume requests
-	TRACE_IRQS_ON
-	ENABLE_INTERRUPTS(CLBR_NONE)
-	movl	%esp, %eax
-	xorl	%edx, %edx
-	call	do_notify_resume
-	jmp	resume_userspace
-END(work_pending)
-
 	# perform syscall exit tracing
 	ALIGN
 syscall_trace_entry:
@@ -503,15 +464,14 @@ END(syscall_trace_entry)
 
 	# perform syscall exit tracing
 	ALIGN
-syscall_exit_work:
-	testl	$_TIF_WORK_SYSCALL_EXIT, %ecx
-	jz	work_pending
+syscall_exit_work_irqs_off:
 	TRACE_IRQS_ON
-	ENABLE_INTERRUPTS(CLBR_ANY)		# could let syscall_trace_leave() call
-						# schedule() instead
+	ENABLE_INTERRUPTS(CLBR_ANY)
+
+syscall_exit_work:
 	movl	%esp, %eax
-	call	syscall_trace_leave
-	jmp	resume_userspace
+	call	syscall_return_slowpath
+	jmp	restore_all
 END(syscall_exit_work)
 
 syscall_fault:
-- 
2.4.3

[PATCH 3/3] x86/entry: Remove do_notify_resume, syscall_trace_leave, and their TIF masks

[Andy Lutomirski]

They are no longer used.  Good riddance!

Deleting the TIF_ macros is really nice.  It was never clear why
there were so many variants.

Signed-off-by: Andy Lutomirski <luto@kernel.org>
---
 arch/x86/entry/common.c            | 57 --------------------------------------
 arch/x86/include/asm/ptrace.h      |  1 -
 arch/x86/include/asm/signal.h      |  1 -
 arch/x86/include/asm/thread_info.h | 16 -----------
 4 files changed, 75 deletions(-)

diff --git a/arch/x86/entry/common.c b/arch/x86/entry/common.c
index a3e9c7fa15d9..80dcc9261ca3 100644
--- a/arch/x86/entry/common.c
+++ b/arch/x86/entry/common.c
@@ -207,37 +207,6 @@ long syscall_trace_enter(struct pt_regs *regs)
 		return syscall_trace_enter_phase2(regs, arch, phase1_result);
 }
 
-/* Deprecated. */
-void syscall_trace_leave(struct pt_regs *regs)
-{
-	bool step;
-
-	/*
-	 * We may come here right after calling schedule_user()
-	 * or do_notify_resume(), in which case we can be in RCU
-	 * user mode.
-	 */
-	user_exit();
-
-	audit_syscall_exit(regs);
-
-	if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT)))
-		trace_sys_exit(regs, regs->ax);
-
-	/*
-	 * If TIF_SYSCALL_EMU is set, we only get here because of
-	 * TIF_SINGLESTEP (i.e. this is PTRACE_SYSEMU_SINGLESTEP).
-	 * We already reported this syscall instruction in
-	 * syscall_trace_enter().
-	 */
-	step = unlikely(test_thread_flag(TIF_SINGLESTEP)) &&
-			!test_thread_flag(TIF_SYSCALL_EMU);
-	if (step || test_thread_flag(TIF_SYSCALL_TRACE))
-		tracehook_report_syscall_exit(regs, step);
-
-	user_enter();
-}
-
 static struct thread_info *pt_regs_to_thread_info(struct pt_regs *regs)
 {
 	unsigned long top_of_stack =
@@ -347,29 +316,3 @@ __visible void syscall_return_slowpath(struct pt_regs *regs)
 	local_irq_disable();
 	prepare_exit_to_usermode(regs);
 }
-
-/*
- * Deprecated notification of userspace execution resumption
- * - triggered by the TIF_WORK_MASK flags
- */
-__visible void
-do_notify_resume(struct pt_regs *regs, void *unused, __u32 thread_info_flags)
-{
-	user_exit();
-
-	if (thread_info_flags & _TIF_UPROBE)
-		uprobe_notify_resume(regs);
-
-	/* deal with pending signal delivery */
-	if (thread_info_flags & _TIF_SIGPENDING)
-		do_signal(regs);
-
-	if (thread_info_flags & _TIF_NOTIFY_RESUME) {
-		clear_thread_flag(TIF_NOTIFY_RESUME);
-		tracehook_notify_resume(regs);
-	}
-	if (thread_info_flags & _TIF_USER_RETURN_NOTIFY)
-		fire_user_return_notifiers();
-
-	user_enter();
-}
diff --git a/arch/x86/include/asm/ptrace.h b/arch/x86/include/asm/ptrace.h
index 5fabf1362942..6271281f947d 100644
--- a/arch/x86/include/asm/ptrace.h
+++ b/arch/x86/include/asm/ptrace.h
@@ -88,7 +88,6 @@ extern long syscall_trace_enter_phase2(struct pt_regs *, u32 arch,
 				       unsigned long phase1_result);
 
 extern long syscall_trace_enter(struct pt_regs *);
-extern void syscall_trace_leave(struct pt_regs *);
 
 static inline unsigned long regs_return_value(struct pt_regs *regs)
 {
diff --git a/arch/x86/include/asm/signal.h b/arch/x86/include/asm/signal.h
index b42408bcf6b5..c481be78fcf1 100644
--- a/arch/x86/include/asm/signal.h
+++ b/arch/x86/include/asm/signal.h
@@ -31,7 +31,6 @@ typedef sigset_t compat_sigset_t;
 #include <uapi/asm/signal.h>
 #ifndef __ASSEMBLY__
 extern void do_signal(struct pt_regs *regs);
-extern void do_notify_resume(struct pt_regs *, void *, __u32);
 
 #define __ARCH_HAS_SA_RESTORER
 
diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h
index fdad5c244350..8afdc3e44247 100644
--- a/arch/x86/include/asm/thread_info.h
+++ b/arch/x86/include/asm/thread_info.h
@@ -143,27 +143,11 @@ struct thread_info {
 	 _TIF_SECCOMP | _TIF_SINGLESTEP | _TIF_SYSCALL_TRACEPOINT |	\
 	 _TIF_NOHZ)
 
-/* work to do in syscall_trace_leave() */
-#define _TIF_WORK_SYSCALL_EXIT	\
-	(_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT | _TIF_SINGLESTEP |	\
-	 _TIF_SYSCALL_TRACEPOINT | _TIF_NOHZ)
-
-/* work to do on interrupt/exception return */
-#define _TIF_WORK_MASK							\
-	(0x0000FFFF &							\
-	 ~(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|			\
-	   _TIF_SINGLESTEP|_TIF_SECCOMP|_TIF_SYSCALL_EMU))
-
 /* work to do on any return to user space */
 #define _TIF_ALLWORK_MASK						\
 	((0x0000FFFF & ~_TIF_SECCOMP) | _TIF_SYSCALL_TRACEPOINT |	\
 	_TIF_NOHZ)
 
-/* Only used for 64 bit */
-#define _TIF_DO_NOTIFY_MASK						\
-	(_TIF_SIGPENDING | _TIF_NOTIFY_RESUME |				\
-	 _TIF_USER_RETURN_NOTIFY | _TIF_UPROBE)
-
 /* flags to check in __switch_to() */
 #define _TIF_WORK_CTXSW							\
 	(_TIF_IO_BITMAP|_TIF_NOTSC|_TIF_BLOCKSTEP)
-- 
2.4.3

[tip:x86/asm] x86/entry/32: Remove 32-bit syscall audit optimizations

[tip-bot for Andy Lutomirski]

Commit-ID:  c5f69fde26d1581ee495f68bb9de4049c8168a04
Gitweb:     http://git.kernel.org/tip/c5f69fde26d1581ee495f68bb9de4049c8168a04
Author:     Andy Lutomirski <luto@kernel.org>
AuthorDate: Fri, 31 Jul 2015 14:41:08 -0700
Committer:  Ingo Molnar <mingo@kernel.org>
CommitDate: Wed, 5 Aug 2015 10:54:35 +0200

x86/entry/32: Remove 32-bit syscall audit optimizations

The asm audit optimizations are ugly and obfuscate the code too
much. Remove them.

This will regress performance if syscall auditing is enabled on
32-bit kernels and SYSENTER is in use. If this becomes a
problem, interested parties are encouraged to implement the
equivalent of the 64-bit opportunistic SYSRET optimization.

Alternatively, a case could be made that, on 32-bit kernels, a
less messy asm audit optimization could be done. 32-bit kernels
don't have the complicated partial register saving tricks that
64-bit kernels have, so the SYSENTER post-syscall path could
just call the audit hooks directly.  Any reimplementation of
this ought to demonstrate that it only calls the audit hook once
per syscall, though, which does not currently appear to be true.

Someone would have to make the case that doing so would be
better than implementing opportunistic SYSEXIT, though.

Signed-off-by: Andy Lutomirski <luto@kernel.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Eric Paris <eparis@parisplace.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/212be39dd8c90b44c4b7bbc678128d6b88bdb9912.1438378274.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/entry/entry_32.S | 48 ++---------------------------------------------
 1 file changed, 2 insertions(+), 46 deletions(-)

diff --git a/arch/x86/entry/entry_32.S b/arch/x86/entry/entry_32.S
index f940e24..a3c307a 100644
--- a/arch/x86/entry/entry_32.S
+++ b/arch/x86/entry/entry_32.S
@@ -45,16 +45,6 @@
 #include <asm/asm.h>
 #include <asm/smap.h>
 
-/* Avoid __ASSEMBLER__'ifying <linux/audit.h> just for this.  */
-#include <linux/elf-em.h>
-#define AUDIT_ARCH_I386		(EM_386|__AUDIT_ARCH_LE)
-#define __AUDIT_ARCH_LE		0x40000000
-
-#ifndef CONFIG_AUDITSYSCALL
-# define sysenter_audit		syscall_trace_entry
-# define sysexit_audit		syscall_exit_work
-#endif
-
 	.section .entry.text, "ax"
 
 /*
@@ -339,7 +329,7 @@ sysenter_past_esp:
 	GET_THREAD_INFO(%ebp)
 
 	testl	$_TIF_WORK_SYSCALL_ENTRY, TI_flags(%ebp)
-	jnz	sysenter_audit
+	jnz	syscall_trace_entry
 sysenter_do_call:
 	cmpl	$(NR_syscalls), %eax
 	jae	sysenter_badsys
@@ -351,7 +341,7 @@ sysenter_after_call:
 	TRACE_IRQS_OFF
 	movl	TI_flags(%ebp), %ecx
 	testl	$_TIF_ALLWORK_MASK, %ecx
-	jnz	sysexit_audit
+	jnz	syscall_exit_work
 sysenter_exit:
 /* if something modifies registers it must also disable sysexit */
 	movl	PT_EIP(%esp), %edx
@@ -362,40 +352,6 @@ sysenter_exit:
 	PTGS_TO_GS
 	ENABLE_INTERRUPTS_SYSEXIT
 
-#ifdef CONFIG_AUDITSYSCALL
-sysenter_audit:
-	testl	$(_TIF_WORK_SYSCALL_ENTRY & ~_TIF_SYSCALL_AUDIT), TI_flags(%ebp)
-	jnz	syscall_trace_entry
-	/* movl	PT_EAX(%esp), %eax already set, syscall number: 1st arg to audit */
-	movl	PT_EBX(%esp), %edx		/* ebx/a0: 2nd arg to audit */
-	/* movl	PT_ECX(%esp), %ecx already set, a1: 3nd arg to audit */
-	pushl	PT_ESI(%esp)			/* a3: 5th arg */
-	pushl	PT_EDX+4(%esp)			/* a2: 4th arg */
-	call	__audit_syscall_entry
-	popl	%ecx				/* get that remapped edx off the stack */
-	popl	%ecx				/* get that remapped esi off the stack */
-	movl	PT_EAX(%esp), %eax		/* reload syscall number */
-	jmp	sysenter_do_call
-
-sysexit_audit:
-	testl	$(_TIF_ALLWORK_MASK & ~_TIF_SYSCALL_AUDIT), %ecx
-	jnz	syscall_exit_work
-	TRACE_IRQS_ON
-	ENABLE_INTERRUPTS(CLBR_ANY)
-	movl	%eax, %edx			/* second arg, syscall return value */
-	cmpl	$-MAX_ERRNO, %eax		/* is it an error ? */
-	setbe %al				/* 1 if so, 0 if not */
-	movzbl %al, %eax			/* zero-extend that */
-	call	__audit_syscall_exit
-	DISABLE_INTERRUPTS(CLBR_ANY)
-	TRACE_IRQS_OFF
-	movl	TI_flags(%ebp), %ecx
-	testl	$(_TIF_ALLWORK_MASK & ~_TIF_SYSCALL_AUDIT), %ecx
-	jnz	syscall_exit_work
-	movl	PT_EAX(%esp), %eax		/* reload syscall return value */
-	jmp	sysenter_exit
-#endif
-
 .pushsection .fixup, "ax"
 2:	movl	$0, PT_FS(%esp)
 	jmp	1b

[tip:x86/asm] x86/entry/32: Migrate to C exit path

[tip-bot for Andy Lutomirski]

Commit-ID:  5d73fc70996d9de0d1b2fc87e62dc51153204eba
Gitweb:     http://git.kernel.org/tip/5d73fc70996d9de0d1b2fc87e62dc51153204eba
Author:     Andy Lutomirski <luto@kernel.org>
AuthorDate: Fri, 31 Jul 2015 14:41:09 -0700
Committer:  Ingo Molnar <mingo@kernel.org>
CommitDate: Wed, 5 Aug 2015 10:54:35 +0200

x86/entry/32: Migrate to C exit path

This removes the hybrid asm-and-C implementation of exit work.

Signed-off-by: Andy Lutomirski <luto@kernel.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Eric Paris <eparis@parisplace.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/2baa438619ea6c027b40ec9fceacca52f09c74d09.1438378274.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/entry/entry_32.S | 62 +++++++++--------------------------------------
 1 file changed, 11 insertions(+), 51 deletions(-)

diff --git a/arch/x86/entry/entry_32.S b/arch/x86/entry/entry_32.S
index a3c307a..b2909bf 100644
--- a/arch/x86/entry/entry_32.S
+++ b/arch/x86/entry/entry_32.S
@@ -256,14 +256,10 @@ ret_from_intr:
 
 ENTRY(resume_userspace)
 	LOCKDEP_SYS_EXIT
-	DISABLE_INTERRUPTS(CLBR_ANY)		# make sure we don't miss an interrupt
-						# setting need_resched or sigpending
-						# between sampling and the iret
+	DISABLE_INTERRUPTS(CLBR_ANY)
 	TRACE_IRQS_OFF
-	movl	TI_flags(%ebp), %ecx
-	andl	$_TIF_WORK_MASK, %ecx		# is there any work to be done on
-						# int/exception return?
-	jne	work_pending
+	movl	%esp, %eax
+	call	prepare_exit_to_usermode
 	jmp	restore_all
 END(ret_from_exception)
 
@@ -341,7 +337,7 @@ sysenter_after_call:
 	TRACE_IRQS_OFF
 	movl	TI_flags(%ebp), %ecx
 	testl	$_TIF_ALLWORK_MASK, %ecx
-	jnz	syscall_exit_work
+	jnz	syscall_exit_work_irqs_off
 sysenter_exit:
 /* if something modifies registers it must also disable sysexit */
 	movl	PT_EIP(%esp), %edx
@@ -377,13 +373,7 @@ syscall_after_call:
 	movl	%eax, PT_EAX(%esp)		# store the return value
 syscall_exit:
 	LOCKDEP_SYS_EXIT
-	DISABLE_INTERRUPTS(CLBR_ANY)		# make sure we don't miss an interrupt
-						# setting need_resched or sigpending
-						# between sampling and the iret
-	TRACE_IRQS_OFF
-	movl	TI_flags(%ebp), %ecx
-	testl	$_TIF_ALLWORK_MASK, %ecx	# current->work
-	jnz	syscall_exit_work
+	jmp	syscall_exit_work
 
 restore_all:
 	TRACE_IRQS_IRET
@@ -460,35 +450,6 @@ ldt_ss:
 #endif
 ENDPROC(entry_INT80_32)
 
-	# perform work that needs to be done immediately before resumption
-	ALIGN
-work_pending:
-	testb	$_TIF_NEED_RESCHED, %cl
-	jz	work_notifysig
-work_resched:
-	call	schedule
-	LOCKDEP_SYS_EXIT
-	DISABLE_INTERRUPTS(CLBR_ANY)		# make sure we don't miss an interrupt
-						# setting need_resched or sigpending
-						# between sampling and the iret
-	TRACE_IRQS_OFF
-	movl	TI_flags(%ebp), %ecx
-	andl	$_TIF_WORK_MASK, %ecx		# is there any work to be done other
-						# than syscall tracing?
-	jz	restore_all
-	testb	$_TIF_NEED_RESCHED, %cl
-	jnz	work_resched
-
-work_notifysig:					# deal with pending signals and
-						# notify-resume requests
-	TRACE_IRQS_ON
-	ENABLE_INTERRUPTS(CLBR_NONE)
-	movl	%esp, %eax
-	xorl	%edx, %edx
-	call	do_notify_resume
-	jmp	resume_userspace
-END(work_pending)
-
 	# perform syscall exit tracing
 	ALIGN
 syscall_trace_entry:
@@ -503,15 +464,14 @@ END(syscall_trace_entry)
 
 	# perform syscall exit tracing
 	ALIGN
-syscall_exit_work:
-	testl	$_TIF_WORK_SYSCALL_EXIT, %ecx
-	jz	work_pending
+syscall_exit_work_irqs_off:
 	TRACE_IRQS_ON
-	ENABLE_INTERRUPTS(CLBR_ANY)		# could let syscall_trace_leave() call
-						# schedule() instead
+	ENABLE_INTERRUPTS(CLBR_ANY)
+
+syscall_exit_work:
 	movl	%esp, %eax
-	call	syscall_trace_leave
-	jmp	resume_userspace
+	call	syscall_return_slowpath
+	jmp	restore_all
 END(syscall_exit_work)
 
 syscall_fault:

[tip:x86/asm] x86/entry: Remove do_notify_resume(), syscall_trace_leave(), and their TIF masks

[tip-bot for Andy Lutomirski]

Commit-ID:  88cd622f9299c4c9e61e978bb9ef9d7599769ed0
Gitweb:     http://git.kernel.org/tip/88cd622f9299c4c9e61e978bb9ef9d7599769ed0
Author:     Andy Lutomirski <luto@kernel.org>
AuthorDate: Fri, 31 Jul 2015 14:41:10 -0700
Committer:  Ingo Molnar <mingo@kernel.org>
CommitDate: Wed, 5 Aug 2015 10:54:35 +0200

x86/entry: Remove do_notify_resume(), syscall_trace_leave(), and their TIF masks

They are no longer used. Good riddance!

Deleting the TIF_ macros is really nice. It was never clear why
there were so many variants.

Signed-off-by: Andy Lutomirski <luto@kernel.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Eric Paris <eparis@parisplace.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/22c61682f446628573dde0f1d573ab821677e06da.1438378274.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/entry/common.c            | 57 --------------------------------------
 arch/x86/include/asm/ptrace.h      |  1 -
 arch/x86/include/asm/signal.h      |  1 -
 arch/x86/include/asm/thread_info.h | 16 -----------
 4 files changed, 75 deletions(-)

diff --git a/arch/x86/entry/common.c b/arch/x86/entry/common.c
index a3e9c7f..80dcc92 100644
--- a/arch/x86/entry/common.c
+++ b/arch/x86/entry/common.c
@@ -207,37 +207,6 @@ long syscall_trace_enter(struct pt_regs *regs)
 		return syscall_trace_enter_phase2(regs, arch, phase1_result);
 }
 
-/* Deprecated. */
-void syscall_trace_leave(struct pt_regs *regs)
-{
-	bool step;
-
-	/*
-	 * We may come here right after calling schedule_user()
-	 * or do_notify_resume(), in which case we can be in RCU
-	 * user mode.
-	 */
-	user_exit();
-
-	audit_syscall_exit(regs);
-
-	if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT)))
-		trace_sys_exit(regs, regs->ax);
-
-	/*
-	 * If TIF_SYSCALL_EMU is set, we only get here because of
-	 * TIF_SINGLESTEP (i.e. this is PTRACE_SYSEMU_SINGLESTEP).
-	 * We already reported this syscall instruction in
-	 * syscall_trace_enter().
-	 */
-	step = unlikely(test_thread_flag(TIF_SINGLESTEP)) &&
-			!test_thread_flag(TIF_SYSCALL_EMU);
-	if (step || test_thread_flag(TIF_SYSCALL_TRACE))
-		tracehook_report_syscall_exit(regs, step);
-
-	user_enter();
-}
-
 static struct thread_info *pt_regs_to_thread_info(struct pt_regs *regs)
 {
 	unsigned long top_of_stack =
@@ -347,29 +316,3 @@ __visible void syscall_return_slowpath(struct pt_regs *regs)
 	local_irq_disable();
 	prepare_exit_to_usermode(regs);
 }
-
-/*
- * Deprecated notification of userspace execution resumption
- * - triggered by the TIF_WORK_MASK flags
- */
-__visible void
-do_notify_resume(struct pt_regs *regs, void *unused, __u32 thread_info_flags)
-{
-	user_exit();
-
-	if (thread_info_flags & _TIF_UPROBE)
-		uprobe_notify_resume(regs);
-
-	/* deal with pending signal delivery */
-	if (thread_info_flags & _TIF_SIGPENDING)
-		do_signal(regs);
-
-	if (thread_info_flags & _TIF_NOTIFY_RESUME) {
-		clear_thread_flag(TIF_NOTIFY_RESUME);
-		tracehook_notify_resume(regs);
-	}
-	if (thread_info_flags & _TIF_USER_RETURN_NOTIFY)
-		fire_user_return_notifiers();
-
-	user_enter();
-}
diff --git a/arch/x86/include/asm/ptrace.h b/arch/x86/include/asm/ptrace.h
index 5fabf13..6271281 100644
--- a/arch/x86/include/asm/ptrace.h
+++ b/arch/x86/include/asm/ptrace.h
@@ -88,7 +88,6 @@ extern long syscall_trace_enter_phase2(struct pt_regs *, u32 arch,
 				       unsigned long phase1_result);
 
 extern long syscall_trace_enter(struct pt_regs *);
-extern void syscall_trace_leave(struct pt_regs *);
 
 static inline unsigned long regs_return_value(struct pt_regs *regs)
 {
diff --git a/arch/x86/include/asm/signal.h b/arch/x86/include/asm/signal.h
index b42408b..c481be7 100644
--- a/arch/x86/include/asm/signal.h
+++ b/arch/x86/include/asm/signal.h
@@ -31,7 +31,6 @@ typedef sigset_t compat_sigset_t;
 #include <uapi/asm/signal.h>
 #ifndef __ASSEMBLY__
 extern void do_signal(struct pt_regs *regs);
-extern void do_notify_resume(struct pt_regs *, void *, __u32);
 
 #define __ARCH_HAS_SA_RESTORER
 
diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h
index fdad5c2..8afdc3e 100644
--- a/arch/x86/include/asm/thread_info.h
+++ b/arch/x86/include/asm/thread_info.h
@@ -143,27 +143,11 @@ struct thread_info {
 	 _TIF_SECCOMP | _TIF_SINGLESTEP | _TIF_SYSCALL_TRACEPOINT |	\
 	 _TIF_NOHZ)
 
-/* work to do in syscall_trace_leave() */
-#define _TIF_WORK_SYSCALL_EXIT	\
-	(_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT | _TIF_SINGLESTEP |	\
-	 _TIF_SYSCALL_TRACEPOINT | _TIF_NOHZ)
-
-/* work to do on interrupt/exception return */
-#define _TIF_WORK_MASK							\
-	(0x0000FFFF &							\
-	 ~(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|			\
-	   _TIF_SINGLESTEP|_TIF_SECCOMP|_TIF_SYSCALL_EMU))
-
 /* work to do on any return to user space */
 #define _TIF_ALLWORK_MASK						\
 	((0x0000FFFF & ~_TIF_SECCOMP) | _TIF_SYSCALL_TRACEPOINT |	\
 	_TIF_NOHZ)
 
-/* Only used for 64 bit */
-#define _TIF_DO_NOTIFY_MASK						\
-	(_TIF_SIGPENDING | _TIF_NOTIFY_RESUME |				\
-	 _TIF_USER_RETURN_NOTIFY | _TIF_UPROBE)
-
 /* flags to check in __switch_to() */
 #define _TIF_WORK_CTXSW							\
 	(_TIF_IO_BITMAP|_TIF_NOTSC|_TIF_BLOCKSTEP)