x86/asm/entry/64: Save user RSP in pt_regs->sp on SYSCALL64 fastpath

Prepare for the removal of 'usersp', by simplifying PER_CPU(old_rsp) usage:

  - use it only as temp storage

  - store the userspace stack pointer immediately in pt_regs->sp
    on syscall entry, instead of using it later, on syscall exit.

  - change C code to use pt_regs->sp only, instead of PER_CPU(old_rsp)
    and task->thread.usersp.

FIXUP/RESTORE_TOP_OF_STACK are simplified as well.

Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Alexei Starovoitov <ast@plumgrid.com>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Will Drewry <wad@chromium.org>
Link: http://lkml.kernel.org/r/1425926364-9526-4-git-send-email-dvlasenk@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>

5 files changed, 12 insertions, 21 deletions

diff --git a/arch/x86/include/asm/compat.h b/arch/x86/include/asm/compat.h
index 59c6c40..acdee09 100644
--- a/arch/x86/include/asm/compat.h
+++ b/arch/x86/include/asm/compat.h
@@ -301,7 +301,7 @@ static inline void __user *arch_compat_alloc_user_space(long len)
 		sp = task_pt_regs(current)->sp;
 	} else {
 		/* -128 for the x32 ABI redzone */
-		sp = this_cpu_read(old_rsp) - 128;
+		sp = task_pt_regs(current)->sp - 128;
 	}
 
 	return (void __user *)round_down(sp - len, 16);
diff --git a/arch/x86/include/asm/ptrace.h b/arch/x86/include/asm/ptrace.h
index 4077d96..74bb2e0 100644
--- a/arch/x86/include/asm/ptrace.h
+++ b/arch/x86/include/asm/ptrace.h
@@ -145,12 +145,8 @@ static inline bool user_64bit_mode(struct pt_regs *regs)
 #endif
 }
 
-#define current_user_stack_pointer()	this_cpu_read(old_rsp)
-/* ia32 vs. x32 difference */
-#define compat_user_stack_pointer()	\
-	(test_thread_flag(TIF_IA32) 	\
-	 ? current_pt_regs()->sp 	\
-	 : this_cpu_read(old_rsp))
+#define current_user_stack_pointer()	current_pt_regs()->sp
+#define compat_user_stack_pointer()	current_pt_regs()->sp
 #endif
 
 #ifdef CONFIG_X86_32
diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S
index 703ced0..d86788c 100644
--- a/arch/x86/kernel/entry_64.S
+++ b/arch/x86/kernel/entry_64.S
@@ -128,8 +128,6 @@ ENDPROC(native_usergs_sysret64)
  * manipulation.
  */
 	.macro FIXUP_TOP_OF_STACK tmp offset=0
-	movq PER_CPU_VAR(old_rsp),\tmp
-	movq \tmp,RSP+\offset(%rsp)
 	movq $__USER_DS,SS+\offset(%rsp)
 	movq $__USER_CS,CS+\offset(%rsp)
 	movq RIP+\offset(%rsp),\tmp  /* get rip */
@@ -139,8 +137,7 @@ ENDPROC(native_usergs_sysret64)
 	.endm
 
 	.macro RESTORE_TOP_OF_STACK tmp offset=0
-	movq RSP+\offset(%rsp),\tmp
-	movq \tmp,PER_CPU_VAR(old_rsp)
+	/* nothing to do */
 	.endm
 
 /*
@@ -222,9 +219,6 @@ ENDPROC(native_usergs_sysret64)
  * Interrupts are off on entry.
  * Only called from user space.
  *
- * XXX	if we had a free scratch register we could save the RSP into the stack frame
- *      and report it properly in ps. Unfortunately we haven't.
- *
  * When user can change the frames always force IRET. That is because
  * it deals with uncanonical addresses better. SYSRET has trouble
  * with them due to bugs in both AMD and Intel CPUs.
@@ -253,11 +247,13 @@ GLOBAL(system_call_after_swapgs)
 	 */
 	ENABLE_INTERRUPTS(CLBR_NONE)
 	ALLOC_PT_GPREGS_ON_STACK 8		/* +8: space for orig_ax */
+	movq	%rcx,RIP(%rsp)
+	movq	PER_CPU_VAR(old_rsp),%rcx
+	movq	%r11,EFLAGS(%rsp)
+	movq	%rcx,RSP(%rsp)
+	movq_cfi rax,ORIG_RAX
 	SAVE_C_REGS_EXCEPT_RAX_RCX_R11
 	movq	$-ENOSYS,RAX(%rsp)
-	movq_cfi rax,ORIG_RAX
-	movq	%r11,EFLAGS(%rsp)
-	movq	%rcx,RIP(%rsp)
 	CFI_REL_OFFSET rip,RIP
 	testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags+THREAD_INFO(%rsp,RIP)
 	jnz tracesys
@@ -293,7 +289,7 @@ ret_from_sys_call:
 	CFI_REGISTER	rip,rcx
 	movq	EFLAGS(%rsp),%r11
 	/*CFI_REGISTER	rflags,r11*/
-	movq	PER_CPU_VAR(old_rsp), %rsp
+	movq	RSP(%rsp),%rsp
 	/*
 	 * 64bit SYSRET restores rip from rcx,
 	 * rflags from r11 (but RF and VM bits are forced to 0),
diff --git a/arch/x86/kernel/perf_regs.c b/arch/x86/kernel/perf_regs.c
index 781861c..02a8720 100644
--- a/arch/x86/kernel/perf_regs.c
+++ b/arch/x86/kernel/perf_regs.c
@@ -177,7 +177,7 @@ void perf_get_regs_user(struct perf_regs *regs_user,
 		 * than just blindly copying user_regs.
 		 */
 		regs_user->abi = PERF_SAMPLE_REGS_ABI_64;
-		regs_user_copy->sp = this_cpu_read(old_rsp);
+		regs_user_copy->sp = user_regs->sp;
 		regs_user_copy->cs = __USER_CS;
 		regs_user_copy->ss = __USER_DS;
 		regs_user_copy->cx = -1;  /* usually contains garbage */
diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c
index 1e393d2..e8c124a 100644
--- a/arch/x86/kernel/process_64.c
+++ b/arch/x86/kernel/process_64.c
@@ -602,6 +602,5 @@ long sys_arch_prctl(int code, unsigned long addr)
 
 unsigned long KSTK_ESP(struct task_struct *task)
 {
-	return (test_tsk_thread_flag(task, TIF_IA32)) ?
-			(task_pt_regs(task)->sp) : ((task)->thread.usersp);
+	return task_pt_regs(task)->sp;
 }