x86, ptrace: Fix block-step

Implement ptrace-block-step using TIF_BLOCKSTEP which will set
DEBUGCTLMSR_BTF when set for a task while preserving any other
DEBUGCTLMSR bits.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <20100325135414.017536066@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h
index 5bec21a..32428b4 100644
--- a/arch/x86/include/asm/processor.h
+++ b/arch/x86/include/asm/processor.h
@@ -799,7 +799,7 @@
 
 static inline unsigned long get_debugctlmsr(void)
 {
-    unsigned long debugctlmsr = 0;
+	unsigned long debugctlmsr = 0;
 
 #ifndef CONFIG_X86_DEBUGCTLMSR
 	if (boot_cpu_data.x86 < 6)
@@ -807,7 +807,7 @@
 #endif
 	rdmsrl(MSR_IA32_DEBUGCTLMSR, debugctlmsr);
 
-    return debugctlmsr;
+	return debugctlmsr;
 }
 
 static inline void update_debugctlmsr(unsigned long debugctlmsr)
diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h
index dc85e12..d017ed55 100644
--- a/arch/x86/include/asm/thread_info.h
+++ b/arch/x86/include/asm/thread_info.h
@@ -92,6 +92,7 @@
 #define TIF_IO_BITMAP		22	/* uses I/O bitmap */
 #define TIF_FREEZE		23	/* is freezing for suspend */
 #define TIF_FORCED_TF		24	/* true if TF in eflags artificially */
+#define TIF_BLOCKSTEP		25	/* set when we want DEBUGCTLMSR_BTF */
 #define TIF_LAZY_MMU_UPDATES	27	/* task is updating the mmu lazily */
 #define TIF_SYSCALL_TRACEPOINT	28	/* syscall tracepoint instrumentation */
 
@@ -113,6 +114,7 @@
 #define _TIF_IO_BITMAP		(1 << TIF_IO_BITMAP)
 #define _TIF_FREEZE		(1 << TIF_FREEZE)
 #define _TIF_FORCED_TF		(1 << TIF_FORCED_TF)
+#define _TIF_BLOCKSTEP		(1 << TIF_BLOCKSTEP)
 #define _TIF_LAZY_MMU_UPDATES	(1 << TIF_LAZY_MMU_UPDATES)
 #define _TIF_SYSCALL_TRACEPOINT	(1 << TIF_SYSCALL_TRACEPOINT)
 
@@ -143,7 +145,7 @@
 
 /* flags to check in __switch_to() */
 #define _TIF_WORK_CTXSW							\
-	(_TIF_IO_BITMAP|_TIF_NOTSC)
+	(_TIF_IO_BITMAP|_TIF_NOTSC|_TIF_BLOCKSTEP)
 
 #define _TIF_WORK_CTXSW_PREV (_TIF_WORK_CTXSW|_TIF_USER_RETURN_NOTIFY)
 #define _TIF_WORK_CTXSW_NEXT (_TIF_WORK_CTXSW|_TIF_DEBUG)
diff --git a/arch/x86/kernel/kprobes.c b/arch/x86/kernel/kprobes.c
index 7a880ad..f2f56c0 100644
--- a/arch/x86/kernel/kprobes.c
+++ b/arch/x86/kernel/kprobes.c
@@ -422,12 +422,22 @@
 
 static void __kprobes clear_btf(void)
 {
-	/* XXX */
+	if (test_thread_flag(TIF_BLOCKSTEP)) {
+		unsigned long debugctl = get_debugctlmsr();
+
+		debugctl &= ~DEBUGCTLMSR_BTF;
+		update_debugctlmsr(debugctl);
+	}
 }
 
 static void __kprobes restore_btf(void)
 {
-	/* XXX */
+	if (test_thread_flag(TIF_BLOCKSTEP)) {
+		unsigned long debugctl = get_debugctlmsr();
+
+		debugctl |= DEBUGCTLMSR_BTF;
+		update_debugctlmsr(debugctl);
+	}
 }
 
 void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri,
diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c
index 1a60beb..8328009 100644
--- a/arch/x86/kernel/process.c
+++ b/arch/x86/kernel/process.c
@@ -195,6 +195,17 @@
 	prev = &prev_p->thread;
 	next = &next_p->thread;
 
+	if (test_tsk_thread_flag(prev_p, TIF_BLOCKSTEP) ^
+	    test_tsk_thread_flag(next_p, TIF_BLOCKSTEP)) {
+		unsigned long debugctl = get_debugctlmsr();
+
+		debugctl &= ~DEBUGCTLMSR_BTF;
+		if (test_tsk_thread_flag(next_p, TIF_BLOCKSTEP))
+			debugctl |= DEBUGCTLMSR_BTF;
+
+		update_debugctlmsr(debugctl);
+	}
+
 	if (test_tsk_thread_flag(prev_p, TIF_NOTSC) ^
 	    test_tsk_thread_flag(next_p, TIF_NOTSC)) {
 		/* prev and next are different */
diff --git a/arch/x86/kernel/step.c b/arch/x86/kernel/step.c
index 7beba07..58de45e 100644
--- a/arch/x86/kernel/step.c
+++ b/arch/x86/kernel/step.c
@@ -169,9 +169,19 @@
 	 * So noone should try to use debugger block stepping in a program
 	 * that uses user-mode single stepping itself.
 	 */
-	if (!enable_single_step(child))
-		return;
-	/* XXX */
+	if (enable_single_step(child) && block) {
+		unsigned long debugctl = get_debugctlmsr();
+
+		debugctl |= DEBUGCTLMSR_BTF;
+		update_debugctlmsr(debugctl);
+		set_tsk_thread_flag(child, TIF_BLOCKSTEP);
+	} else if (test_tsk_thread_flag(child, TIF_BLOCKSTEP)) {
+		unsigned long debugctl = get_debugctlmsr();
+
+		debugctl &= ~DEBUGCTLMSR_BTF;
+		update_debugctlmsr(debugctl);
+		clear_tsk_thread_flag(child, TIF_BLOCKSTEP);
+	}
 }
 
 void user_enable_single_step(struct task_struct *child)
@@ -189,7 +199,13 @@
 	/*
 	 * Make sure block stepping (BTF) is disabled.
 	 */
-	/* XXX */
+	if (test_tsk_thread_flag(child, TIF_BLOCKSTEP)) {
+		unsigned long debugctl = get_debugctlmsr();
+
+		debugctl &= ~DEBUGCTLMSR_BTF;
+		update_debugctlmsr(debugctl);
+		clear_tsk_thread_flag(child, TIF_BLOCKSTEP);
+	}
 
 	/* Always clear TIF_SINGLESTEP... */
 	clear_tsk_thread_flag(child, TIF_SINGLESTEP);
diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c
index e3da5d7..36f1bd9 100644
--- a/arch/x86/kernel/traps.c
+++ b/arch/x86/kernel/traps.c
@@ -544,6 +544,11 @@
 	/* DR6 may or may not be cleared by the CPU */
 	set_debugreg(0, 6);
 
+	/*
+	 * The processor cleared BTF, so don't mark that we need it set.
+	 */
+	clear_tsk_thread_flag(tsk, TIF_BLOCKSTEP);
+
 	/* Store the virtualized DR6 value */
 	tsk->thread.debugreg6 = dr6;