Merge tag 'for-linus' of git://git.armlinux.org.uk/~rmk/linux-arm

Pull ARM updates from Russell King:

 - handle inexact watchpoint addresses (Douglas Anderson)

 - decompressor serial debug cleanups (Linus Walleij)

 - update L2 cache prefetch bits (Guillaume Tucker)

 - add text offset and malloc size to the decompressor kexec data

* tag 'for-linus' of git://git.armlinux.org.uk/~rmk/linux-arm:
  ARM: add malloc size to decompressor kexec size structure
  ARM: add TEXT_OFFSET to decompressor kexec image structure
  ARM: 9007/1: l2c: fix prefetch bits init in L2X0_AUX_CTRL using DT values
  ARM: 9010/1: uncompress: Print the location of appended DTB
  ARM: 9009/1: uncompress: Enable debug in head.S
  ARM: 9008/1: uncompress: Drop excess whitespace print
  ARM: 9006/1: uncompress: Wait for ready and busy in debug prints
  ARM: 9005/1: debug: Select flow control for all debug UARTs
  ARM: 9004/1: debug: Split waituart to CTS and TXRDY
  ARM: 9003/1: uncompress: Delete unused debug macros
  ARM: 8997/2: hw_breakpoint: Handle inexact watchpoint addresses

2 files changed, 81 insertions, 30 deletions

diff --git a/arch/arm/kernel/debug.S b/arch/arm/kernel/debug.S
index e112072b579d..d92f44bdf438 100644
--- a/arch/arm/kernel/debug.S
+++ b/arch/arm/kernel/debug.S
@@ -89,11 +89,18 @@ ENTRY(printascii)
 2:		teq     r1, #'\n'
 		bne	3f
 		mov	r1, #'\r'
-		waituart r2, r3
+#ifdef CONFIG_DEBUG_UART_FLOW_CONTROL
+		waituartcts r2, r3
+#endif
+		waituarttxrdy r2, r3
 		senduart r1, r3
 		busyuart r2, r3
 		mov	r1, #'\n'
-3:		waituart r2, r3
+3:
+#ifdef CONFIG_DEBUG_UART_FLOW_CONTROL
+		waituartcts r2, r3
+#endif
+		waituarttxrdy r2, r3
 		senduart r1, r3
 		busyuart r2, r3
 		b	1b
diff --git a/arch/arm/kernel/hw_breakpoint.c b/arch/arm/kernel/hw_breakpoint.c
index 7a4853b1213a..08660ae9dcbc 100644
--- a/arch/arm/kernel/hw_breakpoint.c
+++ b/arch/arm/kernel/hw_breakpoint.c
@@ -683,6 +683,40 @@ static void disable_single_step(struct perf_event *bp)
 	arch_install_hw_breakpoint(bp);
 }
 
+/*
+ * Arm32 hardware does not always report a watchpoint hit address that matches
+ * one of the watchpoints set. It can also report an address "near" the
+ * watchpoint if a single instruction access both watched and unwatched
+ * addresses. There is no straight-forward way, short of disassembling the
+ * offending instruction, to map that address back to the watchpoint. This
+ * function computes the distance of the memory access from the watchpoint as a
+ * heuristic for the likelyhood that a given access triggered the watchpoint.
+ *
+ * See this same function in the arm64 platform code, which has the same
+ * problem.
+ *
+ * The function returns the distance of the address from the bytes watched by
+ * the watchpoint. In case of an exact match, it returns 0.
+ */
+static u32 get_distance_from_watchpoint(unsigned long addr, u32 val,
+					struct arch_hw_breakpoint_ctrl *ctrl)
+{
+	u32 wp_low, wp_high;
+	u32 lens, lene;
+
+	lens = __ffs(ctrl->len);
+	lene = __fls(ctrl->len);
+
+	wp_low = val + lens;
+	wp_high = val + lene;
+	if (addr < wp_low)
+		return wp_low - addr;
+	else if (addr > wp_high)
+		return addr - wp_high;
+	else
+		return 0;
+}
+
 static int watchpoint_fault_on_uaccess(struct pt_regs *regs,
 				       struct arch_hw_breakpoint *info)
 {
@@ -692,23 +726,25 @@ static int watchpoint_fault_on_uaccess(struct pt_regs *regs,
 static void watchpoint_handler(unsigned long addr, unsigned int fsr,
 			       struct pt_regs *regs)
 {
-	int i, access;
-	u32 val, ctrl_reg, alignment_mask;
+	int i, access, closest_match = 0;
+	u32 min_dist = -1, dist;
+	u32 val, ctrl_reg;
 	struct perf_event *wp, **slots;
 	struct arch_hw_breakpoint *info;
 	struct arch_hw_breakpoint_ctrl ctrl;
 
 	slots = this_cpu_ptr(wp_on_reg);
 
+	/*
+	 * Find all watchpoints that match the reported address. If no exact
+	 * match is found. Attribute the hit to the closest watchpoint.
+	 */
+	rcu_read_lock();
 	for (i = 0; i < core_num_wrps; ++i) {
-		rcu_read_lock();
-
 		wp = slots[i];
-
 		if (wp == NULL)
-			goto unlock;
+			continue;
 
-		info = counter_arch_bp(wp);
 		/*
 		 * The DFAR is an unknown value on debug architectures prior
 		 * to 7.1. Since we only allow a single watchpoint on these
@@ -717,33 +753,31 @@ static void watchpoint_handler(unsigned long addr, unsigned int fsr,
 		 */
 		if (debug_arch < ARM_DEBUG_ARCH_V7_1) {
 			BUG_ON(i > 0);
+			info = counter_arch_bp(wp);
 			info->trigger = wp->attr.bp_addr;
 		} else {
-			if (info->ctrl.len == ARM_BREAKPOINT_LEN_8)
-				alignment_mask = 0x7;
-			else
-				alignment_mask = 0x3;
-
-			/* Check if the watchpoint value matches. */
-			val = read_wb_reg(ARM_BASE_WVR + i);
-			if (val != (addr & ~alignment_mask))
-				goto unlock;
-
-			/* Possible match, check the byte address select. */
-			ctrl_reg = read_wb_reg(ARM_BASE_WCR + i);
-			decode_ctrl_reg(ctrl_reg, &ctrl);
-			if (!((1 << (addr & alignment_mask)) & ctrl.len))
-				goto unlock;
-
 			/* Check that the access type matches. */
 			if (debug_exception_updates_fsr()) {
 				access = (fsr & ARM_FSR_ACCESS_MASK) ?
 					  HW_BREAKPOINT_W : HW_BREAKPOINT_R;
 				if (!(access & hw_breakpoint_type(wp)))
-					goto unlock;
+					continue;
 			}
 
+			val = read_wb_reg(ARM_BASE_WVR + i);
+			ctrl_reg = read_wb_reg(ARM_BASE_WCR + i);
+			decode_ctrl_reg(ctrl_reg, &ctrl);
+			dist = get_distance_from_watchpoint(addr, val, &ctrl);
+			if (dist < min_dist) {
+				min_dist = dist;
+				closest_match = i;
+			}
+			/* Is this an exact match? */
+			if (dist != 0)
+				continue;
+
 			/* We have a winner. */
+			info = counter_arch_bp(wp);
 			info->trigger = addr;
 		}
 
@@ -765,13 +799,23 @@ static void watchpoint_handler(unsigned long addr, unsigned int fsr,
 		 * we can single-step over the watchpoint trigger.
 		 */
 		if (!is_default_overflow_handler(wp))
-			goto unlock;
-
+			continue;
 step:
 		enable_single_step(wp, instruction_pointer(regs));
-unlock:
-		rcu_read_unlock();
 	}
+
+	if (min_dist > 0 && min_dist != -1) {
+		/* No exact match found. */
+		wp = slots[closest_match];
+		info = counter_arch_bp(wp);
+		info->trigger = addr;
+		pr_debug("watchpoint fired: address = 0x%x\n", info->trigger);
+		perf_bp_event(wp, regs);
+		if (is_default_overflow_handler(wp))
+			enable_single_step(wp, instruction_pointer(regs));
+	}
+
+	rcu_read_unlock();
 }
 
 static void watchpoint_single_step_handler(unsigned long pc)