s390/dumpstack: fix call chain walking

dumpstack() did not always print a sane callchain when being called.
The reason is that show_trace() accessed register 15 directly to get
the current stack pointer and passed that pointer to __show_trace()
which expects a valid stack frame pointer as argument.
However due to tail call optimization the stack frame may not exist
anymore when __show_trace() gets called and therefore an invalid
stack frame pointer gets passed.
To prevent that disable tail call optimization for call chain walking
functions.
So move all the show_* functions to a dumpstack.c file like other
architectures have it already and add a -fno-optimize-sibling-calls
compile flag to both dumpstack.c and stacktrace.c to prevent tail
call optimization.

Fixes callchains that looked e.g. like this:

[   12.868258] Call Trace:
[   12.868262] ([<0000000000008000>] 0x8000)

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>

arch/s390/kernel/Makefile

@@ -13,17 +13,26 @@ endif
 #
 CFLAGS_smp.o	:= -Wno-nonnull
 
+#
+# Disable tailcall optimizations for stack / callchain walking functions
+# since this might generate broken code when accessing register 15 and
+# passing its content to other functions.
+#
+CFLAGS_stacktrace.o	+= -fno-optimize-sibling-calls
+CFLAGS_dumpstack.o	+= -fno-optimize-sibling-calls
+
 #
 # Pass UTS_MACHINE for user_regset definition
 #
 CFLAGS_ptrace.o		+= -DUTS_MACHINE='"$(UTS_MACHINE)"'
 
 CFLAGS_sysinfo.o += -Iinclude/math-emu -Iarch/s390/math-emu -w
 
-obj-y	:=  bitmap.o traps.o time.o process.o base.o early.o setup.o vtime.o \
-	    processor.o sys_s390.o ptrace.o signal.o cpcmd.o ebcdic.o nmi.o \
-	    debug.o irq.o ipl.o dis.o diag.o mem_detect.o sclp.o vdso.o \
-	    sysinfo.o jump_label.o lgr.o os_info.o machine_kexec.o pgm_check.o
+obj-y	:= bitmap.o traps.o time.o process.o base.o early.o setup.o vtime.o
+obj-y	+= processor.o sys_s390.o ptrace.o signal.o cpcmd.o ebcdic.o nmi.o
+obj-y	+= debug.o irq.o ipl.o dis.o diag.o mem_detect.o sclp.o vdso.o
+obj-y	+= sysinfo.o jump_label.o lgr.o os_info.o machine_kexec.o pgm_check.o
+obj-y	+= dumpstack.o
 
 obj-y	+= $(if $(CONFIG_64BIT),entry64.o,entry.o)
 obj-y	+= $(if $(CONFIG_64BIT),reipl64.o,reipl.o)

arch/s390/kernel/dumpstack.c

@@ -0,0 +1,236 @@
+/*
+ * Stack dumping functions
+ *
+ *  Copyright IBM Corp. 1999, 2013
+ */
+
+#include <linux/kallsyms.h>
+#include <linux/hardirq.h>
+#include <linux/kprobes.h>
+#include <linux/utsname.h>
+#include <linux/export.h>
+#include <linux/kdebug.h>
+#include <linux/ptrace.h>
+#include <linux/module.h>
+#include <linux/sched.h>
+#include <asm/processor.h>
+#include <asm/debug.h>
+#include <asm/ipl.h>
+
+#ifndef CONFIG_64BIT
+#define LONG "%08lx "
+#define FOURLONG "%08lx %08lx %08lx %08lx\n"
+static int kstack_depth_to_print = 12;
+#else /* CONFIG_64BIT */
+#define LONG "%016lx "
+#define FOURLONG "%016lx %016lx %016lx %016lx\n"
+static int kstack_depth_to_print = 20;
+#endif /* CONFIG_64BIT */
+
+/*
+ * For show_trace we have tree different stack to consider:
+ *   - the panic stack which is used if the kernel stack has overflown
+ *   - the asynchronous interrupt stack (cpu related)
+ *   - the synchronous kernel stack (process related)
+ * The stack trace can start at any of the three stack and can potentially
+ * touch all of them. The order is: panic stack, async stack, sync stack.
+ */
+static unsigned long
+__show_trace(unsigned long sp, unsigned long low, unsigned long high)
+{
+	struct stack_frame *sf;
+	struct pt_regs *regs;
+
+	while (1) {
+		sp = sp & PSW_ADDR_INSN;
+		if (sp < low || sp > high - sizeof(*sf))
+			return sp;
+		sf = (struct stack_frame *) sp;
+		printk("([<%016lx>] ", sf->gprs[8] & PSW_ADDR_INSN);
+		print_symbol("%s)\n", sf->gprs[8] & PSW_ADDR_INSN);
+		/* Follow the backchain. */
+		while (1) {
+			low = sp;
+			sp = sf->back_chain & PSW_ADDR_INSN;
+			if (!sp)
+				break;
+			if (sp <= low || sp > high - sizeof(*sf))
+				return sp;
+			sf = (struct stack_frame *) sp;
+			printk(" [<%016lx>] ", sf->gprs[8] & PSW_ADDR_INSN);
+			print_symbol("%s\n", sf->gprs[8] & PSW_ADDR_INSN);
+		}
+		/* Zero backchain detected, check for interrupt frame. */
+		sp = (unsigned long) (sf + 1);
+		if (sp <= low || sp > high - sizeof(*regs))
+			return sp;
+		regs = (struct pt_regs *) sp;
+		printk(" [<%016lx>] ", regs->psw.addr & PSW_ADDR_INSN);
+		print_symbol("%s\n", regs->psw.addr & PSW_ADDR_INSN);
+		low = sp;
+		sp = regs->gprs[15];
+	}
+}
+
+static void show_trace(struct task_struct *task, unsigned long *stack)
+{
+	register unsigned long __r15 asm ("15");
+	unsigned long sp;
+
+	sp = (unsigned long) stack;
+	if (!sp)
+		sp = task ? task->thread.ksp : __r15;
+	printk("Call Trace:\n");
+#ifdef CONFIG_CHECK_STACK
+	sp = __show_trace(sp, S390_lowcore.panic_stack - 4096,
+			  S390_lowcore.panic_stack);
+#endif
+	sp = __show_trace(sp, S390_lowcore.async_stack - ASYNC_SIZE,
+			  S390_lowcore.async_stack);
+	if (task)
+		__show_trace(sp, (unsigned long) task_stack_page(task),
+			     (unsigned long) task_stack_page(task) + THREAD_SIZE);
+	else
+		__show_trace(sp, S390_lowcore.thread_info,
+			     S390_lowcore.thread_info + THREAD_SIZE);
+	if (!task)
+		task = current;
+	debug_show_held_locks(task);
+}
+
+void show_stack(struct task_struct *task, unsigned long *sp)
+{
+	register unsigned long *__r15 asm ("15");
+	unsigned long *stack;
+	int i;
+
+	if (!sp)
+		stack = task ? (unsigned long *) task->thread.ksp : __r15;
+	else
+		stack = sp;
+
+	for (i = 0; i < kstack_depth_to_print; i++) {
+		if (((addr_t) stack & (THREAD_SIZE-1)) == 0)
+			break;
+		if ((i * sizeof(long) % 32) == 0)
+			printk("%s       ", i == 0 ? "" : "\n");
+		printk(LONG, *stack++);
+	}
+	printk("\n");
+	show_trace(task, sp);
+}
+
+static void show_last_breaking_event(struct pt_regs *regs)
+{
+#ifdef CONFIG_64BIT
+	printk("Last Breaking-Event-Address:\n");
+	printk(" [<%016lx>] ", regs->args[0] & PSW_ADDR_INSN);
+	print_symbol("%s\n", regs->args[0] & PSW_ADDR_INSN);
+#endif
+}
+
+/*
+ * The architecture-independent dump_stack generator
+ */
+void dump_stack(void)
+{
+	printk("CPU: %d %s %s %.*s\n",
+	       task_thread_info(current)->cpu, print_tainted(),
+	       init_utsname()->release,
+	       (int)strcspn(init_utsname()->version, " "),
+	       init_utsname()->version);
+	printk("Process %s (pid: %d, task: %p, ksp: %p)\n",
+	       current->comm, current->pid, current,
+	       (void *) current->thread.ksp);
+	show_stack(NULL, NULL);
+}
+EXPORT_SYMBOL(dump_stack);
+
+static inline int mask_bits(struct pt_regs *regs, unsigned long bits)
+{
+	return (regs->psw.mask & bits) / ((~bits + 1) & bits);
+}
+
+void show_registers(struct pt_regs *regs)
+{
+	char *mode;
+
+	mode = user_mode(regs) ? "User" : "Krnl";
+	printk("%s PSW : %p %p",
+	       mode, (void *) regs->psw.mask,
+	       (void *) regs->psw.addr);
+	print_symbol(" (%s)\n", regs->psw.addr & PSW_ADDR_INSN);
+	printk("           R:%x T:%x IO:%x EX:%x Key:%x M:%x W:%x "
+	       "P:%x AS:%x CC:%x PM:%x", mask_bits(regs, PSW_MASK_PER),
+	       mask_bits(regs, PSW_MASK_DAT), mask_bits(regs, PSW_MASK_IO),
+	       mask_bits(regs, PSW_MASK_EXT), mask_bits(regs, PSW_MASK_KEY),
+	       mask_bits(regs, PSW_MASK_MCHECK), mask_bits(regs, PSW_MASK_WAIT),
+	       mask_bits(regs, PSW_MASK_PSTATE), mask_bits(regs, PSW_MASK_ASC),
+	       mask_bits(regs, PSW_MASK_CC), mask_bits(regs, PSW_MASK_PM));
+#ifdef CONFIG_64BIT
+	printk(" EA:%x", mask_bits(regs, PSW_MASK_EA | PSW_MASK_BA));
+#endif
+	printk("\n%s GPRS: " FOURLONG, mode,
+	       regs->gprs[0], regs->gprs[1], regs->gprs[2], regs->gprs[3]);
+	printk("           " FOURLONG,
+	       regs->gprs[4], regs->gprs[5], regs->gprs[6], regs->gprs[7]);
+	printk("           " FOURLONG,
+	       regs->gprs[8], regs->gprs[9], regs->gprs[10], regs->gprs[11]);
+	printk("           " FOURLONG,
+	       regs->gprs[12], regs->gprs[13], regs->gprs[14], regs->gprs[15]);
+	show_code(regs);
+}
+
+void show_regs(struct pt_regs *regs)
+{
+	printk("CPU: %d %s %s %.*s\n",
+	       task_thread_info(current)->cpu, print_tainted(),
+	       init_utsname()->release,
+	       (int)strcspn(init_utsname()->version, " "),
+	       init_utsname()->version);
+	printk("Process %s (pid: %d, task: %p, ksp: %p)\n",
+	       current->comm, current->pid, current,
+	       (void *) current->thread.ksp);
+	show_registers(regs);
+	/* Show stack backtrace if pt_regs is from kernel mode */
+	if (!user_mode(regs))
+		show_trace(NULL, (unsigned long *) regs->gprs[15]);
+	show_last_breaking_event(regs);
+}
+
+static DEFINE_SPINLOCK(die_lock);
+
+void die(struct pt_regs *regs, const char *str)
+{
+	static int die_counter;
+
+	oops_enter();
+	lgr_info_log();
+	debug_stop_all();
+	console_verbose();
+	spin_lock_irq(&die_lock);
+	bust_spinlocks(1);
+	printk("%s: %04x [#%d] ", str, regs->int_code & 0xffff, ++die_counter);
+#ifdef CONFIG_PREEMPT
+	printk("PREEMPT ");
+#endif
+#ifdef CONFIG_SMP
+	printk("SMP ");
+#endif
+#ifdef CONFIG_DEBUG_PAGEALLOC
+	printk("DEBUG_PAGEALLOC");
+#endif
+	printk("\n");
+	notify_die(DIE_OOPS, str, regs, 0, regs->int_code & 0xffff, SIGSEGV);
+	print_modules();
+	show_regs(regs);
+	bust_spinlocks(0);
+	add_taint(TAINT_DIE, LOCKDEP_NOW_UNRELIABLE);
+	spin_unlock_irq(&die_lock);
+	if (in_interrupt())
+		panic("Fatal exception in interrupt");
+	if (panic_on_oops)
+		panic("Fatal exception: panic_on_oops");
+	oops_exit();
+	do_exit(SIGSEGV);
+}