[PATCH] powerpc: Add arch dependent basic infrastructure for Kdump.

Implementing the machine_crash_shutdown which will be called by
crash_kexec (called in case of a panic, sysrq etc.). Disable the
interrupts, shootdown cpus using debugger IPI and collect regs
for all CPUs.

elfcorehdr= specifies the location of elf core header stored by
the crashed kernel. This command line option will be passed by
the kexec-tools to capture kernel.

savemaxmem= specifies the actual memory size that the first kernel
has and this value will be used for dumping in the capture kernel.
This command line option will be passed by the kexec-tools to
capture kernel.

Signed-off-by: Haren Myneni <haren@us.ibm.com>
Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Paul Mackerras <paulus@samba.org>

arch/powerpc/kernel/Makefile

@@ -66,7 +66,7 @@ pci64-$(CONFIG_PPC64)		+= pci_64.o pci_dn.o pci_iommu.o \
 obj-$(CONFIG_PCI)		+= $(pci64-y)
 kexec-$(CONFIG_PPC64)		:= machine_kexec_64.o
 kexec-$(CONFIG_PPC32)		:= machine_kexec_32.o
-obj-$(CONFIG_KEXEC)		+= machine_kexec.o $(kexec-y)
+obj-$(CONFIG_KEXEC)		+= machine_kexec.o crash.o $(kexec-y)
 
 ifeq ($(CONFIG_PPC_ISERIES),y)
 $(obj)/head_64.o: $(obj)/lparmap.s

arch/powerpc/kernel/crash.c

@@ -0,0 +1,264 @@
+/*
+ * Architecture specific (PPC64) functions for kexec based crash dumps.
+ *
+ * Copyright (C) 2005, IBM Corp.
+ *
+ * Created by: Haren Myneni
+ *
+ * This source code is licensed under the GNU General Public License,
+ * Version 2.  See the file COPYING for more details.
+ *
+ */
+
+#undef DEBUG
+
+#include <linux/kernel.h>
+#include <linux/smp.h>
+#include <linux/reboot.h>
+#include <linux/kexec.h>
+#include <linux/bootmem.h>
+#include <linux/crash_dump.h>
+#include <linux/irq.h>
+#include <linux/delay.h>
+#include <linux/elf.h>
+#include <linux/elfcore.h>
+#include <linux/init.h>
+#include <linux/types.h>
+
+#include <asm/processor.h>
+#include <asm/machdep.h>
+#include <asm/kdump.h>
+#include <asm/lmb.h>
+#include <asm/firmware.h>
+
+#ifdef DEBUG
+#include <asm/udbg.h>
+#define DBG(fmt...) udbg_printf(fmt)
+#else
+#define DBG(fmt...)
+#endif
+
+/* This keeps a track of which one is crashing cpu. */
+int crashing_cpu = -1;
+
+static u32 *append_elf_note(u32 *buf, char *name, unsigned type, void *data,
+							       size_t data_len)
+{
+	struct elf_note note;
+
+	note.n_namesz = strlen(name) + 1;
+	note.n_descsz = data_len;
+	note.n_type   = type;
+	memcpy(buf, &note, sizeof(note));
+	buf += (sizeof(note) +3)/4;
+	memcpy(buf, name, note.n_namesz);
+	buf += (note.n_namesz + 3)/4;
+	memcpy(buf, data, note.n_descsz);
+	buf += (note.n_descsz + 3)/4;
+
+	return buf;
+}
+
+static void final_note(u32 *buf)
+{
+	struct elf_note note;
+
+	note.n_namesz = 0;
+	note.n_descsz = 0;
+	note.n_type   = 0;
+	memcpy(buf, &note, sizeof(note));
+}
+
+static void crash_save_this_cpu(struct pt_regs *regs, int cpu)
+{
+	struct elf_prstatus prstatus;
+	u32 *buf;
+
+	if ((cpu < 0) || (cpu >= NR_CPUS))
+		return;
+
+	/* Using ELF notes here is opportunistic.
+	 * I need a well defined structure format
+	 * for the data I pass, and I need tags
+	 * on the data to indicate what information I have
+	 * squirrelled away.  ELF notes happen to provide
+	 * all of that that no need to invent something new.
+	 */
+	buf = &crash_notes[cpu][0];
+	memset(&prstatus, 0, sizeof(prstatus));
+	prstatus.pr_pid = current->pid;
+	elf_core_copy_regs(&prstatus.pr_reg, regs);
+	buf = append_elf_note(buf, "CORE", NT_PRSTATUS, &prstatus,
+			sizeof(prstatus));
+	final_note(buf);
+}
+
+/* FIXME Merge this with xmon_save_regs ?? */
+static inline void crash_get_current_regs(struct pt_regs *regs)
+{
+	unsigned long tmp1, tmp2;
+
+	__asm__ __volatile__ (
+		"std    0,0(%2)\n"
+		"std    1,8(%2)\n"
+		"std    2,16(%2)\n"
+		"std    3,24(%2)\n"
+		"std    4,32(%2)\n"
+		"std    5,40(%2)\n"
+		"std    6,48(%2)\n"
+		"std    7,56(%2)\n"
+		"std    8,64(%2)\n"
+		"std    9,72(%2)\n"
+		"std    10,80(%2)\n"
+		"std    11,88(%2)\n"
+		"std    12,96(%2)\n"
+		"std    13,104(%2)\n"
+		"std    14,112(%2)\n"
+		"std    15,120(%2)\n"
+		"std    16,128(%2)\n"
+		"std    17,136(%2)\n"
+		"std    18,144(%2)\n"
+		"std    19,152(%2)\n"
+		"std    20,160(%2)\n"
+		"std    21,168(%2)\n"
+		"std    22,176(%2)\n"
+		"std    23,184(%2)\n"
+		"std    24,192(%2)\n"
+		"std    25,200(%2)\n"
+		"std    26,208(%2)\n"
+		"std    27,216(%2)\n"
+		"std    28,224(%2)\n"
+		"std    29,232(%2)\n"
+		"std    30,240(%2)\n"
+		"std    31,248(%2)\n"
+		"mfmsr  %0\n"
+		"std    %0, 264(%2)\n"
+		"mfctr  %0\n"
+		"std    %0, 280(%2)\n"
+		"mflr   %0\n"
+		"std    %0, 288(%2)\n"
+		"bl     1f\n"
+	"1:      mflr   %1\n"
+		"std    %1, 256(%2)\n"
+		"mtlr   %0\n"
+		"mfxer  %0\n"
+		"std    %0, 296(%2)\n"
+		: "=&r" (tmp1), "=&r" (tmp2)
+		: "b" (regs));
+}
+
+/* We may have saved_regs from where the error came from
+ * or it is NULL if via a direct panic().
+ */
+static void crash_save_self(struct pt_regs *saved_regs)
+{
+	struct pt_regs regs;
+	int cpu;
+
+	cpu = smp_processor_id();
+	if (saved_regs)
+		memcpy(&regs, saved_regs, sizeof(regs));
+	else
+		crash_get_current_regs(&regs);
+	crash_save_this_cpu(&regs, cpu);
+}
+
+#ifdef CONFIG_SMP
+static atomic_t waiting_for_crash_ipi;
+
+void crash_ipi_callback(struct pt_regs *regs)
+{
+	int cpu = smp_processor_id();
+
+	if (cpu == crashing_cpu)
+		return;
+
+	if (!cpu_online(cpu))
+		return;
+
+	if (ppc_md.kexec_cpu_down)
+		ppc_md.kexec_cpu_down(1, 1);
+
+	local_irq_disable();
+
+	crash_save_this_cpu(regs, cpu);
+	atomic_dec(&waiting_for_crash_ipi);
+	kexec_smp_wait();
+	/* NOTREACHED */
+}
+
+static void crash_kexec_prepare_cpus(void)
+{
+	unsigned int msecs;
+
+	atomic_set(&waiting_for_crash_ipi, num_online_cpus() - 1);
+
+	crash_send_ipi(crash_ipi_callback);
+	smp_wmb();
+
+	/*
+	 * FIXME: Until we will have the way to stop other CPUSs reliabally,
+	 * the crash CPU will send an IPI and wait for other CPUs to
+	 * respond. If not, proceed the kexec boot even though we failed to
+	 * capture other CPU states.
+	 */
+	msecs = 1000000;
+	while ((atomic_read(&waiting_for_crash_ipi) > 0) && (--msecs > 0)) {
+		barrier();
+		mdelay(1);
+	}
+
+	/* Would it be better to replace the trap vector here? */
+
+	/*
+	 * FIXME: In case if we do not get all CPUs, one possibility: ask the
+	 * user to do soft reset such that we get all.
+	 * IPI handler is already set by the panic cpu initially. Therefore,
+	 * all cpus could invoke this handler from die() and the panic CPU
+	 * will call machine_kexec() directly from this handler to do
+	 * kexec boot.
+	 */
+	if (atomic_read(&waiting_for_crash_ipi))
+		printk(KERN_ALERT "done waiting: %d cpus not responding\n",
+			atomic_read(&waiting_for_crash_ipi));
+	/* Leave the IPI callback set */
+}
+#else
+static void crash_kexec_prepare_cpus(void)
+{
+	/*
+	 * move the secondarys to us so that we can copy
+	 * the new kernel 0-0x100 safely
+	 *
+	 * do this if kexec in setup.c ?
+	 */
+	smp_release_cpus();
+}
+
+#endif
+
+void default_machine_crash_shutdown(struct pt_regs *regs)
+{
+	/*
+	 * This function is only called after the system
+	 * has paniced or is otherwise in a critical state.
+	 * The minimum amount of code to allow a kexec'd kernel
+	 * to run successfully needs to happen here.
+	 *
+	 * In practice this means stopping other cpus in
+	 * an SMP system.
+	 * The kernel is broken so disable interrupts.
+	 */
+	local_irq_disable();
+
+	if (ppc_md.kexec_cpu_down)
+		ppc_md.kexec_cpu_down(1, 0);
+
+	/*
+	 * Make a note of crashing cpu. Will be used in machine_kexec
+	 * such that another IPI will not be sent.
+	 */
+	crashing_cpu = smp_processor_id();
+	crash_kexec_prepare_cpus();
+	crash_save_self(regs);
+}

arch/powerpc/kernel/crash_dump.c

@@ -11,6 +11,8 @@
 
 #undef DEBUG
 
+#include <linux/crash_dump.h>
+#include <linux/bootmem.h>
 #include <asm/kdump.h>
 #include <asm/lmb.h>
 #include <asm/firmware.h>
@@ -51,3 +53,21 @@ void __init kdump_setup(void)
 
 	DBG(" <- kdump_setup()\n");
 }
+
+static int __init parse_elfcorehdr(char *p)
+{
+	if (p)
+		elfcorehdr_addr = memparse(p, &p);
+
+	return 0;
+}
+__setup("elfcorehdr=", parse_elfcorehdr);
+
+static int __init parse_savemaxmem(char *p)
+{
+	if (p)
+		saved_max_pfn = (memparse(p, &p) >> PAGE_SHIFT) - 1;
+
+	return 0;
+}
+__setup("savemaxmem=", parse_savemaxmem);

arch/powerpc/kernel/machine_kexec_64.c

@@ -265,11 +265,18 @@ extern NORET_TYPE void kexec_sequence(void *newstack, unsigned long start,
 /* too late to fail here */
 void default_machine_kexec(struct kimage *image)
 {
-
 	/* prepare control code if any */
 
-	/* shutdown other cpus into our wait loop and quiesce interrupts */
-	kexec_prepare_cpus();
+	/*
+        * If the kexec boot is the normal one, need to shutdown other cpus
+        * into our wait loop and quiesce interrupts.
+        * Otherwise, in the case of crashed mode (crashing_cpu >= 0),
+        * stopping other CPUs and collecting their pt_regs is done before
+        * using debugger IPI.
+        */
+
+       if (crashing_cpu == -1)
+               kexec_prepare_cpus();
 
 	/* switch to a staticly allocated stack.  Based on irq stack code.
 	 * XXX: the task struct will likely be invalid once we do the copy!

arch/powerpc/kernel/smp.c

@@ -75,6 +75,8 @@ void smp_call_function_interrupt(void);
 
 int smt_enabled_at_boot = 1;
 
+static void (*crash_ipi_function_ptr)(struct pt_regs *) = NULL;
+
 #ifdef CONFIG_MPIC
 int __init smp_mpic_probe(void)
 {
@@ -123,11 +125,16 @@ void smp_message_recv(int msg, struct pt_regs *regs)
 		/* XXX Do we have to do this? */
 		set_need_resched();
 		break;
-#ifdef CONFIG_DEBUGGER
 	case PPC_MSG_DEBUGGER_BREAK:
+		if (crash_ipi_function_ptr) {
+			crash_ipi_function_ptr(regs);
+			break;
+		}
+#ifdef CONFIG_DEBUGGER
 		debugger_ipi(regs);
 		break;
-#endif
+#endif /* CONFIG_DEBUGGER */
+		/* FALLTHROUGH */
 	default:
 		printk("SMP %d: smp_message_recv(): unknown msg %d\n",
 		       smp_processor_id(), msg);
@@ -147,6 +154,17 @@ void smp_send_debugger_break(int cpu)
 }
 #endif
 
+#ifdef CONFIG_KEXEC
+void crash_send_ipi(void (*crash_ipi_callback)(struct pt_regs *))
+{
+	crash_ipi_function_ptr = crash_ipi_callback;
+	if (crash_ipi_callback) {
+		mb();
+		smp_ops->message_pass(MSG_ALL_BUT_SELF, PPC_MSG_DEBUGGER_BREAK);
+	}
+}
+#endif
+
 static void stop_this_cpu(void *dummy)
 {
 	local_irq_disable();

arch/powerpc/kernel/traps.c

@@ -31,6 +31,7 @@
 #include <linux/prctl.h>
 #include <linux/delay.h>
 #include <linux/kprobes.h>
+#include <linux/kexec.h>
 
 #include <asm/kdebug.h>
 #include <asm/pgtable.h>
@@ -95,7 +96,7 @@ static DEFINE_SPINLOCK(die_lock);
 
 int die(const char *str, struct pt_regs *regs, long err)
 {
-	static int die_counter;
+	static int die_counter, crash_dump_start = 0;
 	int nl = 0;
 
 	if (debugger(regs))
@@ -156,7 +157,21 @@ int die(const char *str, struct pt_regs *regs, long err)
 	print_modules();
 	show_regs(regs);
 	bust_spinlocks(0);
+
+	if (!crash_dump_start && kexec_should_crash(current)) {
+		crash_dump_start = 1;
+		spin_unlock_irq(&die_lock);
+		crash_kexec(regs);
+		/* NOTREACHED */
+	}
 	spin_unlock_irq(&die_lock);
+	if (crash_dump_start)
+		/*
+		 * Only for soft-reset: Other CPUs will be responded to an IPI
+		 * sent by first kexec CPU.
+		 */
+		for(;;)
+			;
 
 	if (in_interrupt())
 		panic("Fatal exception in interrupt");

arch/powerpc/platforms/cell/setup.c

@@ -217,5 +217,6 @@ struct machdep_calls __initdata cell_md = {
 #ifdef CONFIG_KEXEC
 	.machine_kexec		= default_machine_kexec,
 	.machine_kexec_prepare	= default_machine_kexec_prepare,
+	.machine_crash_shutdown	= default_machine_crash_shutdown,
 #endif
 };

arch/powerpc/platforms/maple/setup.c

@@ -282,5 +282,6 @@ struct machdep_calls __initdata maple_md = {
 #ifdef CONFIG_KEXEC
 	.machine_kexec		= default_machine_kexec,
 	.machine_kexec_prepare	= default_machine_kexec_prepare,
+	.machine_crash_shutdown	= default_machine_crash_shutdown,
 #endif
 };

arch/powerpc/platforms/powermac/setup.c

@@ -771,6 +771,7 @@ struct machdep_calls __initdata pmac_md = {
 #ifdef CONFIG_KEXEC
 	.machine_kexec		= default_machine_kexec,
 	.machine_kexec_prepare	= default_machine_kexec_prepare,
+	.machine_crash_shutdown	= default_machine_crash_shutdown,
 #endif
 #endif /* CONFIG_PPC64 */
 #ifdef CONFIG_PPC32

arch/powerpc/platforms/pseries/setup.c

@@ -629,5 +629,6 @@ struct machdep_calls __initdata pSeries_md = {
 	.kexec_cpu_down		= pseries_kexec_cpu_down,
 	.machine_kexec		= default_machine_kexec,
 	.machine_kexec_prepare	= default_machine_kexec_prepare,
+	.machine_crash_shutdown	= default_machine_crash_shutdown,
 #endif
 };

arch/powerpc/platforms/pseries/xics.c

@@ -407,7 +407,7 @@ irqreturn_t xics_ipi_action(int irq, void *dev_id, struct pt_regs *regs)
 			smp_message_recv(PPC_MSG_MIGRATE_TASK, regs);
 		}
 #endif
-#ifdef CONFIG_DEBUGGER
+#if defined(CONFIG_DEBUGGER) || defined(CONFIG_KEXEC)
 		if (test_and_clear_bit(PPC_MSG_DEBUGGER_BREAK,
 				       &xics_ipi_message[cpu].value)) {
 			mb();

include/asm-powerpc/kexec.h

@@ -32,6 +32,8 @@
 
 #ifndef __ASSEMBLY__
 
+#ifdef CONFIG_KEXEC
+
 #define MAX_NOTE_BYTES 1024
 typedef u32 note_buf_t[MAX_NOTE_BYTES / sizeof(u32)];
 
@@ -41,11 +43,17 @@ extern note_buf_t crash_notes[];
 extern void kexec_smp_wait(void);	/* get and clear naca physid, wait for
 					  master to copy new code to 0 */
 extern void __init kexec_setup(void);
-#endif
+extern int crashing_cpu;
+extern void crash_send_ipi(void (*crash_ipi_callback)(struct pt_regs *));
+#endif /* __powerpc64 __ */
 
 struct kimage;
+struct pt_regs;
 extern void default_machine_kexec(struct kimage *image);
 extern int default_machine_kexec_prepare(struct kimage *image);
+extern void default_machine_crash_shutdown(struct pt_regs *regs);
+
+#endif /* !CONFIG_KEXEC */
 
 #endif /* ! __ASSEMBLY__ */
 #endif /* _ASM_POWERPC_KEXEC_H */