[PATCH 1/5] powerpc: introduce a new helper to obtain function entry points

[PATCH 1/5] powerpc: introduce a new helper to obtain function entry points

[Naveen N. Rao]

kprobe_lookup_name() is specific to the kprobe subsystem and may not
always return the function entry point (in a subsequent patch for
KPROBES_ON_FTRACE). For looking up function entry points, introduce a
separate helper and use the same in optprobes.c

Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
---
 arch/powerpc/include/asm/code-patching.h | 37 ++++++++++++++++++++++++++++++++
 arch/powerpc/kernel/optprobes.c          |  6 +++---
 2 files changed, 40 insertions(+), 3 deletions(-)

diff --git a/arch/powerpc/include/asm/code-patching.h b/arch/powerpc/include/asm/code-patching.h
index 8ab937771068..3e994f404434 100644
--- a/arch/powerpc/include/asm/code-patching.h
+++ b/arch/powerpc/include/asm/code-patching.h
@@ -12,6 +12,8 @@
 
 #include <asm/types.h>
 #include <asm/ppc-opcode.h>
+#include <linux/string.h>
+#include <linux/kallsyms.h>
 
 /* Flags for create_branch:
  * "b"   == create_branch(addr, target, 0);
@@ -99,6 +101,41 @@ static inline unsigned long ppc_global_function_entry(void *func)
 #endif
 }
 
+/*
+ * Wrapper around kallsyms_lookup() to return function entry address:
+ * - For ABIv1, we lookup the dot variant.
+ * - For ABIv2, we return the local entry point.
+ */
+static inline unsigned long ppc_kallsyms_lookup_name(const char *name)
+{
+	unsigned long addr;
+#ifdef PPC64_ELF_ABI_v1
+	/* check for dot variant */
+	char dot_name[1 + KSYM_NAME_LEN];
+	bool dot_appended = false;
+	if (name[0] != '.') {
+		dot_name[0] = '.';
+		dot_name[1] = '\0';
+		strncat(dot_name, name, KSYM_NAME_LEN - 2);
+		dot_appended = true;
+	} else {
+		dot_name[0] = '\0';
+		strncat(dot_name, name, KSYM_NAME_LEN - 1);
+	}
+	addr = kallsyms_lookup_name(dot_name);
+	if (!addr && dot_appended)
+		/* Let's try the original non-dot symbol lookup	*/
+		addr = kallsyms_lookup_name(name);
+#elif defined(PPC64_ELF_ABI_v2)
+	addr = kallsyms_lookup_name(name);
+	if (addr)
+		addr = ppc_function_entry((void *)addr);
+#else
+	addr = kallsyms_lookup_name(name);
+#endif
+	return addr;
+}
+
 #ifdef CONFIG_PPC64
 /*
  * Some instruction encodings commonly used in dynamic ftracing
diff --git a/arch/powerpc/kernel/optprobes.c b/arch/powerpc/kernel/optprobes.c
index 2282bf4e63cd..ec60ed0d4aad 100644
--- a/arch/powerpc/kernel/optprobes.c
+++ b/arch/powerpc/kernel/optprobes.c
@@ -243,10 +243,10 @@ int arch_prepare_optimized_kprobe(struct optimized_kprobe *op, struct kprobe *p)
 	/*
 	 * 2. branch to optimized_callback() and emulate_step()
 	 */
-	kprobe_lookup_name("optimized_callback", op_callback_addr);
-	kprobe_lookup_name("emulate_step", emulate_step_addr);
+	op_callback_addr = (kprobe_opcode_t *)ppc_kallsyms_lookup_name("optimized_callback");
+	emulate_step_addr = (kprobe_opcode_t *)ppc_kallsyms_lookup_name("emulate_step");
 	if (!op_callback_addr || !emulate_step_addr) {
-		WARN(1, "kprobe_lookup_name() failed\n");
+		WARN(1, "Unable to lookup optimized_callback()/emulate_step()\n");
 		goto error;
 	}
 
-- 
2.11.0

[PATCH 2/5] kprobes: convert kprobe_lookup_name() to a function

[Naveen N. Rao]

The macro is now pretty long and ugly on powerpc. In the light of
further changes needed here, convert it to a __weak variant to be
over-ridden with a nicer looking function.

Suggested-by: Masami Hiramatsu <mhiramat@kernel.org>
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
---
 arch/powerpc/include/asm/kprobes.h | 53 ----------------------------------
 arch/powerpc/kernel/kprobes.c      | 58 ++++++++++++++++++++++++++++++++++++++
 include/linux/kprobes.h            |  1 +
 kernel/kprobes.c                   | 20 ++++++-------
 4 files changed, 67 insertions(+), 65 deletions(-)

diff --git a/arch/powerpc/include/asm/kprobes.h b/arch/powerpc/include/asm/kprobes.h
index d821835ade86..ab5bd200bb48 100644
--- a/arch/powerpc/include/asm/kprobes.h
+++ b/arch/powerpc/include/asm/kprobes.h
@@ -58,59 +58,6 @@ extern kprobe_opcode_t optprobe_template_end[];
 #define MAX_OPTINSN_SIZE	(optprobe_template_end - optprobe_template_entry)
 #define RELATIVEJUMP_SIZE	sizeof(kprobe_opcode_t)	/* 4 bytes */
 
-#ifdef PPC64_ELF_ABI_v2
-/* PPC64 ABIv2 needs local entry point */
-#define kprobe_lookup_name(name, addr)					\
-{									\
-	addr = (kprobe_opcode_t *)kallsyms_lookup_name(name);		\
-	if (addr)							\
-		addr = (kprobe_opcode_t *)ppc_function_entry(addr);	\
-}
-#elif defined(PPC64_ELF_ABI_v1)
-/*
- * 64bit powerpc ABIv1 uses function descriptors:
- * - Check for the dot variant of the symbol first.
- * - If that fails, try looking up the symbol provided.
- *
- * This ensures we always get to the actual symbol and not the descriptor.
- * Also handle <module:symbol> format.
- */
-#define kprobe_lookup_name(name, addr)					\
-{									\
-	char dot_name[MODULE_NAME_LEN + 1 + KSYM_NAME_LEN];		\
-	const char *modsym;							\
-	bool dot_appended = false;					\
-	if ((modsym = strchr(name, ':')) != NULL) {			\
-		modsym++;						\
-		if (*modsym != '\0' && *modsym != '.') {		\
-			/* Convert to <module:.symbol> */		\
-			strncpy(dot_name, name, modsym - name);		\
-			dot_name[modsym - name] = '.';			\
-			dot_name[modsym - name + 1] = '\0';		\
-			strncat(dot_name, modsym,			\
-				sizeof(dot_name) - (modsym - name) - 2);\
-			dot_appended = true;				\
-		} else {						\
-			dot_name[0] = '\0';				\
-			strncat(dot_name, name, sizeof(dot_name) - 1);	\
-		}							\
-	} else if (name[0] != '.') {					\
-		dot_name[0] = '.';					\
-		dot_name[1] = '\0';					\
-		strncat(dot_name, name, KSYM_NAME_LEN - 2);		\
-		dot_appended = true;					\
-	} else {							\
-		dot_name[0] = '\0';					\
-		strncat(dot_name, name, KSYM_NAME_LEN - 1);		\
-	}								\
-	addr = (kprobe_opcode_t *)kallsyms_lookup_name(dot_name);	\
-	if (!addr && dot_appended) {					\
-		/* Let's try the original non-dot symbol lookup	*/	\
-		addr = (kprobe_opcode_t *)kallsyms_lookup_name(name);	\
-	}								\
-}
-#endif
-
 #define flush_insn_slot(p)	do { } while (0)
 #define kretprobe_blacklist_size 0
 
diff --git a/arch/powerpc/kernel/kprobes.c b/arch/powerpc/kernel/kprobes.c
index 331751701fed..a7aa7394954d 100644
--- a/arch/powerpc/kernel/kprobes.c
+++ b/arch/powerpc/kernel/kprobes.c
@@ -42,6 +42,64 @@ DEFINE_PER_CPU(struct kprobe_ctlblk, kprobe_ctlblk);
 
 struct kretprobe_blackpoint kretprobe_blacklist[] = {{NULL, NULL}};
 
+kprobe_opcode_t *kprobe_lookup_name(const char *name)
+{
+	kprobe_opcode_t *addr;
+
+#ifdef PPC64_ELF_ABI_v2
+	/* PPC64 ABIv2 needs local entry point */
+	addr = (kprobe_opcode_t *)kallsyms_lookup_name(name);
+	if (addr)
+		addr = (kprobe_opcode_t *)ppc_function_entry(addr);
+#elif defined(PPC64_ELF_ABI_v1)
+	/*
+	 * 64bit powerpc ABIv1 uses function descriptors:
+	 * - Check for the dot variant of the symbol first.
+	 * - If that fails, try looking up the symbol provided.
+	 *
+	 * This ensures we always get to the actual symbol and not
+	 * the descriptor.
+	 *
+	 * Also handle <module:symbol> format.
+	 */
+	char dot_name[MODULE_NAME_LEN + 1 + KSYM_NAME_LEN];
+	const char *modsym;
+	bool dot_appended = false;
+	if ((modsym = strchr(name, ':')) != NULL) {
+		modsym++;
+		if (*modsym != '\0' && *modsym != '.') {
+			/* Convert to <module:.symbol> */
+			strncpy(dot_name, name, modsym - name);
+			dot_name[modsym - name] = '.';
+			dot_name[modsym - name + 1] = '\0';
+			strncat(dot_name, modsym,
+				sizeof(dot_name) - (modsym - name) - 2);
+			dot_appended = true;
+		} else {
+			dot_name[0] = '\0';
+			strncat(dot_name, name, sizeof(dot_name) - 1);
+		}
+	} else if (name[0] != '.') {
+		dot_name[0] = '.';
+		dot_name[1] = '\0';
+		strncat(dot_name, name, KSYM_NAME_LEN - 2);
+		dot_appended = true;
+	} else {
+		dot_name[0] = '\0';
+		strncat(dot_name, name, KSYM_NAME_LEN - 1);
+	}
+	addr = (kprobe_opcode_t *)kallsyms_lookup_name(dot_name);
+	if (!addr && dot_appended) {
+		/* Let's try the original non-dot symbol lookup	*/
+		addr = (kprobe_opcode_t *)kallsyms_lookup_name(name);
+	}
+#else
+	addr = (kprobe_opcode_t *)kallsyms_lookup_name(dot_name);
+#endif
+
+	return addr;
+}
+
 int __kprobes arch_prepare_kprobe(struct kprobe *p)
 {
 	int ret = 0;
diff --git a/include/linux/kprobes.h b/include/linux/kprobes.h
index 0c2489435117..34b7d00765d2 100644
--- a/include/linux/kprobes.h
+++ b/include/linux/kprobes.h
@@ -364,6 +364,7 @@ static inline struct kprobe_ctlblk *get_kprobe_ctlblk(void)
 	return this_cpu_ptr(&kprobe_ctlblk);
 }
 
+kprobe_opcode_t *kprobe_lookup_name(const char *name);
 int register_kprobe(struct kprobe *p);
 void unregister_kprobe(struct kprobe *p);
 int register_kprobes(struct kprobe **kps, int num);
diff --git a/kernel/kprobes.c b/kernel/kprobes.c
index ab35ed802389..30f1d732dd74 100644
--- a/kernel/kprobes.c
+++ b/kernel/kprobes.c
@@ -58,15 +58,6 @@
 #define KPROBE_TABLE_SIZE (1 << KPROBE_HASH_BITS)
 
 
-/*
- * Some oddball architectures like 64bit powerpc have function descriptors
- * so this must be overridable.
- */
-#ifndef kprobe_lookup_name
-#define kprobe_lookup_name(name, addr) \
-	addr = ((kprobe_opcode_t *)(kallsyms_lookup_name(name)))
-#endif
-
 static int kprobes_initialized;
 static struct hlist_head kprobe_table[KPROBE_TABLE_SIZE];
 static struct hlist_head kretprobe_inst_table[KPROBE_TABLE_SIZE];
@@ -81,6 +72,11 @@ static struct {
 	raw_spinlock_t lock ____cacheline_aligned_in_smp;
 } kretprobe_table_locks[KPROBE_TABLE_SIZE];
 
+kprobe_opcode_t * __weak kprobe_lookup_name(const char *name)
+{
+	return ((kprobe_opcode_t *)(kallsyms_lookup_name(name)));
+}
+
 static raw_spinlock_t *kretprobe_table_lock_ptr(unsigned long hash)
 {
 	return &(kretprobe_table_locks[hash].lock);
@@ -1365,7 +1361,7 @@ static kprobe_opcode_t *kprobe_addr(struct kprobe *p)
 		goto invalid;
 
 	if (p->symbol_name) {
-		kprobe_lookup_name(p->symbol_name, addr);
+		addr = kprobe_lookup_name(p->symbol_name);
 		if (!addr)
 			return ERR_PTR(-ENOENT);
 	}
@@ -2170,8 +2166,8 @@ static int __init init_kprobes(void)
 	if (kretprobe_blacklist_size) {
 		/* lookup the function address from its name */
 		for (i = 0; kretprobe_blacklist[i].name != NULL; i++) {
-			kprobe_lookup_name(kretprobe_blacklist[i].name,
-					   kretprobe_blacklist[i].addr);
+			kretprobe_blacklist[i].addr =
+				kprobe_lookup_name(kretprobe_blacklist[i].name);
 			if (!kretprobe_blacklist[i].addr)
 				printk("kretprobe: lookup failed: %s\n",
 				       kretprobe_blacklist[i].name);
-- 
2.11.0

[PATCH 3/5] powerpc: kprobes: fix handling of function offsets on ABIv2

[Naveen N. Rao]

commit 239aeba76409 ("perf powerpc: Fix kprobe and kretprobe handling
with kallsyms on ppc64le") changed how we use the offset field in struct
kprobe on ABIv2. perf now offsets from the GEP (Global entry point) if an
offset is specified and otherwise chooses the LEP (Local entry point).

Fix the same in kernel for kprobe API users. We do this by extending
kprobe_lookup_name() to accept an additional parameter to indicate the
offset specified with the kprobe registration. If offset is 0, we return
the local function entry and return the global entry point otherwise.

With:
	# cd /sys/kernel/debug/tracing/
	# echo "p _do_fork" >> kprobe_events
	# echo "p _do_fork+0x10" >> kprobe_events

before this patch:
	# cat ../kprobes/list
	c0000000000d0748  k  _do_fork+0x8    [DISABLED]
	c0000000000d0758  k  _do_fork+0x18    [DISABLED]
	c0000000000412b0  k  kretprobe_trampoline+0x0    [OPTIMIZED]

and after:
	# cat ../kprobes/list
	c0000000000d04c8  k  _do_fork+0x8    [DISABLED]
	c0000000000d04d0  k  _do_fork+0x10    [DISABLED]
	c0000000000412b0  k  kretprobe_trampoline+0x0    [OPTIMIZED]

Acked-by: Ananth N Mavinakayanahalli <ananth@linux.vnet.ibm.com>
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
---
 arch/powerpc/kernel/kprobes.c | 4 ++--
 include/linux/kprobes.h       | 2 +-
 kernel/kprobes.c              | 7 ++++---
 3 files changed, 7 insertions(+), 6 deletions(-)

diff --git a/arch/powerpc/kernel/kprobes.c b/arch/powerpc/kernel/kprobes.c
index a7aa7394954d..0732a0291ace 100644
--- a/arch/powerpc/kernel/kprobes.c
+++ b/arch/powerpc/kernel/kprobes.c
@@ -42,14 +42,14 @@ DEFINE_PER_CPU(struct kprobe_ctlblk, kprobe_ctlblk);
 
 struct kretprobe_blackpoint kretprobe_blacklist[] = {{NULL, NULL}};
 
-kprobe_opcode_t *kprobe_lookup_name(const char *name)
+kprobe_opcode_t *kprobe_lookup_name(const char *name, unsigned int offset)
 {
 	kprobe_opcode_t *addr;
 
 #ifdef PPC64_ELF_ABI_v2
 	/* PPC64 ABIv2 needs local entry point */
 	addr = (kprobe_opcode_t *)kallsyms_lookup_name(name);
-	if (addr)
+	if (addr && !offset)
 		addr = (kprobe_opcode_t *)ppc_function_entry(addr);
 #elif defined(PPC64_ELF_ABI_v1)
 	/*
diff --git a/include/linux/kprobes.h b/include/linux/kprobes.h
index 34b7d00765d2..442ecaeddebe 100644
--- a/include/linux/kprobes.h
+++ b/include/linux/kprobes.h
@@ -364,7 +364,7 @@ static inline struct kprobe_ctlblk *get_kprobe_ctlblk(void)
 	return this_cpu_ptr(&kprobe_ctlblk);
 }
 
-kprobe_opcode_t *kprobe_lookup_name(const char *name);
+kprobe_opcode_t *kprobe_lookup_name(const char *name, unsigned int offset);
 int register_kprobe(struct kprobe *p);
 void unregister_kprobe(struct kprobe *p);
 int register_kprobes(struct kprobe **kps, int num);
diff --git a/kernel/kprobes.c b/kernel/kprobes.c
index 30f1d732dd74..fbc7a70ff33e 100644
--- a/kernel/kprobes.c
+++ b/kernel/kprobes.c
@@ -72,7 +72,8 @@ static struct {
 	raw_spinlock_t lock ____cacheline_aligned_in_smp;
 } kretprobe_table_locks[KPROBE_TABLE_SIZE];
 
-kprobe_opcode_t * __weak kprobe_lookup_name(const char *name)
+kprobe_opcode_t * __weak kprobe_lookup_name(const char *name,
+					unsigned int __unused)
 {
 	return ((kprobe_opcode_t *)(kallsyms_lookup_name(name)));
 }
@@ -1361,7 +1362,7 @@ static kprobe_opcode_t *kprobe_addr(struct kprobe *p)
 		goto invalid;
 
 	if (p->symbol_name) {
-		addr = kprobe_lookup_name(p->symbol_name);
+		addr = kprobe_lookup_name(p->symbol_name, p->offset);
 		if (!addr)
 			return ERR_PTR(-ENOENT);
 	}
@@ -2167,7 +2168,7 @@ static int __init init_kprobes(void)
 		/* lookup the function address from its name */
 		for (i = 0; kretprobe_blacklist[i].name != NULL; i++) {
 			kretprobe_blacklist[i].addr =
-				kprobe_lookup_name(kretprobe_blacklist[i].name);
+				kprobe_lookup_name(kretprobe_blacklist[i].name, 0);
 			if (!kretprobe_blacklist[i].addr)
 				printk("kretprobe: lookup failed: %s\n",
 				       kretprobe_blacklist[i].name);
-- 
2.11.0

[PATCH 4/5] powerpc: kprobes: factor out code to emulate instruction into a helper

[Naveen N. Rao]

This helper will be used in a subsequent patch to emulate instructions
on re-entering the kprobe handler. No functional change.

Acked-by: Ananth N Mavinakayanahalli <ananth@linux.vnet.ibm.com>
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
---
 arch/powerpc/kernel/kprobes.c | 52 ++++++++++++++++++++++++++-----------------
 1 file changed, 31 insertions(+), 21 deletions(-)

diff --git a/arch/powerpc/kernel/kprobes.c b/arch/powerpc/kernel/kprobes.c
index 0732a0291ace..8b48f7d046bd 100644
--- a/arch/powerpc/kernel/kprobes.c
+++ b/arch/powerpc/kernel/kprobes.c
@@ -207,6 +207,35 @@ void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri,
 	regs->link = (unsigned long)kretprobe_trampoline;
 }
 
+int __kprobes try_to_emulate(struct kprobe *p, struct pt_regs *regs)
+{
+	int ret;
+	unsigned int insn = *p->ainsn.insn;
+
+	/* regs->nip is also adjusted if emulate_step returns 1 */
+	ret = emulate_step(regs, insn);
+	if (ret > 0) {
+		/*
+		 * Once this instruction has been boosted
+		 * successfully, set the boostable flag
+		 */
+		if (unlikely(p->ainsn.boostable == 0))
+			p->ainsn.boostable = 1;
+	} else if (ret < 0) {
+		/*
+		 * We don't allow kprobes on mtmsr(d)/rfi(d), etc.
+		 * So, we should never get here... but, its still
+		 * good to catch them, just in case...
+		 */
+		printk("Can't step on instruction %x\n", insn);
+		BUG();
+	} else if (ret == 0)
+		/* This instruction can't be boosted */
+		p->ainsn.boostable = -1;
+
+	return ret;
+}
+
 int __kprobes kprobe_handler(struct pt_regs *regs)
 {
 	struct kprobe *p;
@@ -302,18 +331,9 @@ int __kprobes kprobe_handler(struct pt_regs *regs)
 
 ss_probe:
 	if (p->ainsn.boostable >= 0) {
-		unsigned int insn = *p->ainsn.insn;
+		ret = try_to_emulate(p, regs);
 
-		/* regs->nip is also adjusted if emulate_step returns 1 */
-		ret = emulate_step(regs, insn);
 		if (ret > 0) {
-			/*
-			 * Once this instruction has been boosted
-			 * successfully, set the boostable flag
-			 */
-			if (unlikely(p->ainsn.boostable == 0))
-				p->ainsn.boostable = 1;
-
 			if (p->post_handler)
 				p->post_handler(p, regs, 0);
 
@@ -321,17 +341,7 @@ int __kprobes kprobe_handler(struct pt_regs *regs)
 			reset_current_kprobe();
 			preempt_enable_no_resched();
 			return 1;
-		} else if (ret < 0) {
-			/*
-			 * We don't allow kprobes on mtmsr(d)/rfi(d), etc.
-			 * So, we should never get here... but, its still
-			 * good to catch them, just in case...
-			 */
-			printk("Can't step on instruction %x\n", insn);
-			BUG();
-		} else if (ret == 0)
-			/* This instruction can't be boosted */
-			p->ainsn.boostable = -1;
+		}
 	}
 	prepare_singlestep(p, regs);
 	kcb->kprobe_status = KPROBE_HIT_SS;
-- 
2.11.0

[PATCH 5/5] powerpc: kprobes: emulate instructions on kprobe handler re-entry

[Naveen N. Rao]

On kprobe handler re-entry, try to emulate the instruction rather than
single stepping always.

As a related change, remove the duplicate saving of msr as that is
already done in set_current_kprobe()

Acked-by: Ananth N Mavinakayanahalli <ananth@linux.vnet.ibm.com>
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
---
 arch/powerpc/kernel/kprobes.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/arch/powerpc/kernel/kprobes.c b/arch/powerpc/kernel/kprobes.c
index 8b48f7d046bd..005bd4a75902 100644
--- a/arch/powerpc/kernel/kprobes.c
+++ b/arch/powerpc/kernel/kprobes.c
@@ -273,10 +273,17 @@ int __kprobes kprobe_handler(struct pt_regs *regs)
 			 */
 			save_previous_kprobe(kcb);
 			set_current_kprobe(p, regs, kcb);
-			kcb->kprobe_saved_msr = regs->msr;
 			kprobes_inc_nmissed_count(p);
 			prepare_singlestep(p, regs);
 			kcb->kprobe_status = KPROBE_REENTER;
+			if (p->ainsn.boostable >= 0) {
+				ret = try_to_emulate(p, regs);
+
+				if (ret > 0) {
+					restore_previous_kprobe(kcb);
+					return 1;
+				}
+			}
 			return 1;
 		} else {
 			if (*addr != BREAKPOINT_INSTRUCTION) {
-- 
2.11.0