Merge tag 'probes-v6.9' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace

Pull probes updates from Masami Hiramatsu:
 "x86 kprobes:

   - Use boolean for some function return instead of 0 and 1

   - Prohibit probing on INT/UD. This prevents user to put kprobe on
     INTn/INT1/INT3/INTO and UD0/UD1/UD2 because these are used for a
     special purpose in the kernel

   - Boost Grp instructions. Because a few percent of kernel
     instructions are Grp 2/3/4/5 and those are safe to be executed
     without ip register fixup, allow those to be boosted (direct
     execution on the trampoline buffer with a JMP)

  tracing:

   - Add function argument access from return events (kretprobe and
     fprobe). This allows user to compare how a data structure field is
     changed after executing a function. With BTF, return event also
     accepts function argument access by name.

   - Fix a wrong comment (using "Kretprobe" in fprobe)

   - Cleanup a big probe argument parser function into three parts, type
     parser, post-processing function, and main parser

   - Cleanup to set nr_args field when initializing trace_probe instead
     of counting up it while parsing

   - Cleanup a redundant #else block from tracefs/README source code

   - Update selftests to check entry argument access from return probes

   - Documentation update about entry argument access from return
     probes"

* tag 'probes-v6.9' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:
  Documentation: tracing: Add entry argument access at function exit
  selftests/ftrace: Add test cases for entry args at function exit
  tracing/probes: Support $argN in return probe (kprobe and fprobe)
  tracing: Remove redundant #else block for BTF args from README
  tracing/probes: cleanup: Set trace_probe::nr_args at trace_probe_init
  tracing/probes: Cleanup probe argument parser
  tracing/fprobe-event: cleanup: Fix a wrong comment in fprobe event
  x86/kprobes: Boost more instructions from grp2/3/4/5
  x86/kprobes: Prohibit kprobing on INT and UD
  x86/kprobes: Refactor can_{probe,boost} return type to bool

Documentation/trace/fprobetrace.rst

@@ -70,6 +70,14 @@ Synopsis of fprobe-events
 
 For the details of TYPE, see :ref:`kprobetrace documentation <kprobetrace_types>`.
 
+Function arguments at exit
+--------------------------
+Function arguments can be accessed at exit probe using $arg<N> fetcharg. This
+is useful to record the function parameter and return value at once, and
+trace the difference of structure fields (for debuging a function whether it
+correctly updates the given data structure or not)
+See the :ref:`sample<fprobetrace_exit_args_sample>` below for how it works.
+
 BTF arguments
 -------------
 BTF (BPF Type Format) argument allows user to trace function and tracepoint
@@ -218,3 +226,26 @@ traceprobe event, you can trace that field as below.
            <idle>-0       [000] d..3.  5606.690317: sched_switch: (__probestub_sched_switch+0x4/0x10) comm="kworker/0:1" usage=1 start_time=137000000
       kworker/0:1-14      [000] d..3.  5606.690339: sched_switch: (__probestub_sched_switch+0x4/0x10) comm="swapper/0" usage=2 start_time=0
            <idle>-0       [000] d..3.  5606.692368: sched_switch: (__probestub_sched_switch+0x4/0x10) comm="kworker/0:1" usage=1 start_time=137000000
+
+.. _fprobetrace_exit_args_sample:
+
+The return probe allows us to access the results of some functions, which returns
+the error code and its results are passed via function parameter, such as an
+structure-initialization function.
+
+For example, vfs_open() will link the file structure to the inode and update
+mode. You can trace that changes with return probe.
+::
+
+ # echo 'f vfs_open mode=file->f_mode:x32 inode=file->f_inode:x64' >> dynamic_events
+ # echo 'f vfs_open%%return mode=file->f_mode:x32 inode=file->f_inode:x64' >> dynamic_events
+ # echo 1 > events/fprobes/enable
+ # cat trace
+              sh-131     [006] ...1.  1945.714346: vfs_open__entry: (vfs_open+0x4/0x40) mode=0x2 inode=0x0
+              sh-131     [006] ...1.  1945.714358: vfs_open__exit: (do_open+0x274/0x3d0 <- vfs_open) mode=0x4d801e inode=0xffff888008470168
+             cat-143     [007] ...1.  1945.717949: vfs_open__entry: (vfs_open+0x4/0x40) mode=0x1 inode=0x0
+             cat-143     [007] ...1.  1945.717956: vfs_open__exit: (do_open+0x274/0x3d0 <- vfs_open) mode=0x4a801d inode=0xffff888005f78d28
+             cat-143     [007] ...1.  1945.720616: vfs_open__entry: (vfs_open+0x4/0x40) mode=0x1 inode=0x0
+             cat-143     [007] ...1.  1945.728263: vfs_open__exit: (do_open+0x274/0x3d0 <- vfs_open) mode=0xa800d inode=0xffff888004ada8d8
+
+You can see the `file::f_mode` and `file::f_inode` are upated in `vfs_open()`.

Documentation/trace/kprobetrace.rst

@@ -70,6 +70,15 @@ Synopsis of kprobe_events
   (\*3) this is useful for fetching a field of data structures.
   (\*4) "u" means user-space dereference. See :ref:`user_mem_access`.
 
+Function arguments at kretprobe
+-------------------------------
+Function arguments can be accessed at kretprobe using $arg<N> fetcharg. This
+is useful to record the function parameter and return value at once, and
+trace the difference of structure fields (for debuging a function whether it
+correctly updates the given data structure or not).
+See the :ref:`sample<fprobetrace_exit_args_sample>` in fprobe event for how
+it works.
+
 .. _kprobetrace_types:
 
 Types

arch/x86/kernel/kprobes/common.h

@@ -78,7 +78,7 @@
 #endif
 
 /* Ensure if the instruction can be boostable */
-extern int can_boost(struct insn *insn, void *orig_addr);
+extern bool can_boost(struct insn *insn, void *orig_addr);
 /* Recover instruction if given address is probed */
 extern unsigned long recover_probed_instruction(kprobe_opcode_t *buf,
 					 unsigned long addr);

arch/x86/kernel/kprobes/core.c

@@ -137,14 +137,14 @@ NOKPROBE_SYMBOL(synthesize_relcall);
  * Returns non-zero if INSN is boostable.
  * RIP relative instructions are adjusted at copying time in 64 bits mode
  */
-int can_boost(struct insn *insn, void *addr)
+bool can_boost(struct insn *insn, void *addr)
 {
 	kprobe_opcode_t opcode;
 	insn_byte_t prefix;
 	int i;
 
 	if (search_exception_tables((unsigned long)addr))
-		return 0;	/* Page fault may occur on this address. */
+		return false;	/* Page fault may occur on this address. */
 
 	/* 2nd-byte opcode */
 	if (insn->opcode.nbytes == 2)
@@ -152,7 +152,7 @@ int can_boost(struct insn *insn, void *addr)
 				(unsigned long *)twobyte_is_boostable);
 
 	if (insn->opcode.nbytes != 1)
-		return 0;
+		return false;
 
 	for_each_insn_prefix(insn, i, prefix) {
 		insn_attr_t attr;
@@ -160,7 +160,7 @@ int can_boost(struct insn *insn, void *addr)
 		attr = inat_get_opcode_attribute(prefix);
 		/* Can't boost Address-size override prefix and CS override prefix */
 		if (prefix == 0x2e || inat_is_address_size_prefix(attr))
-			return 0;
+			return false;
 	}
 
 	opcode = insn->opcode.bytes[0];
@@ -169,24 +169,35 @@ int can_boost(struct insn *insn, void *addr)
 	case 0x62:		/* bound */
 	case 0x70 ... 0x7f:	/* Conditional jumps */
 	case 0x9a:		/* Call far */
-	case 0xc0 ... 0xc1:	/* Grp2 */
 	case 0xcc ... 0xce:	/* software exceptions */
-	case 0xd0 ... 0xd3:	/* Grp2 */
 	case 0xd6:		/* (UD) */
 	case 0xd8 ... 0xdf:	/* ESC */
 	case 0xe0 ... 0xe3:	/* LOOP*, JCXZ */
 	case 0xe8 ... 0xe9:	/* near Call, JMP */
 	case 0xeb:		/* Short JMP */
 	case 0xf0 ... 0xf4:	/* LOCK/REP, HLT */
+		/* ... are not boostable */
+		return false;
+	case 0xc0 ... 0xc1:	/* Grp2 */
+	case 0xd0 ... 0xd3:	/* Grp2 */
+		/*
+		 * AMD uses nnn == 110 as SHL/SAL, but Intel makes it reserved.
+		 */
+		return X86_MODRM_REG(insn->modrm.bytes[0]) != 0b110;
 	case 0xf6 ... 0xf7:	/* Grp3 */
+		/* AMD uses nnn == 001 as TEST, but Intel makes it reserved. */
+		return X86_MODRM_REG(insn->modrm.bytes[0]) != 0b001;
 	case 0xfe:		/* Grp4 */
-		/* ... are not boostable */
-		return 0;
+		/* Only INC and DEC are boostable */
+		return X86_MODRM_REG(insn->modrm.bytes[0]) == 0b000 ||
+		       X86_MODRM_REG(insn->modrm.bytes[0]) == 0b001;
 	case 0xff:		/* Grp5 */
-		/* Only indirect jmp is boostable */
-		return X86_MODRM_REG(insn->modrm.bytes[0]) == 4;
+		/* Only INC, DEC, and indirect JMP are boostable */
+		return X86_MODRM_REG(insn->modrm.bytes[0]) == 0b000 ||
+		       X86_MODRM_REG(insn->modrm.bytes[0]) == 0b001 ||
+		       X86_MODRM_REG(insn->modrm.bytes[0]) == 0b100;
 	default:
-		return 1;
+		return true;
 	}
 }
 
@@ -252,21 +263,40 @@ unsigned long recover_probed_instruction(kprobe_opcode_t *buf, unsigned long add
 	return __recover_probed_insn(buf, addr);
 }
 
-/* Check if paddr is at an instruction boundary */
-static int can_probe(unsigned long paddr)
+/* Check if insn is INT or UD */
+static inline bool is_exception_insn(struct insn *insn)
+{
+	/* UD uses 0f escape */
+	if (insn->opcode.bytes[0] == 0x0f) {
+		/* UD0 / UD1 / UD2 */
+		return insn->opcode.bytes[1] == 0xff ||
+		       insn->opcode.bytes[1] == 0xb9 ||
+		       insn->opcode.bytes[1] == 0x0b;
+	}
+
+	/* INT3 / INT n / INTO / INT1 */
+	return insn->opcode.bytes[0] == 0xcc ||
+	       insn->opcode.bytes[0] == 0xcd ||
+	       insn->opcode.bytes[0] == 0xce ||
+	       insn->opcode.bytes[0] == 0xf1;
+}
+
+/*
+ * Check if paddr is at an instruction boundary and that instruction can
+ * be probed
+ */
+static bool can_probe(unsigned long paddr)
 {
 	unsigned long addr, __addr, offset = 0;
 	struct insn insn;
 	kprobe_opcode_t buf[MAX_INSN_SIZE];
 
 	if (!kallsyms_lookup_size_offset(paddr, NULL, &offset))
-		return 0;
+		return false;
 
 	/* Decode instructions */
 	addr = paddr - offset;
 	while (addr < paddr) {
-		int ret;
-
 		/*
 		 * Check if the instruction has been modified by another
 		 * kprobe, in which case we replace the breakpoint by the
@@ -277,11 +307,10 @@ static int can_probe(unsigned long paddr)
 		 */
 		__addr = recover_probed_instruction(buf, addr);
 		if (!__addr)
-			return 0;
+			return false;
 
-		ret = insn_decode_kernel(&insn, (void *)__addr);
-		if (ret < 0)
-			return 0;
+		if (insn_decode_kernel(&insn, (void *)__addr) < 0)
+			return false;
 
 #ifdef CONFIG_KGDB
 		/*
@@ -290,10 +319,26 @@ static int can_probe(unsigned long paddr)
 		 */
 		if (insn.opcode.bytes[0] == INT3_INSN_OPCODE &&
 		    kgdb_has_hit_break(addr))
-			return 0;
+			return false;
 #endif
 		addr += insn.length;
 	}
+
+	/* Check if paddr is at an instruction boundary */
+	if (addr != paddr)
+		return false;
+
+	__addr = recover_probed_instruction(buf, addr);
+	if (!__addr)
+		return false;
+
+	if (insn_decode_kernel(&insn, (void *)__addr) < 0)
+		return false;
+
+	/* INT and UD are special and should not be kprobed */
+	if (is_exception_insn(&insn))
+		return false;
+
 	if (IS_ENABLED(CONFIG_CFI_CLANG)) {
 		/*
 		 * The compiler generates the following instruction sequence
@@ -308,13 +353,6 @@ static int can_probe(unsigned long paddr)
 		 * Also, these movl and addl are used for showing expected
 		 * type. So those must not be touched.
 		 */
-		__addr = recover_probed_instruction(buf, addr);
-		if (!__addr)
-			return 0;
-
-		if (insn_decode_kernel(&insn, (void *)__addr) < 0)
-			return 0;
-
 		if (insn.opcode.value == 0xBA)
 			offset = 12;
 		else if (insn.opcode.value == 0x3)
@@ -324,11 +362,11 @@ static int can_probe(unsigned long paddr)
 
 		/* This movl/addl is used for decoding CFI. */
 		if (is_cfi_trap(addr + offset))
-			return 0;
+			return false;
 	}
 
 out:
-	return (addr == paddr);
+	return true;
 }
 
 /* If x86 supports IBT (ENDBR) it must be skipped. */

kernel/trace/trace.c

@@ -5747,16 +5747,15 @@ static const char readme_msg[] =
 	"\t     args: <name>=fetcharg[:type]\n"
 	"\t fetcharg: (%<register>|$<efield>), @<address>, @<symbol>[+|-<offset>],\n"
 #ifdef CONFIG_HAVE_FUNCTION_ARG_ACCESS_API
-#ifdef CONFIG_PROBE_EVENTS_BTF_ARGS
 	"\t           $stack<index>, $stack, $retval, $comm, $arg<N>,\n"
+#ifdef CONFIG_PROBE_EVENTS_BTF_ARGS
 	"\t           <argname>[->field[->field|.field...]],\n"
-#else
-	"\t           $stack<index>, $stack, $retval, $comm, $arg<N>,\n"
 #endif
 #else
 	"\t           $stack<index>, $stack, $retval, $comm,\n"
 #endif
 	"\t           +|-[u]<offset>(<fetcharg>), \\imm-value, \\\"imm-string\"\n"
+	"\t     kernel return probes support: $retval, $arg<N>, $comm\n"
 	"\t     type: s8/16/32/64, u8/16/32/64, x8/16/32/64, char, string, symbol,\n"
 	"\t           b<bit-width>@<bit-offset>/<container-size>, ustring,\n"
 	"\t           symstr, <type>\\[<array-size>\\]\n"

kernel/trace/trace_eprobe.c

@@ -220,7 +220,7 @@ static struct trace_eprobe *alloc_event_probe(const char *group,
 	if (!ep->event_system)
 		goto error;
 
-	ret = trace_probe_init(&ep->tp, this_event, group, false);
+	ret = trace_probe_init(&ep->tp, this_event, group, false, nargs);
 	if (ret < 0)
 		goto error;
 
@@ -390,8 +390,8 @@ static int get_eprobe_size(struct trace_probe *tp, void *rec)
 
 /* Note that we don't verify it, since the code does not come from user space */
 static int
-process_fetch_insn(struct fetch_insn *code, void *rec, void *dest,
-		   void *base)
+process_fetch_insn(struct fetch_insn *code, void *rec, void *edata,
+		   void *dest, void *base)
 {
 	unsigned long val;
 	int ret;
@@ -438,7 +438,7 @@ __eprobe_trace_func(struct eprobe_data *edata, void *rec)
 		return;
 
 	entry = fbuffer.entry = ring_buffer_event_data(fbuffer.event);
-	store_trace_args(&entry[1], &edata->ep->tp, rec, sizeof(*entry), dsize);
+	store_trace_args(&entry[1], &edata->ep->tp, rec, NULL, sizeof(*entry), dsize);
 
 	trace_event_buffer_commit(&fbuffer);
 }

kernel/trace/trace_fprobe.c

@@ -4,6 +4,7 @@
  * Copyright (C) 2022 Google LLC.
  */
 #define pr_fmt(fmt)	"trace_fprobe: " fmt
+#include <asm/ptrace.h>
 
 #include <linux/fprobe.h>
 #include <linux/module.h>
@@ -129,8 +130,8 @@ static bool trace_fprobe_is_registered(struct trace_fprobe *tf)
  * from user space.
  */
 static int
-process_fetch_insn(struct fetch_insn *code, void *rec, void *dest,
-		   void *base)
+process_fetch_insn(struct fetch_insn *code, void *rec, void *edata,
+		   void *dest, void *base)
 {
 	struct pt_regs *regs = rec;
 	unsigned long val;
@@ -152,6 +153,9 @@ process_fetch_insn(struct fetch_insn *code, void *rec, void *dest,
 	case FETCH_OP_ARG:
 		val = regs_get_kernel_argument(regs, code->param);
 		break;
+	case FETCH_OP_EDATA:
+		val = *(unsigned long *)((unsigned long)edata + code->offset);
+		break;
 #endif
 	case FETCH_NOP_SYMBOL:	/* Ignore a place holder */
 		code++;
@@ -184,7 +188,7 @@ __fentry_trace_func(struct trace_fprobe *tf, unsigned long entry_ip,
 	if (trace_trigger_soft_disabled(trace_file))
 		return;
 
-	dsize = __get_data_size(&tf->tp, regs);
+	dsize = __get_data_size(&tf->tp, regs, NULL);
 
 	entry = trace_event_buffer_reserve(&fbuffer, trace_file,
 					   sizeof(*entry) + tf->tp.size + dsize);
@@ -194,7 +198,7 @@ __fentry_trace_func(struct trace_fprobe *tf, unsigned long entry_ip,
 	fbuffer.regs = regs;
 	entry = fbuffer.entry = ring_buffer_event_data(fbuffer.event);
 	entry->ip = entry_ip;
-	store_trace_args(&entry[1], &tf->tp, regs, sizeof(*entry), dsize);
+	store_trace_args(&entry[1], &tf->tp, regs, NULL, sizeof(*entry), dsize);
 
 	trace_event_buffer_commit(&fbuffer);
 }
@@ -210,11 +214,24 @@ fentry_trace_func(struct trace_fprobe *tf, unsigned long entry_ip,
 }
 NOKPROBE_SYMBOL(fentry_trace_func);
 
-/* Kretprobe handler */
+/* function exit handler */
+static int trace_fprobe_entry_handler(struct fprobe *fp, unsigned long entry_ip,
+				unsigned long ret_ip, struct pt_regs *regs,
+				void *entry_data)
+{
+	struct trace_fprobe *tf = container_of(fp, struct trace_fprobe, fp);
+
+	if (tf->tp.entry_arg)
+		store_trace_entry_data(entry_data, &tf->tp, regs);
+
+	return 0;
+}
+NOKPROBE_SYMBOL(trace_fprobe_entry_handler)
+
 static nokprobe_inline void
 __fexit_trace_func(struct trace_fprobe *tf, unsigned long entry_ip,
 		   unsigned long ret_ip, struct pt_regs *regs,
-		   struct trace_event_file *trace_file)
+		   void *entry_data, struct trace_event_file *trace_file)
 {
 	struct fexit_trace_entry_head *entry;
 	struct trace_event_buffer fbuffer;
@@ -227,7 +244,7 @@ __fexit_trace_func(struct trace_fprobe *tf, unsigned long entry_ip,
 	if (trace_trigger_soft_disabled(trace_file))
 		return;
 
-	dsize = __get_data_size(&tf->tp, regs);
+	dsize = __get_data_size(&tf->tp, regs, entry_data);
 
 	entry = trace_event_buffer_reserve(&fbuffer, trace_file,
 					   sizeof(*entry) + tf->tp.size + dsize);
@@ -238,19 +255,19 @@ __fexit_trace_func(struct trace_fprobe *tf, unsigned long entry_ip,
 	entry = fbuffer.entry = ring_buffer_event_data(fbuffer.event);
 	entry->func = entry_ip;
 	entry->ret_ip = ret_ip;
-	store_trace_args(&entry[1], &tf->tp, regs, sizeof(*entry), dsize);
+	store_trace_args(&entry[1], &tf->tp, regs, entry_data, sizeof(*entry), dsize);
 
 	trace_event_buffer_commit(&fbuffer);
 }
 
 static void
 fexit_trace_func(struct trace_fprobe *tf, unsigned long entry_ip,
-		 unsigned long ret_ip, struct pt_regs *regs)
+		 unsigned long ret_ip, struct pt_regs *regs, void *entry_data)
 {
 	struct event_file_link *link;
 
 	trace_probe_for_each_link_rcu(link, &tf->tp)
-		__fexit_trace_func(tf, entry_ip, ret_ip, regs, link->file);
+		__fexit_trace_func(tf, entry_ip, ret_ip, regs, entry_data, link->file);
 }
 NOKPROBE_SYMBOL(fexit_trace_func);
 
@@ -269,7 +286,7 @@ static int fentry_perf_func(struct trace_fprobe *tf, unsigned long entry_ip,
 	if (hlist_empty(head))
 		return 0;
 
-	dsize = __get_data_size(&tf->tp, regs);
+	dsize = __get_data_size(&tf->tp, regs, NULL);
 	__size = sizeof(*entry) + tf->tp.size + dsize;
 	size = ALIGN(__size + sizeof(u32), sizeof(u64));
 	size -= sizeof(u32);
@@ -280,7 +297,7 @@ static int fentry_perf_func(struct trace_fprobe *tf, unsigned long entry_ip,
 
 	entry->ip = entry_ip;
 	memset(&entry[1], 0, dsize);
-	store_trace_args(&entry[1], &tf->tp, regs, sizeof(*entry), dsize);
+	store_trace_args(&entry[1], &tf->tp, regs, NULL, sizeof(*entry), dsize);
 	perf_trace_buf_submit(entry, size, rctx, call->event.type, 1, regs,
 			      head, NULL);
 	return 0;
@@ -289,7 +306,8 @@ NOKPROBE_SYMBOL(fentry_perf_func);
 
 static void
 fexit_perf_func(struct trace_fprobe *tf, unsigned long entry_ip,
-		unsigned long ret_ip, struct pt_regs *regs)
+		unsigned long ret_ip, struct pt_regs *regs,
+		void *entry_data)
 {
 	struct trace_event_call *call = trace_probe_event_call(&tf->tp);
 	struct fexit_trace_entry_head *entry;
@@ -301,7 +319,7 @@ fexit_perf_func(struct trace_fprobe *tf, unsigned long entry_ip,
 	if (hlist_empty(head))
 		return;
 
-	dsize = __get_data_size(&tf->tp, regs);
+	dsize = __get_data_size(&tf->tp, regs, entry_data);
 	__size = sizeof(*entry) + tf->tp.size + dsize;
 	size = ALIGN(__size + sizeof(u32), sizeof(u64));
 	size -= sizeof(u32);
@@ -312,7 +330,7 @@ fexit_perf_func(struct trace_fprobe *tf, unsigned long entry_ip,
 
 	entry->func = entry_ip;
 	entry->ret_ip = ret_ip;
-	store_trace_args(&entry[1], &tf->tp, regs, sizeof(*entry), dsize);
+	store_trace_args(&entry[1], &tf->tp, regs, entry_data, sizeof(*entry), dsize);
 	perf_trace_buf_submit(entry, size, rctx, call->event.type, 1, regs,
 			      head, NULL);
 }
@@ -343,10 +361,10 @@ static void fexit_dispatcher(struct fprobe *fp, unsigned long entry_ip,
 	struct trace_fprobe *tf = container_of(fp, struct trace_fprobe, fp);
 
 	if (trace_probe_test_flag(&tf->tp, TP_FLAG_TRACE))
-		fexit_trace_func(tf, entry_ip, ret_ip, regs);
+		fexit_trace_func(tf, entry_ip, ret_ip, regs, entry_data);
 #ifdef CONFIG_PERF_EVENTS
 	if (trace_probe_test_flag(&tf->tp, TP_FLAG_PROFILE))
-		fexit_perf_func(tf, entry_ip, ret_ip, regs);
+		fexit_perf_func(tf, entry_ip, ret_ip, regs, entry_data);
 #endif
 }
 NOKPROBE_SYMBOL(fexit_dispatcher);
@@ -389,7 +407,7 @@ static struct trace_fprobe *alloc_trace_fprobe(const char *group,
 	tf->tpoint = tpoint;
 	tf->fp.nr_maxactive = maxactive;
 
-	ret = trace_probe_init(&tf->tp, event, group, false);
+	ret = trace_probe_init(&tf->tp, event, group, false, nargs);
 	if (ret < 0)
 		goto error;
 
@@ -1109,6 +1127,11 @@ static int __trace_fprobe_create(int argc, const char *argv[])
 			goto error;	/* This can be -ENOMEM */
 	}
 
+	if (is_return && tf->tp.entry_arg) {
+		tf->fp.entry_handler = trace_fprobe_entry_handler;
+		tf->fp.entry_data_size = traceprobe_get_entry_data_size(&tf->tp);
+	}
+
 	ret = traceprobe_set_print_fmt(&tf->tp,
 			is_return ? PROBE_PRINT_RETURN : PROBE_PRINT_NORMAL);
 	if (ret < 0)

kernel/trace/trace_kprobe.c

@@ -290,7 +290,7 @@ static struct trace_kprobe *alloc_trace_kprobe(const char *group,
 	INIT_HLIST_NODE(&tk->rp.kp.hlist);
 	INIT_LIST_HEAD(&tk->rp.kp.list);
 
-	ret = trace_probe_init(&tk->tp, event, group, false);
+	ret = trace_probe_init(&tk->tp, event, group, false, nargs);
 	if (ret < 0)
 		goto error;
 
@@ -740,6 +740,9 @@ static unsigned int number_of_same_symbols(char *func_name)
 	return ctx.count;
 }
 
+static int trace_kprobe_entry_handler(struct kretprobe_instance *ri,
+				      struct pt_regs *regs);
+
 static int __trace_kprobe_create(int argc, const char *argv[])
 {
 	/*
@@ -948,6 +951,11 @@ static int __trace_kprobe_create(int argc, const char *argv[])
 		if (ret)
 			goto error;	/* This can be -ENOMEM */
 	}
+	/* entry handler for kretprobe */
+	if (is_return && tk->tp.entry_arg) {
+		tk->rp.entry_handler = trace_kprobe_entry_handler;
+		tk->rp.data_size = traceprobe_get_entry_data_size(&tk->tp);
+	}
 
 	ptype = is_return ? PROBE_PRINT_RETURN : PROBE_PRINT_NORMAL;
 	ret = traceprobe_set_print_fmt(&tk->tp, ptype);
@@ -1303,8 +1311,8 @@ static const struct file_operations kprobe_profile_ops = {
 
 /* Note that we don't verify it, since the code does not come from user space */
 static int
-process_fetch_insn(struct fetch_insn *code, void *rec, void *dest,
-		   void *base)
+process_fetch_insn(struct fetch_insn *code, void *rec, void *edata,
+		   void *dest, void *base)
 {
 	struct pt_regs *regs = rec;
 	unsigned long val;
@@ -1329,6 +1337,9 @@ process_fetch_insn(struct fetch_insn *code, void *rec, void *dest,
 	case FETCH_OP_ARG:
 		val = regs_get_kernel_argument(regs, code->param);
 		break;
+	case FETCH_OP_EDATA:
+		val = *(unsigned long *)((unsigned long)edata + code->offset);
+		break;
 #endif
 	case FETCH_NOP_SYMBOL:	/* Ignore a place holder */
 		code++;
@@ -1359,7 +1370,7 @@ __kprobe_trace_func(struct trace_kprobe *tk, struct pt_regs *regs,
 	if (trace_trigger_soft_disabled(trace_file))
 		return;
 
-	dsize = __get_data_size(&tk->tp, regs);
+	dsize = __get_data_size(&tk->tp, regs, NULL);
 
 	entry = trace_event_buffer_reserve(&fbuffer, trace_file,
 					   sizeof(*entry) + tk->tp.size + dsize);
@@ -1368,7 +1379,7 @@ __kprobe_trace_func(struct trace_kprobe *tk, struct pt_regs *regs,
 
 	fbuffer.regs = regs;
 	entry->ip = (unsigned long)tk->rp.kp.addr;
-	store_trace_args(&entry[1], &tk->tp, regs, sizeof(*entry), dsize);
+	store_trace_args(&entry[1], &tk->tp, regs, NULL, sizeof(*entry), dsize);
 
 	trace_event_buffer_commit(&fbuffer);
 }
@@ -1384,6 +1395,31 @@ kprobe_trace_func(struct trace_kprobe *tk, struct pt_regs *regs)
 NOKPROBE_SYMBOL(kprobe_trace_func);
 
 /* Kretprobe handler */
+
+static int trace_kprobe_entry_handler(struct kretprobe_instance *ri,
+				      struct pt_regs *regs)
+{
+	struct kretprobe *rp = get_kretprobe(ri);
+	struct trace_kprobe *tk;
+
+	/*
+	 * There is a small chance that get_kretprobe(ri) returns NULL when
+	 * the kretprobe is unregister on another CPU between kretprobe's
+	 * trampoline_handler and this function.
+	 */
+	if (unlikely(!rp))
+		return -ENOENT;
+
+	tk = container_of(rp, struct trace_kprobe, rp);
+
+	/* store argument values into ri->data as entry data */
+	if (tk->tp.entry_arg)
+		store_trace_entry_data(ri->data, &tk->tp, regs);
+
+	return 0;
+}
+
+
 static nokprobe_inline void
 __kretprobe_trace_func(struct trace_kprobe *tk, struct kretprobe_instance *ri,
 		       struct pt_regs *regs,
@@ -1399,7 +1435,7 @@ __kretprobe_trace_func(struct trace_kprobe *tk, struct kretprobe_instance *ri,
 	if (trace_trigger_soft_disabled(trace_file))
 		return;
 
-	dsize = __get_data_size(&tk->tp, regs);
+	dsize = __get_data_size(&tk->tp, regs, ri->data);
 
 	entry = trace_event_buffer_reserve(&fbuffer, trace_file,
 					   sizeof(*entry) + tk->tp.size + dsize);
@@ -1409,7 +1445,7 @@ __kretprobe_trace_func(struct trace_kprobe *tk, struct kretprobe_instance *ri,
 	fbuffer.regs = regs;
 	entry->func = (unsigned long)tk->rp.kp.addr;
 	entry->ret_ip = get_kretprobe_retaddr(ri);
-	store_trace_args(&entry[1], &tk->tp, regs, sizeof(*entry), dsize);
+	store_trace_args(&entry[1], &tk->tp, regs, ri->data, sizeof(*entry), dsize);
 
 	trace_event_buffer_commit(&fbuffer);
 }
@@ -1557,7 +1593,7 @@ kprobe_perf_func(struct trace_kprobe *tk, struct pt_regs *regs)
 	if (hlist_empty(head))
 		return 0;
 
-	dsize = __get_data_size(&tk->tp, regs);
+	dsize = __get_data_size(&tk->tp, regs, NULL);
 	__size = sizeof(*entry) + tk->tp.size + dsize;
 	size = ALIGN(__size + sizeof(u32), sizeof(u64));
 	size -= sizeof(u32);
@@ -1568,7 +1604,7 @@ kprobe_perf_func(struct trace_kprobe *tk, struct pt_regs *regs)
 
 	entry->ip = (unsigned long)tk->rp.kp.addr;
 	memset(&entry[1], 0, dsize);
-	store_trace_args(&entry[1], &tk->tp, regs, sizeof(*entry), dsize);
+	store_trace_args(&entry[1], &tk->tp, regs, NULL, sizeof(*entry), dsize);
 	perf_trace_buf_submit(entry, size, rctx, call->event.type, 1, regs,
 			      head, NULL);
 	return 0;
@@ -1593,7 +1629,7 @@ kretprobe_perf_func(struct trace_kprobe *tk, struct kretprobe_instance *ri,
 	if (hlist_empty(head))
 		return;
 
-	dsize = __get_data_size(&tk->tp, regs);
+	dsize = __get_data_size(&tk->tp, regs, ri->data);
 	__size = sizeof(*entry) + tk->tp.size + dsize;
 	size = ALIGN(__size + sizeof(u32), sizeof(u64));
 	size -= sizeof(u32);
@@ -1604,7 +1640,7 @@ kretprobe_perf_func(struct trace_kprobe *tk, struct kretprobe_instance *ri,
 
 	entry->func = (unsigned long)tk->rp.kp.addr;
 	entry->ret_ip = get_kretprobe_retaddr(ri);
-	store_trace_args(&entry[1], &tk->tp, regs, sizeof(*entry), dsize);
+	store_trace_args(&entry[1], &tk->tp, regs, ri->data, sizeof(*entry), dsize);
 	perf_trace_buf_submit(entry, size, rctx, call->event.type, 1, regs,
 			      head, NULL);
 }

kernel/trace/trace_probe.h

@@ -92,6 +92,7 @@ enum fetch_op {
 	FETCH_OP_ARG,		/* Function argument : .param */
 	FETCH_OP_FOFFS,		/* File offset: .immediate */
 	FETCH_OP_DATA,		/* Allocated data: .data */
+	FETCH_OP_EDATA,		/* Entry data: .offset */
 	// Stage 2 (dereference) op
 	FETCH_OP_DEREF,		/* Dereference: .offset */
 	FETCH_OP_UDEREF,	/* User-space Dereference: .offset */
@@ -102,6 +103,7 @@ enum fetch_op {
 	FETCH_OP_ST_STRING,	/* String: .offset, .size */
 	FETCH_OP_ST_USTRING,	/* User String: .offset, .size */
 	FETCH_OP_ST_SYMSTR,	/* Kernel Symbol String: .offset, .size */
+	FETCH_OP_ST_EDATA,	/* Store Entry Data: .offset */
 	// Stage 4 (modify) op
 	FETCH_OP_MOD_BF,	/* Bitfield: .basesize, .lshift, .rshift */
 	// Stage 5 (loop) op
@@ -232,6 +234,11 @@ struct probe_arg {
 	const struct fetch_type	*type;	/* Type of this argument */
 };
 
+struct probe_entry_arg {
+	struct fetch_insn	*code;
+	unsigned int		size;	/* The entry data size */
+};
+
 struct trace_uprobe_filter {
 	rwlock_t		rwlock;
 	int			nr_systemwide;
@@ -253,6 +260,7 @@ struct trace_probe {
 	struct trace_probe_event	*event;
 	ssize_t				size;	/* trace entry size */
 	unsigned int			nr_args;
+	struct probe_entry_arg		*entry_arg;	/* This is only for return probe */
 	struct probe_arg		args[];
 };
 
@@ -338,7 +346,7 @@ static inline bool trace_probe_has_single_file(struct trace_probe *tp)
 }
 
 int trace_probe_init(struct trace_probe *tp, const char *event,
-		     const char *group, bool alloc_filter);
+		     const char *group, bool alloc_filter, int nargs);
 void trace_probe_cleanup(struct trace_probe *tp);
 int trace_probe_append(struct trace_probe *tp, struct trace_probe *to);
 void trace_probe_unlink(struct trace_probe *tp);
@@ -355,6 +363,18 @@ int trace_probe_create(const char *raw_command, int (*createfn)(int, const char
 int trace_probe_print_args(struct trace_seq *s, struct probe_arg *args, int nr_args,
 		 u8 *data, void *field);
 
+#ifdef CONFIG_HAVE_FUNCTION_ARG_ACCESS_API
+int traceprobe_get_entry_data_size(struct trace_probe *tp);
+/* This is a runtime function to store entry data */
+void store_trace_entry_data(void *edata, struct trace_probe *tp, struct pt_regs *regs);
+#else /* !CONFIG_HAVE_FUNCTION_ARG_ACCESS_API */
+static inline int traceprobe_get_entry_data_size(struct trace_probe *tp)
+{
+	return 0;
+}
+#define store_trace_entry_data(edata, tp, regs) do { } while (0)
+#endif
+
 #define trace_probe_for_each_link(pos, tp)	\
 	list_for_each_entry(pos, &(tp)->event->files, list)
 #define trace_probe_for_each_link_rcu(pos, tp)	\
@@ -381,6 +401,11 @@ static inline bool tparg_is_function_entry(unsigned int flags)
 	return (flags & TPARG_FL_LOC_MASK) == (TPARG_FL_KERNEL | TPARG_FL_FENTRY);
 }
 
+static inline bool tparg_is_function_return(unsigned int flags)
+{
+	return (flags & TPARG_FL_LOC_MASK) == (TPARG_FL_KERNEL | TPARG_FL_RETURN);
+}
+
 struct traceprobe_parse_context {
 	struct trace_event_call *event;
 	/* BTF related parameters */
@@ -392,6 +417,7 @@ struct traceprobe_parse_context {
 	const struct btf_type *last_type;	/* Saved type */
 	u32 last_bitoffs;		/* Saved bitoffs */
 	u32 last_bitsize;		/* Saved bitsize */
+	struct trace_probe *tp;
 	unsigned int flags;
 	int offset;
 };
@@ -506,7 +532,7 @@ extern int traceprobe_define_arg_fields(struct trace_event_call *event_call,
 	C(NO_BTFARG,		"This variable is not found at this probe point"),\
 	C(NO_BTF_ENTRY,		"No BTF entry for this probe point"),	\
 	C(BAD_VAR_ARGS,		"$arg* must be an independent parameter without name etc."),\
-	C(NOFENTRY_ARGS,	"$arg* can be used only on function entry"),	\
+	C(NOFENTRY_ARGS,	"$arg* can be used only on function entry or exit"),	\
 	C(DOUBLE_ARGS,		"$arg* can be used only once in the parameters"),	\
 	C(ARGS_2LONG,		"$arg* failed because the argument list is too long"),	\
 	C(ARGIDX_2BIG,		"$argN index is too big"),		\

kernel/trace/trace_probe_tmpl.h

@@ -54,7 +54,7 @@ fetch_apply_bitfield(struct fetch_insn *code, void *buf)
  * If dest is NULL, don't store result and return required dynamic data size.
  */
 static int
-process_fetch_insn(struct fetch_insn *code, void *rec,
+process_fetch_insn(struct fetch_insn *code, void *rec, void *edata,
 		   void *dest, void *base);
 static nokprobe_inline int fetch_store_strlen(unsigned long addr);
 static nokprobe_inline int
@@ -232,7 +232,7 @@ process_fetch_insn_bottom(struct fetch_insn *code, unsigned long val,
 
 /* Sum up total data length for dynamic arrays (strings) */
 static nokprobe_inline int
-__get_data_size(struct trace_probe *tp, struct pt_regs *regs)
+__get_data_size(struct trace_probe *tp, struct pt_regs *regs, void *edata)
 {
 	struct probe_arg *arg;
 	int i, len, ret = 0;
@@ -240,7 +240,7 @@ __get_data_size(struct trace_probe *tp, struct pt_regs *regs)
 	for (i = 0; i < tp->nr_args; i++) {
 		arg = tp->args + i;
 		if (unlikely(arg->dynamic)) {
-			len = process_fetch_insn(arg->code, regs, NULL, NULL);
+			len = process_fetch_insn(arg->code, regs, edata, NULL, NULL);
 			if (len > 0)
 				ret += len;
 		}
@@ -251,7 +251,7 @@ __get_data_size(struct trace_probe *tp, struct pt_regs *regs)
 
 /* Store the value of each argument */
 static nokprobe_inline void
-store_trace_args(void *data, struct trace_probe *tp, void *rec,
+store_trace_args(void *data, struct trace_probe *tp, void *rec, void *edata,
 		 int header_size, int maxlen)
 {
 	struct probe_arg *arg;
@@ -266,7 +266,7 @@ store_trace_args(void *data, struct trace_probe *tp, void *rec,
 		/* Point the dynamic data area if needed */
 		if (unlikely(arg->dynamic))
 			*dl = make_data_loc(maxlen, dyndata - base);
-		ret = process_fetch_insn(arg->code, rec, dl, base);
+		ret = process_fetch_insn(arg->code, rec, edata, dl, base);
 		if (arg->dynamic && likely(ret > 0)) {
 			dyndata += ret;
 			maxlen -= ret;

kernel/trace/trace_uprobe.c

@@ -211,8 +211,8 @@ static unsigned long translate_user_vaddr(unsigned long file_offset)
 
 /* Note that we don't verify it, since the code does not come from user space */
 static int
-process_fetch_insn(struct fetch_insn *code, void *rec, void *dest,
-		   void *base)
+process_fetch_insn(struct fetch_insn *code, void *rec, void *edata,
+		   void *dest, void *base)
 {
 	struct pt_regs *regs = rec;
 	unsigned long val;
@@ -337,7 +337,7 @@ alloc_trace_uprobe(const char *group, const char *event, int nargs, bool is_ret)
 	if (!tu)
 		return ERR_PTR(-ENOMEM);
 
-	ret = trace_probe_init(&tu->tp, event, group, true);
+	ret = trace_probe_init(&tu->tp, event, group, true, nargs);
 	if (ret < 0)
 		goto error;
 
@@ -1490,11 +1490,11 @@ static int uprobe_dispatcher(struct uprobe_consumer *con, struct pt_regs *regs)
 	if (WARN_ON_ONCE(!uprobe_cpu_buffer))
 		return 0;
 
-	dsize = __get_data_size(&tu->tp, regs);
+	dsize = __get_data_size(&tu->tp, regs, NULL);
 	esize = SIZEOF_TRACE_ENTRY(is_ret_probe(tu));
 
 	ucb = uprobe_buffer_get();
-	store_trace_args(ucb->buf, &tu->tp, regs, esize, dsize);
+	store_trace_args(ucb->buf, &tu->tp, regs, NULL, esize, dsize);
 
 	if (trace_probe_test_flag(&tu->tp, TP_FLAG_TRACE))
 		ret |= uprobe_trace_func(tu, regs, ucb, dsize);
@@ -1525,11 +1525,11 @@ static int uretprobe_dispatcher(struct uprobe_consumer *con,
 	if (WARN_ON_ONCE(!uprobe_cpu_buffer))
 		return 0;
 
-	dsize = __get_data_size(&tu->tp, regs);
+	dsize = __get_data_size(&tu->tp, regs, NULL);
 	esize = SIZEOF_TRACE_ENTRY(is_ret_probe(tu));
 
 	ucb = uprobe_buffer_get();
-	store_trace_args(ucb->buf, &tu->tp, regs, esize, dsize);
+	store_trace_args(ucb->buf, &tu->tp, regs, NULL, esize, dsize);
 
 	if (trace_probe_test_flag(&tu->tp, TP_FLAG_TRACE))
 		uretprobe_trace_func(tu, func, regs, ucb, dsize);

tools/testing/selftests/ftrace/test.d/dynevent/fprobe_entry_arg.tc

+#!/bin/sh
+# SPDX-License-Identifier: GPL-2.0
+# description: Function return probe entry argument access
+# requires: dynamic_events 'f[:[<group>/][<event>]] <func-name>':README 'kernel return probes support:':README
+
+echo 'f:tests/myevent1 vfs_open arg=$arg1' >> dynamic_events
+echo 'f:tests/myevent2 vfs_open%return arg=$arg1' >> dynamic_events
+
+echo 1 > events/tests/enable
+
+echo > trace
+cat trace > /dev/null
+
+function streq() {
+	test $1 = $2
+}
+
+streq `grep -A 1 -m 1 myevent1 trace | sed -r 's/^.*(arg=.*)/\1/' `

tools/testing/selftests/ftrace/test.d/dynevent/fprobe_syntax_errors.tc

@@ -34,7 +34,9 @@ check_error 'f vfs_read ^$stack10000'	# BAD_STACK_NUM
 
 check_error 'f vfs_read ^$arg10000'	# BAD_ARG_NUM
 
+if !grep -q 'kernel return probes support:' README; then
 check_error 'f vfs_read $retval ^$arg1' # BAD_VAR
+fi
 check_error 'f vfs_read ^$none_var'	# BAD_VAR
 check_error 'f vfs_read ^'$REG		# BAD_VAR
 
@@ -99,7 +101,9 @@ if grep -q "<argname>" README; then
 check_error 'f vfs_read args=^$arg*'		# BAD_VAR_ARGS
 check_error 'f vfs_read +0(^$arg*)'		# BAD_VAR_ARGS
 check_error 'f vfs_read $arg* ^$arg*'		# DOUBLE_ARGS
+if !grep -q 'kernel return probes support:' README; then
 check_error 'f vfs_read%return ^$arg*'		# NOFENTRY_ARGS
+fi
 check_error 'f vfs_read ^hoge'			# NO_BTFARG
 check_error 'f kfree ^$arg10'			# NO_BTFARG (exceed the number of parameters)
 check_error 'f kfree%return ^$retval'		# NO_RETVAL

tools/testing/selftests/ftrace/test.d/kprobe/kprobe_syntax_errors.tc

@@ -108,7 +108,9 @@ if grep -q "<argname>" README; then
 check_error 'p vfs_read args=^$arg*'		# BAD_VAR_ARGS
 check_error 'p vfs_read +0(^$arg*)'		# BAD_VAR_ARGS
 check_error 'p vfs_read $arg* ^$arg*'		# DOUBLE_ARGS
+if !grep -q 'kernel return probes support:' README; then
 check_error 'r vfs_read ^$arg*'			# NOFENTRY_ARGS
+fi
 check_error 'p vfs_read+8 ^$arg*'		# NOFENTRY_ARGS
 check_error 'p vfs_read ^hoge'			# NO_BTFARG
 check_error 'p kfree ^$arg10'			# NO_BTFARG (exceed the number of parameters)

tools/testing/selftests/ftrace/test.d/kprobe/kretprobe_entry_arg.tc

+#!/bin/sh
+# SPDX-License-Identifier: GPL-2.0
+# description: Kretprobe entry argument access
+# requires: kprobe_events 'kernel return probes support:':README
+
+echo 'p:myevent1 vfs_open arg=$arg1' >> kprobe_events
+echo 'r:myevent2 vfs_open arg=$arg1' >> kprobe_events
+
+echo 1 > events/kprobes/enable
+
+echo > trace
+cat trace > /dev/null
+
+function streq() {
+	test $1 = $2
+}
+
+streq `grep -A 1 -m 1 myevent1 trace | sed -r 's/^.*(arg=.*)/\1/' `