2020-07-09 02:20:40

by Guo Ren

Subject: [PATCH v2 0/6] riscv: Add k/uprobe supported

From: Guo Ren <[email protected]>

The patchset includes kprobe/uprobe support and some related fixups.
Patrick provides HAVE_REGS_AND_STACK_ACCESS_API support and some
kprobe's code. The framework of k/uprobe is from csky but also refers
to other arches'. kprobes on ftrace is also supported in the patchset.

There is no single step exception in riscv ISA, only single-step
facility for jtag. See riscv-Privileged spec:

Interrupt Exception Code-Description
1 0 Reserved
1 1 Supervisor software interrupt
1 2–4 Reserved
1 5 Supervisor timer interrupt
1 6–8 Reserved
1 9 Supervisor external interrupt
1 10–15 Reserved
1 ≥16 Available for platform use
0 0 Instruction address misaligned
0 1 Instruction access fault
0 2 Illegal instruction
0 3 Breakpoint
0 4 Load address misaligned
0 5 Load access fault
0 6 Store/AMO address misaligned
0 7 Store/AMO access fault
0 8 Environment call from U-mode
0 9 Environment call from S-mode
0 10–11 Reserved
0 12 Instruction page fault
0 13 Load page fault
0 14 Reserved
0 15 Store/AMO page fault
0 16–23 Reserved
0 24–31 Available for custom use
0 32–47 Reserved
0 48–63 Available for custom use
0 ≥64 Reserved

No single step!

Other arches use hardware single-step exception for k/uprobe, eg:
- powerpc: regs->msr |= MSR_SINGLESTEP
- arm/arm64: PSTATE.D for enabling software step exceptions
- s390: Set PER control regs, turns on single step for the given address
- x86: regs->flags |= X86_EFLAGS_TF
- csky: of course use hw single step :)

All the above arches use a hardware single-step exception
mechanism to execute the instruction that was replaced with a probe
breakpoint. So utilize ebreak to simulate.

Some pc related instructions couldn't be executed out of line and some
system/fence instructions couldn't be a trace site at all. So we give
out a reject list and simulate list in decode-insn.c.

You could use uprobe to test simulate code like this:

echo 'p:enter_current_state_one /hello:0x6e4 a0=%a0 a1=%a1' >> /sys/kernel/debug/tracing/uprobe_events
echo 1 > /sys/kernel/debug/tracing/events/uprobes/enable
/hello
^C
cat /sys/kernel/debug/tracing/trace
tracer: nop

entries-in-buffer/entries-written: 1/1 #P:1

_-----=> irqs-off
/ _----=> need-resched
| / _---=> hardirq/softirq
|| / _--=> preempt-depth
||| / delay
TASK-PID CPU# |||| TIMESTAMP FUNCTION
| | | |||| | |
hello-94 [000] d... 55.404242: enter_current_state_one: (0x106e4) a0=0x1 a1=0x3fffa8ada8

Be care /hello:0x6e4 is the file offset in elf and it relate to 0x106e4
in memory and hello is your target elf program.

Try kprobe like this:

echo 'p:myprobe _do_fork dfd=%a0 filename=%a1 flags=%a2 mode=+4($stack)' > /sys/kernel/debug/tracing/kprobe_events
echo 'r:myretprobe _do_fork $retval' >> /sys/kernel/debug/tracing/kprobe_event

echo 1 >/sys/kernel/debug/tracing/events/kprobes/enable
cat /sys/kernel/debug/tracing/trace
tracer: nop

entries-in-buffer/entries-written: 2/2 #P:1

_-----=> irqs-off
/ _----=> need-resched
| / _---=> hardirq/softirq
|| / _--=> preempt-depth
||| / delay
TASK-PID CPU# |||| TIMESTAMP FUNCTION
| | | |||| | |
sh-92 [000] .n.. 131.804230: myprobe: (_do_fork+0x0/0x2e6) dfd=0xffffffe03929fdf8 filename=0x0 flags=0x101000 mode=0x1200000ffffffe0
sh-92 [000] d... 131.806607: myretprobe: (__do_sys_clone+0x70/0x82 <- _do_fork) arg1=0x5f
cat /sys/kernel/debug/tracing/trace

Changes v2:
- Add Reviewed-by, Tested-by, Acked-by, thx for all of you
- Add kprobes on ftrace feature

Guo Ren (5):
riscv: Fixup __vdso_gettimeofday broke dynamic ftrace
riscv: Fixup compile error BUILD_BUG_ON failed
riscv: Add kprobes supported
riscv: Add uprobes supported
riscv: Add KPROBES_ON_FTRACE supported

Patrick Stählin (1):
RISC-V: Implement ptrace regs and stack API

arch/riscv/Kconfig | 7 +
arch/riscv/include/asm/kprobes.h | 40 +++
arch/riscv/include/asm/probes.h | 24 ++
arch/riscv/include/asm/processor.h | 1 +
arch/riscv/include/asm/ptrace.h | 29 ++
arch/riscv/include/asm/thread_info.h | 4 +-
arch/riscv/include/asm/uprobes.h | 40 +++
arch/riscv/kernel/Makefile | 1 +
arch/riscv/kernel/patch.c | 8 +-
arch/riscv/kernel/probes/Makefile | 6 +
arch/riscv/kernel/probes/decode-insn.c | 48 +++
arch/riscv/kernel/probes/decode-insn.h | 18 +
arch/riscv/kernel/probes/ftrace.c | 53 +++
arch/riscv/kernel/probes/kprobes.c | 471 ++++++++++++++++++++++++++
arch/riscv/kernel/probes/kprobes_trampoline.S | 93 +++++
arch/riscv/kernel/probes/simulate-insn.c | 85 +++++
arch/riscv/kernel/probes/simulate-insn.h | 47 +++
arch/riscv/kernel/probes/uprobes.c | 186 ++++++++++
arch/riscv/kernel/ptrace.c | 99 ++++++
arch/riscv/kernel/signal.c | 3 +
arch/riscv/kernel/traps.c | 19 ++
arch/riscv/kernel/vdso/Makefile | 3 +
arch/riscv/mm/fault.c | 11 +
23 files changed, 1293 insertions(+), 3 deletions(-)
create mode 100644 arch/riscv/include/asm/probes.h
create mode 100644 arch/riscv/include/asm/uprobes.h
create mode 100644 arch/riscv/kernel/probes/Makefile
create mode 100644 arch/riscv/kernel/probes/decode-insn.c
create mode 100644 arch/riscv/kernel/probes/decode-insn.h
create mode 100644 arch/riscv/kernel/probes/ftrace.c
create mode 100644 arch/riscv/kernel/probes/kprobes.c
create mode 100644 arch/riscv/kernel/probes/kprobes_trampoline.S
create mode 100644 arch/riscv/kernel/probes/simulate-insn.c
create mode 100644 arch/riscv/kernel/probes/simulate-insn.h
create mode 100644 arch/riscv/kernel/probes/uprobes.c

--
2.7.4

2020-07-09 02:22:49

by Guo Ren

Subject: [PATCH v2 1/6] riscv: Fixup __vdso_gettimeofday broke dynamic ftrace

From: Guo Ren <[email protected]>

For linux-5.8-rc1, enable ftrace of riscv will cause boot panic:

[ 2.388980] Run /sbin/init as init process
[ 2.529938] init[39]: unhandled signal 4 code 0x1 at 0x0000003ff449e000
[ 2.531078] CPU: 0 PID: 39 Comm: init Not tainted 5.8.0-rc1-dirty #13
[ 2.532719] epc: 0000003ff449e000 ra : 0000003ff449e954 sp : 0000003fffedb900
[ 2.534005] gp : 00000000000e8528 tp : 0000003ff449d800 t0 : 000000000000001e
[ 2.534965] t1 : 000000000000000a t2 : 0000003fffedb89e s0 : 0000003fffedb920
[ 2.536279] s1 : 0000003fffedb940 a0 : 0000003ff43d4b2c a1 : 0000000000000000
[ 2.537334] a2 : 0000000000000001 a3 : 0000000000000000 a4 : fffffffffbad8000
[ 2.538466] a5 : 0000003ff449e93a a6 : 0000000000000000 a7 : 0000000000000000
[ 2.539511] s2 : 0000000000000000 s3 : 0000003ff448412c s4 : 0000000000000010
[ 2.541260] s5 : 0000000000000016 s6 : 00000000000d0a30 s7 : 0000003fffedba70
[ 2.542152] s8 : 0000000000000000 s9 : 0000000000000000 s10: 0000003fffedb960
[ 2.543335] s11: 0000000000000000 t3 : 0000000000000000 t4 : 0000003fffedb8a0
[ 2.544471] t5 : 0000000000000000 t6 : 0000000000000000
[ 2.545730] status: 0000000000004020 badaddr: 00000000464c457f cause: 0000000000000002
[ 2.549867] Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000004
[ 2.551267] CPU: 0 PID: 1 Comm: init Not tainted 5.8.0-rc1-dirty #13
[ 2.552061] Call Trace:
[ 2.552626] [<ffffffe00020374a>] walk_stackframe+0x0/0xc4
[ 2.553486] [<ffffffe0002039f4>] show_stack+0x40/0x4c
[ 2.553995] [<ffffffe00054a6ae>] dump_stack+0x7a/0x98
[ 2.554615] [<ffffffe00020b9b8>] panic+0x114/0x2f4
[ 2.555395] [<ffffffe00020ebd6>] do_exit+0x89c/0x8c2
[ 2.555949] [<ffffffe00020f930>] do_group_exit+0x3a/0x90
[ 2.556715] [<ffffffe000219e08>] get_signal+0xe2/0x6e6
[ 2.557388] [<ffffffe000202d72>] do_notify_resume+0x6a/0x37a
[ 2.558089] [<ffffffe000201c16>] ret_from_exception+0x0/0xc

"ra:0x3ff449e954" is the return address of "call _mcount" in the
prologue of __vdso_gettimeofday(). Without proper relocate, pc jmp
to 0x0000003ff449e000 (vdso map base) with a illegal instruction
trap.

The solution comes from arch/arm64/kernel/vdso/Makefile:

CFLAGS_REMOVE_vgettimeofday.o = $(CC_FLAGS_FTRACE) -Os $(CC_FLAGS_SCS)

- CC_FLAGS_SCS is ShadowCallStack feature in Clang and only
implemented for arm64, no use for riscv.

The bug comes from the following commit:

ad5d1122b82f ("riscv: use vDSO common flow to reduce the latency of the time-related functions")

Signed-off-by: Guo Ren <[email protected]>
Reviewed-by: Pekka Enberg <[email protected]>
Cc: Vincent Chen <[email protected]>
Cc: Atish Patra <[email protected]>
Cc: Palmer Dabbelt <[email protected]>
Cc: Alan Kao <[email protected]>
Cc: Greentime Hu <[email protected]>
---
arch/riscv/kernel/vdso/Makefile | 3 +++
1 file changed, 3 insertions(+)

diff --git a/arch/riscv/kernel/vdso/Makefile b/arch/riscv/kernel/vdso/Makefile
index 38ba55b..3079935 100644
--- a/arch/riscv/kernel/vdso/Makefile
+++ b/arch/riscv/kernel/vdso/Makefile
@@ -27,6 +27,9 @@ obj-vdso := $(addprefix $(obj)/, $(obj-vdso))
obj-y += vdso.o vdso-syms.o
CPPFLAGS_vdso.lds += -P -C -U$(ARCH)

+# Disable -pg to prevent insert call site
+CFLAGS_REMOVE_vgettimeofday.o = $(CC_FLAGS_FTRACE) -Os
+
# Disable gcov profiling for VDSO code
GCOV_PROFILE := n

--
2.7.4

2020-07-09 02:22:53

by Guo Ren

Subject: [PATCH v2 2/6] RISC-V: Implement ptrace regs and stack API

From: Patrick Stählin <[email protected]>

Needed for kprobes support. Copied and adapted from arm64 code.

Guo Ren fixup pt_regs type for linux-5.8-rc1.

Signed-off-by: Patrick Stählin <[email protected]>
Signed-off-by: Guo Ren <[email protected]>
Reviewed-by: Pekka Enberg <[email protected]>
Reviewed-by: Zong Li <[email protected]>
---
arch/riscv/Kconfig | 1 +
arch/riscv/include/asm/ptrace.h | 29 ++++++++++++
arch/riscv/kernel/ptrace.c | 99 +++++++++++++++++++++++++++++++++++++++++
3 files changed, 129 insertions(+)

diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig
index 128192e..58d6f66 100644
--- a/arch/riscv/Kconfig
+++ b/arch/riscv/Kconfig
@@ -76,6 +76,7 @@ config RISCV
select SPARSE_IRQ
select SYSCTL_EXCEPTION_TRACE
select THREAD_INFO_IN_TASK
+ select HAVE_REGS_AND_STACK_ACCESS_API

config ARCH_MMAP_RND_BITS_MIN
default 18 if 64BIT
diff --git a/arch/riscv/include/asm/ptrace.h b/arch/riscv/include/asm/ptrace.h
index ee49f80..23372bb 100644
--- a/arch/riscv/include/asm/ptrace.h
+++ b/arch/riscv/include/asm/ptrace.h
@@ -8,6 +8,7 @@

#include <uapi/asm/ptrace.h>
#include <asm/csr.h>
+#include <linux/compiler.h>

#ifndef __ASSEMBLY__

@@ -60,6 +61,7 @@ struct pt_regs {

#define user_mode(regs) (((regs)->status & SR_PP) == 0)

+#define MAX_REG_OFFSET offsetof(struct pt_regs, orig_a0)

/* Helpers for working with the instruction pointer */
static inline unsigned long instruction_pointer(struct pt_regs *regs)
@@ -85,6 +87,12 @@ static inline void user_stack_pointer_set(struct pt_regs *regs,
regs->sp = val;
}

+/* Valid only for Kernel mode traps. */
+static inline unsigned long kernel_stack_pointer(struct pt_regs *regs)
+{
+ return regs->sp;
+}
+
/* Helpers for working with the frame pointer */
static inline unsigned long frame_pointer(struct pt_regs *regs)
{
@@ -101,6 +109,27 @@ static inline unsigned long regs_return_value(struct pt_regs *regs)
return regs->a0;
}

+extern int regs_query_register_offset(const char *name);
+extern unsigned long regs_get_kernel_stack_nth(struct pt_regs *regs,
+ unsigned int n);
+
+/**
+ * regs_get_register() - get register value from its offset
+ * @regs: pt_regs from which register value is gotten
+ * @offset: offset of the register.
+ *
+ * regs_get_register returns the value of a register whose offset from @regs.
+ * The @offset is the offset of the register in struct pt_regs.
+ * If @offset is bigger than MAX_REG_OFFSET, this returns 0.
+ */
+static inline unsigned long regs_get_register(struct pt_regs *regs,
+ unsigned int offset)
+{
+ if (unlikely(offset > MAX_REG_OFFSET))
+ return 0;
+
+ return *(unsigned long *)((unsigned long)regs + offset);
+}
#endif /* __ASSEMBLY__ */

#endif /* _ASM_RISCV_PTRACE_H */
diff --git a/arch/riscv/kernel/ptrace.c b/arch/riscv/kernel/ptrace.c
index 444dc7b..a11c692 100644
--- a/arch/riscv/kernel/ptrace.c
+++ b/arch/riscv/kernel/ptrace.c
@@ -125,6 +125,105 @@ const struct user_regset_view *task_user_regset_view(struct task_struct *task)
return &riscv_user_native_view;
}

+struct pt_regs_offset {
+ const char *name;
+ int offset;
+};
+
+#define REG_OFFSET_NAME(r) {.name = #r, .offset = offsetof(struct pt_regs, r)}
+#define REG_OFFSET_END {.name = NULL, .offset = 0}
+
+static const struct pt_regs_offset regoffset_table[] = {
+ REG_OFFSET_NAME(epc),
+ REG_OFFSET_NAME(ra),
+ REG_OFFSET_NAME(sp),
+ REG_OFFSET_NAME(gp),
+ REG_OFFSET_NAME(tp),
+ REG_OFFSET_NAME(t0),
+ REG_OFFSET_NAME(t1),
+ REG_OFFSET_NAME(t2),
+ REG_OFFSET_NAME(s0),
+ REG_OFFSET_NAME(s1),
+ REG_OFFSET_NAME(a0),
+ REG_OFFSET_NAME(a1),
+ REG_OFFSET_NAME(a2),
+ REG_OFFSET_NAME(a3),
+ REG_OFFSET_NAME(a4),
+ REG_OFFSET_NAME(a5),
+ REG_OFFSET_NAME(a6),
+ REG_OFFSET_NAME(a7),
+ REG_OFFSET_NAME(s2),
+ REG_OFFSET_NAME(s3),
+ REG_OFFSET_NAME(s4),
+ REG_OFFSET_NAME(s5),
+ REG_OFFSET_NAME(s6),
+ REG_OFFSET_NAME(s7),
+ REG_OFFSET_NAME(s8),
+ REG_OFFSET_NAME(s9),
+ REG_OFFSET_NAME(s10),
+ REG_OFFSET_NAME(s11),
+ REG_OFFSET_NAME(t3),
+ REG_OFFSET_NAME(t4),
+ REG_OFFSET_NAME(t5),
+ REG_OFFSET_NAME(t6),
+ REG_OFFSET_NAME(status),
+ REG_OFFSET_NAME(badaddr),
+ REG_OFFSET_NAME(cause),
+ REG_OFFSET_NAME(orig_a0),
+ REG_OFFSET_END,
+};
+
+/**
+ * regs_query_register_offset() - query register offset from its name
+ * @name: the name of a register
+ *
+ * regs_query_register_offset() returns the offset of a register in struct
+ * pt_regs from its name. If the name is invalid, this returns -EINVAL;
+ */
+int regs_query_register_offset(const char *name)
+{
+ const struct pt_regs_offset *roff;
+
+ for (roff = regoffset_table; roff->name != NULL; roff++)
+ if (!strcmp(roff->name, name))
+ return roff->offset;
+ return -EINVAL;
+}
+
+/**
+ * regs_within_kernel_stack() - check the address in the stack
+ * @regs: pt_regs which contains kernel stack pointer.
+ * @addr: address which is checked.
+ *
+ * regs_within_kernel_stack() checks @addr is within the kernel stack page(s).
+ * If @addr is within the kernel stack, it returns true. If not, returns false.
+ */
+static bool regs_within_kernel_stack(struct pt_regs *regs, unsigned long addr)
+{
+ return (addr & ~(THREAD_SIZE - 1)) ==
+ (kernel_stack_pointer(regs) & ~(THREAD_SIZE - 1));
+}
+
+/**
+ * regs_get_kernel_stack_nth() - get Nth entry of the stack
+ * @regs: pt_regs which contains kernel stack pointer.
+ * @n: stack entry number.
+ *
+ * regs_get_kernel_stack_nth() returns @n th entry of the kernel stack which
+ * is specified by @regs. If the @n th entry is NOT in the kernel stack,
+ * this returns 0.
+ */
+unsigned long regs_get_kernel_stack_nth(struct pt_regs *regs, unsigned int n)
+{
+ unsigned long *addr = (unsigned long *)kernel_stack_pointer(regs);
+
+ addr += n;
+ if (regs_within_kernel_stack(regs, (unsigned long)addr))
+ return *addr;
+ else
+ return 0;
+}
+
void ptrace_disable(struct task_struct *child)
{
clear_tsk_thread_flag(child, TIF_SYSCALL_TRACE);
--
2.7.4

2020-07-09 02:22:59

by Guo Ren

Subject: [PATCH v2 3/6] riscv: Fixup compile error BUILD_BUG_ON failed

From: Guo Ren <[email protected]>

Unfortunately, the current code couldn't be compiled:

CC arch/riscv/kernel/patch.o
In file included from ./include/linux/kernel.h:11,
from ./include/linux/list.h:9,
from ./include/linux/preempt.h:11,
from ./include/linux/spinlock.h:51,
from arch/riscv/kernel/patch.c:6:
In function ‘fix_to_virt’,
inlined from ‘patch_map’ at arch/riscv/kernel/patch.c:37:17:
./include/linux/compiler.h:392:38: error: call to ‘__compiletime_assert_205’ declared with attribute error: BUILD_BUG_ON failed: idx >= __end_of_fixed_addresses
_compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
^
./include/linux/compiler.h:373:4: note: in definition of macro ‘__compiletime_assert’
prefix ## suffix(); \
^~~~~~
./include/linux/compiler.h:392:2: note: in expansion of macro ‘_compiletime_assert’
_compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
^~~~~~~~~~~~~~~~~~~
./include/linux/build_bug.h:39:37: note: in expansion of macro ‘compiletime_assert’
#define BUILD_BUG_ON_MSG(cond, msg) compiletime_assert(!(cond), msg)
^~~~~~~~~~~~~~~~~~
./include/linux/build_bug.h:50:2: note: in expansion of macro ‘BUILD_BUG_ON_MSG’
BUILD_BUG_ON_MSG(condition, "BUILD_BUG_ON failed: " #condition)
^~~~~~~~~~~~~~~~
./include/asm-generic/fixmap.h:32:2: note: in expansion of macro ‘BUILD_BUG_ON’
BUILD_BUG_ON(idx >= __end_of_fixed_addresses);
^~~~~~~~~~~~

Because fix_to_virt(, idx) needs a const value, not a dynamic variable of
reg-a0 or BUILD_BUG_ON failed with "idx >= __end_of_fixed_addresses".

Signed-off-by: Guo Ren <[email protected]>
Reviewed-by: Masami Hiramatsu <[email protected]>
Reviewed-by: Pekka Enberg <[email protected]>
Cc: Zong Li <[email protected]>
---
Changelog V2:
- Use __always_inline as same as fix_to_virt
- Use const "const unsigned int" for 2th param

Signed-off-by: Guo Ren <[email protected]>
---
arch/riscv/kernel/patch.c | 8 ++++++--
1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/arch/riscv/kernel/patch.c b/arch/riscv/kernel/patch.c
index d4a64df..3179a4e 100644
--- a/arch/riscv/kernel/patch.c
+++ b/arch/riscv/kernel/patch.c
@@ -20,7 +20,12 @@ struct patch_insn {
};

#ifdef CONFIG_MMU
-static void *patch_map(void *addr, int fixmap)
+/*
+ * The fix_to_virt(, idx) needs a const value (not a dynamic variable of
+ * reg-a0) or BUILD_BUG_ON failed with "idx >= __end_of_fixed_addresses".
+ * So use '__always_inline' and 'const unsigned int fixmap' here.
+ */
+static __always_inline void *patch_map(void *addr, const unsigned int fixmap)
{
uintptr_t uintaddr = (uintptr_t) addr;
struct page *page;
@@ -37,7 +42,6 @@ static void *patch_map(void *addr, int fixmap)
return (void *)set_fixmap_offset(fixmap, page_to_phys(page) +
(uintaddr & ~PAGE_MASK));
}
-NOKPROBE_SYMBOL(patch_map);

static void patch_unmap(int fixmap)
{
--
2.7.4

2020-07-09 02:23:11

by Guo Ren

Subject: [PATCH v2 4/6] riscv: Add kprobes supported

From: Guo Ren <[email protected]>

This patch enables "kprobe & kretprobe" to work with ftrace
interface. It utilized software breakpoint as single-step
mechanism.

Some instructions which can't be single-step executed must be
simulated in kernel execution slot, such as: branch, jal, auipc,
la ...

Some instructions should be rejected for probing and we use a
blacklist to filter, such as: ecall, ebreak, ...

We use ebreak & c.ebreak to replace origin instruction and the
kprobe handler prepares an executable memory slot for out-of-line
execution with a copy of the original instruction being probed.
In execution slot we add ebreak behind original instruction to
simulate a single-setp mechanism.

The patch is based on packi's work [1] and csky's work [2].
- The kprobes_trampoline.S is all from packi's patch
- The single-step mechanism is new designed for riscv without hw
single-step trap
- The simulation codes are from csky
- Frankly, all codes refer to other archs' implementation

[1] https://lore.kernel.org/linux-riscv/[email protected]/
[2] https://lore.kernel.org/linux-csky/[email protected]/

Signed-off-by: Guo Ren <[email protected]>
Co-Developed-by: Patrick Stählin <[email protected]>
Acked-by: Masami Hiramatsu <[email protected]>
Tested-by: Zong Li <[email protected]>
Reviewed-by: Pekka Enberg <[email protected]>
Cc: Patrick Stählin <[email protected]>
Cc: Palmer Dabbelt <[email protected]>
Cc: Björn Töpel <[email protected]>
---
arch/riscv/Kconfig | 2 +
arch/riscv/include/asm/kprobes.h | 40 +++
arch/riscv/include/asm/probes.h | 24 ++
arch/riscv/kernel/Makefile | 1 +
arch/riscv/kernel/probes/Makefile | 4 +
arch/riscv/kernel/probes/decode-insn.c | 48 +++
arch/riscv/kernel/probes/decode-insn.h | 18 +
arch/riscv/kernel/probes/kprobes.c | 471 ++++++++++++++++++++++++++
arch/riscv/kernel/probes/kprobes_trampoline.S | 93 +++++
arch/riscv/kernel/probes/simulate-insn.c | 85 +++++
arch/riscv/kernel/probes/simulate-insn.h | 47 +++
arch/riscv/kernel/traps.c | 9 +
arch/riscv/mm/fault.c | 4 +
13 files changed, 846 insertions(+)
create mode 100644 arch/riscv/include/asm/probes.h
create mode 100644 arch/riscv/kernel/probes/Makefile
create mode 100644 arch/riscv/kernel/probes/decode-insn.c
create mode 100644 arch/riscv/kernel/probes/decode-insn.h
create mode 100644 arch/riscv/kernel/probes/kprobes.c
create mode 100644 arch/riscv/kernel/probes/kprobes_trampoline.S
create mode 100644 arch/riscv/kernel/probes/simulate-insn.c
create mode 100644 arch/riscv/kernel/probes/simulate-insn.h

diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig
index 58d6f66..a295f0b 100644
--- a/arch/riscv/Kconfig
+++ b/arch/riscv/Kconfig
@@ -57,6 +57,8 @@ config RISCV
select HAVE_EBPF_JIT if MMU
select HAVE_FUTEX_CMPXCHG if FUTEX
select HAVE_GENERIC_VDSO if MMU && 64BIT
+ select HAVE_KPROBES
+ select HAVE_KRETPROBES
select HAVE_PCI
select HAVE_PERF_EVENTS
select HAVE_PERF_REGS
diff --git a/arch/riscv/include/asm/kprobes.h b/arch/riscv/include/asm/kprobes.h
index 56a98ea3..4647d38 100644
--- a/arch/riscv/include/asm/kprobes.h
+++ b/arch/riscv/include/asm/kprobes.h
@@ -11,4 +11,44 @@

#include <asm-generic/kprobes.h>

+#ifdef CONFIG_KPROBES
+#include <linux/types.h>
+#include <linux/ptrace.h>
+#include <linux/percpu.h>
+
+#define __ARCH_WANT_KPROBES_INSN_SLOT
+#define MAX_INSN_SIZE 2
+
+#define flush_insn_slot(p) do { } while (0)
+#define kretprobe_blacklist_size 0
+
+#include <asm/probes.h>
+
+struct prev_kprobe {
+ struct kprobe *kp;
+ unsigned int status;
+};
+
+/* Single step context for kprobe */
+struct kprobe_step_ctx {
+ unsigned long ss_pending;
+ unsigned long match_addr;
+};
+
+/* per-cpu kprobe control block */
+struct kprobe_ctlblk {
+ unsigned int kprobe_status;
+ unsigned long saved_status;
+ struct prev_kprobe prev_kprobe;
+ struct kprobe_step_ctx ss_ctx;
+};
+
+void arch_remove_kprobe(struct kprobe *p);
+int kprobe_fault_handler(struct pt_regs *regs, unsigned int trapnr);
+bool kprobe_breakpoint_handler(struct pt_regs *regs);
+bool kprobe_single_step_handler(struct pt_regs *regs);
+void kretprobe_trampoline(void);
+void __kprobes *trampoline_probe_handler(struct pt_regs *regs);
+
+#endif /* CONFIG_KPROBES */
#endif /* _ASM_RISCV_KPROBES_H */
diff --git a/arch/riscv/include/asm/probes.h b/arch/riscv/include/asm/probes.h
new file mode 100644
index 00000000..a787e6d
--- /dev/null
+++ b/arch/riscv/include/asm/probes.h
@@ -0,0 +1,24 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+
+#ifndef _ASM_RISCV_PROBES_H
+#define _ASM_RISCV_PROBES_H
+
+typedef u32 probe_opcode_t;
+typedef bool (probes_handler_t) (u32 opcode, unsigned long addr, struct pt_regs *);
+
+/* architecture specific copy of original instruction */
+struct arch_probe_insn {
+ probe_opcode_t *insn;
+ probes_handler_t *handler;
+ /* restore address after simulation */
+ unsigned long restore;
+};
+
+#ifdef CONFIG_KPROBES
+typedef u32 kprobe_opcode_t;
+struct arch_specific_insn {
+ struct arch_probe_insn api;
+};
+#endif
+
+#endif /* _ASM_RISCV_PROBES_H */
diff --git a/arch/riscv/kernel/Makefile b/arch/riscv/kernel/Makefile
index b355cf4..c3fff3e 100644
--- a/arch/riscv/kernel/Makefile
+++ b/arch/riscv/kernel/Makefile
@@ -29,6 +29,7 @@ obj-y += riscv_ksyms.o
obj-y += stacktrace.o
obj-y += cacheinfo.o
obj-y += patch.o
+obj-y += probes/
obj-$(CONFIG_MMU) += vdso.o vdso/

obj-$(CONFIG_RISCV_M_MODE) += clint.o traps_misaligned.o
diff --git a/arch/riscv/kernel/probes/Makefile b/arch/riscv/kernel/probes/Makefile
new file mode 100644
index 00000000..8a39507
--- /dev/null
+++ b/arch/riscv/kernel/probes/Makefile
@@ -0,0 +1,4 @@
+# SPDX-License-Identifier: GPL-2.0
+obj-$(CONFIG_KPROBES) += kprobes.o decode-insn.o simulate-insn.o
+obj-$(CONFIG_KPROBES) += kprobes_trampoline.o
+CFLAGS_REMOVE_simulate-insn.o = $(CC_FLAGS_FTRACE)
diff --git a/arch/riscv/kernel/probes/decode-insn.c b/arch/riscv/kernel/probes/decode-insn.c
new file mode 100644
index 00000000..0876c30
--- /dev/null
+++ b/arch/riscv/kernel/probes/decode-insn.c
@@ -0,0 +1,48 @@
+// SPDX-License-Identifier: GPL-2.0+
+
+#include <linux/kernel.h>
+#include <linux/kprobes.h>
+#include <linux/module.h>
+#include <linux/kallsyms.h>
+#include <asm/sections.h>
+
+#include "decode-insn.h"
+#include "simulate-insn.h"
+
+/* Return:
+ * INSN_REJECTED If instruction is one not allowed to kprobe,
+ * INSN_GOOD_NO_SLOT If instruction is supported but doesn't use its slot.
+ */
+enum probe_insn __kprobes
+riscv_probe_decode_insn(probe_opcode_t *addr, struct arch_probe_insn *api)
+{
+ probe_opcode_t insn = le32_to_cpu(*addr);
+
+ /*
+ * Reject instructions list:
+ */
+ RISCV_INSN_REJECTED(system, insn);
+ RISCV_INSN_REJECTED(fence, insn);
+
+ /*
+ * Simulate instructions list:
+ * TODO: the REJECTED ones below need to be implemented
+ */
+#ifdef CONFIG_RISCV_ISA_C
+ RISCV_INSN_REJECTED(c_j, insn);
+ RISCV_INSN_REJECTED(c_jr, insn);
+ RISCV_INSN_REJECTED(c_jal, insn);
+ RISCV_INSN_REJECTED(c_jalr, insn);
+ RISCV_INSN_REJECTED(c_beqz, insn);
+ RISCV_INSN_REJECTED(c_bnez, insn);
+ RISCV_INSN_REJECTED(c_ebreak, insn);
+#endif
+
+ RISCV_INSN_REJECTED(auipc, insn);
+ RISCV_INSN_REJECTED(branch, insn);
+
+ RISCV_INSN_SET_SIMULATE(jal, insn);
+ RISCV_INSN_SET_SIMULATE(jalr, insn);
+
+ return INSN_GOOD;
+}
diff --git a/arch/riscv/kernel/probes/decode-insn.h b/arch/riscv/kernel/probes/decode-insn.h
new file mode 100644
index 00000000..42269a7
--- /dev/null
+++ b/arch/riscv/kernel/probes/decode-insn.h
@@ -0,0 +1,18 @@
+/* SPDX-License-Identifier: GPL-2.0+ */
+
+#ifndef _RISCV_KERNEL_KPROBES_DECODE_INSN_H
+#define _RISCV_KERNEL_KPROBES_DECODE_INSN_H
+
+#include <asm/sections.h>
+#include <asm/kprobes.h>
+
+enum probe_insn {
+ INSN_REJECTED,
+ INSN_GOOD_NO_SLOT,
+ INSN_GOOD,
+};
+
+enum probe_insn __kprobes
+riscv_probe_decode_insn(probe_opcode_t *addr, struct arch_probe_insn *asi);
+
+#endif /* _RISCV_KERNEL_KPROBES_DECODE_INSN_H */
diff --git a/arch/riscv/kernel/probes/kprobes.c b/arch/riscv/kernel/probes/kprobes.c
new file mode 100644
index 00000000..31b6196
--- /dev/null
+++ b/arch/riscv/kernel/probes/kprobes.c
@@ -0,0 +1,471 @@
+// SPDX-License-Identifier: GPL-2.0+
+
+#include <linux/kprobes.h>
+#include <linux/extable.h>
+#include <linux/slab.h>
+#include <linux/stop_machine.h>
+#include <asm/ptrace.h>
+#include <linux/uaccess.h>
+#include <asm/sections.h>
+#include <asm/cacheflush.h>
+#include <asm/bug.h>
+#include <asm/patch.h>
+
+#include "decode-insn.h"
+
+DEFINE_PER_CPU(struct kprobe *, current_kprobe) = NULL;
+DEFINE_PER_CPU(struct kprobe_ctlblk, kprobe_ctlblk);
+
+static void __kprobes
+post_kprobe_handler(struct kprobe_ctlblk *, struct pt_regs *);
+
+static void __kprobes arch_prepare_ss_slot(struct kprobe *p)
+{
+ unsigned long offset = GET_INSN_LENGTH(p->opcode);
+
+ p->ainsn.api.restore = (unsigned long)p->addr + offset;
+
+ patch_text(p->ainsn.api.insn, p->opcode);
+ patch_text((void *)((unsigned long)(p->ainsn.api.insn) + offset),
+ __BUG_INSN_32);
+}
+
+static void __kprobes arch_prepare_simulate(struct kprobe *p)
+{
+ p->ainsn.api.restore = 0;
+}
+
+static void __kprobes arch_simulate_insn(struct kprobe *p, struct pt_regs *regs)
+{
+ struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
+
+ if (p->ainsn.api.handler)
+ p->ainsn.api.handler((u32)p->opcode,
+ (unsigned long)p->addr, regs);
+
+ post_kprobe_handler(kcb, regs);
+}
+
+int __kprobes arch_prepare_kprobe(struct kprobe *p)
+{
+ unsigned long probe_addr = (unsigned long)p->addr;
+
+ if (probe_addr & 0x1) {
+ pr_warn("Address not aligned.\n");
+
+ return -EINVAL;
+ }
+
+ /* copy instruction */
+ p->opcode = le32_to_cpu(*p->addr);
+
+ /* decode instruction */
+ switch (riscv_probe_decode_insn(p->addr, &p->ainsn.api)) {
+ case INSN_REJECTED: /* insn not supported */
+ return -EINVAL;
+
+ case INSN_GOOD_NO_SLOT: /* insn need simulation */
+ p->ainsn.api.insn = NULL;
+ break;
+
+ case INSN_GOOD: /* instruction uses slot */
+ p->ainsn.api.insn = get_insn_slot();
+ if (!p->ainsn.api.insn)
+ return -ENOMEM;
+ break;
+ }
+
+ /* prepare the instruction */
+ if (p->ainsn.api.insn)
+ arch_prepare_ss_slot(p);
+ else
+ arch_prepare_simulate(p);
+
+ return 0;
+}
+
+/* install breakpoint in text */
+void __kprobes arch_arm_kprobe(struct kprobe *p)
+{
+ if ((p->opcode & __INSN_LENGTH_MASK) == __INSN_LENGTH_32)
+ patch_text(p->addr, __BUG_INSN_32);
+ else
+ patch_text(p->addr, __BUG_INSN_16);
+}
+
+/* remove breakpoint from text */
+void __kprobes arch_disarm_kprobe(struct kprobe *p)
+{
+ patch_text(p->addr, p->opcode);
+}
+
+void __kprobes arch_remove_kprobe(struct kprobe *p)
+{
+}
+
+static void __kprobes save_previous_kprobe(struct kprobe_ctlblk *kcb)
+{
+ kcb->prev_kprobe.kp = kprobe_running();
+ kcb->prev_kprobe.status = kcb->kprobe_status;
+}
+
+static void __kprobes restore_previous_kprobe(struct kprobe_ctlblk *kcb)
+{
+ __this_cpu_write(current_kprobe, kcb->prev_kprobe.kp);
+ kcb->kprobe_status = kcb->prev_kprobe.status;
+}
+
+static void __kprobes set_current_kprobe(struct kprobe *p)
+{
+ __this_cpu_write(current_kprobe, p);
+}
+
+/*
+ * Interrupts need to be disabled before single-step mode is set, and not
+ * reenabled until after single-step mode ends.
+ * Without disabling interrupt on local CPU, there is a chance of
+ * interrupt occurrence in the period of exception return and start of
+ * out-of-line single-step, that result in wrongly single stepping
+ * into the interrupt handler.
+ */
+static void __kprobes kprobes_save_local_irqflag(struct kprobe_ctlblk *kcb,
+ struct pt_regs *regs)
+{
+ kcb->saved_status = regs->status;
+ regs->status &= ~SR_SPIE;
+}
+
+static void __kprobes kprobes_restore_local_irqflag(struct kprobe_ctlblk *kcb,
+ struct pt_regs *regs)
+{
+ regs->status = kcb->saved_status;
+}
+
+static void __kprobes
+set_ss_context(struct kprobe_ctlblk *kcb, unsigned long addr, struct kprobe *p)
+{
+ unsigned long offset = GET_INSN_LENGTH(p->opcode);
+
+ kcb->ss_ctx.ss_pending = true;
+ kcb->ss_ctx.match_addr = addr + offset;
+}
+
+static void __kprobes clear_ss_context(struct kprobe_ctlblk *kcb)
+{
+ kcb->ss_ctx.ss_pending = false;
+ kcb->ss_ctx.match_addr = 0;
+}
+
+static void __kprobes setup_singlestep(struct kprobe *p,
+ struct pt_regs *regs,
+ struct kprobe_ctlblk *kcb, int reenter)
+{
+ unsigned long slot;
+
+ if (reenter) {
+ save_previous_kprobe(kcb);
+ set_current_kprobe(p);
+ kcb->kprobe_status = KPROBE_REENTER;
+ } else {
+ kcb->kprobe_status = KPROBE_HIT_SS;
+ }
+
+ if (p->ainsn.api.insn) {
+ /* prepare for single stepping */
+ slot = (unsigned long)p->ainsn.api.insn;
+
+ set_ss_context(kcb, slot, p); /* mark pending ss */
+
+ /* IRQs and single stepping do not mix well. */
+ kprobes_save_local_irqflag(kcb, regs);
+
+ instruction_pointer_set(regs, slot);
+ } else {
+ /* insn simulation */
+ arch_simulate_insn(p, regs);
+ }
+}
+
+static int __kprobes reenter_kprobe(struct kprobe *p,
+ struct pt_regs *regs,
+ struct kprobe_ctlblk *kcb)
+{
+ switch (kcb->kprobe_status) {
+ case KPROBE_HIT_SSDONE:
+ case KPROBE_HIT_ACTIVE:
+ kprobes_inc_nmissed_count(p);
+ setup_singlestep(p, regs, kcb, 1);
+ break;
+ case KPROBE_HIT_SS:
+ case KPROBE_REENTER:
+ pr_warn("Unrecoverable kprobe detected.\n");
+ dump_kprobe(p);
+ BUG();
+ break;
+ default:
+ WARN_ON(1);
+ return 0;
+ }
+
+ return 1;
+}
+
+static void __kprobes
+post_kprobe_handler(struct kprobe_ctlblk *kcb, struct pt_regs *regs)
+{
+ struct kprobe *cur = kprobe_running();
+
+ if (!cur)
+ return;
+
+ /* return addr restore if non-branching insn */
+ if (cur->ainsn.api.restore != 0)
+ regs->epc = cur->ainsn.api.restore;
+
+ /* restore back original saved kprobe variables and continue */
+ if (kcb->kprobe_status == KPROBE_REENTER) {
+ restore_previous_kprobe(kcb);
+ return;
+ }
+
+ /* call post handler */
+ kcb->kprobe_status = KPROBE_HIT_SSDONE;
+ if (cur->post_handler) {
+ /* post_handler can hit breakpoint and single step
+ * again, so we enable D-flag for recursive exception.
+ */
+ cur->post_handler(cur, regs, 0);
+ }
+
+ reset_current_kprobe();
+}
+
+int __kprobes kprobe_fault_handler(struct pt_regs *regs, unsigned int trapnr)
+{
+ struct kprobe *cur = kprobe_running();
+ struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
+
+ switch (kcb->kprobe_status) {
+ case KPROBE_HIT_SS:
+ case KPROBE_REENTER:
+ /*
+ * We are here because the instruction being single
+ * stepped caused a page fault. We reset the current
+ * kprobe and the ip points back to the probe address
+ * and allow the page fault handler to continue as a
+ * normal page fault.
+ */
+ regs->epc = (unsigned long) cur->addr;
+ if (!instruction_pointer(regs))
+ BUG();
+
+ if (kcb->kprobe_status == KPROBE_REENTER)
+ restore_previous_kprobe(kcb);
+ else
+ reset_current_kprobe();
+
+ break;
+ case KPROBE_HIT_ACTIVE:
+ case KPROBE_HIT_SSDONE:
+ /*
+ * We increment the nmissed count for accounting,
+ * we can also use npre/npostfault count for accounting
+ * these specific fault cases.
+ */
+ kprobes_inc_nmissed_count(cur);
+
+ /*
+ * We come here because instructions in the pre/post
+ * handler caused the page_fault, this could happen
+ * if handler tries to access user space by
+ * copy_from_user(), get_user() etc. Let the
+ * user-specified handler try to fix it first.
+ */
+ if (cur->fault_handler && cur->fault_handler(cur, regs, trapnr))
+ return 1;
+
+ /*
+ * In case the user-specified fault handler returned
+ * zero, try to fix up.
+ */
+ if (fixup_exception(regs))
+ return 1;
+ }
+ return 0;
+}
+
+bool __kprobes
+kprobe_breakpoint_handler(struct pt_regs *regs)
+{
+ struct kprobe *p, *cur_kprobe;
+ struct kprobe_ctlblk *kcb;
+ unsigned long addr = instruction_pointer(regs);
+
+ kcb = get_kprobe_ctlblk();
+ cur_kprobe = kprobe_running();
+
+ p = get_kprobe((kprobe_opcode_t *) addr);
+
+ if (p) {
+ if (cur_kprobe) {
+ if (reenter_kprobe(p, regs, kcb))
+ return true;
+ } else {
+ /* Probe hit */
+ set_current_kprobe(p);
+ kcb->kprobe_status = KPROBE_HIT_ACTIVE;
+
+ /*
+ * If we have no pre-handler or it returned 0, we
+ * continue with normal processing. If we have a
+ * pre-handler and it returned non-zero, it will
+ * modify the execution path and no need to single
+ * stepping. Let's just reset current kprobe and exit.
+ *
+ * pre_handler can hit a breakpoint and can step thru
+ * before return.
+ */
+ if (!p->pre_handler || !p->pre_handler(p, regs))
+ setup_singlestep(p, regs, kcb, 0);
+ else
+ reset_current_kprobe();
+ }
+ return true;
+ }
+
+ /*
+ * The breakpoint instruction was removed right
+ * after we hit it. Another cpu has removed
+ * either a probepoint or a debugger breakpoint
+ * at this address. In either case, no further
+ * handling of this interrupt is appropriate.
+ * Return back to original instruction, and continue.
+ */
+ return false;
+}
+
+bool __kprobes
+kprobe_single_step_handler(struct pt_regs *regs)
+{
+ struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
+
+ if ((kcb->ss_ctx.ss_pending)
+ && (kcb->ss_ctx.match_addr == instruction_pointer(regs))) {
+ clear_ss_context(kcb); /* clear pending ss */
+
+ kprobes_restore_local_irqflag(kcb, regs);
+
+ post_kprobe_handler(kcb, regs);
+ return true;
+ }
+ return false;
+}
+
+/*
+ * Provide a blacklist of symbols identifying ranges which cannot be kprobed.
+ * This blacklist is exposed to userspace via debugfs (kprobes/blacklist).
+ */
+int __init arch_populate_kprobe_blacklist(void)
+{
+ int ret;
+
+ ret = kprobe_add_area_blacklist((unsigned long)__irqentry_text_start,
+ (unsigned long)__irqentry_text_end);
+ return ret;
+}
+
+void __kprobes __used *trampoline_probe_handler(struct pt_regs *regs)
+{
+ struct kretprobe_instance *ri = NULL;
+ struct hlist_head *head, empty_rp;
+ struct hlist_node *tmp;
+ unsigned long flags, orig_ret_address = 0;
+ unsigned long trampoline_address =
+ (unsigned long)&kretprobe_trampoline;
+ kprobe_opcode_t *correct_ret_addr = NULL;
+
+ INIT_HLIST_HEAD(&empty_rp);
+ kretprobe_hash_lock(current, &head, &flags);
+
+ /*
+ * It is possible to have multiple instances associated with a given
+ * task either because multiple functions in the call path have
+ * return probes installed on them, and/or more than one
+ * return probe was registered for a target function.
+ *
+ * We can handle this because:
+ * - instances are always pushed into the head of the list
+ * - when multiple return probes are registered for the same
+ * function, the (chronologically) first instance's ret_addr
+ * will be the real return address, and all the rest will
+ * point to kretprobe_trampoline.
+ */
+ hlist_for_each_entry_safe(ri, tmp, head, hlist) {
+ if (ri->task != current)
+ /* another task is sharing our hash bucket */
+ continue;
+
+ orig_ret_address = (unsigned long)ri->ret_addr;
+
+ if (orig_ret_address != trampoline_address)
+ /*
+ * This is the real return address. Any other
+ * instances associated with this task are for
+ * other calls deeper on the call stack
+ */
+ break;
+ }
+
+ kretprobe_assert(ri, orig_ret_address, trampoline_address);
+
+ correct_ret_addr = ri->ret_addr;
+ hlist_for_each_entry_safe(ri, tmp, head, hlist) {
+ if (ri->task != current)
+ /* another task is sharing our hash bucket */
+ continue;
+
+ orig_ret_address = (unsigned long)ri->ret_addr;
+ if (ri->rp && ri->rp->handler) {
+ __this_cpu_write(current_kprobe, &ri->rp->kp);
+ get_kprobe_ctlblk()->kprobe_status = KPROBE_HIT_ACTIVE;
+ ri->ret_addr = correct_ret_addr;
+ ri->rp->handler(ri, regs);
+ __this_cpu_write(current_kprobe, NULL);
+ }
+
+ recycle_rp_inst(ri, &empty_rp);
+
+ if (orig_ret_address != trampoline_address)
+ /*
+ * This is the real return address. Any other
+ * instances associated with this task are for
+ * other calls deeper on the call stack
+ */
+ break;
+ }
+
+ kretprobe_hash_unlock(current, &flags);
+
+ hlist_for_each_entry_safe(ri, tmp, &empty_rp, hlist) {
+ hlist_del(&ri->hlist);
+ kfree(ri);
+ }
+ return (void *)orig_ret_address;
+}
+
+void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri,
+ struct pt_regs *regs)
+{
+ ri->ret_addr = (kprobe_opcode_t *)regs->ra;
+ regs->ra = (unsigned long) &kretprobe_trampoline;
+}
+
+int __kprobes arch_trampoline_kprobe(struct kprobe *p)
+{
+ return 0;
+}
+
+int __init arch_init_kprobes(void)
+{
+ return 0;
+}
diff --git a/arch/riscv/kernel/probes/kprobes_trampoline.S b/arch/riscv/kernel/probes/kprobes_trampoline.S
new file mode 100644
index 00000000..6e85d02
--- /dev/null
+++ b/arch/riscv/kernel/probes/kprobes_trampoline.S
@@ -0,0 +1,93 @@
+/* SPDX-License-Identifier: GPL-2.0+ */
+/*
+ * Author: Patrick Stählin <[email protected]>
+ */
+#include <linux/linkage.h>
+
+#include <asm/asm.h>
+#include <asm/asm-offsets.h>
+
+ .text
+ .altmacro
+
+ .macro save_all_base_regs
+ REG_S x1, PT_RA(sp)
+ REG_S x3, PT_GP(sp)
+ REG_S x4, PT_TP(sp)
+ REG_S x5, PT_T0(sp)
+ REG_S x6, PT_T1(sp)
+ REG_S x7, PT_T2(sp)
+ REG_S x8, PT_S0(sp)
+ REG_S x9, PT_S1(sp)
+ REG_S x10, PT_A0(sp)
+ REG_S x11, PT_A1(sp)
+ REG_S x12, PT_A2(sp)
+ REG_S x13, PT_A3(sp)
+ REG_S x14, PT_A4(sp)
+ REG_S x15, PT_A5(sp)
+ REG_S x16, PT_A6(sp)
+ REG_S x17, PT_A7(sp)
+ REG_S x18, PT_S2(sp)
+ REG_S x19, PT_S3(sp)
+ REG_S x20, PT_S4(sp)
+ REG_S x21, PT_S5(sp)
+ REG_S x22, PT_S6(sp)
+ REG_S x23, PT_S7(sp)
+ REG_S x24, PT_S8(sp)
+ REG_S x25, PT_S9(sp)
+ REG_S x26, PT_S10(sp)
+ REG_S x27, PT_S11(sp)
+ REG_S x28, PT_T3(sp)
+ REG_S x29, PT_T4(sp)
+ REG_S x30, PT_T5(sp)
+ REG_S x31, PT_T6(sp)
+ .endm
+
+ .macro restore_all_base_regs
+ REG_L x3, PT_GP(sp)
+ REG_L x4, PT_TP(sp)
+ REG_L x5, PT_T0(sp)
+ REG_L x6, PT_T1(sp)
+ REG_L x7, PT_T2(sp)
+ REG_L x8, PT_S0(sp)
+ REG_L x9, PT_S1(sp)
+ REG_L x10, PT_A0(sp)
+ REG_L x11, PT_A1(sp)
+ REG_L x12, PT_A2(sp)
+ REG_L x13, PT_A3(sp)
+ REG_L x14, PT_A4(sp)
+ REG_L x15, PT_A5(sp)
+ REG_L x16, PT_A6(sp)
+ REG_L x17, PT_A7(sp)
+ REG_L x18, PT_S2(sp)
+ REG_L x19, PT_S3(sp)
+ REG_L x20, PT_S4(sp)
+ REG_L x21, PT_S5(sp)
+ REG_L x22, PT_S6(sp)
+ REG_L x23, PT_S7(sp)
+ REG_L x24, PT_S8(sp)
+ REG_L x25, PT_S9(sp)
+ REG_L x26, PT_S10(sp)
+ REG_L x27, PT_S11(sp)
+ REG_L x28, PT_T3(sp)
+ REG_L x29, PT_T4(sp)
+ REG_L x30, PT_T5(sp)
+ REG_L x31, PT_T6(sp)
+ .endm
+
+ENTRY(kretprobe_trampoline)
+ addi sp, sp, -(PT_SIZE_ON_STACK)
+ save_all_base_regs
+
+ move a0, sp /* pt_regs */
+
+ call trampoline_probe_handler
+
+ /* use the result as the return-address */
+ move ra, a0
+
+ restore_all_base_regs
+ addi sp, sp, PT_SIZE_ON_STACK
+
+ ret
+ENDPROC(kretprobe_trampoline)
diff --git a/arch/riscv/kernel/probes/simulate-insn.c b/arch/riscv/kernel/probes/simulate-insn.c
new file mode 100644
index 00000000..2519ce2
--- /dev/null
+++ b/arch/riscv/kernel/probes/simulate-insn.c
@@ -0,0 +1,85 @@
+// SPDX-License-Identifier: GPL-2.0+
+
+#include <linux/bitops.h>
+#include <linux/kernel.h>
+#include <linux/kprobes.h>
+
+#include "decode-insn.h"
+#include "simulate-insn.h"
+
+static inline bool rv_insn_reg_get_val(struct pt_regs *regs, u32 index,
+ unsigned long *ptr)
+{
+ if (index == 0)
+ *ptr = 0;
+ else if (index <= 31)
+ *ptr = *((unsigned long *)regs + index);
+ else
+ return false;
+
+ return true;
+}
+
+static inline bool rv_insn_reg_set_val(struct pt_regs *regs, u32 index,
+ unsigned long val)
+{
+ if (index == 0)
+ return false;
+ else if (index <= 31)
+ *((unsigned long *)regs + index) = val;
+ else
+ return false;
+
+ return true;
+}
+
+bool __kprobes simulate_jal(u32 opcode, unsigned long addr, struct pt_regs *regs)
+{
+ /*
+ * 31 30 21 20 19 12 11 7 6 0
+ * imm [20] | imm[10:1] | imm[11] | imm[19:12] | rd | opcode
+ * 1 10 1 8 5 JAL/J
+ */
+ bool ret;
+ u32 imm;
+ u32 index = (opcode >> 7) & 0x1f;
+
+ ret = rv_insn_reg_set_val(regs, index, addr + 4);
+ if (!ret)
+ return ret;
+
+ imm = ((opcode >> 21) & 0x3ff) << 1;
+ imm |= ((opcode >> 20) & 0x1) << 11;
+ imm |= ((opcode >> 12) & 0xff) << 12;
+ imm |= ((opcode >> 31) & 0x1) << 20;
+
+ instruction_pointer_set(regs, addr + sign_extend32((imm), 20));
+
+ return ret;
+}
+
+bool __kprobes simulate_jalr(u32 opcode, unsigned long addr, struct pt_regs *regs)
+{
+ /*
+ * 31 20 19 15 14 12 11 7 6 0
+ * offset[11:0] | rs1 | 010 | rd | opcode
+ * 12 5 3 5 JALR/JR
+ */
+ bool ret;
+ unsigned long base_addr;
+ u32 imm = (opcode >> 20) & 0xfff;
+ u32 rd_index = (opcode >> 7) & 0x1f;
+ u32 rs1_index = (opcode >> 15) & 0x1f;
+
+ ret = rv_insn_reg_set_val(regs, rd_index, addr + 4);
+ if (!ret)
+ return ret;
+
+ ret = rv_insn_reg_get_val(regs, rs1_index, &base_addr);
+ if (!ret)
+ return ret;
+
+ instruction_pointer_set(regs, (base_addr + sign_extend32((imm), 11))&~1);
+
+ return ret;
+}
diff --git a/arch/riscv/kernel/probes/simulate-insn.h b/arch/riscv/kernel/probes/simulate-insn.h
new file mode 100644
index 00000000..a62d784
--- /dev/null
+++ b/arch/riscv/kernel/probes/simulate-insn.h
@@ -0,0 +1,47 @@
+/* SPDX-License-Identifier: GPL-2.0+ */
+
+#ifndef _RISCV_KERNEL_PROBES_SIMULATE_INSN_H
+#define _RISCV_KERNEL_PROBES_SIMULATE_INSN_H
+
+#define __RISCV_INSN_FUNCS(name, mask, val) \
+static __always_inline bool riscv_insn_is_##name(probe_opcode_t code) \
+{ \
+ BUILD_BUG_ON(~(mask) & (val)); \
+ return (code & (mask)) == (val); \
+} \
+bool simulate_##name(u32 opcode, unsigned long addr, \
+ struct pt_regs *regs);
+
+#define RISCV_INSN_REJECTED(name, code) \
+ do { \
+ if (riscv_insn_is_##name(code)) { \
+ return INSN_REJECTED; \
+ } \
+ } while (0)
+
+__RISCV_INSN_FUNCS(system, 0x7f, 0x73)
+__RISCV_INSN_FUNCS(fence, 0x7f, 0x0f)
+
+#define RISCV_INSN_SET_SIMULATE(name, code) \
+ do { \
+ if (riscv_insn_is_##name(code)) { \
+ api->handler = simulate_##name; \
+ return INSN_GOOD_NO_SLOT; \
+ } \
+ } while (0)
+
+__RISCV_INSN_FUNCS(c_j, 0xe003, 0xa001)
+__RISCV_INSN_FUNCS(c_jr, 0xf007, 0x8002)
+__RISCV_INSN_FUNCS(c_jal, 0xe003, 0x2001)
+__RISCV_INSN_FUNCS(c_jalr, 0xf007, 0x9002)
+__RISCV_INSN_FUNCS(c_beqz, 0xe003, 0xc001)
+__RISCV_INSN_FUNCS(c_bnez, 0xe003, 0xe001)
+__RISCV_INSN_FUNCS(c_ebreak, 0xffff, 0x9002)
+
+__RISCV_INSN_FUNCS(auipc, 0x7f, 0x17)
+__RISCV_INSN_FUNCS(branch, 0x7f, 0x63)
+
+__RISCV_INSN_FUNCS(jal, 0x7f, 0x6f)
+__RISCV_INSN_FUNCS(jalr, 0x707f, 0x67)
+
+#endif /* _RISCV_KERNEL_PROBES_SIMULATE_INSN_H */
diff --git a/arch/riscv/kernel/traps.c b/arch/riscv/kernel/traps.c
index ecec177..ac2e786 100644
--- a/arch/riscv/kernel/traps.c
+++ b/arch/riscv/kernel/traps.c
@@ -12,6 +12,7 @@
#include <linux/signal.h>
#include <linux/kdebug.h>
#include <linux/uaccess.h>
+#include <linux/kprobes.h>
#include <linux/mm.h>
#include <linux/module.h>
#include <linux/irq.h>
@@ -145,6 +146,14 @@ static inline unsigned long get_break_insn_length(unsigned long pc)

asmlinkage __visible void do_trap_break(struct pt_regs *regs)
{
+#ifdef CONFIG_KPROBES
+ if (kprobe_single_step_handler(regs))
+ return;
+
+ if (kprobe_breakpoint_handler(regs))
+ return;
+#endif
+
if (user_mode(regs))
force_sig_fault(SIGTRAP, TRAP_BRKPT, (void __user *)regs->epc);
#ifdef CONFIG_KGDB
diff --git a/arch/riscv/mm/fault.c b/arch/riscv/mm/fault.c
index ae7b7fe..da0c08c 100644
--- a/arch/riscv/mm/fault.c
+++ b/arch/riscv/mm/fault.c
@@ -13,6 +13,7 @@
#include <linux/perf_event.h>
#include <linux/signal.h>
#include <linux/uaccess.h>
+#include <linux/kprobes.h>

#include <asm/pgalloc.h>
#include <asm/ptrace.h>
@@ -40,6 +41,9 @@ asmlinkage void do_page_fault(struct pt_regs *regs)
tsk = current;
mm = tsk->mm;

+ if (kprobe_page_fault(regs, cause))
+ return;
+
/*
* Fault-in kernel-space virtual memory on-demand.
* The 'reference' page table is init_mm.pgd.
--
2.7.4

2020-07-09 02:24:37

by Guo Ren

Subject: [PATCH v2 5/6] riscv: Add uprobes supported

From: Guo Ren <[email protected]>

This patch adds support for uprobes on riscv architecture.

Just like kprobe, it support single-step and simulate instructions.

Signed-off-by: Guo Ren <[email protected]>
Reviewed-by: Pekka Enberg <[email protected]>
Cc: Patrick Stählin <[email protected]>
Cc: Masami Hiramatsu <[email protected]>
Cc: Palmer Dabbelt <[email protected]>
Cc: Björn Töpel <[email protected]>
---
arch/riscv/Kconfig | 3 +
arch/riscv/include/asm/processor.h | 1 +
arch/riscv/include/asm/thread_info.h | 4 +-
arch/riscv/include/asm/uprobes.h | 40 ++++++++
arch/riscv/kernel/probes/Makefile | 1 +
arch/riscv/kernel/probes/uprobes.c | 186 +++++++++++++++++++++++++++++++++++
arch/riscv/kernel/signal.c | 3 +
arch/riscv/kernel/traps.c | 10 ++
arch/riscv/mm/fault.c | 7 ++
9 files changed, 254 insertions(+), 1 deletion(-)
create mode 100644 arch/riscv/include/asm/uprobes.h
create mode 100644 arch/riscv/kernel/probes/uprobes.c

diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig
index a295f0b..f927a91 100644
--- a/arch/riscv/Kconfig
+++ b/arch/riscv/Kconfig
@@ -146,6 +146,9 @@ config ARCH_WANT_GENERAL_HUGETLB
config ARCH_SUPPORTS_DEBUG_PAGEALLOC
def_bool y

+config ARCH_SUPPORTS_UPROBES
+ def_bool y
+
config SYS_SUPPORTS_HUGETLBFS
depends on MMU
def_bool y
diff --git a/arch/riscv/include/asm/processor.h b/arch/riscv/include/asm/processor.h
index bdddcd5..3a24003 100644
--- a/arch/riscv/include/asm/processor.h
+++ b/arch/riscv/include/asm/processor.h
@@ -34,6 +34,7 @@ struct thread_struct {
unsigned long sp; /* Kernel mode stack */
unsigned long s[12]; /* s[0]: frame pointer */
struct __riscv_d_ext_state fstate;
+ unsigned long bad_cause;
};

#define INIT_THREAD { \
diff --git a/arch/riscv/include/asm/thread_info.h b/arch/riscv/include/asm/thread_info.h
index 1dd12a0..b3a7eb6 100644
--- a/arch/riscv/include/asm/thread_info.h
+++ b/arch/riscv/include/asm/thread_info.h
@@ -76,6 +76,7 @@ struct thread_info {
#define TIF_SYSCALL_TRACEPOINT 6 /* syscall tracepoint instrumentation */
#define TIF_SYSCALL_AUDIT 7 /* syscall auditing */
#define TIF_SECCOMP 8 /* syscall secure computing */
+#define TIF_UPROBE 9 /* uprobe breakpoint or singlestep */

#define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE)
#define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME)
@@ -84,9 +85,10 @@ struct thread_info {
#define _TIF_SYSCALL_TRACEPOINT (1 << TIF_SYSCALL_TRACEPOINT)
#define _TIF_SYSCALL_AUDIT (1 << TIF_SYSCALL_AUDIT)
#define _TIF_SECCOMP (1 << TIF_SECCOMP)
+#define _TIF_UPROBE (1 << TIF_UPROBE)

#define _TIF_WORK_MASK \
- (_TIF_NOTIFY_RESUME | _TIF_SIGPENDING | _TIF_NEED_RESCHED)
+ (_TIF_NOTIFY_RESUME | _TIF_SIGPENDING | _TIF_NEED_RESCHED | _TIF_UPROBE)

#define _TIF_SYSCALL_WORK \
(_TIF_SYSCALL_TRACE | _TIF_SYSCALL_TRACEPOINT | _TIF_SYSCALL_AUDIT | \
diff --git a/arch/riscv/include/asm/uprobes.h b/arch/riscv/include/asm/uprobes.h
new file mode 100644
index 00000000..f2183e0
--- /dev/null
+++ b/arch/riscv/include/asm/uprobes.h
@@ -0,0 +1,40 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+
+#ifndef _ASM_RISCV_UPROBES_H
+#define _ASM_RISCV_UPROBES_H
+
+#include <asm/probes.h>
+#include <asm/patch.h>
+#include <asm/bug.h>
+
+#define MAX_UINSN_BYTES 8
+
+#ifdef CONFIG_RISCV_ISA_C
+#define UPROBE_SWBP_INSN __BUG_INSN_16
+#define UPROBE_SWBP_INSN_SIZE 2
+#else
+#define UPROBE_SWBP_INSN __BUG_INSN_32
+#define UPROBE_SWBP_INSN_SIZE 4
+#endif
+#define UPROBE_XOL_SLOT_BYTES MAX_UINSN_BYTES
+
+typedef u32 uprobe_opcode_t;
+
+struct arch_uprobe_task {
+ unsigned long saved_cause;
+};
+
+struct arch_uprobe {
+ union {
+ u8 insn[MAX_UINSN_BYTES];
+ u8 ixol[MAX_UINSN_BYTES];
+ };
+ struct arch_probe_insn api;
+ unsigned long insn_size;
+ bool simulate;
+};
+
+bool uprobe_breakpoint_handler(struct pt_regs *regs);
+bool uprobe_single_step_handler(struct pt_regs *regs);
+
+#endif /* _ASM_RISCV_UPROBES_H */
diff --git a/arch/riscv/kernel/probes/Makefile b/arch/riscv/kernel/probes/Makefile
index 8a39507..cb62991 100644
--- a/arch/riscv/kernel/probes/Makefile
+++ b/arch/riscv/kernel/probes/Makefile
@@ -1,4 +1,5 @@
# SPDX-License-Identifier: GPL-2.0
obj-$(CONFIG_KPROBES) += kprobes.o decode-insn.o simulate-insn.o
obj-$(CONFIG_KPROBES) += kprobes_trampoline.o
+obj-$(CONFIG_UPROBES) += uprobes.o decode-insn.o simulate-insn.o
CFLAGS_REMOVE_simulate-insn.o = $(CC_FLAGS_FTRACE)
diff --git a/arch/riscv/kernel/probes/uprobes.c b/arch/riscv/kernel/probes/uprobes.c
new file mode 100644
index 00000000..7a057b5
--- /dev/null
+++ b/arch/riscv/kernel/probes/uprobes.c
@@ -0,0 +1,186 @@
+// SPDX-License-Identifier: GPL-2.0-only
+
+#include <linux/highmem.h>
+#include <linux/ptrace.h>
+#include <linux/uprobes.h>
+
+#include "decode-insn.h"
+
+#define UPROBE_TRAP_NR UINT_MAX
+
+bool is_swbp_insn(uprobe_opcode_t *insn)
+{
+#ifdef CONFIG_RISCV_ISA_C
+ return (*insn & 0xffff) == UPROBE_SWBP_INSN;
+#else
+ return *insn == UPROBE_SWBP_INSN;
+#endif
+}
+
+unsigned long uprobe_get_swbp_addr(struct pt_regs *regs)
+{
+ return instruction_pointer(regs);
+}
+
+int arch_uprobe_analyze_insn(struct arch_uprobe *auprobe, struct mm_struct *mm,
+ unsigned long addr)
+{
+ probe_opcode_t opcode;
+
+ opcode = *(probe_opcode_t *)(&auprobe->insn[0]);
+
+ auprobe->insn_size = GET_INSN_LENGTH(opcode);
+
+ switch (riscv_probe_decode_insn(&opcode, &auprobe->api)) {
+ case INSN_REJECTED:
+ return -EINVAL;
+
+ case INSN_GOOD_NO_SLOT:
+ auprobe->simulate = true;
+ break;
+
+ case INSN_GOOD:
+ auprobe->simulate = false;
+ break;
+
+ default:
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+int arch_uprobe_pre_xol(struct arch_uprobe *auprobe, struct pt_regs *regs)
+{
+ struct uprobe_task *utask = current->utask;
+
+ utask->autask.saved_cause = current->thread.bad_cause;
+ current->thread.bad_cause = UPROBE_TRAP_NR;
+
+ instruction_pointer_set(regs, utask->xol_vaddr);
+
+ regs->status &= ~SR_SPIE;
+
+ return 0;
+}
+
+int arch_uprobe_post_xol(struct arch_uprobe *auprobe, struct pt_regs *regs)
+{
+ struct uprobe_task *utask = current->utask;
+
+ WARN_ON_ONCE(current->thread.bad_cause != UPROBE_TRAP_NR);
+
+ instruction_pointer_set(regs, utask->vaddr + auprobe->insn_size);
+
+ regs->status |= SR_SPIE;
+
+ return 0;
+}
+
+bool arch_uprobe_xol_was_trapped(struct task_struct *t)
+{
+ if (t->thread.bad_cause != UPROBE_TRAP_NR)
+ return true;
+
+ return false;
+}
+
+bool arch_uprobe_skip_sstep(struct arch_uprobe *auprobe, struct pt_regs *regs)
+{
+ probe_opcode_t insn;
+ unsigned long addr;
+
+ if (!auprobe->simulate)
+ return false;
+
+ insn = *(probe_opcode_t *)(&auprobe->insn[0]);
+ addr = instruction_pointer(regs);
+
+ if (auprobe->api.handler)
+ auprobe->api.handler(insn, addr, regs);
+
+ return true;
+}
+
+void arch_uprobe_abort_xol(struct arch_uprobe *auprobe, struct pt_regs *regs)
+{
+ struct uprobe_task *utask = current->utask;
+
+ /*
+ * Task has received a fatal signal, so reset back to probbed
+ * address.
+ */
+ instruction_pointer_set(regs, utask->vaddr);
+
+ regs->status &= ~SR_SPIE;
+}
+
+bool arch_uretprobe_is_alive(struct return_instance *ret, enum rp_check ctx,
+ struct pt_regs *regs)
+{
+ if (ctx == RP_CHECK_CHAIN_CALL)
+ return regs->sp <= ret->stack;
+ else
+ return regs->sp < ret->stack;
+}
+
+unsigned long
+arch_uretprobe_hijack_return_addr(unsigned long trampoline_vaddr,
+ struct pt_regs *regs)
+{
+ unsigned long ra;
+
+ ra = regs->ra;
+
+ regs->ra = trampoline_vaddr;
+
+ return ra;
+}
+
+int arch_uprobe_exception_notify(struct notifier_block *self,
+ unsigned long val, void *data)
+{
+ return NOTIFY_DONE;
+}
+
+bool uprobe_breakpoint_handler(struct pt_regs *regs)
+{
+ if (uprobe_pre_sstep_notifier(regs))
+ return true;
+
+ return false;
+}
+
+bool uprobe_single_step_handler(struct pt_regs *regs)
+{
+ if (uprobe_post_sstep_notifier(regs))
+ return true;
+
+ return false;
+}
+
+void arch_uprobe_copy_ixol(struct page *page, unsigned long vaddr,
+ void *src, unsigned long len)
+{
+ /* Initialize the slot */
+ void *kaddr = kmap_atomic(page);
+ void *dst = kaddr + (vaddr & ~PAGE_MASK);
+
+ memcpy(dst, src, len);
+
+ /* Add ebreak behind opcode to simulate singlestep */
+ if (vaddr) {
+ dst += GET_INSN_LENGTH(*(probe_opcode_t *)src);
+ *(uprobe_opcode_t *)dst = __BUG_INSN_32;
+ }
+
+ kunmap_atomic(kaddr);
+
+ /*
+ * We probably need flush_icache_user_page() but it needs vma.
+ * This should work on most of architectures by default. If
+ * architecture needs to do something different it can define
+ * its own version of the function.
+ */
+ flush_dcache_page(page);
+}
diff --git a/arch/riscv/kernel/signal.c b/arch/riscv/kernel/signal.c
index 17ba190..a96db83b 100644
--- a/arch/riscv/kernel/signal.c
+++ b/arch/riscv/kernel/signal.c
@@ -309,6 +309,9 @@ static void do_signal(struct pt_regs *regs)
asmlinkage __visible void do_notify_resume(struct pt_regs *regs,
unsigned long thread_info_flags)
{
+ if (thread_info_flags & _TIF_UPROBE)
+ uprobe_notify_resume(regs);
+
/* Handle pending signal delivery */
if (thread_info_flags & _TIF_SIGPENDING)
do_signal(regs);
diff --git a/arch/riscv/kernel/traps.c b/arch/riscv/kernel/traps.c
index ac2e786..6981276 100644
--- a/arch/riscv/kernel/traps.c
+++ b/arch/riscv/kernel/traps.c
@@ -76,6 +76,8 @@ void do_trap(struct pt_regs *regs, int signo, int code, unsigned long addr)
static void do_trap_error(struct pt_regs *regs, int signo, int code,
unsigned long addr, const char *str)
{
+ current->thread.bad_cause = regs->cause;
+
if (user_mode(regs)) {
do_trap(regs, signo, code, addr);
} else {
@@ -153,6 +155,14 @@ asmlinkage __visible void do_trap_break(struct pt_regs *regs)
if (kprobe_breakpoint_handler(regs))
return;
#endif
+#ifdef CONFIG_UPROBES
+ if (uprobe_single_step_handler(regs))
+ return;
+
+ if (uprobe_breakpoint_handler(regs))
+ return;
+#endif
+ current->thread.bad_cause = regs->cause;

if (user_mode(regs))
force_sig_fault(SIGTRAP, TRAP_BRKPT, (void __user *)regs->epc);
diff --git a/arch/riscv/mm/fault.c b/arch/riscv/mm/fault.c
index da0c08c..ac96d93 100644
--- a/arch/riscv/mm/fault.c
+++ b/arch/riscv/mm/fault.c
@@ -170,11 +170,14 @@ asmlinkage void do_page_fault(struct pt_regs *regs)
mmap_read_unlock(mm);
/* User mode accesses just cause a SIGSEGV */
if (user_mode(regs)) {
+ tsk->thread.bad_cause = cause;
do_trap(regs, SIGSEGV, code, addr);
return;
}

no_context:
+ tsk->thread.bad_cause = cause;
+
/* Are we prepared to handle this kernel fault? */
if (fixup_exception(regs))
return;
@@ -195,6 +198,8 @@ asmlinkage void do_page_fault(struct pt_regs *regs)
* (which will retry the fault, or kill us if we got oom-killed).
*/
out_of_memory:
+ tsk->thread.bad_cause = cause;
+
mmap_read_unlock(mm);
if (!user_mode(regs))
goto no_context;
@@ -202,6 +207,8 @@ asmlinkage void do_page_fault(struct pt_regs *regs)
return;

do_sigbus:
+ tsk->thread.bad_cause = cause;
+
mmap_read_unlock(mm);
/* Kernel mode? Handle exceptions or die */
if (!user_mode(regs))
--
2.7.4

2020-07-09 02:26:55

by Guo Ren

Subject: [PATCH v2 6/6] riscv: Add KPROBES_ON_FTRACE supported

From: Guo Ren <[email protected]>

This patch adds support for kprobes on ftrace call sites to avoids
much of the overhead with regular kprobes. Try it with simple
steps:

1. Get _do_fork ftrace call site.
Dump of assembler code for function _do_fork:
0xffffffe00020af64 <+0>: addi sp,sp,-128
0xffffffe00020af66 <+2>: sd s0,112(sp)
0xffffffe00020af68 <+4>: sd ra,120(sp)
0xffffffe00020af6a <+6>: addi s0,sp,128
0xffffffe00020af6c <+8>: sd s1,104(sp)
0xffffffe00020af6e <+10>: sd s2,96(sp)
0xffffffe00020af70 <+12>: sd s3,88(sp)
0xffffffe00020af72 <+14>: sd s4,80(sp)
0xffffffe00020af74 <+16>: sd s5,72(sp)
0xffffffe00020af76 <+18>: sd s6,64(sp)
0xffffffe00020af78 <+20>: sd s7,56(sp)
0xffffffe00020af7a <+22>: mv s4,a0
0xffffffe00020af7c <+24>: mv a0,ra
0xffffffe00020af7e <+26>: nop <<<<<<<< here!
0xffffffe00020af82 <+30>: nop
0xffffffe00020af86 <+34>: ld s3,0(s4)

2. Set _do_fork+26 as the kprobe.
echo 'p:myprobe _do_fork+26 dfd=%a0 filename=%a1 flags=%a2 mode=+4($stack)' > /sys/kernel/debug/tracing/kprobe_events
echo 1 > /sys/kernel/debug/tracing/events/kprobes/enable
cat /sys/kernel/debug/tracing/trace
tracer: nop

entries-in-buffer/entries-written: 3/3 #P:1

_-----=> irqs-off
/ _----=> need-resched
| / _---=> hardirq/softirq
|| / _--=> preempt-depth
||| / delay
TASK-PID CPU# |||| TIMESTAMP FUNCTION
| | | |||| | |
sh-87 [000] .... 551.557031: myprobe: (_do_fork+0x1a/0x2e6) dfd=0xffffffe00020af7e filename=0xffffffe00020b34e flags=0xffffffe00101e7c0 mode=0x20af86ffffffe0

cat /sys/kernel/debug/kprobes/list
ffffffe00020af7e k _do_fork+0x1a [FTRACE]
^^^^^^

Signed-off-by: Guo Ren <[email protected]>
Cc: Masami Hiramatsu <[email protected]>
Cc: Palmer Dabbelt <[email protected]>
Cc: Paul Walmsley <[email protected]>
Cc: Björn Töpel <[email protected]>
Cc: Zong Li <[email protected]>
Cc: Pekka Enberg <[email protected]>
---
arch/riscv/Kconfig | 1 +
arch/riscv/kernel/probes/Makefile | 1 +
arch/riscv/kernel/probes/ftrace.c | 53 +++++++++++++++++++++++++++++++++++++++
3 files changed, 55 insertions(+)
create mode 100644 arch/riscv/kernel/probes/ftrace.c

diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig
index f927a91..91bfc6c 100644
--- a/arch/riscv/Kconfig
+++ b/arch/riscv/Kconfig
@@ -58,6 +58,7 @@ config RISCV
select HAVE_FUTEX_CMPXCHG if FUTEX
select HAVE_GENERIC_VDSO if MMU && 64BIT
select HAVE_KPROBES
+ select HAVE_KPROBES_ON_FTRACE
select HAVE_KRETPROBES
select HAVE_PCI
select HAVE_PERF_EVENTS
diff --git a/arch/riscv/kernel/probes/Makefile b/arch/riscv/kernel/probes/Makefile
index cb62991..7f0840d 100644
--- a/arch/riscv/kernel/probes/Makefile
+++ b/arch/riscv/kernel/probes/Makefile
@@ -1,5 +1,6 @@
# SPDX-License-Identifier: GPL-2.0
obj-$(CONFIG_KPROBES) += kprobes.o decode-insn.o simulate-insn.o
obj-$(CONFIG_KPROBES) += kprobes_trampoline.o
+obj-$(CONFIG_KPROBES_ON_FTRACE) += ftrace.o
obj-$(CONFIG_UPROBES) += uprobes.o decode-insn.o simulate-insn.o
CFLAGS_REMOVE_simulate-insn.o = $(CC_FLAGS_FTRACE)
diff --git a/arch/riscv/kernel/probes/ftrace.c b/arch/riscv/kernel/probes/ftrace.c
new file mode 100644
index 00000000..08fc1cc
--- /dev/null
+++ b/arch/riscv/kernel/probes/ftrace.c
@@ -0,0 +1,53 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#include <linux/kprobes.h>
+
+/* Ftrace callback handler for kprobes -- called under preepmt disabed */
+void kprobe_ftrace_handler(unsigned long ip, unsigned long parent_ip,
+ struct ftrace_ops *ops, struct pt_regs *regs)
+{
+ struct kprobe *p;
+ struct kprobe_ctlblk *kcb;
+
+ p = get_kprobe((kprobe_opcode_t *)ip);
+ if (unlikely(!p) || kprobe_disabled(p))
+ return;
+
+ kcb = get_kprobe_ctlblk();
+ if (kprobe_running()) {
+ kprobes_inc_nmissed_count(p);
+ } else {
+ /*
+ * The regs->epc hasn't been saved by SAVE_ALL in mcount-dyn.S
+ * So no need to resume it, just for kprobe handler.
+ */
+ instruction_pointer_set(regs, ip);
+ __this_cpu_write(current_kprobe, p);
+ kcb->kprobe_status = KPROBE_HIT_ACTIVE;
+ if (!p->pre_handler || !p->pre_handler(p, regs)) {
+ /*
+ * Emulate singlestep (and also recover regs->pc)
+ * as if there is a nop
+ */
+ instruction_pointer_set(regs,
+ (unsigned long)p->addr + MCOUNT_INSN_SIZE);
+ if (unlikely(p->post_handler)) {
+ kcb->kprobe_status = KPROBE_HIT_SSDONE;
+ p->post_handler(p, regs, 0);
+ }
+ }
+
+ /*
+ * If pre_handler returns !0, it changes regs->pc. We have to
+ * skip emulating post_handler.
+ */
+ __this_cpu_write(current_kprobe, NULL);
+ }
+}
+NOKPROBE_SYMBOL(kprobe_ftrace_handler);
+
+int arch_prepare_kprobe_ftrace(struct kprobe *p)
+{
+ p->ainsn.api.insn = NULL;
+ return 0;
+}
--
2.7.4

2020-07-10 13:50:51

by Masami Hiramatsu

Subject: Re: [PATCH v2 6/6] riscv: Add KPROBES_ON_FTRACE supported

Hi Guo,

On Thu, 9 Jul 2020 02:19:14 +0000
[email protected] wrote:

> +/* Ftrace callback handler for kprobes -- called under preepmt disabed */
> +void kprobe_ftrace_handler(unsigned long ip, unsigned long parent_ip,
> + struct ftrace_ops *ops, struct pt_regs *regs)
> +{
> + struct kprobe *p;
> + struct kprobe_ctlblk *kcb;
> +
> + p = get_kprobe((kprobe_opcode_t *)ip);
> + if (unlikely(!p) || kprobe_disabled(p))
> + return;
> +
> + kcb = get_kprobe_ctlblk();
> + if (kprobe_running()) {
> + kprobes_inc_nmissed_count(p);
> + } else {
> + /*
> + * The regs->epc hasn't been saved by SAVE_ALL in mcount-dyn.S
> + * So no need to resume it, just for kprobe handler.
> + */
> + instruction_pointer_set(regs, ip);
> + __this_cpu_write(current_kprobe, p);
> + kcb->kprobe_status = KPROBE_HIT_ACTIVE;
> + if (!p->pre_handler || !p->pre_handler(p, regs)) {
> + /*
> + * Emulate singlestep (and also recover regs->pc)
> + * as if there is a nop
> + */
> + instruction_pointer_set(regs,
> + (unsigned long)p->addr + MCOUNT_INSN_SIZE);
> + if (unlikely(p->post_handler)) {
> + kcb->kprobe_status = KPROBE_HIT_SSDONE;
> + p->post_handler(p, regs, 0);
> + }

Hmm, don't you need restoring the previous instruction pointer here?
If you don't support modifying the instruction pointer in the handler,
it must not be compatible with kprobes.

Now BPF function override and function error injection depends on
this behevior, so could you consider to support it in the "ftrace"
implementation at first? (And if it is enabled, you can enable the
livepatch on RISCV too)

Thank you,

> + }
> +
> + /*
> + * If pre_handler returns !0, it changes regs->pc. We have to
> + * skip emulating post_handler.
> + */
> + __this_cpu_write(current_kprobe, NULL);
> + }
> +}
> +NOKPROBE_SYMBOL(kprobe_ftrace_handler);
> +
> +int arch_prepare_kprobe_ftrace(struct kprobe *p)
> +{
> + p->ainsn.api.insn = NULL;
> + return 0;
> +}
> --
> 2.7.4
>


--
Masami Hiramatsu <[email protected]>

2020-07-11 01:35:14

by Guo Ren

Subject: Re: [PATCH v2 6/6] riscv: Add KPROBES_ON_FTRACE supported

Thx Masami,

On Fri, Jul 10, 2020 at 9:50 PM Masami Hiramatsu <[email protected]> wrote:
>
> Hi Guo,
>
> On Thu, 9 Jul 2020 02:19:14 +0000
> [email protected] wrote:
>
> > +/* Ftrace callback handler for kprobes -- called under preepmt disabed */
> > +void kprobe_ftrace_handler(unsigned long ip, unsigned long parent_ip,
> > + struct ftrace_ops *ops, struct pt_regs *regs)
> > +{
> > + struct kprobe *p;
> > + struct kprobe_ctlblk *kcb;
> > +
> > + p = get_kprobe((kprobe_opcode_t *)ip);
> > + if (unlikely(!p) || kprobe_disabled(p))
> > + return;
> > +
> > + kcb = get_kprobe_ctlblk();
> > + if (kprobe_running()) {
> > + kprobes_inc_nmissed_count(p);
> > + } else {
> > + /*
> > + * The regs->epc hasn't been saved by SAVE_ALL in mcount-dyn.S
> > + * So no need to resume it, just for kprobe handler.
> > + */
> > + instruction_pointer_set(regs, ip);
> > + __this_cpu_write(current_kprobe, p);
> > + kcb->kprobe_status = KPROBE_HIT_ACTIVE;
> > + if (!p->pre_handler || !p->pre_handler(p, regs)) {
> > + /*
> > + * Emulate singlestep (and also recover regs->pc)
> > + * as if there is a nop
> > + */
> > + instruction_pointer_set(regs,
> > + (unsigned long)p->addr + MCOUNT_INSN_SIZE);
> > + if (unlikely(p->post_handler)) {
> > + kcb->kprobe_status = KPROBE_HIT_SSDONE;
> > + p->post_handler(p, regs, 0);
> > + }
>
> Hmm, don't you need restoring the previous instruction pointer here?
look at riscv mcount-dyn.S SAVE_ALL function, sp frame lay out like this:
-----------------------
| return address |
-----------------------
| frame pointer |
-----------------------
| pt_regs x1-x31|
-----------------------
It's not a complete pt_regs for the handler, so modifing regs->ip is no use.

> If you don't support modifying the instruction pointer in the handler,
We can modify ip like this if necessary:
*(unsigned long *)((unsigned long)regs + sizeof(struct pt_regs) + 8) = xxx;

> it must not be compatible with kprobes.
Why, can you show related codes? thank you very much.

>
> Now BPF function override and function error injection depends on
> this behevior, so could you consider to support it in the "ftrace"
> implementation at first? (And if it is enabled, you can enable the
> livepatch on RISCV too)
Great message!

But can you show me codes that bpf and err-jnject using the behavior? Thx

I'll try to fix up it :)

--
Best Regards
Guo Ren

ML: https://lore.kernel.org/linux-csky/

2020-07-12 13:39:19

by Masami Hiramatsu

Subject: Re: [PATCH v2 6/6] riscv: Add KPROBES_ON_FTRACE supported

On Sat, 11 Jul 2020 09:32:01 +0800
Guo Ren <[email protected]> wrote:

> Thx Masami,
>
> On Fri, Jul 10, 2020 at 9:50 PM Masami Hiramatsu <[email protected]> wrote:
> >
> > Hi Guo,
> >
> > On Thu, 9 Jul 2020 02:19:14 +0000
> > [email protected] wrote:
> >
> > > +/* Ftrace callback handler for kprobes -- called under preepmt disabed */
> > > +void kprobe_ftrace_handler(unsigned long ip, unsigned long parent_ip,
> > > + struct ftrace_ops *ops, struct pt_regs *regs)
> > > +{
> > > + struct kprobe *p;
> > > + struct kprobe_ctlblk *kcb;
> > > +
> > > + p = get_kprobe((kprobe_opcode_t *)ip);
> > > + if (unlikely(!p) || kprobe_disabled(p))
> > > + return;
> > > +
> > > + kcb = get_kprobe_ctlblk();
> > > + if (kprobe_running()) {
> > > + kprobes_inc_nmissed_count(p);
> > > + } else {
> > > + /*
> > > + * The regs->epc hasn't been saved by SAVE_ALL in mcount-dyn.S
> > > + * So no need to resume it, just for kprobe handler.
> > > + */
> > > + instruction_pointer_set(regs, ip);
> > > + __this_cpu_write(current_kprobe, p);
> > > + kcb->kprobe_status = KPROBE_HIT_ACTIVE;
> > > + if (!p->pre_handler || !p->pre_handler(p, regs)) {
> > > + /*
> > > + * Emulate singlestep (and also recover regs->pc)
> > > + * as if there is a nop
> > > + */
> > > + instruction_pointer_set(regs,
> > > + (unsigned long)p->addr + MCOUNT_INSN_SIZE);
> > > + if (unlikely(p->post_handler)) {
> > > + kcb->kprobe_status = KPROBE_HIT_SSDONE;
> > > + p->post_handler(p, regs, 0);
> > > + }
> >
> > Hmm, don't you need restoring the previous instruction pointer here?
> look at riscv mcount-dyn.S SAVE_ALL function, sp frame lay out like this:
> -----------------------
> | return address |
> -----------------------
> | frame pointer |
> -----------------------
> | pt_regs x1-x31|
> -----------------------
> It's not a complete pt_regs for the handler, so modifing regs->ip is no use.

Yes, that is same on x86. But ftrace regs_caller on x86-64 modifies the
return address on the stack by regs->ip.

See arch/x86/kernel/ftrace_64.S:

-----
SYM_INNER_LABEL(ftrace_regs_call, SYM_L_GLOBAL)
call ftrace_stub

/* Copy flags back to SS, to restore them */
movq EFLAGS(%rsp), %rax
movq %rax, MCOUNT_REG_SIZE(%rsp)

/* Handlers can change the RIP */
movq RIP(%rsp), %rax
movq %rax, MCOUNT_REG_SIZE+8(%rsp)
-----

The last part update the return address on the stack.

> > If you don't support modifying the instruction pointer in the handler,
> We can modify ip like this if necessary:
> *(unsigned long *)((unsigned long)regs + sizeof(struct pt_regs) + 8) = xxx;
>
> > it must not be compatible with kprobes.
> Why, can you show related codes? thank you very much.

The "Changing Execution Path" section in the Documentation/kprobes.txt said

-----
Since kprobes can probe into a running kernel code, it can change the
register set, including instruction pointer.
-----

As you can see, this is the expected behavior on kprobes.
Since user will not know the kprobes is working on a breakpoint or
ftrace, we have to support this behavior transparently. So, could you
enable this feature at first on ftrace@RISCV?

>
> >
> > Now BPF function override and function error injection depends on
> > this behevior, so could you consider to support it in the "ftrace"
> > implementation at first? (And if it is enabled, you can enable the
> > livepatch on RISCV too)
> Great message!
>
> But can you show me codes that bpf and err-jnject using the behavior? Thx

In kernel/fail_function.c

-----
static int fei_kprobe_handler(struct kprobe *kp, struct pt_regs *regs)
{
struct fei_attr *attr = container_of(kp, struct fei_attr, kp);

if (should_fail(&fei_fault_attr, 1)) {
regs_set_return_value(regs, attr->retval);
override_function_with_return(regs);
return 1;
}

return 0;
}
-----

And follow the code related to CONFIG_BPF_KPROBE_OVERRIDE in
kernel/trace/bpf_trace.c. (doing similar thing, but fail_function.c
is simpler.)

Thank you,


--
Masami Hiramatsu <[email protected]>

2020-07-13 23:48:10

by Guo Ren

Subject: Re: [PATCH v2 6/6] riscv: Add KPROBES_ON_FTRACE supported

Hi Masami,

On Sun, Jul 12, 2020 at 9:37 PM Masami Hiramatsu <[email protected]> wrote:
>
> On Sat, 11 Jul 2020 09:32:01 +0800
> Guo Ren <[email protected]> wrote:
>
> > Thx Masami,
> >
> > On Fri, Jul 10, 2020 at 9:50 PM Masami Hiramatsu <[email protected]> wrote:
> > >
> > > Hi Guo,
> > >
> > > On Thu, 9 Jul 2020 02:19:14 +0000
> > > [email protected] wrote:
> > >
> > > > +/* Ftrace callback handler for kprobes -- called under preepmt disabed */
> > > > +void kprobe_ftrace_handler(unsigned long ip, unsigned long parent_ip,
> > > > + struct ftrace_ops *ops, struct pt_regs *regs)
> > > > +{
> > > > + struct kprobe *p;
> > > > + struct kprobe_ctlblk *kcb;
> > > > +
> > > > + p = get_kprobe((kprobe_opcode_t *)ip);
> > > > + if (unlikely(!p) || kprobe_disabled(p))
> > > > + return;
> > > > +
> > > > + kcb = get_kprobe_ctlblk();
> > > > + if (kprobe_running()) {
> > > > + kprobes_inc_nmissed_count(p);
> > > > + } else {
> > > > + /*
> > > > + * The regs->epc hasn't been saved by SAVE_ALL in mcount-dyn.S
> > > > + * So no need to resume it, just for kprobe handler.
> > > > + */
> > > > + instruction_pointer_set(regs, ip);
> > > > + __this_cpu_write(current_kprobe, p);
> > > > + kcb->kprobe_status = KPROBE_HIT_ACTIVE;
> > > > + if (!p->pre_handler || !p->pre_handler(p, regs)) {
> > > > + /*
> > > > + * Emulate singlestep (and also recover regs->pc)
> > > > + * as if there is a nop
> > > > + */
> > > > + instruction_pointer_set(regs,
> > > > + (unsigned long)p->addr + MCOUNT_INSN_SIZE);
> > > > + if (unlikely(p->post_handler)) {
> > > > + kcb->kprobe_status = KPROBE_HIT_SSDONE;
> > > > + p->post_handler(p, regs, 0);
> > > > + }
> > >
> > > Hmm, don't you need restoring the previous instruction pointer here?
> > look at riscv mcount-dyn.S SAVE_ALL function, sp frame lay out like this:
> > -----------------------
> > | return address |
> > -----------------------
> > | frame pointer |
> > -----------------------
> > | pt_regs x1-x31|
> > -----------------------
> > It's not a complete pt_regs for the handler, so modifing regs->ip is no use.
>
> Yes, that is same on x86. But ftrace regs_caller on x86-64 modifies the
> return address on the stack by regs->ip.
>
> See arch/x86/kernel/ftrace_64.S:
>
> -----
> SYM_INNER_LABEL(ftrace_regs_call, SYM_L_GLOBAL)
> call ftrace_stub
>
> /* Copy flags back to SS, to restore them */
> movq EFLAGS(%rsp), %rax
> movq %rax, MCOUNT_REG_SIZE(%rsp)
>
> /* Handlers can change the RIP */
> movq RIP(%rsp), %rax
> movq %rax, MCOUNT_REG_SIZE+8(%rsp)
> -----
Yes, and in save_mcount_regs, it prepare the origin pt_regs RIP.

/* Move RIP to its proper location */
movq MCOUNT_REG_SIZE+\added(%rsp), %rdi
movq %rdi, RIP(%rsp)

I've fixed up riscv's: [1]
[1]: https://lore.kernel.org/linux-riscv/[email protected]/T/#u

>
> The last part update the return address on the stack.
>
> > > If you don't support modifying the instruction pointer in the handler,
> > We can modify ip like this if necessary:
> > *(unsigned long *)((unsigned long)regs + sizeof(struct pt_regs) + 8) = xxx;
> >
> > > it must not be compatible with kprobes.
> > Why, can you show related codes? thank you very much.
>
> The "Changing Execution Path" section in the Documentation/kprobes.txt said
>
> -----
> Since kprobes can probe into a running kernel code, it can change the
> register set, including instruction pointer.
> -----
Got it, thx for correct me.

>
> As you can see, this is the expected behavior on kprobes.
> Since user will not know the kprobes is working on a breakpoint or
> ftrace, we have to support this behavior transparently. So, could you
> enable this feature at first on ftrace@RISCV?
Ditto

>
> >
> > >
> > > Now BPF function override and function error injection depends on
> > > this behevior, so could you consider to support it in the "ftrace"
> > > implementation at first? (And if it is enabled, you can enable the
> > > livepatch on RISCV too)
> > Great message!
> >
> > But can you show me codes that bpf and err-jnject using the behavior? Thx
>
> In kernel/fail_function.c
Nice tip, thx. And I've tested err-jnject: [2]
[2]: https://lore.kernel.org/linux-riscv/[email protected]/T/#md66e4e58f3463c1369e11ab9e5d646343f7d250d

>
> -----
> static int fei_kprobe_handler(struct kprobe *kp, struct pt_regs *regs)
> {
> struct fei_attr *attr = container_of(kp, struct fei_attr, kp);
>
> if (should_fail(&fei_fault_attr, 1)) {
> regs_set_return_value(regs, attr->retval);
> override_function_with_return(regs);
> return 1;
> }
>
> return 0;
> }
> -----
>
> And follow the code related to CONFIG_BPF_KPROBE_OVERRIDE in
> kernel/trace/bpf_trace.c. (doing similar thing, but fail_function.c
> is simpler.)
Nice tip, thx

--
Best Regards
Guo Ren

ML: https://lore.kernel.org/linux-csky/

2020-07-14 11:33:37

by Masami Hiramatsu

Subject: Re: [PATCH v2 6/6] riscv: Add KPROBES_ON_FTRACE supported

Hi Guo,

On Tue, 14 Jul 2020 07:47:20 +0800
Guo Ren <[email protected]> wrote:

> Hi Masami,
>
> On Sun, Jul 12, 2020 at 9:37 PM Masami Hiramatsu <[email protected]> wrote:
> >
> > On Sat, 11 Jul 2020 09:32:01 +0800
> > Guo Ren <[email protected]> wrote:
> >
> > > Thx Masami,
> > >
> > > On Fri, Jul 10, 2020 at 9:50 PM Masami Hiramatsu <[email protected]> wrote:
> > > >
> > > > Hi Guo,
> > > >
> > > > On Thu, 9 Jul 2020 02:19:14 +0000
> > > > [email protected] wrote:
> > > >
> > > > > +/* Ftrace callback handler for kprobes -- called under preepmt disabed */
> > > > > +void kprobe_ftrace_handler(unsigned long ip, unsigned long parent_ip,
> > > > > + struct ftrace_ops *ops, struct pt_regs *regs)
> > > > > +{
> > > > > + struct kprobe *p;
> > > > > + struct kprobe_ctlblk *kcb;
> > > > > +
> > > > > + p = get_kprobe((kprobe_opcode_t *)ip);
> > > > > + if (unlikely(!p) || kprobe_disabled(p))
> > > > > + return;
> > > > > +
> > > > > + kcb = get_kprobe_ctlblk();
> > > > > + if (kprobe_running()) {
> > > > > + kprobes_inc_nmissed_count(p);
> > > > > + } else {
> > > > > + /*
> > > > > + * The regs->epc hasn't been saved by SAVE_ALL in mcount-dyn.S
> > > > > + * So no need to resume it, just for kprobe handler.
> > > > > + */
> > > > > + instruction_pointer_set(regs, ip);
> > > > > + __this_cpu_write(current_kprobe, p);
> > > > > + kcb->kprobe_status = KPROBE_HIT_ACTIVE;
> > > > > + if (!p->pre_handler || !p->pre_handler(p, regs)) {
> > > > > + /*
> > > > > + * Emulate singlestep (and also recover regs->pc)
> > > > > + * as if there is a nop
> > > > > + */
> > > > > + instruction_pointer_set(regs,
> > > > > + (unsigned long)p->addr + MCOUNT_INSN_SIZE);
> > > > > + if (unlikely(p->post_handler)) {
> > > > > + kcb->kprobe_status = KPROBE_HIT_SSDONE;
> > > > > + p->post_handler(p, regs, 0);
> > > > > + }
> > > >
> > > > Hmm, don't you need restoring the previous instruction pointer here?
> > > look at riscv mcount-dyn.S SAVE_ALL function, sp frame lay out like this:
> > > -----------------------
> > > | return address |
> > > -----------------------
> > > | frame pointer |
> > > -----------------------
> > > | pt_regs x1-x31|
> > > -----------------------
> > > It's not a complete pt_regs for the handler, so modifing regs->ip is no use.
> >
> > Yes, that is same on x86. But ftrace regs_caller on x86-64 modifies the
> > return address on the stack by regs->ip.
> >
> > See arch/x86/kernel/ftrace_64.S:
> >
> > -----
> > SYM_INNER_LABEL(ftrace_regs_call, SYM_L_GLOBAL)
> > call ftrace_stub
> >
> > /* Copy flags back to SS, to restore them */
> > movq EFLAGS(%rsp), %rax
> > movq %rax, MCOUNT_REG_SIZE(%rsp)
> >
> > /* Handlers can change the RIP */
> > movq RIP(%rsp), %rax
> > movq %rax, MCOUNT_REG_SIZE+8(%rsp)
> > -----
> Yes, and in save_mcount_regs, it prepare the origin pt_regs RIP.
>
> /* Move RIP to its proper location */
> movq MCOUNT_REG_SIZE+\added(%rsp), %rdi
> movq %rdi, RIP(%rsp)
>
> I've fixed up riscv's: [1]
> [1]: https://lore.kernel.org/linux-riscv/[email protected]/T/#u

OK, looks good :)

> >
> > The last part update the return address on the stack.
> >
> > > > If you don't support modifying the instruction pointer in the handler,
> > > We can modify ip like this if necessary:
> > > *(unsigned long *)((unsigned long)regs + sizeof(struct pt_regs) + 8) = xxx;
> > >
> > > > it must not be compatible with kprobes.
> > > Why, can you show related codes? thank you very much.
> >
> > The "Changing Execution Path" section in the Documentation/kprobes.txt said
> >
> > -----
> > Since kprobes can probe into a running kernel code, it can change the
> > register set, including instruction pointer.
> > -----
> Got it, thx for correct me.
>
> >
> > As you can see, this is the expected behavior on kprobes.
> > Since user will not know the kprobes is working on a breakpoint or
> > ftrace, we have to support this behavior transparently. So, could you
> > enable this feature at first on ftrace@RISCV?
> Ditto
>
> >
> > >
> > > >
> > > > Now BPF function override and function error injection depends on
> > > > this behevior, so could you consider to support it in the "ftrace"
> > > > implementation at first? (And if it is enabled, you can enable the
> > > > livepatch on RISCV too)
> > > Great message!
> > >
> > > But can you show me codes that bpf and err-jnject using the behavior? Thx
> >
> > In kernel/fail_function.c
> Nice tip, thx. And I've tested err-jnject: [2]
> [2]: https://lore.kernel.org/linux-riscv/[email protected]/T/#md66e4e58f3463c1369e11ab9e5d646343f7d250d

Good! Thanks for porting it!


>
> >
> > -----
> > static int fei_kprobe_handler(struct kprobe *kp, struct pt_regs *regs)
> > {
> > struct fei_attr *attr = container_of(kp, struct fei_attr, kp);
> >
> > if (should_fail(&fei_fault_attr, 1)) {
> > regs_set_return_value(regs, attr->retval);
> > override_function_with_return(regs);
> > return 1;
> > }
> >
> > return 0;
> > }
> > -----
> >
> > And follow the code related to CONFIG_BPF_KPROBE_OVERRIDE in
> > kernel/trace/bpf_trace.c. (doing similar thing, but fail_function.c
> > is simpler.)
> Nice tip, thx
>
> --
> Best Regards
> Guo Ren
>
> ML: https://lore.kernel.org/linux-csky/


--
Masami Hiramatsu <[email protected]>

2020-07-21 01:12:10

by Palmer Dabbelt

Subject: Re: [PATCH v2 1/6] riscv: Fixup __vdso_gettimeofday broke dynamic ftrace

On Wed, 08 Jul 2020 19:19:09 PDT (-0700), [email protected] wrote:
> From: Guo Ren <[email protected]>
>
> For linux-5.8-rc1, enable ftrace of riscv will cause boot panic:
>
> [ 2.388980] Run /sbin/init as init process
> [ 2.529938] init[39]: unhandled signal 4 code 0x1 at 0x0000003ff449e000
> [ 2.531078] CPU: 0 PID: 39 Comm: init Not tainted 5.8.0-rc1-dirty #13
> [ 2.532719] epc: 0000003ff449e000 ra : 0000003ff449e954 sp : 0000003fffedb900
> [ 2.534005] gp : 00000000000e8528 tp : 0000003ff449d800 t0 : 000000000000001e
> [ 2.534965] t1 : 000000000000000a t2 : 0000003fffedb89e s0 : 0000003fffedb920
> [ 2.536279] s1 : 0000003fffedb940 a0 : 0000003ff43d4b2c a1 : 0000000000000000
> [ 2.537334] a2 : 0000000000000001 a3 : 0000000000000000 a4 : fffffffffbad8000
> [ 2.538466] a5 : 0000003ff449e93a a6 : 0000000000000000 a7 : 0000000000000000
> [ 2.539511] s2 : 0000000000000000 s3 : 0000003ff448412c s4 : 0000000000000010
> [ 2.541260] s5 : 0000000000000016 s6 : 00000000000d0a30 s7 : 0000003fffedba70
> [ 2.542152] s8 : 0000000000000000 s9 : 0000000000000000 s10: 0000003fffedb960
> [ 2.543335] s11: 0000000000000000 t3 : 0000000000000000 t4 : 0000003fffedb8a0
> [ 2.544471] t5 : 0000000000000000 t6 : 0000000000000000
> [ 2.545730] status: 0000000000004020 badaddr: 00000000464c457f cause: 0000000000000002
> [ 2.549867] Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000004
> [ 2.551267] CPU: 0 PID: 1 Comm: init Not tainted 5.8.0-rc1-dirty #13
> [ 2.552061] Call Trace:
> [ 2.552626] [<ffffffe00020374a>] walk_stackframe+0x0/0xc4
> [ 2.553486] [<ffffffe0002039f4>] show_stack+0x40/0x4c
> [ 2.553995] [<ffffffe00054a6ae>] dump_stack+0x7a/0x98
> [ 2.554615] [<ffffffe00020b9b8>] panic+0x114/0x2f4
> [ 2.555395] [<ffffffe00020ebd6>] do_exit+0x89c/0x8c2
> [ 2.555949] [<ffffffe00020f930>] do_group_exit+0x3a/0x90
> [ 2.556715] [<ffffffe000219e08>] get_signal+0xe2/0x6e6
> [ 2.557388] [<ffffffe000202d72>] do_notify_resume+0x6a/0x37a
> [ 2.558089] [<ffffffe000201c16>] ret_from_exception+0x0/0xc
>
> "ra:0x3ff449e954" is the return address of "call _mcount" in the
> prologue of __vdso_gettimeofday(). Without proper relocate, pc jmp
> to 0x0000003ff449e000 (vdso map base) with a illegal instruction
> trap.
>
> The solution comes from arch/arm64/kernel/vdso/Makefile:
>
> CFLAGS_REMOVE_vgettimeofday.o = $(CC_FLAGS_FTRACE) -Os $(CC_FLAGS_SCS)
>
> - CC_FLAGS_SCS is ShadowCallStack feature in Clang and only
> implemented for arm64, no use for riscv.
>
> The bug comes from the following commit:
>
> ad5d1122b82f ("riscv: use vDSO common flow to reduce the latency of the time-related functions")
>
> Signed-off-by: Guo Ren <[email protected]>
> Reviewed-by: Pekka Enberg <[email protected]>
> Cc: Vincent Chen <[email protected]>
> Cc: Atish Patra <[email protected]>
> Cc: Palmer Dabbelt <[email protected]>
> Cc: Alan Kao <[email protected]>
> Cc: Greentime Hu <[email protected]>
> ---
> arch/riscv/kernel/vdso/Makefile | 3 +++
> 1 file changed, 3 insertions(+)
>
> diff --git a/arch/riscv/kernel/vdso/Makefile b/arch/riscv/kernel/vdso/Makefile
> index 38ba55b..3079935 100644
> --- a/arch/riscv/kernel/vdso/Makefile
> +++ b/arch/riscv/kernel/vdso/Makefile
> @@ -27,6 +27,9 @@ obj-vdso := $(addprefix $(obj)/, $(obj-vdso))
> obj-y += vdso.o vdso-syms.o
> CPPFLAGS_vdso.lds += -P -C -U$(ARCH)
>
> +# Disable -pg to prevent insert call site
> +CFLAGS_REMOVE_vgettimeofday.o = $(CC_FLAGS_FTRACE) -Os
> +
> # Disable gcov profiling for VDSO code
> GCOV_PROFILE := n

As far as I can tell this is already in Linus' tree. LMK if I'm missing
something.