Re: [PATCH v11 1/9] arm64: Add HAVE_REGS_AND_STACK_ACCESS_API feature

From: David Long
Date: Mon Mar 21 2016 - 03:15:46 EST


On 03/15/2016 07:04 AM, Marc Zyngier wrote:
David,

On 09/03/16 05:32, David Long wrote:
From: "David A. Long" <dave.long@xxxxxxxxxx>

Add HAVE_REGS_AND_STACK_ACCESS_API feature for arm64.

Signed-off-by: David A. Long <dave.long@xxxxxxxxxx>
---
arch/arm64/Kconfig | 1 +
arch/arm64/include/asm/ptrace.h | 31 +++++++++++
arch/arm64/kernel/ptrace.c | 117 ++++++++++++++++++++++++++++++++++++++++
3 files changed, 149 insertions(+)

diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig
index 8cc6228..4211b0d 100644
--- a/arch/arm64/Kconfig
+++ b/arch/arm64/Kconfig
@@ -78,6 +78,7 @@ config ARM64
select HAVE_PERF_EVENTS
select HAVE_PERF_REGS
select HAVE_PERF_USER_STACK_DUMP
+ select HAVE_REGS_AND_STACK_ACCESS_API
select HAVE_RCU_TABLE_FREE
select HAVE_SYSCALL_TRACEPOINTS
select IOMMU_DMA if IOMMU_SUPPORT
diff --git a/arch/arm64/include/asm/ptrace.h b/arch/arm64/include/asm/ptrace.h
index e9e5467..7bd6445 100644
--- a/arch/arm64/include/asm/ptrace.h
+++ b/arch/arm64/include/asm/ptrace.h
@@ -118,6 +118,8 @@ struct pt_regs {
u64 syscallno;
};

+#define MAX_REG_OFFSET offsetof(struct user_pt_regs, pstate)

So here you're using user_pt_regs...


Changed it to pt_regs, and removed the typecast in the instruction_pointer() define to get things to compile again.

+
#define arch_has_single_step() (1)

#ifdef CONFIG_COMPAT
@@ -146,6 +148,35 @@ struct pt_regs {
#define user_stack_pointer(regs) \
(!compat_user_mode(regs) ? (regs)->sp : (regs)->compat_sp)

+extern int regs_query_register_offset(const char *name);
+extern const char *regs_query_register_name(unsigned int offset);
+extern bool regs_within_kernel_stack(struct pt_regs *regs, unsigned long addr);
+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 number 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.

Is that the offset in pt_regs? Or should it be in the actual regs array
instead? So far, this is the same thing, but that feels pretty fragile.


It's the offset within the entire structure, including sp, pc, and pstate.

+ * If @offset is bigger than MAX_REG_OFFSET, this returns 0.
+ */
+static inline u64 regs_get_register(struct pt_regs *regs,
+ unsigned int offset)

... and here this is pt_regs. I know that the structures are quite
similar, but some uniformity wouldn't hurt. Given that this series is
mostly concerned with kernel space, it should probably the latter rather
than the former.


Yes, pt_regs it is.

+{
+ if (unlikely(offset > MAX_REG_OFFSET))
+ return 0;
+ return *(u64 *)((u64)regs + offset);

Now that's a bit disgusting... You are assuming way too much about the
layout of pt_regs (imagine someone insert a new field right before the
union?). How about:

u64 *reg_array = regs->regs;
return reg_array[offset >> 3];

instead? I know the semantic is not the same, but I'd really like to see
something a bit more robust.

This index is supposed to provide access to useful info saved in pt_regs structure including regs, pc, sp, and pstate. This is how it works on the other architectures that provide this feature. The index is going to be something that was previously looked up with the regs_query_register_offset() call, which ultimately looks up the offset in the structure using the field name and a table created at compile time (see regoffset_table in include/asm/ptrace.h). Additions to the structure will not create a problem since the numeric value is not hardcoded in existing code (although thought should be given about whether new names should be added to regs_query_register_name()'s lookup table to make them accessible).

Changing this mechanism would require changes to the generic code and to each other architecture, while preserving the architecture-independent nature of this API. Changing the API would likely affect performance tools and debugger(s) too.

+}
+
+/* Valid only for Kernel mode traps. */
+static inline unsigned long kernel_stack_pointer(struct pt_regs *regs)
+{
+ return regs->sp;
+}
+
static inline unsigned long regs_return_value(struct pt_regs *regs)
{
return regs->regs[0];
diff --git a/arch/arm64/kernel/ptrace.c b/arch/arm64/kernel/ptrace.c
index ff7f132..efebf0f 100644
--- a/arch/arm64/kernel/ptrace.c
+++ b/arch/arm64/kernel/ptrace.c
@@ -48,6 +48,123 @@
#define CREATE_TRACE_POINTS
#include <trace/events/syscalls.h>

+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}
+#define GPR_OFFSET_NAME(r) \
+ {.name = "x" #r, .offset = offsetof(struct pt_regs, regs[r])}
+
+static const struct pt_regs_offset regoffset_table[] = {
+ GPR_OFFSET_NAME(0),
+ GPR_OFFSET_NAME(1),
+ GPR_OFFSET_NAME(2),
+ GPR_OFFSET_NAME(3),
+ GPR_OFFSET_NAME(4),
+ GPR_OFFSET_NAME(5),
+ GPR_OFFSET_NAME(6),
+ GPR_OFFSET_NAME(7),
+ GPR_OFFSET_NAME(8),
+ GPR_OFFSET_NAME(9),
+ GPR_OFFSET_NAME(10),
+ GPR_OFFSET_NAME(11),
+ GPR_OFFSET_NAME(12),
+ GPR_OFFSET_NAME(13),
+ GPR_OFFSET_NAME(14),
+ GPR_OFFSET_NAME(15),
+ GPR_OFFSET_NAME(16),
+ GPR_OFFSET_NAME(17),
+ GPR_OFFSET_NAME(18),
+ GPR_OFFSET_NAME(19),
+ GPR_OFFSET_NAME(20),
+ GPR_OFFSET_NAME(21),
+ GPR_OFFSET_NAME(22),
+ GPR_OFFSET_NAME(23),
+ GPR_OFFSET_NAME(24),
+ GPR_OFFSET_NAME(25),
+ GPR_OFFSET_NAME(26),
+ GPR_OFFSET_NAME(27),
+ GPR_OFFSET_NAME(28),
+ GPR_OFFSET_NAME(29),
+ GPR_OFFSET_NAME(30),
+ {.name = "lr", .offset = offsetof(struct pt_regs, regs[30])},
+ REG_OFFSET_NAME(sp),
+ REG_OFFSET_NAME(pc),
+ REG_OFFSET_NAME(pstate),
+ 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_query_register_name() - query register name from its offset
+ * @offset: the offset of a register in struct pt_regs.
+ *
+ * regs_query_register_name() returns the name of a register from its
+ * offset in struct pt_regs. If the @offset is invalid, this returns NULL;
+ */
+const char *regs_query_register_name(unsigned int offset)
+{
+ const struct pt_regs_offset *roff;
+
+ for (roff = regoffset_table; roff->name != NULL; roff++)
+ if (roff->offset == offset)
+ return roff->name;
+ return NULL;
+}
+
+/**
+ * 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.
+ */
+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;
+}
+
/*
* TODO: does not yet catch signals sent when the child dies.
* in exit.c or in signal.c.


Thanks,

M.


Thanks for your feedback.

-dl