#include <linux/spinlock.h>
#include <linux/hardirq.h>
+#include <linux/uaccess.h>
#include <linux/ftrace.h>
#include <linux/percpu.h>
#include <linux/init.h>
#include <linux/list.h>
-#include <asm/alternative.h>
+#include <asm/ftrace.h>
+#include <asm/nops.h>
-#define CALL_BACK 5
/* Long is fine, even if it is only 4 bytes ;-) */
-static long *ftrace_nop;
+static unsigned long *ftrace_nop;
union ftrace_code_union {
- char code[5];
+ char code[MCOUNT_INSN_SIZE];
struct {
char e8;
int offset;
} __attribute__((packed));
};
+
static int notrace ftrace_calc_offset(long ip, long addr)
{
return (int)(addr - ip);
static union ftrace_code_union calc;
calc.e8 = 0xe8;
- calc.offset = ftrace_calc_offset(ip, addr);
+ calc.offset = ftrace_calc_offset(ip + MCOUNT_INSN_SIZE, addr);
/*
* No locking needed, this must be called via kstop_machine
ftrace_modify_code(unsigned long ip, unsigned char *old_code,
unsigned char *new_code)
{
- unsigned replaced;
- unsigned old = *(unsigned *)old_code; /* 4 bytes */
- unsigned new = *(unsigned *)new_code; /* 4 bytes */
- unsigned char newch = new_code[4];
- int faulted = 0;
-
- /* move the IP back to the start of the call */
- ip -= CALL_BACK;
+ unsigned char replaced[MCOUNT_INSN_SIZE];
/*
* Note: Due to modules and __init, code can
* disappear and change, we need to protect against faulting
- * as well as code changing.
+ * as well as code changing. We do this by using the
+ * probe_kernel_* functions.
*
* No real locking needed, this code is run through
- * kstop_machine.
+ * kstop_machine, or before SMP starts.
*/
- asm volatile (
- "1: lock\n"
- " cmpxchg %3, (%2)\n"
- " jnz 2f\n"
- " movb %b4, 4(%2)\n"
- "2:\n"
- ".section .fixup, \"ax\"\n"
- "3: movl $1, %0\n"
- " jmp 2b\n"
- ".previous\n"
- _ASM_EXTABLE(1b, 3b)
- : "=r"(faulted), "=a"(replaced)
- : "r"(ip), "r"(new), "r"(newch),
- "0"(faulted), "a"(old)
- : "memory");
- sync_core();
- if (replaced != old && replaced != new)
- faulted = 2;
+ /* read the text we want to modify */
+ if (probe_kernel_read(replaced, (void *)ip, MCOUNT_INSN_SIZE))
+ return -EFAULT;
+
+ /* Make sure it is what we expect it to be */
+ if (memcmp(replaced, old_code, MCOUNT_INSN_SIZE) != 0)
+ return -EINVAL;
- return faulted;
+ /* replace the text with the new text */
+ if (probe_kernel_write((void *)ip, new_code, MCOUNT_INSN_SIZE))
+ return -EPERM;
+
+ sync_core();
+
+ return 0;
}
notrace int ftrace_update_ftrace_func(ftrace_func_t func)
{
unsigned long ip = (unsigned long)(&ftrace_call);
- unsigned char old[5], *new;
+ unsigned char old[MCOUNT_INSN_SIZE], *new;
int ret;
- ip += CALL_BACK;
-
- memcpy(old, &ftrace_call, 5);
+ memcpy(old, &ftrace_call, MCOUNT_INSN_SIZE);
new = ftrace_call_replace(ip, (unsigned long)func);
ret = ftrace_modify_code(ip, old, new);
return ret;
}
-notrace int ftrace_mcount_set(unsigned long *data)
+int __init ftrace_dyn_arch_init(void *data)
{
- unsigned long ip = (long)(&mcount_call);
- unsigned long *addr = data;
- unsigned char old[5], *new;
-
- /* ip is at the location, but modify code will subtact this */
- ip += CALL_BACK;
+ extern const unsigned char ftrace_test_p6nop[];
+ extern const unsigned char ftrace_test_nop5[];
+ extern const unsigned char ftrace_test_jmp[];
+ int faulted = 0;
/*
- * Replace the mcount stub with a pointer to the
- * ip recorder function.
+ * There is no good nop for all x86 archs.
+ * We will default to using the P6_NOP5, but first we
+ * will test to make sure that the nop will actually
+ * work on this CPU. If it faults, we will then
+ * go to a lesser efficient 5 byte nop. If that fails
+ * we then just use a jmp as our nop. This isn't the most
+ * efficient nop, but we can not use a multi part nop
+ * since we would then risk being preempted in the middle
+ * of that nop, and if we enabled tracing then, it might
+ * cause a system crash.
+ *
+ * TODO: check the cpuid to determine the best nop.
*/
- memcpy(old, &mcount_call, 5);
- new = ftrace_call_replace(ip, *addr);
- *addr = ftrace_modify_code(ip, old, new);
-
- return 0;
-}
-
-int __init ftrace_dyn_arch_init(void *data)
-{
- const unsigned char *const *noptable = find_nop_table();
-
- /* This is running in kstop_machine */
-
- ftrace_mcount_set(data);
-
- ftrace_nop = (unsigned long *)noptable[CALL_BACK];
+ asm volatile (
+ "jmp ftrace_test_jmp\n"
+ /* This code needs to stay around */
+ ".section .text, \"ax\"\n"
+ "ftrace_test_jmp:"
+ "jmp ftrace_test_p6nop\n"
+ "nop\n"
+ "nop\n"
+ "nop\n" /* 2 byte jmp + 3 bytes */
+ "ftrace_test_p6nop:"
+ P6_NOP5
+ "jmp 1f\n"
+ "ftrace_test_nop5:"
+ ".byte 0x66,0x66,0x66,0x66,0x90\n"
+ "jmp 1f\n"
+ ".previous\n"
+ "1:"
+ ".section .fixup, \"ax\"\n"
+ "2: movl $1, %0\n"
+ " jmp ftrace_test_nop5\n"
+ "3: movl $2, %0\n"
+ " jmp 1b\n"
+ ".previous\n"
+ _ASM_EXTABLE(ftrace_test_p6nop, 2b)
+ _ASM_EXTABLE(ftrace_test_nop5, 3b)
+ : "=r"(faulted) : "0" (faulted));
+
+ switch (faulted) {
+ case 0:
+ pr_info("ftrace: converting mcount calls to 0f 1f 44 00 00\n");
+ ftrace_nop = (unsigned long *)ftrace_test_p6nop;
+ break;
+ case 1:
+ pr_info("ftrace: converting mcount calls to 66 66 66 66 90\n");
+ ftrace_nop = (unsigned long *)ftrace_test_nop5;
+ break;
+ case 2:
+ pr_info("ftrace: converting mcount calls to jmp . + 5\n");
+ ftrace_nop = (unsigned long *)ftrace_test_jmp;
+ break;
+ }
+
+ /* The return code is retured via data */
+ *(unsigned long *)data = 0;
return 0;
}
-
git://ftp.safe.ca / safe / jmp / linux-2.6 / blobdiff