-
-
未解决 [求助] 将Windows端的 x64 vt-x框架迁移至Linux时，出现Triple fault 50雪币
-
发表于: 2024-11-15 17:22
-
vmcs的填充源码如下：
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 | struct page *vmcs_page, *msr_bitmap_page;void *vmcs, *vmcs_phys, *host_stack, *msr_bitmap_phys, *msr_bitmap;unsigned short cs, ds, es, fs, gs, ss, tr, ldtr;unsigned int cr0, cr3, cr4;/* vm-control */VmxPinBasedControls pin_base;VmxProcessorBasedControls process_base;VmxSecondaryProcessorBasedControls secondary_process_base;VmxVmEntryControls entry;VmxVmExitControls exit;VmxExceptionBitmap exception_bitmap = {0};// 1.申请内存vmcs_page = alloc_pages(GFP_KERNEL, 0);vmcs_phys = (void *)page_to_phys(vmcs_page);vmcs = page_address(vmcs_page);memset(vmcs, 0, 0x1000);*(char *)vmcs = 1;vmx_config.vmcs_region_page = vmcs_page;// 2. vmclear&vmptrld// asm_vmclear(vmcs_phys);asm_vmptrld(vmcs_phys);// 3.初始化VMCS//// 3.1 Guest State Area//cr0 = asm_read_cr0();cr3 = asm_read_cr3();cr4 = asm_read_cr4();asm_vmwrite(GUEST_CR0, cr0);asm_vmwrite(GUEST_CR3, cr3);asm_vmwrite(GUEST_CR4, cr4);asm_vmwrite(CR0_READ_SHADOW, cr0);asm_vmwrite(CR4_READ_SHADOW, cr4);asm_vmwrite(GUEST_DR7, 0x400);/*段寄存器*/es = asm_read_es();cs = asm_read_cs();ds = asm_read_ds();fs = asm_read_fs();gs = asm_read_gs();ss = asm_read_ss();tr = asm_read_tr();ldtr = asm_read_ldtr();asm_vmwrite(GUEST_ES_SELECTOR, es);asm_vmwrite(GUEST_ES_BASE, get_segment_base(es));asm_vmwrite(GUEST_ES_LIMIT, get_segment_limit(es));asm_vmwrite(GUEST_ES_AR_BYTES, get_segment_access_rights(es));printk(KERN_INFO "[*] es: %#llx %#llx %#llx %#llx\n", (__u64)es, (__u64)get_segment_base(es), (__u64)get_segment_limit(es), (__u64)(get_segment_access_rights(es) & 0xffffffff));asm_vmwrite(GUEST_CS_SELECTOR, cs);asm_vmwrite(GUEST_CS_BASE, get_segment_base(cs));asm_vmwrite(GUEST_CS_LIMIT, get_segment_limit(cs));asm_vmwrite(GUEST_CS_AR_BYTES, get_segment_access_rights(cs));printk(KERN_INFO "[*] cs: %#llx %#llx %#llx %#llx\n", (__u64)cs, (__u64)get_segment_base(cs), (__u64)get_segment_limit(cs), (__u64)(get_segment_access_rights(cs) & 0xffffffff));asm_vmwrite(GUEST_DS_SELECTOR, ds);asm_vmwrite(GUEST_DS_BASE, get_segment_base(ds));asm_vmwrite(GUEST_DS_LIMIT, get_segment_limit(ds));asm_vmwrite(GUEST_DS_AR_BYTES, get_segment_access_rights(ds));printk(KERN_INFO "[*] ds: %#llx %#llx %#llx %#llx\n", (__u64)ds, (__u64)get_segment_base(ds), (__u64)get_segment_limit(ds), (__u64)(get_segment_access_rights(ds) & 0xffffffff));asm_vmwrite(GUEST_FS_SELECTOR, fs);asm_vmwrite(GUEST_FS_BASE, asm_rdmsr2(IA32_FS_BASE));asm_vmwrite(GUEST_FS_LIMIT, get_segment_limit(fs));asm_vmwrite(GUEST_FS_AR_BYTES, get_segment_access_rights(fs));printk(KERN_INFO "[*] fs: %#llx %#llx %#llx %#llx\n", (__u64)fs, (__u64)get_segment_base(fs), (__u64)get_segment_limit(fs), (__u64)(get_segment_access_rights(fs) & 0xffffffff));asm_vmwrite(GUEST_GS_SELECTOR, gs);asm_vmwrite(GUEST_GS_BASE, asm_rdmsr2(IA32_GS_BASE));asm_vmwrite(GUEST_GS_LIMIT, get_segment_limit(gs));asm_vmwrite(GUEST_GS_AR_BYTES, get_segment_access_rights(gs));printk(KERN_INFO "[*] gs: %#llx %#llx %#llx %#llx\n", (__u64)gs, (__u64)get_segment_base(gs), (__u64)get_segment_limit(gs), (__u64)(get_segment_access_rights(gs) & 0xffffffff));asm_vmwrite(GUEST_SS_SELECTOR, ss);asm_vmwrite(GUEST_SS_BASE, get_segment_base(ss));asm_vmwrite(GUEST_SS_LIMIT, get_segment_limit(ss));asm_vmwrite(GUEST_SS_AR_BYTES, get_segment_access_rights(ss));printk(KERN_INFO "[*] ss: %#llx %#llx %#llx %#llx\n", (__u64)ss, (__u64)get_segment_base(ss), (__u64)get_segment_limit(ss), (__u64)(get_segment_access_rights(ss) & 0xffffffff));asm_vmwrite(GUEST_TR_SELECTOR, tr);asm_vmwrite(GUEST_TR_BASE, get_segment_base(tr));asm_vmwrite(GUEST_TR_LIMIT, get_segment_limit(tr));asm_vmwrite(GUEST_TR_AR_BYTES, get_segment_access_rights(tr));printk(KERN_INFO "[*] tr: %#llx %#llx %#llx %#llx\n", (__u64)tr, (__u64)get_segment_base(tr), (__u64)get_segment_limit(tr), (__u64)(get_segment_access_rights(tr) & 0xffffffff));asm_vmwrite(GUEST_LDTR_SELECTOR, ldtr); // 必填项asm_vmwrite(GUEST_LDTR_BASE, get_segment_base(ldtr));asm_vmwrite(GUEST_LDTR_LIMIT, get_segment_limit(ldtr));asm_vmwrite(GUEST_LDTR_AR_BYTES, get_segment_access_rights(ldtr));printk(KERN_INFO "[*] ldtr: %#llx %#llx %#llx %#llx\n", (__u64)ldtr, (__u64)get_segment_base(ldtr), (__u64)get_segment_limit(ldtr), (__u64)(get_segment_access_rights(ldtr) & 0xffffffff));asm_vmwrite(GUEST_GDTR_BASE, (__u64)asm_get_gdtr_base());asm_vmwrite(GUEST_GDTR_LIMIT, asm_get_gdtr_limit());printk(KERN_INFO "[*] gdtr: %#llx %#llx\n", (__u64)asm_get_gdtr_base(), (__u64)asm_get_gdtr_limit());asm_vmwrite(GUEST_IDTR_BASE, (__u64)asm_get_idtr_base());asm_vmwrite(GUEST_IDTR_LIMIT, asm_get_idtr_limit());printk(KERN_INFO "[*] idtr: %#llx %#llx\n", (__u64)asm_get_idtr_base(), (__u64)asm_get_idtr_limit());/*MSR寄存器组*/asm_vmwrite(GUEST_IA32_DEBUGCTL, asm_rdmsr2(IA32_DEBUGCTL)); // 分支跟踪或性能监控asm_vmwrite(GUEST_SYSENTER_CS, asm_rdmsr2(IA32_SYSENTER_CS));asm_vmwrite(GUEST_SYSENTER_ESP, asm_rdmsr2(IA32_SYSENTER_ESP));asm_vmwrite(GUEST_SYSENTER_EIP, asm_rdmsr2(IA32_SYSENTER_EIP)); // KiFastCallEntryasm_vmwrite(GUEST_IA32_EFER, asm_rdmsr2(IA32_EFER));/* 设置guest.rip/rsp/rfalgs */asm_vmwrite(GUEST_RSP, guest_regs.rsp);asm_vmwrite(GUEST_RIP, (__u64)&__guest_entry);asm_vmwrite(GUEST_RFLAGS, guest_regs.rflags);// printk(KERN_INFO "[+] guest rip: %pK, rsp: %pK, rflags: %pK\n", (void *)&__guest_entry, (void *)guest_regs.rsp, (void *)guest_regs.rflags);// VMCS link pointer, 嵌套虚拟化asm_vmwrite(VMCS_LINK_POINTER, -1);//// 3.2 Host State Area//asm_vmwrite(HOST_CR0, cr0);asm_vmwrite(HOST_CR3, cr3);asm_vmwrite(HOST_CR4, cr4);asm_vmwrite(HOST_ES_SELECTOR, es);asm_vmwrite(HOST_CS_SELECTOR, cs);asm_vmwrite(HOST_SS_SELECTOR, ss);asm_vmwrite(HOST_DS_SELECTOR, ds);asm_vmwrite(HOST_FS_SELECTOR, fs);asm_vmwrite(HOST_GS_SELECTOR, gs);asm_vmwrite(HOST_TR_SELECTOR, tr);asm_vmwrite(HOST_FS_BASE, asm_rdmsr2(IA32_FS_BASE));asm_vmwrite(HOST_GS_BASE, asm_rdmsr2(IA32_GS_BASE));asm_vmwrite(HOST_TR_BASE, get_segment_base(tr));asm_vmwrite(HOST_GDTR_BASE, (__u64)asm_get_gdtr_base());asm_vmwrite(HOST_IDTR_BASE, (__u64)asm_get_idtr_base());/*MSR寄存器组*/asm_vmwrite(HOST_IA32_SYSENTER_CS, asm_rdmsr2(IA32_SYSENTER_CS));asm_vmwrite(HOST_IA32_SYSENTER_ESP, asm_rdmsr2(IA32_SYSENTER_ESP));asm_vmwrite(HOST_IA32_SYSENTER_EIP, asm_rdmsr2(IA32_SYSENTER_EIP)); // KiFastCallEntry// asm_vmwrite(HOST_IA32_PAT, asm_rdmsr2(IA32_PAT));asm_vmwrite(HOST_IA32_EFER, asm_rdmsr2(IA32_EFER));/*申请VMX栈空间*/host_stack = vmalloc(0x2000);memset(host_stack, 0, 0x2000);vmx_config.host_stack = host_stack;asm_vmwrite(HOST_RSP, (__u64)host_stack + 0x2000);asm_vmwrite(HOST_RIP, (__u64)&__host_entry);// printk(KERN_INFO "[+] host ==> rip: %pK, rsp: %pK\n", &__host_entry, host_stack + 0x2000);//// 3.3 vm-control///*检查IA32_VMX_BASIC[55]是否置位, 填充保留位*/if (asm_rdmsr2(IA32_VMX_BASIC) & 0x0080000000000000){ /*vm execution controls*/ pin_base.all = asm_rdmsr2(IA32_VMX_TRUE_PINBASED_CTLS) & 0xffffffff; process_base.all = asm_rdmsr2(IA32_VMX_TRUE_PROCBASED_CTLS) & 0xffffffff; /*vm entry control*/ entry.all = asm_rdmsr2(IA32_VMX_TRUE_ENTRY_CTLS) & 0xffffffff; /*vm exit control*/ exit.all = asm_rdmsr2(IA32_VMX_TRUE_EXIT_CTLS) & 0xffffffff;}else{ /*vm execution controls*/ pin_base.all = asm_rdmsr2(IA32_VMX_PINBASED_CTLS) & 0xffffffff; process_base.all = asm_rdmsr2(IA32_VMX_PROCBASED_CTLS) & 0xffffffff; /*vm entry control*/ entry.all = asm_rdmsr2(IA32_VMX_ENTRY_CTLS) & 0xffffffff; /*vm exit control*/ exit.all = asm_rdmsr2(IA32_VMX_EXIT_CTLS) & 0xffffffff;}/*检查是否启用Secondary Processor-Based VM-Execution Controls*/if (asm_rdmsr2(IA32_VMX_PROCBASED_CTLS) & 0x8000000000000000){ printk(KERN_INFO "[+] 启用Secondary Processor-Based VM-Execution Controls\n"); secondary_process_base.all = asm_rdmsr2(IA32_VMX_PROCBASED_CTLS2);}/* * VMCS controls域属性添加 */// process_base.Bits.interrupt_window_exiting = true;process_base.Bits.activate_secondary_control = true;/* msr bitmap */if (vmx_config.vmcs_controls.enable_msr_bitmap){ process_base.Bits.use_msr_bitmaps = true; msr_bitmap_page = alloc_pages(GFP_KERNEL, 0); msr_bitmap = page_address(msr_bitmap_page); memset(msr_bitmap, 0, 0x1000); msr_bitmap_phys = (void *)page_to_phys(msr_bitmap_page); asm_vmwrite(MSR_BITMAP_ADDRESS, (__u64)msr_bitmap_phys); vmx_config.msr_bitmap_page = msr_bitmap_page;}secondary_process_base.Bits.enable_rdtscp = true;secondary_process_base.Bits.enable_invpcid = true;secondary_process_base.Bits.enable_xsaves_xstors = true;entry.Bits.ia32e_mode_guest = true; // x64必填【重要】entry.Bits.load_ia32_efer = true; // 启用 EFER 寄存器// entry.Bits.load_ia32_pat = true; // 页面管理相关【可选】entry.Bits.load_debug_controls = true;exit.Bits.host_address_space_size = true; // x64必填【重要】, 进入IA-32e模式exit.Bits.load_ia32_efer = true;exit.Bits.save_ia32_efer = true;exit.Bits.acknowledge_interrupt_on_exit = true;exit.Bits.save_debug_controls = true;/* VMCS controls域写入*/asm_vmwrite(PIN_BASED_VM_EXEC_CONTROL, pin_base.all);asm_vmwrite(CPU_BASED_VM_EXEC_CONTROL, process_base.all);asm_vmwrite(SECONDARY_VM_EXEC_CONTROL, secondary_process_base.all);asm_vmwrite(VM_ENTRY_CONTROLS, entry.all);asm_vmwrite(VM_EXIT_CONTROLS, exit.all); |
当执行流返回后，很快就会触发三重中断(VMware的log日志)，但是我无法确定具体错误发生在哪里？vmcs的设置存在问题吗？vm-exit部分屏蔽了msr的读写触发的退出，此外只处理了cpuid指令，其余通过default进行打印。但是，在cpu进入关闭状态前，并未有任何的日志输出。
寄存器的属性值如上图示，以此为(selector, base, limit, rights)。
[培训]内核驱动高级班，冲击BAT一流互联网大厂工作，每周日13:00-18:00直播授课
|
|
|---|---|
|
|
|
|
|
guest.rip返回后，系统实现虚拟化。关键是，发生错误时似乎并没有进入vm-exit，无法捕捉异常。 |
|
|
我在host的处理例程中设置了int3断点，正常情况下，gdb可以在断点处断下来。但是在发生三重错误时，host并没有任何输出日志或者捕捉到中断 |
|
|
虽然2号exit-reason对应于Triple fault，但是似乎无法捕获而陷入vm-exit。
|
|
|
我设置了exception-bitmap的double fault的拦截，但是似乎无法触发(kprint没有任何输出消息)... |
|
|
Double Fault 一般是OS处理，不是你的Vmm处理。所以你需要找到First Fault的RIP来猜测哪里出了问题，因此你需要在内存中保存Dbg数据，这样出了问题你才能好调试，在某些条件下Int 3会卡死或者hung住。所以在确定int 3不会影响正常执行流程的情况下在使用。 |
|
|
|
|
|
|
|
|
参考https://bbs.kanxue.com/thread-284530.htm |
