正在分析一个软件，其中涉及大数运算。在该软件中，大数使用如下结构表示：
struct BigInt
{
char Sign; /*表示符号。1为负；0为非负*/
unsigned long Alloced; /*为大整数数据所分配空间的元素数*/
unsigned long Used; /*实际使用空间的元素数*/
int *Data[]; /*大整数数据空间指针*/
}
该结构元素定位于四字节边界。
需要说明的是大整数的实际数据存放。大整数被转换成十六进制后，每十六位占据Data的一个元素，按低位在先的顺序存放。这样，软件中所使用的1024位的大整数所对应的实体数组就有64个元素。
在软件中，不论何时使用大整数，都去申请上述结构体以及数据实体所用的内存空间。
现在的问题是，当遇到其中一个函数的时候，我一时搞不清其算法，希望各位大牛指点一二。
传入该函数的共有三个大整数，即上述结构体的指针，其汇编代码如下，其中的n2就是第二个参数的实体数据的元素个数：
.text:10241C1E
.text:10241C1E ; =============== S U B R O U T I N E =======================================
.text:10241C1E
.text:10241C1E
.text:10241C1E ; int __usercall sub_10241C1E<eax>(struct BigInt *pBigInt1<eax>, struct BigInt *pBigInt2<edx>, struct BigInt *pBigInt3<ecx>)
.text:10241C1E sub_10241C1E proc near ; CODE XREF: sub_10241788+222
.text:10241C1E ; sub_10241788+251 ...
.text:10241C1E
.text:10241C1E TmpBigInt = BigInt ptr -20h
.text:10241C1E
.text:10241C1E push edi
.text:10241C1F push esi
.text:10241C20 push ebp
.text:10241C21 push ebx
.text:10241C22 sub esp, 1Ch
.text:10241C25 mov edi, ecx
.text:10241C27 mov ebp, edx
.text:10241C29 mov ebx, eax
.text:10241C2B lea eax, [esp+2Ch+TmpBigInt]
.text:10241C2F push ebx ; pSrcBigInt
.text:10241C30 push eax ; pDstBigInt
.text:10241C31 mov esi, [ebp+BigInt.Used] ; ESI=BigInt2.Used
.text:10241C34 call BigIntInitCopy ; TmpBigInt=BigInt1
.text:10241C39 add esp, 8
.text:10241C3C test eax, eax
.text:10241C3E jz short loc_10241C48
.text:10241C40 add esp, 1Ch
.text:10241C43 pop ebx
.text:10241C44 pop ebp
.text:10241C45 pop esi
.text:10241C46 pop edi
.text:10241C47 retn
.text:10241C48 ; ---------------------------------------------------------------------------
.text:10241C48
.text:10241C48 loc_10241C48: ; CODE XREF: sub_10241C1E+20
.text:10241C48 lea eax, [esp+2Ch+TmpBigInt] ; pBigInt
.text:10241C4C lea edx, [esi-1] ; n
.text:10241C4F call BigIntDiv_4nP2 ; TmpBigInt=TmpBigInt/(16^(n2-1))
.text:10241C54 lea eax, [esp+2Ch+TmpBigInt] ; pBigInt1
.text:10241C58 mov edx, edi ; pBigInt2
.text:10241C5A call BigIntUnsignedMul ; TmpBigInt=TmpBigInt*BigInt3
.text:10241C5F lea eax, [esp+2Ch+TmpBigInt] ; pBigInt
.text:10241C63 add esi, 1
.text:10241C66 mov edx, esi ; n
.text:10241C68 call BigIntDiv_4nP2 ; TmpBigInt=TmpBigInt/(16^(n2+1))
.text:10241C6D mov edi, esi
.text:10241C6F shl edi, 4
.text:10241C72 movzx edi, di
.text:10241C75 mov eax, ebx ; pBigInt
.text:10241C77 mov edx, edi ; n
.text:10241C79 call BigIntMod_nP2 ; BigInt1=BigInt1 Mod (16^(n2+1))
.text:10241C7E lea eax, [esp+2Ch+TmpBigInt] ; pBigInt1
.text:10241C82 mov edx, ebp ; pBigInt2
.text:10241C84 call BigIntUnsignedMul ; TmpBigInt=TmpBigInt*BigInt2
.text:10241C89 lea eax, [esp+2Ch+TmpBigInt] ; pBigInt
.text:10241C8D mov edx, edi ; n
.text:10241C8F call BigIntMod_nP2 ; TmpBigInt=TmpBigInt Mod (16^(n2+1))
.text:10241C94 lea edx, [esp+2Ch+TmpBigInt] ; BigInt2
.text:10241C98 mov eax, ebx ; BigInt1
.text:10241C9A mov ecx, ebx ; BigIntDiff
.text:10241C9C call BigIntSignedSub ; BigInt1=BigInt1-TmpBigInt
.text:10241CA1 mov edi, eax
.text:10241CA3 test edi, edi
.text:10241CA5 jnz short loc_10241CDB
.text:10241CA7 mov eax, ebx ; pBigInt
.text:10241CA9 call BigIntSgn
.text:10241CAE test eax, eax
.text:10241CB0 jl short loc_10241CEE ; BigInt1<0则转
.text:10241CB2
.text:10241CB2 loc_10241CB2: ; CODE XREF: sub_10241C1E+109
.text:10241CB2 mov eax, ebx ; BigInt1
.text:10241CB4 mov edx, ebp ; BigInt2
.text:10241CB6 call BigIntSignedCmp ; 比较BigInt1和BigInt2
.text:10241CBB test eax, eax
.text:10241CBD jl short loc_10241CDB ; BigInt1<BigInt2则转
.text:10241CBF
.text:10241CBF loc_10241CBF: ; CODE XREF: sub_10241C1E+BB
.text:10241CBF mov eax, ebx ; pDstBigInt
.text:10241CC1 mov edx, ebp ; pSrcBigInt
.text:10241CC3 call BigIntUnsignedSub ; BigInt1=BigInt1-BigInt2
.text:10241CC8 mov edi, eax
.text:10241CCA test edi, edi
.text:10241CCC jnz short loc_10241CDB ; 出错则转
.text:10241CCE mov eax, ebx ; BigInt1
.text:10241CD0 mov edx, ebp ; BigInt2
.text:10241CD2 call BigIntSignedCmp ; 比较BigInt1和BigInt2
.text:10241CD7 test eax, eax
.text:10241CD9 jge short loc_10241CBF ; BigInt1<BigInt2则循环
.text:10241CDB
.text:10241CDB loc_10241CDB: ; CODE XREF: sub_10241C1E+87
.text:10241CDB ; sub_10241C1E+9F ...
.text:10241CDB lea eax, [esp+2Ch+TmpBigInt] ; pBigInt
.text:10241CDF call UC_ReleaseBigInt ; 释放TmpBigInt
.text:10241CE4 mov eax, edi
.text:10241CE6 add esp, 1Ch
.text:10241CE9 pop ebx
.text:10241CEA pop ebp
.text:10241CEB pop esi
.text:10241CEC pop edi
.text:10241CED retn
.text:10241CEE ; ---------------------------------------------------------------------------
.text:10241CEE
.text:10241CEE loc_10241CEE: ; CODE XREF: sub_10241C1E+92
.text:10241CEE lea eax, [esp+2Ch+TmpBigInt] ; pBigInt
.text:10241CF2 call BigIntSetZ ; TmpBigInt清零
.text:10241CF7 mov eax, [esp+2Ch+TmpBigInt.DataPtr]
.text:10241CFB mov edi, 1
.text:10241D00 mov [eax], di ; TmpBigInt=1
.text:10241D03 lea eax, [esp+2Ch+TmpBigInt] ; SrcBigInt
.text:10241D07 mov edx, esi ; n
.text:10241D09 call BigIntMul_4nP2 ; TmpBigInt=TmpBigInt*(16^(n2+1))
.text:10241D0E mov edi, eax
.text:10241D10 test edi, edi
.text:10241D12 jnz short loc_10241CDB
.text:10241D14 lea edx, [esp+2Ch+TmpBigInt] ; BigInt2
.text:10241D18 mov eax, ebx ; BigInt1
.text:10241D1A mov ecx, ebx ; BigIntSum
.text:10241D1C call BigIntSignedAdd ; BigInt1=BigInt1+TmpBigInt
.text:10241D21 mov edi, eax
.text:10241D23 test edi, edi
.text:10241D25 jnz short loc_10241CDB ; 出错则转
.text:10241D27 jmp short loc_10241CB2
.text:10241D27 sub_10241C1E endp
.text:10241D27
.text:10241D27 ; ---------------------------------------------------------------------------
[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
