|
1 2 3 |
php_source_dir="/usr/src/php-7.2.12" sed -i "s/#define HAVE_LQ_TCP_INFO 1/\/\/ #define HAVE_LQ_TCP_INFO 1/g" ${php_source_dir}/main/php_config.h make -j32 clean && make -j$(cat /proc/cpuinfo | grep "processor" | wc -l) && make -j32 install |
Libraries:
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 |
apt -y install \ libcurl4-openssl-dev \ libgd-dev \ libsodium-dev \ libxml2-dev \ libpcre3-dev \ libssl-dev \ libsqlite3-dev \ libzip-dev \ libbz2-dev \ libedit-dev \ libxmlrpc-epi-dev \ libxmlrpc-core-c3-dev \ libxslt1-dev \ libwebp-dev \ libreadline-dev |
Configure:
|
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 |
./configure \ --prefix=/usr/local/php-7.2 \ --build=x86_64-linux-gnu \ --host=x86_64-linux-gnu \ --config-cache --cache-file=$(pwd -P)/config.cache \ --with-libdir=lib/x86_64-linux-gnu \ --enable-cli \ --enable-fpm \ --with-config-file-path=\${prefix}/etc \ --with-config-file-scan-dir=\${prefix}/etc/conf.d \ --with-fpm-user=www-data \ --with-fpm-group=www-data \ --disable-debug \ --with-pic \ --with-layout=GNU \ --with-pear \ --enable-filter \ --with-openssl=yes \ --with-pcre-regex=/usr \ --enable-hash \ --with-mhash=/usr \ --enable-libxml \ --enable-session \ --with-sodium \ --with-zlib=/usr \ --with-zlib-dir=/usr \ \ --enable-opcache \ --enable-opcache-file \ --enable-huge-code-pages \ \ --enable-mysqlnd \ --enable-mysqlnd-compression-support \ --with-zlib-dir=/usr \ --with-mysqli=mysqlnd \ --with-pdo-mysql=mysqlnd \ \ --with-sqlite3=/usr \ --with-pdo-sqlite=/usr \ \ --enable-json \ \ --enable-intl \ \ --with-curl=/usr \ \ --with-bz2=/usr \ \ --with-gd=/usr \ --with-jpeg-dir=/usr \ --with-xpm-dir=/usr/X11R6 \ --with-png-dir=/usr \ --with-freetype-dir=/usr \ --with-webp-dir=/usr \ \ --enable-mbstring \ --enable-mbregex \ --enable-mbregex-backtrack \ \ --with-libedit=/usr \ \ --enable-wddx \ --enable-xml \ --with-xsl=/usr \ \ --with-xmlrpc=/usr \ --with-libxml-dir=/usr \ \ --enable-zip \ --with-zlib-dir=/usr \ --with-libzip=/usr \ \ --enable-calendar \ --enable-ctype \ --enable-exif \ --enable-fileinfo \ --enable-ftp \ --with-openssl-dir=/usr \ --with-gettext=/usr \ --with-iconv \ --enable-pdo \ --enable-phar \ --enable-pcntl \ --enable-posix \ --enable-shmop \ --enable-sockets \ --enable-sysvmsg \ --enable-sysvsem \ --enable-sysvshm \ --enable-tokenizer |
Make xmlrpc error 1:
|
1 2 3 4 5 6 7 8 9 |
/usr/src/php-7.2.12/ext/xmlrpc/xmlrpc-epi-php.c:187:2: error: unknown type name 'XMLRPC_SERVER' XMLRPC_SERVER server_ptr; ^~~~~~~~~~~~~ /usr/src/php-7.2.12/ext/xmlrpc/xmlrpc-epi-php.c:195:2: error: unknown type name 'STRUCT_XMLRPC_REQUEST_OUTPUT_OPTIONS' STRUCT_XMLRPC_REQUEST_OUTPUT_OPTIONS xmlrpc_out; ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ... Makefile:1933: recipe for target 'ext/xmlrpc/xmlrpc-epi-php.lo' failed |
Open Makefile, add
|
1 |
-I/usr/include/xmlrpc-epi |
to the line(1933) where the error occured. xmlrpc error 2:
|
1 2 3 4 5 6 7 8 9 |
ext/xmlrpc/.libs/xmlrpc-epi-php.o: In function `destroy_server_data': /usr/src/php-7.2.12/ext/xmlrpc/xmlrpc-epi-php.c:268: undefined reference to `XMLRPC_ServerDestroy' ext/xmlrpc/.libs/xmlrpc-epi-php.o: In function `zm_info_xmlrpc': ... collect2: error: ld returned 1 exit status Makefile:296: recipe for target 'sapi/cli/php' failed make: *** [sapi/cli/php] Error 1 |
[…]
在x86_64的Debian Jessie上交叉编译一个mipsel架构的程序，能静态链接的都静态链接了。 程序编译好传送到mipsel架构的设备上时，ldd查看库的依赖，却提示
|
1 |
/lib/mipsel-linux-gnu/libc.so.6: version `GLIBC_2.17' not found |
(⊙﹏⊙)b mipsel设备上运行着Debian Wheezy，使用的是glibc 2.13：
|
1 2 |
# /lib/mipsel-linux-gnu/libc.so.6 GNU C Library (Debian EGLIBC 2.13-38+deb7u11) stable release version 2.13, by Roland McGrath et al. |
尝试了更新到glibc 2.17，结果运行所有动态链接glibc的程序都提示Segmentation fault或Bus error，眼看要变砖!!!∑(ﾟДﾟノ)ノ，最后用一个静态链接版本 […]
C语言基本类型的返回通过寄存器实现。 寄存器容量不大，一个结构体的大小不定，多塞几个成员，很容易就会超过寄存器的容量，不太可能通过寄存器一次性返回一个结构体。 把以下代码转为汇编，探个究竟:
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 |
// return_struct.c struct the_struct { int num1; int num2; double num3; double num4; int *num_ptr1; int *num_ptr2; }; struct the_struct fun() { struct the_struct the_struct1; the_struct1.num1 = 1; the_struct1.num2 = 2; return the_struct1; } int get_struct() { struct the_struct the_struct1 = fun(); return the_struct1.num1 + 1; } |
优化后的代码可能比较晦涩，64位寄存器数量是32位的一倍，为了简便理解，关闭代码优化，并且转为32位汇编，命令:
|
1 |
bash # gcc -O0 -S -masm=intel -m32 return_struct.c |
结果:
|
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 |
.file "return_struct.c" .intel_syntax noprefix .text .globl _fun .def _fun; .scl 2; .type 32; .endef _fun: push ebp mov ebp, esp sub esp, 32 ; 为结构体在栈帧中分配32字节内存 mov DWORD PTR [ebp-32], 1 ; *(ebp - 32) = 1 mov DWORD PTR [ebp-28], 2 ; *(ebp - 28) = 1 ; ebp指向的位置储存着get_struct()函数的ebp值 ; ebp+4指向返回地址 ; ebp+8是get_struct()函数中结构体的指针 mov eax, DWORD PTR [ebp+8] ; eax = get_struct()函数中结构体的指针 mov edx, DWORD PTR [ebp-32] ; edx是fun()函数中结构体首DWORD的值 mov DWORD PTR [eax], edx ; *eax = edx，让get_struct()中的结构体第一个DWORD = edx ; 结构体剩余的其余值传值 mov edx, DWORD PTR [ebp-28] mov DWORD PTR [eax+4], edx mov edx, DWORD PTR [ebp-24] mov DWORD PTR [eax+8], edx mov edx, DWORD PTR [ebp-20] mov DWORD PTR [eax+12], edx mov edx, DWORD PTR [ebp-16] mov DWORD PTR [eax+16], edx mov edx, DWORD PTR [ebp-12] mov DWORD PTR [eax+20], edx mov edx, DWORD PTR [ebp-8] mov DWORD PTR [eax+24], edx mov edx, DWORD PTR [ebp-4] mov DWORD PTR [eax+28], edx mov eax, DWORD PTR [ebp+8] leave ret 4 .globl _get_struct .def _get_struct; .scl 2; .type 32; .endef _get_struct: push ebp mov ebp, esp sub esp, 40 ; 分配40字节内存 lea eax, [ebp-32] ; eax = ebp - 32，eax是结构体的指针 mov DWORD PTR [esp], eax ; *esp储存结构体的指针，即get_struct()函数栈顶是结构体的指针 call _fun sub esp, 4 mov eax, DWORD PTR [ebp-32] ; (ebp - 32)是结构体的地址 add eax, 1 leave ret .ident "GCC: (GNU) 5.3.0" |
汇编代码已写上注释。 在get_struct()函数中，调用fun()之前，会首先把ge […]
当初看到Visual Studio 2010的这个提示，我真当VS实现了C LANG的函数重载。 今天即兴尝试VS下C LANG的函数重载，虽然IDE没报语法错误，但编译器编译时会报错。无奈之下只好打开长长的math.h一探究竟。 通过全文搜索很容易找到math.h声明的pow()原型：
|
1 2 3 4 5 6 7 8 9 10 11 |
double __cdecl pow(_In_ double _X, _In_ double _Y); inline double __CRTDECL pow(_In_ double _X, _In_ int _Y) {return (_Pow_int(_X, _Y)); } inline float __CRTDECL pow(_In_ float _X, _In_ float _Y) {return (powf(_X, _Y)); } inline float __CRTDECL pow(_In_ float _X, _In_ int _Y) {return (_Pow_int(_X, _Y)); } inline long double __CRTDECL pow(_In_ long double _X, _In_ long double _Y) {return (powl(_X, _Y)); } inline long double __CRTDECL pow(_In_ long double _X, _In_ int _Y) {return (_Pow_int(_X, _Y)); } |
声明的这些原型一眼看上去没什么问题，挺正常的重载方式。不过math.h内容有点多，不细心点还是看不出来，实际上pow()的重载使用了条件编译：
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 |
... #ifdef __cplusplus extern "C" { #endif ... double __cdecl pow(_In_ double _X, _In_ double _Y); ... #ifdef __cplusplus } extern "C++" { ... inline double __CRTDECL pow(_In_ double _X, _In_ int _Y) {return (_Pow_int(_X, _Y)); } inline float __CRTDECL pow(_In_ float _X, _In_ float _Y) {return (powf(_X, _Y)); } inline float __CRTDECL pow(_In_ float _X, _In_ int _Y) {return (_Pow_int(_X, _Y)); } inline long double __CRTDECL pow(_In_ long double _X, _In_ long double _Y) {return (powl(_X, _Y)); } inline long double __CRTDECL pow(_In_ long double _X, _In_ int _Y) {return (_Pow_int(_X, _Y)); } ... } #endif |
查找了一下资料，得知” […]