{"id":7923,"date":"2012-04-05T07:00:00","date_gmt":"2012-04-05T07:00:00","guid":{"rendered":"https:\/\/blogs.msdn.microsoft.com\/oldnewthing\/2012\/04\/05\/you-can-use-an-overlapped-structure-with-synchronous-io-too\/"},"modified":"2012-04-05T07:00:00","modified_gmt":"2012-04-05T07:00:00","slug":"you-can-use-an-overlapped-structure-with-synchronous-io-too","status":"publish","type":"post","link":"https:\/\/devblogs.microsoft.com\/oldnewthing\/20120405-00\/?p=7923","title":{"rendered":"You can use an OVERLAPPED structure with synchronous I\/O, too"},"content":{"rendered":"

\nEven if you didn’t open a file with\nFILE_FLAG_OVERLAPPED<\/code>,\nyou can still use the OVERLAPPED<\/code> structure\nwhen you issue reads and writes.\nMind you,\nthe I\/O will still complete synchronously,\nbut you can take advantage of the other stuff\nthat OVERLAPPED<\/code> has to offer.\n<\/p>\n

\nSpecifically, you can take advantage of the\nOffset<\/code> and OffsetHigh<\/code> members\nto issue the I\/O against a file location different from\nthe current file pointer.\n(This is a file pointer in the sense of\nSet­File­Pointer<\/code>\nand not in the sense of the C runtime FILE*<\/code>.)\nIf your program does a lot of reads and writes to random locations\nin a file, using the synchronous OVERLAPPED<\/code>\nstructure saves you a call to\nSet­File­Pointer<\/code> at each I\/O.\n<\/p>\n

\nLet’s illustrate this by writing some code to walk through\na file format that contains a lot of offsets to other\nparts of the file:\n\nThe ICO file format<\/a>.\nFirst, the old-fashioned way:\n<\/p>\n

\n#define UNICODE<\/a>\n#define _UNICODE\n#include <windows.h>\n#include <pshpack1.h>\nstruct ICONDIRHEADER {\n    WORD idReserved;\n    WORD idType;\n    WORD idCount;\n};\nstruct ICONDIRENTRY {\n    BYTE bWidth;\n    BYTE bHeight;\n    BYTE bColorCount;\n    BYTE  bReserved;\n    WORD  wPlanes;\n    WORD  wBitCount;\n    DWORD dwBytesInRes;\n    DWORD dwImageOffset;\n};\n#include <poppack.h>\nBOOL ReadBufferAt(__in HANDLE hFile,\n    __out_bcount(cbBuffer) void *pvBuffer,\n    DWORD cbBuffer,\n    DWORD64 offset)\n{\n LARGE_INTEGER li;\n DWORD cbRead;\n li.QuadPart = offset;\n return SetFilePointerEx(hFile, li, nullptr, FILE_BEGIN) &&\n        ReadFile(hFile, pvBuffer, cbBuffer, &cbRead, nullptr) &&\n        cbBuffer == cbRead;\n}\nint __cdecl wmain(int argc, wchar_t **argv)\n{\n HANDLE hFile = CreateFile(argv[1], GENERIC_READ,\n  FILE_SHARE_READ | FILE_SHARE_WRITE | FILE_SHARE_DELETE,\n  nullptr, OPEN_EXISTING, 0, nullptr);\n if (hFile != INVALID_HANDLE_VALUE) {\n  ICONDIRHEADER hdr;\n  if (ReadBufferAt(hFile, &hdr, sizeof(hdr), 0) &&\n      hdr.idReserved == 0 && hdr.idType == 1) {\n   for (UINT uiIcon = 0; uiIcon < hdr.idCount; uiIcon++) {\n    ICONDIRENTRY entry;\n    if (ReadBufferAt(hFile, &entry, sizeof(entry),\n                     sizeof(hdr) + uiIcon * sizeof(entry))) {\n     void *pvData = LocalAlloc(LMEM_FIXED, entry.dwBytesInRes);\n     if (pvData) {\n      if (ReadBufferAt(hFile, pvData,\n                       entry.dwBytesInRes, entry.dwImageOffset)) {\n       \/\/ process one image in the icon\n      }\n      LocalFree(pvData);\n     }\n    }\n   }\n  }\n  CloseHandle(hFile);\n }\n return 0;\n}\n<\/pre>\n
\nRun this program with the name of an icon file on the command line,\nand nothing interesting happens because the program doesn’t\ngenerate any output.\nBut if you step through it, you can see that we start by\nreading the ICON­DIR­HEADER<\/code>\nto verify that it’s an icon and determine the number of images.\nWe then loop through the images: For each one,\nwe read the ICON­DIR­ENTRY<\/code>\n(specifying the explicit file offset),\nthen read the image data (again, specifying the explicit\nfile offset).\n<\/p>\n
\nWe use the Read­Buffer­At<\/code> function\nto read data from the file.\nFor each read, we first call Set­File­Pointer<\/code>\nto position the file pointer at the byte we want to read,\nthen call Read­File<\/code> to read it.\n<\/p>\n
\nLet’s change this program to take advantage of our newfound knowledge:\n<\/p>\n
\nBOOL ReadBufferAt(__in HANDLE hFile,\n    __out_bcount(cbBuffer) void *pvBuffer,\n    DWORD cbBuffer,\n    DWORD64 offset)\n{\n OVERLAPPED o = { 0 };\n o.Offset = static_cast<DWORD>(offset);\n o.OffsetHigh = static_cast<DWORD>(offset >> 32);\n DWORD cbRead;\n return ReadFile(hFile, pvBuffer, cbBuffer, &cbRead, &o) &&\n        cbBuffer == cbRead;\n}\n<\/pre>\n
\nWe merge the\nSet­File­Pointer<\/code> call into the\nRead­File<\/code> by specifying the desired byte\noffset in the optional OVERLAPPED<\/code> structure.\nThe I\/O will still complete synchronously\n(since we opened the handle synchronously),\nbut we saved ourselves the hassle of having to call\ntwo functions when it could be done with just one.<\/p>\n","protected":false},"excerpt":{"rendered":"
Even if you didn’t open a file with FILE_FLAG_OVERLAPPED, you can still use the OVERLAPPED structure when you issue reads and writes. Mind you, the I\/O will still complete synchronously, but you can take advantage of the other stuff that OVERLAPPED has to offer. Specifically, you can take advantage of the Offset and OffsetHigh members […]<\/p>\n","protected":false},"author":1069,"featured_media":111744,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[1],"tags":[25],"class_list":["post-7923","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-oldnewthing","tag-code"],"acf":[],"blog_post_summary":"
Even if you didn’t open a file with FILE_FLAG_OVERLAPPED, you can still use the OVERLAPPED structure when you issue reads and writes. Mind you, the I\/O will still complete synchronously, but you can take advantage of the other stuff that OVERLAPPED has to offer. Specifically, you can take advantage of the Offset and OffsetHigh members […]<\/p>\n","_links":{"self":[{"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/posts\/7923","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/users\/1069"}],"replies":[{"embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/comments?post=7923"}],"version-history":[{"count":0,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/posts\/7923\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/media\/111744"}],"wp:attachment":[{"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/media?parent=7923"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/categories?post=7923"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devblogs.microsoft.com\/oldnewthing\/wp-json\/wp\/v2\/tags?post=7923"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}