Could the experts in these areas please comment on this patch of mine? I will add a ChangeLog after I resolve any remaining issues. Please see http://www.xemacs.org/list-archives/xemacs-beta/9911/msg00193.html for my assessment of the fstat/stat problems on Windows NT. I'd like to hear from you if I've drawn any wrong conclusions. Changes to src/callproc.c src/ntproc.c src/sysdep.c and src/sysdep.h implement the call-process patch which is in 21.1.8 already. However, following Ben's request, I have replaced stderr_out () with warn_when_safe (). The part I am concerned about is my ENCAPSULATION for fstat, which is required on Windows NT to avoid inconsistencies between the already encapsulated stat system call and the broken fstat on NT. Do I need to qualify it somehow, so that it only affects native Windows NT builds? I didn't understand how stat is only encapsualted on NT native. The files changed are: src/nt.c src/sysdep.c src/sysfile.h Please advise about any shortcomings of this patch and directions for possible fixes. Many thanks! Adrian I have verified a patched XEmacs to report consistent file times cd d:\tmp\21.2\xemacs\ cvs diff Compilation started at Wed Nov 10 03:09:36 1999 +0100 (W. Europe Standard Time) Index: src/callproc.c =================================================================== RCS file: /usr/CVSroot/XEmacs/xemacs/src/callproc.c,v retrieving revision 1.29.2.6 diff -u -r1.29.2.6 callproc.c --- callproc.c 1999/05/01 05:13:23 1.29.2.6 +++ callproc.c 1999/11/10 02:13:47 ＠＠ -104,6 +104,9 ＠＠ { int fd = XINT (Fcar (fdpid)); int pid = XINT (Fcdr (fdpid)); +#ifdef WINDOWSNT + HANDLE pHandle; +#endif if (!call_process_exited && EMACS_KILLPG (pid, SIGINT) == 0) ＠＠ -114,7 +117,17 ＠＠ /* #### "c-G" -- need non-consing Single-key-description */ message ("Waiting for process to die...(type C-g again to kill it instantly)"); +#ifdef WINDOWSNT + pHandle = OpenProcess(PROCESS_ALL_ACCESS, 0, pid); + if (pHandle == NULL) + { + warn_when_safe (Qprocess, Qerror, + "OpenProcess returns NULL process handle."); + } + wait_for_termination (pHandle); +#else wait_for_termination (pid); +#endif /* "Discard" the unwind protect. */ XCAR (fdpid) = Qnil; ＠＠ -170,6 +183,9 ＠＠ Lisp_Object infile, buffer, current_dir, display, path; int fd[2]; int filefd; +#ifdef WINDOWSNT + HANDLE pHandle; +#endif int pid; char buf[16384]; char *bufptr = buf; ＠＠ -359,6 +375,20 ＠＠ #ifdef WINDOWSNT pid = child_setup (filefd, fd1, fd_error, new_argv, (char *) XSTRING_DATA (current_dir)); + if (!INTP (buffer)) + { + /* OpenProcess() as soon after child_setup as possible. It's too + late once the process terminated. */ + pHandle = OpenProcess(PROCESS_ALL_ACCESS, 0, pid); + if (pHandle == NULL) + { + warn_when_safe (Qprocess, Qerror, + "OpenProcess returns NULL process handle."); + } + } + /* Close STDERR into the parent process. We no longer need it. */ + if (fd_error >= 0) + close (fd_error); #else /* not WINDOWSNT */ pid = fork (); ＠＠ -394,12 +424,14 ＠＠ if (!NILP (fork_error)) signal_error (Qfile_error, fork_error); +#ifndef WINDOWSNT if (pid < 0) { if (fd[0] >= 0) close (fd[0]); report_file_error ("Doing fork", Qnil); } +#endif if (INTP (buffer)) { ＠＠ -507,7 +539,11 ＠＠ QUIT; /* Wait for it to terminate, unless it already has. */ +#ifdef WINDOWSNT + wait_for_termination (pHandle); +#else wait_for_termination (pid); +#endif /* Don't kill any children that the subprocess may have left behind when exiting. */ Index: src/nt.c =================================================================== RCS file: /usr/CVSroot/XEmacs/xemacs/src/nt.c,v retrieving revision 1.17.2.4 diff -u -r1.17.2.4 nt.c --- nt.c 1999/07/05 05:56:43 1.17.2.4 +++ nt.c 1999/11/10 02:14:24 ＠＠ -1313,6 +1313,40 ＠＠ #endif +/* Since stat is encapsulated on Windows NT, we need to encapsulate + the equally broken fstat as well. */ +int +fstat (int handle, struct stat *buffer) +{ + int ret; + BY_HANDLE_FILE_INFORMATION lpFileInfo; + /* Initialize values */ + buffer->st_mode = 0; + buffer->st_size = 0; + buffer->st_dev = 0; + buffer->st_rdev = 0; + buffer->st_atime = 0; + buffer->st_ctime = 0; + buffer->st_mtime = 0; + buffer->st_nlink = 0; + ret = GetFileInformationByHandle((HANDLE) handle, &lpFileInfo); + if (!ret) + { + return -1; + } + else + { + buffer->st_mtime = convert_time (lpFileInfo.ftLastWriteTime); + buffer->st_atime = convert_time (lpFileInfo.ftLastAccessTime); + if (buffer->st_atime == 0) buffer->st_atime = buffer->st_mtime; + buffer->st_ctime = convert_time (lpFileInfo.ftCreationTime); + if (buffer->st_ctime == 0) buffer->st_ctime = buffer->st_mtime; + buffer->st_size = lpFileInfo.nFileSizeLow; + buffer->st_nlink = (short) lpFileInfo.nNumberOfLinks; + return 0; + } +} + /* MSVC stat function can't cope with UNC names and has other bugs, so replace it with our own. This also allows us to calculate consistent inode values without hacks in the main Emacs code. */ Index: src/ntproc.c =================================================================== RCS file: /usr/CVSroot/XEmacs/xemacs/src/ntproc.c,v retrieving revision 1.14.2.10 diff -u -r1.14.2.10 ntproc.c --- ntproc.c 1999/10/10 12:40:13 1.14.2.10 +++ ntproc.c 1999/11/10 02:14:30 ＠＠ -445,16 +445,8 ＠＠ cp->procinfo.hThread=NULL; cp->procinfo.hProcess=NULL; - /* Hack for Windows 95, which assigns large (ie negative) pids */ - if (cp->pid < 0) - cp->pid = -cp->pid; - /* pid must fit in a Lisp_Int */ -#ifdef USE_UNION_TYPE - cp->pid = (cp->pid & ((1U << VALBITS) - 1)); -#else - cp->pid = (cp->pid & VALMASK); -#endif + *pPid = cp->pid; Index: src/sysdep.c =================================================================== RCS file: /usr/CVSroot/XEmacs/xemacs/src/sysdep.c,v retrieving revision 1.32.2.7 diff -u -r1.32.2.7 sysdep.c --- sysdep.c 1999/07/05 05:56:44 1.32.2.7 +++ sysdep.c 1999/11/10 02:14:44 ＠＠ -234,8 +234,11 ＠＠ #endif /* NO_SUBPROCESSES */ -void -wait_for_termination (int pid) +#ifdef WINDOWSNT +void wait_for_termination (HANDLE pHandle) +#else +void wait_for_termination (int pid) +#endif { /* #### With the new improved SIGCHLD handling stuff, there is much less danger of race conditions and some of the comments below ＠＠ -345,6 +348,53 ＠＠ Since implementations may add their own error indicators on top, we ignore it by default. */ +#elif defined (WINDOWSNT) + int ret = 0, status = 0; + if (pHandle == NULL) + { + warn_when_safe (Qprocess, Qerror, "Cannot wait for NULL process handle."); + return; + } + do + { + QUIT; + ret = WaitForSingleObject(pHandle, 100); + } + while (ret == WAIT_TIMEOUT); + if (ret == WAIT_FAILED) + { + warn_when_safe (Qprocess, Qerror, + "WaitForSingleObject returns WAIT_FAILED for process handle %p.", + pHandle); + } + if (ret == WAIT_ABANDONED) + { + warn_when_safe (Qprocess, Qerror, + "WaitForSingleObject returns WAIT_ABANDONED for process handle %p.", + pHandle); + } + if (ret == WAIT_OBJECT_0) + { + ret = GetExitCodeProcess(pHandle, &status); + if (ret) + { + synch_process_alive = 0; + synch_process_retcode = status; + } + else + { + /* GetExitCodeProcess() didn't return a valid exit status, + nothing to do. APA */ + warn_when_safe (Qprocess, Qerror, + "GetExitCodeProcess fails for process handle %p.", + pHandle); + } + } + if (pHandle != NULL && !CloseHandle(pHandle)) + { + warn_when_safe (Qprocess, Qerror, + "CloseHandle fails for process handle %p.", pHandle); + } #elif defined (EMACS_BLOCK_SIGNAL) && !defined (BROKEN_WAIT_FOR_SIGNAL) && defined (SIGCHLD) while (1) { ＠＠ -376,7 +426,7 ＠＠ Try defining BROKEN_WAIT_FOR_SIGNAL. */ EMACS_WAIT_FOR_SIGNAL (SIGCHLD); } -#else /* not HAVE_WAITPID and (not EMACS_BLOCK_SIGNAL or BROKEN_WAIT_FOR_SIGNAL) */ +#else /* not HAVE_WAITPID and not WINDOWSNT and (not EMACS_BLOCK_SIGNAL or BROKEN_WAIT_FOR_SIGNAL) */ /* This approach is kind of cheesy but is guaranteed(?!) to work for all systems. */ while (1) ＠＠ -578,7 +628,11 ＠＠ static void sys_subshell (void) { +#ifdef WINDOWSNT + HANDLE pid; +#else int pid; +#endif struct save_signal saved_handlers[5]; Lisp_Object dir; unsigned char *str = 0; ＠＠ -617,7 +671,7 ＠＠ xyzzy: #ifdef WINDOWSNT - pid = -1; + pid = NULL; #else /* not WINDOWSNT */ pid = fork (); ＠＠ -651,7 +705,7 ＠＠ #ifdef WINDOWSNT /* Waits for process completion */ pid = _spawnlp (_P_WAIT, sh, sh, NULL); - if (pid == -1) + if (pid == NULL) write (1, "Can't execute subshell", 22); #else /* not WINDOWSNT */ ＠＠ -3036,6 +3090,15 ＠＠ return readlink (path, buf, bufsiz); } #endif /* ENCAPSULATE_READLINK */ + + +#ifdef ENCAPSULATE_FSTAT +int +sys_fstat (int fd, struct stat *buf) +{ + return fstat (fd, buf); +} +#endif /* ENCAPSULATE_FSTAT */ #ifdef ENCAPSULATE_STAT Index: src/sysdep.h =================================================================== RCS file: /usr/CVSroot/XEmacs/xemacs/src/sysdep.h,v retrieving revision 1.11.2.1 diff -u -r1.11.2.1 sysdep.h --- sysdep.h 1998/12/05 16:56:31 1.11.2.1 +++ sysdep.h 1999/11/10 02:14:45 ＠＠ -48,7 +48,12 ＠＠ /* Wait for subprocess with process id `pid' to terminate and make sure it will get eliminated (not remain forever as a zombie) */ +#ifdef WINDOWSNT +#include <windows.h> +void wait_for_termination (HANDLE pid); +#else void wait_for_termination (int pid); +#endif /* flush any pending output * (may flush input as well; it does not matter the way we use it) Index: src/sysfile.h =================================================================== RCS file: /usr/CVSroot/XEmacs/xemacs/src/sysfile.h,v retrieving revision 1.7.2.1 diff -u -r1.7.2.1 sysfile.h --- sysfile.h 1998/12/05 16:56:32 1.7.2.1 +++ sysfile.h 1999/11/10 02:14:46 ＠＠ -380,6 +380,18 ＠＠ # define sys_readlink readlink #endif +#ifdef ENCAPSULATE_FSTAT +int sys_fstat (int fd, struct stat *buf); +#endif +#if defined (ENCAPSULATE_FSTAT) && !defined (DONT_ENCAPSULATE) +# undef fstat +/* Need to use arguments to avoid messing with struct stat */ +# define fstat(fd, buf) sys_fstat (fd, buf) +#endif +#if !defined (ENCAPSULATE_FSTAT) && defined (DONT_ENCAPSULATE) +# define sys_fstat fstat +#endif + #ifdef ENCAPSULATE_STAT int sys_stat (CONST char *path, struct stat *buf); #endif cvs server: Diffing src/m cvs server: Diffing src/s cvs server: Diffing tests cvs server: Diffing tests/DLL cvs server: Diffing tests/Dnd cvs server: Diffing tests/automated cvs server: Diffing tests/mule cvs server: Diffing tests/tooltalk Compilation exited abnormally with code 1 at Wed Nov 10 03:14:21