Mercurial > hg > xemacs-beta
comparison src/process-unix.c @ 930:eaedf30d9d76
[xemacs-hg @ 2002-07-23 08:34:59 by youngs]
2002-07-15 Jerry James <james@xemacs.org>
* make-docfile.c: Change whitespace and organization to reduce the
size of the diff against FSF Emacs sources and synch to Emacs 21.2.
Remove unused DO_REALLOC.
Mark XEmacs changes and additions more clearly.
Reintroduce previously deleted Emacs code inside #if 0 ... #endif.
* make-docfile.c (next_extra_elc): Replace goto with do-while.
* make-docfile.c (main): Put XEmacs-only args in one place.
* make-docfile.c (write_c_args): Change buff to buf to match
Emacs. Replace pointer arithmetic with simpler array syntax.
* make-docfile.c (scan_c_file): Note that DEFSIMPLE and DEFPRED no
longer exist. Correct the "name too long" test (off by one).
Die with message if a DEFUN has no docstring instead of hanging.
* make-docfile.c (scan_lisp_file): Introduce while loops used in
Emacs sources to skip consecutive blank lines.
2002-07-21 John Paul Wallington <jpw@xemacs.org>
* process.el (substitute-env-vars): New function; sync with
GNU Emacs 21.1.50.
(setenv): Add optional arg SUBSTITUTE-ENV-VARS; sync with
GNU Emacs 21.1.50.
2002-07-20 Mike Sperber <mike@xemacs.org>
* eval.c (run_post_gc_hook): Use more correct flags when running
post-gc-hook.
2002-07-20 Mike Sperber <mike@xemacs.org>
* process-unix.c (child_setup): Don't try to close file
descriptors for chid process once again---it's already being done
in close_process_descs.
(unix_create_process): Call begin_dont_check_for_quit to inhibit
unwanted interaction (and thus breaking of X event synchronicity)
in the child.
2002-07-15 Jerry James <james@xemacs.org>
* lisp.h: Make Qdll_error visible globally.
* symbols.c (check_sane_subr): Revert 2002-06-26 change.
Check only if !initialized.
* symbols.c (check_module_subr): Add parameter. Duplicate
check_sane_subr checks, but signal an error instead of asserting.
* symbols.c (defsubr): Use check_module_subr parameter.
* symbols.c (defsubr_macro): Ditto.
| author | youngs |
|---|---|
| date | Tue, 23 Jul 2002 08:35:11 +0000 |
| parents | b0c24ea6a2a8 |
| children | 5c7570514cc5 |
comparison
equal
deleted
inserted
replaced
| 929:0c272be3414c | 930:eaedf30d9d76 |
|---|---|
| 1021 | 1021 |
| 1022 retry_close (in); | 1022 retry_close (in); |
| 1023 retry_close (out); | 1023 retry_close (out); |
| 1024 retry_close (err); | 1024 retry_close (err); |
| 1025 | 1025 |
| 1026 /* I can't think of any reason why child processes need any more | |
| 1027 than the standard 3 file descriptors. It would be cleaner to | |
| 1028 close just the ones that need to be, but the following brute | |
| 1029 force approach is certainly effective, and not too slow. | |
| 1030 | |
| 1031 #### Who the hell added this? We already close the descriptors | |
| 1032 by using close_process_descs()!!! --ben */ | |
| 1033 { | |
| 1034 int fd; | |
| 1035 for (fd = 3; fd <= 64; fd++) | |
| 1036 retry_close (fd); | |
| 1037 } | |
| 1038 | |
| 1039 /* we've wrapped execve; it translates its arguments */ | 1026 /* we've wrapped execve; it translates its arguments */ |
| 1040 qxe_execve (new_argv[0], new_argv, env); | 1027 qxe_execve (new_argv[0], new_argv, env); |
| 1041 | 1028 |
| 1042 stdout_out ("Can't exec program %s\n", new_argv[0]); | 1029 stdout_out ("Can't exec program %s\n", new_argv[0]); |
| 1043 _exit (1); | 1030 _exit (1); |
| 1137 { | 1124 { |
| 1138 /**** Now we're in the child process ****/ | 1125 /**** Now we're in the child process ****/ |
| 1139 int xforkin = forkin; | 1126 int xforkin = forkin; |
| 1140 int xforkout = forkout; | 1127 int xforkout = forkout; |
| 1141 int xforkerr = forkerr; | 1128 int xforkerr = forkerr; |
| 1129 | |
| 1130 /* Checking for quit in the child is bad because that will | |
| 1131 cause I/O, and that, in turn, can confuse the X connection. */ | |
| 1132 begin_dont_check_for_quit(); | |
| 1142 | 1133 |
| 1143 /* Disconnect the current controlling terminal, pursuant to | 1134 /* Disconnect the current controlling terminal, pursuant to |
| 1144 making the pty be the controlling terminal of the process. | 1135 making the pty be the controlling terminal of the process. |
| 1145 Also put us in our own process group. */ | 1136 Also put us in our own process group. */ |
| 1146 | 1137 |