Re: building with make -j4
25 years
Gunnar Evermann
Gunnar Evermann
Sean MacLennan <seanm(a)storm.ca> writes:
> Where are you putting the -j4? I have a small cluster I build with
> and I add the -j4 flag to the Makefile/GNUmakefile.
I just used 'make -j4 beta'. Is that too simplistic?
To be honest I wasn't really interested in saving time, I was only
trying to find bugs in the Makefiles :-)
Gunnar
- 2 participants
- 1 comments
OpenBSD 2.7 - no go at all
25 years
Robin S. Socha
Robin S. Socha
No luck today...
(root＠deathwish):(~)# uname -a
OpenBSD deathwish 2.7 DEATHWISH#0 i386
(root＠deathwish):(~)# gcc --version
2.95.2
(root＠deathwish):(~)# /usr/local/bin/gmake --version
GNU Make version 3.78.1, by Richard Stallman and Roland McGrath.
Built for i386-unknown-openbsd2.7
(root＠deathwish):(/usr/local/src/CVS/xemacs-21.2)# cat version.sh
#!/bin/sh
emacs_is_beta=t
emacs_major_version=21
emacs_minor_version=2
emacs_beta_version=35
xemacs_codename="Nike"
infodock_major_version=4
infodock_minor_version=0
infodock_build_version=8
cd ./src && /usr/local/bin/gmake all
gmake[1]: Entering directory `/usr/local/src/CVS/xemacs-21.2/src'
GNUmakefile:279: *** missing separator. Stop.
gmake[1]: Leaving directory `/usr/local/src/CVS/xemacs-21.2/src'
gmake: *** [src] Error 2
(root＠deathwish):(/usr/local/src/CVS/gtk-xemacs)# cat version.sh
#!/bin/sh
emacs_major_version=21
emacs_minor_version=1
emacs_beta_version=11
xemacs_codename="Carlsbad Caverns"
infodock_major_version=4
infodock_minor_version=0
infodock_build_version=7
EMACSBOOTSTRAPLOADPATH="/usr/local/src/CVS/gtk-xemacs/src/../lisp/:/usr/local/src/CVS/gtk-xemacs" ./temacs -batch -l /usr/local/src/CVS/gtk-xemacs/src/../lisp/update-elc.el
Loading /usr/local/src/CVS/gtk-xemacs/src/../lisp/update-elc.el...
Loading very-early-lisp... (1860)
Loading find-paths.el... (10464)
Loading packages.el... (18800)
Loading setup-paths.el... (4120)
Loading dump-paths.el... (0)
Loading /usr/local/src/CVS/gtk-xemacs/src/../lisp/dumped-lisp.el... (136)
Loading loadup-el.el...
Loading loadup.el...
Using load-path (/usr/local/src/CVS/gtk-xemacs/src/../lisp/ /usr/local/src/CVS/gtk-xemacs)
Loading dumped-lisp.el...
Loading backquote.el...
Loading bytecomp-runtime.el...
Loading Installation.el...
Loading find-paths.el...
Loading packages.el...
Loading setup-paths.el...
Loading dump-paths.el...
Loading subr.el...
Loading replace.el...
Loading version.el...
Loading cl.el...
Loading cl-extra.el...
Loading cl-seq.el...
Loading widget.el...
Loading custom.el...
Loading cus-start.el...
Loading cmdloop.el...
Loading keymap.el...
Loading syntax.el...
Loading device.el...
Loading console.el...
Loading obsolete.el...
Loading specifier.el...
Loading faces.el...
Loading cl-macs.el...*** Error in XEmacs initialization
(error "Must be string or vector" global)
*** Backtrace
really-early-error-handler((error "Must be string or vector" global))
check-valid-instantiator(global image)
(if noerror t (check-valid-instantiator spec specifier-type))
)
(if (not (consp spec)) (if noerror t (check-valid-instantiator spec specifier-type)) (if (not ...) (if noerror t ...) (let ... ...)))
)
(if (not (eq result t)) (cons (quote global) result) (if (not ...) (if noerror t ...) (if ... ... ...)))
)
# bind (result)
(let ((result ...)) (if (not ...) (cons ... result) (if ... ... ...)))
)
# bind (noerror specifier-type spec)
canonicalize-spec(global image nil)
(let ((res2 ...)) (if (eq res2 t) (throw ... t) (setq result ...)))
)
(if (not (consp rest)) (if noerror (throw ... t) (signal ... ...)) (let (...) (if ... ... ...)))
)
(while rest (if (not ...) (if noerror ... ...) (let ... ...)) (setq rest (cdr rest)))
)
# bind (result rest)
(let ((rest spec-list) (result nil)) (while rest (if ... ... ...) (setq rest ...)) (nreverse result))
)
# (catch cann-spec-list ...)
(catch (quote cann-spec-list) (let (... ...) (while rest ... ...) (nreverse result)))
)
(if (not (consp spec-list)) (if noerror t (check-valid-instantiator spec-list specifier-type)) (catch (quote cann-spec-list) (let ... ... ...)))
)
(if (not (eq result t)) (list result) (if (not ...) (if noerror t ...) (catch ... ...)))
)
# bind (result)
(let ((result ...)) (if (not ...) (list result) (if ... ... ...)))
)
# bind (noerror specifier-type spec-list)
canonicalize-spec-list((global ((x default mono) . "gray3") ((gtk default mono) . "gray3") ((mswindows default mono) . "gray3")) image)
(add-spec-list-to-specifier specifier (canonicalize-spec-list nval (specifier-type specifier)) how-to-add)
)
(cond ((and ... ...) (copy-specifier nval specifier locale tag-set nil how-to-add)) (t (if tag-set ...) (if locale ...) (add-spec-list-to-specifier specifier ... how-to-add)))
)
# bind (nval is-valid)
(let ((is-valid ...) (nval value)) (cond (... ...) (t ... ... ...)))
)
# bind (how-to-add tag-set locale value specifier)
set-specifier(#<image-specifier global=<unspecified> fallback=#<image-specifier global=<unspecified> 0x198> 0x3e2> (((x default mono) . "gray3") ((gtk default mono) . "gray3") ((mswindows default mono) . "gray3")) global nil nil)
(if (memq property built-in-face-specifiers) (set-specifier (get face property) value locale tag-set how-to-add) (if (not locale) (put face property value) (convert-face-property-into-specifier face property) (add-spec-to-specifier ... value locale tag-set how-to-add)))
)
# bind (how-to-add tag-set locale value property face)
set-face-property(primary-selection background-pixmap (((x default mono) . "gray3") ((gtk default mono) . "gray3") ((mswindows default mono) . "gray3")) global nil nil)
# bind (how-to-add tag-set locale pixmap face)
set-face-background-pixmap(primary-selection (((x default mono) . "gray3") ((gtk default mono) . "gray3") ((mswindows default mono) . "gray3")) global)
# bind (current-load-list)
# (unwind-protect ...)
# bind (load-file-name)
# (unwind-protect ...)
# (unwind-protect ...)
# (unwind-protect ...)
# (unwind-protect ...)
# (unwind-protect ...)
load("/usr/local/src/CVS/gtk-xemacs/lisp/faces.el")
(prog1 (load full-path) (garbage-collect))
)
(if full-path (prog1 (load full-path) (garbage-collect)) (external-debugging-output (format "\nLoad file %s: not found\n" file)) nil)
)
# bind (full-path)
(let ((full-path ...)) (if full-path (prog1 ... ...) (external-debugging-output ...) nil))
)
# bind (file)
pureload("faces")
(or (pureload file) (progn (external-debugging-output "Fatal error during load, aborting") (kill-emacs 1)))
)
(while (setq file (car files)) (or (pureload file) (progn ... ...)) (setq files (cdr files)))
)
# bind (file files)
(let ((files preloaded-file-list) file) (while (setq file ...) (or ... ...) (setq files ...)) (if (not ...) (progn ... ...)) (fmakunbound (quote pureload)))
)
(lambda nil (setq load-path (split-path ...)) (external-debugging-output (format "\nUsing load-path %s" load-path)) (buffer-disable-undo (get-buffer "*scratch*")) (load "very-early-lisp" nil t) (let (...) (setq load-path ...)) (setq load-warn-when-source-newer t load-warn-when-source-only t) (defun pureload (file) (let ... ...)) (load (concat default-directory "../lisp/dumped-lisp.el")) (let (... file) (while ... ... ...) (if ... ...) (fmakunbound ...)) (packages-load-package-dumped-lisps late-package-load-path))()
# (unwind-protect ...)
call-with-condition-handler(really-early-error-handler (lambda nil (setq load-path (split-path ...)) (external-debugging-output (format "\nUsing load-path %s" load-path)) (buffer-disable-undo (get-buffer "*scratch*")) (load "very-early-lisp" nil t) (let (...) (setq load-path ...)) (setq load-warn-when-source-newer t load-warn-when-source-only t) (defun pureload (file) (let ... ...)) (load (concat default-directory "../lisp/dumped-lisp.el")) (let (... file) (while ... ... ...) (if ... ...) (fmakunbound ...)) (packages-load-package-dumped-lisps late-package-load-path)))
# bind (current-load-list)
# (unwind-protect ...)
# bind (load-file-name)
# (unwind-protect ...)
# (unwind-protect ...)
# (unwind-protect ...)
# (unwind-protect ...)
# (unwind-protect ...)
load("loadup.el")
# bind (purify-flag load-ignore-elc-files)
(let ((load-ignore-elc-files t) (purify-flag nil)) (load "loadup.el"))
)
# bind (current-load-list)
# (unwind-protect ...)
# bind (load-file-name)
# (unwind-protect ...)
# (unwind-protect ...)
# (unwind-protect ...)
# (unwind-protect ...)
# (unwind-protect ...)
load("loadup-el.el")
(progn (setq command-line-args (append ... update-elc-files-to-compile)) (load "loadup-el.el"))
)
(if update-elc-files-to-compile (progn (setq command-line-args ...) (load "loadup-el.el")) (condition-case nil (delete-file "./NOBYTECOMPILE") (file-error nil)))
)
# bind (current-load-list)
# (unwind-protect ...)
# bind (load-file-name)
# (unwind-protect ...)
# (unwind-protect ...)
# (unwind-protect ...)
# (unwind-protect ...)
# (unwind-protect ...)
load("/usr/local/src/CVS/gtk-xemacs/src/../lisp/update-elc.el")
# (condition-case ... . error)
# (catch top-level ...)
*** Killing XEmacs
gmake[1]: *** [update-elc.stamp] Error 255
gmake[1]: Leaving directory `/usr/local/src/CVS/gtk-xemacs/src'
gmake: *** [src] Error 2
--
Robin S. Socha
Big. Black. Beautified. <http://socha.net/XEmacs/>
- 2 participants
- 1 comments
building with make -j4
25 years
Gunnar Evermann
Gunnar Evermann
I just tried building 21.2.35 with 'make -j4', i.e. 4 parallel jobs.
It seems to have worked ok (I got a working binary), but during the
build I saw the following:
gcc -c -g -Wall -Wno-switch -Demacs -I. -DHAVE_CONFIG_H -I/usr/X11R6/include -DDEFINE_TOP_LEVEL_EMACS_SHELL /data/scratch/ge204/src/XEmacs/xemacs/src/EmacsShell-sub.c
mv EmacsShell-sub.o TopLevelEmacsShell.o
mv EmacsShell-sub.o TopLevelEmacsShell.o
mv: `EmacsShell-sub.o' and `TopLevelEmacsShell.o' are the same file
make[1]: *** [TopLevelEmacsShell.o] Error 1
gcc -c -g -Wall -Wno-switch -Demacs -I. -DHAVE_CONFIG_H -I/usr/X11R6/include -DDEFINE_TRANSIENT_EMACS_SHELL /data/scratch/ge204/src/XEmacs/xemacs/src/EmacsShell-sub.c
make[1]: Leaving directory `/data/scratch/ge204/src/XEmacs/build-2907-parthenope-j4/src'
make: *** [dump-elcs] Error 2
mv EmacsShell-sub.o TransientEmacsShell.o
gcc -c -g -Wall -Wno-switch -Demacs -I. -DHAVE_CONFIG_H -I/usr/X11R6/include /data/scratch/ge204/src/XEmacs/xemacs/src/EmacsManager.c
I don't quite understand this, but maybe sombody could tweak the
makefiles some to avoid this?
Gunnar
uname -a: Linux parthenope 2.2.14-1smp #1 SMP Mon Feb 21 11:25:50 GMT 2000 i686 unknown
../xemacs/configure '--with-pop' '--prefix=/tools/emacs' '--with-toolbars=no' '--with-dialogs=no' '--cflags=-g -Wall -Wno-switch'
XEmacs 21.2-b35 "Nike" configured for `i686-pc-linux'.
Compilation / Installation:
Source code location: /data/scratch/ge204/src/XEmacs/xemacs
Installation prefix: /tools/emacs
Operating system description file: `s/linux.h'
Machine description file: `m/intel386.h'
Compiler: gcc -g -Wall -Wno-switch
Relocating allocator for buffers: no
GNU version of malloc: yes
- Using Doug Lea's new malloc from the GNU C Library.
Window System:
Compiling in support for the X window system:
- X Windows headers location: /usr/X11R6/include
- X Windows libraries location: /usr/X11R6/lib
- Handling WM_COMMAND properly.
Compiling in support for the Athena widget set:
- Athena headers location: X11/Xaw
- Athena library to link: Xaw
Using Lucid menubars.
Using Lucid scrollbars.
Using Athena native widgets.
TTY:
Compiling in support for ncurses.
Compiling in support for GPM (General Purpose Mouse).
Images:
Compiling in support for GIF images (builtin).
Compiling in support for XPM images.
Compiling in support for PNG images.
Compiling in support for JPEG images.
Compiling in support for TIFF images.
Compiling in support for X-Face message headers.
Sound:
Compiling in support for sound (native).
Databases:
Compiling in support for Berkeley database.
Internationalization:
Mail:
Compiling in support for POP mail retrieval.
Compiling in support for "dot-locking" mail spool file locking method.
Other Features:
Compiling in support for dynamic shared object modules.
Compiling in support for extra debugging code.
WARNING: ---------------------------------------------------------
WARNING: Compiling in support for runtime error checking.
WARNING: XEmacs will run noticeably more slowly as a result.
WARNING: Error checking is on by default for XEmacs beta releases.
WARNING: ---------------------------------------------------------
- 4 participants
- 3 comments
Re: [jgd@lanl.gov] I think the xemacs FTP package directory is fouled up...
25 years
Adrian Aichner
Adrian Aichner
>>>>> "Ben" == Ben Wing <ben(a)666.com> writes:
Ben> have we completely solved this problem now?
Yes, I think so.
I have successfully tested the fix in native NT XEmacs 21.1.11 and
21.2-b35.
I have commited the change to lisp/files.el for release-21-2.
Vin plans to release 21.1.12 soon.
Best regards,
Adrian
Ben> Adrian Aichner wrote:
>>
>> This problem is still alive and kicking.
>>
>> Unfortunately I don't have time to look into it.
>>
>> Adrian
>>
>> --------------------------------------------------------------------------------
>>
>> Subject: I think the xemacs FTP package directory is fouled up...
>> Date: Wed, 26 Jul 2000 10:43:05 -0600
>> From: jgd(a)lanl.gov
>> Organization: CIC-5 Advanced Networking Team, Los Alamos National Laboratory
>> To: webmaster(a)xemacs.org
>>
>> I installed 21.1.11 and the sumo tarball, and when I do a list and
>> install on the manage packages menu, I get:
>>
>> (1) (error/warning) Error in process filter: (ftp-error FTP Error: CWD
>> 550 /pub/xemacs/packages/package-index.LATEST.pgp/: Not a directory
>> failed: )
>>
>> It's also possible that the tarball is fouled up, but I don't think so.
>>
>> The package tool is looking for a different file type than the
>> package-index.LATEST.pgp is.
Ben> --
Ben> Ben
Ben> In order to save my hands, I am cutting back on my mail. I also write
Ben> as succinctly as possible -- please don't be offended. If you send me
Ben> mail, you _will_ get a response, but please be patient, especially for
Ben> XEmacs-related mail. If you need an immediate response and it is not
Ben> apparent in your message, please say so. Thanks for your understanding.
Ben> See also http://www.666.com/ben/chronic-pain/
- 1 participants
- 0 comments
[jgd@lanl.gov] I think the xemacs FTP package directory is fouled up...
25 years
Adrian Aichner
Adrian Aichner
This problem is still alive and kicking.
Unfortunately I don't have time to look into it.
Adrian
- 2 participants
- 1 comments
Re: [xemacsweb PATCH] Update list of review board members
25 years
Adrian Aichner
Adrian Aichner
>>>>> "APA" == Adrian Aichner <aichner(a)ecf.teradyne.com> writes:
APA> BTW: What order is the list in? Must be a very sophisticated
APA> algorithm.
I have regained consciousness and will sort the list by firstname.
Best regards,
Adrian
APA> Has Kyle formally resigned from the board?
APA> Should his name be taken out?
APA> Could we bribe him to come back?
APA> Best regards,
APA> Adrian
- 1 participants
- 0 comments
Re: Security patches to 21.1.10
25 years
Adrian Aichner
Adrian Aichner
>>>>> "Vin" == Vin Shelton <acs(a)xemacs.org> writes:
Vin> Does anyone have any comments on this? As there is currently a
Vin> serious bug in retrieving the package list courtesy of
Vin> insert-file-contents-literally, I am planning on getting 21.1.12 out
Vin> the door quickly. I'm inclined to include the non-package changes.
Wonderful. I can't comment on the security patches, though.
Adrian
Vin> Please advise.
Vin> Thanks,
Vin> vin
--
Adrian Aichner <adrian(a)xemacs.org>
- 1 participants
- 0 comments
Re: Security patches to 21.1.10
25 years
Vin Shelton
Vin Shelton
Does anyone have any comments on this? As there is currently a
serious bug in retrieving the package list courtesy of
insert-file-contents-literally, I am planning on getting 21.1.12 out
the door quickly. I'm inclined to include the non-package changes.
Please advise.
Thanks,
vin
>>>>> On Tue, 25 Jul 2000 14:56:25 +0200, Torsten Duwe <duwe(a)caldera.de> said:
Torsten> Hi,
Torsten> don't know if this stuff is still applicable to the current CVS tree
Torsten> but nevertheless here are the security patches we ship, for your convenience.
Torsten> Credits for the original Unix 98 PTY patch obviously go to Florian Weimer
Torsten> <Florian.Weimer(a)RUS.Uni-Stuttgart.DE>, porting of that FSF-Emacs-patch to
Torsten> XEmacs-21 was done by my colleague Olaf Kirch. The ugly but working TMPDIR
Torsten> and passwd-nonrecording patches were hacked by me. Seems like these two are
Torsten> broken by design (e.g. neither getpass nor mkstemp lisp primitives). Anyway
Torsten> -- here they are:
Torsten> diff -ur xemacs-21.1.orig/configure.in xemacs-21.1/configure.in
Torsten> --- xemacs-21.1.orig/configure.in Tue Apr 25 12:56:09 2000
Torsten> +++ xemacs-21.1/configure.in Tue Apr 25 16:11:54 2000
Torsten> ＠＠ -3105,6 +3105,11 ＠＠
Torsten> AC_CHECK_FUNCS(cbrt closedir dup2 eaccess fmod fpathconf frexp ftime gethostname getpagesize gettimeofday getcwd getwd logb lrand48 matherr mkdir mktime perror poll random rename res_init rint rmdir select setitimer setpgid setlocale setsid sigblock sighold sigprocmask snprintf strcasecmp strerror tzset ulimit usleep utimes waitpid vsnprintf)
Torsten> +# Check for UNIX98 PTYs.
Torsten> +# getpt is a glibc addition which emulates the master device on
Torsten> +# systems without kernel support.
Torsten> +AC_CHECK_FUNCS(grantpt unlockpt getpt ptsname)
Torsten> +
Torsten> dnl realpath is buggy on linux, decosf and aix4
Torsten> dnl The realpath() in linux libc (4.6.27) sometimes fails with ELOOP.
Torsten> diff -ur xemacs-21.1.orig/src/config.h.in xemacs-21.1/src/config.h.in
Torsten> --- xemacs-21.1.orig/src/config.h.in Tue Oct 28 12:36:08 1997
Torsten> +++ xemacs-21.1/src/config.h.in Tue Apr 25 16:11:54 2000
Torsten> ＠＠ -297,6 +297,14 ＠＠
Torsten> #undef HAVE_UTIMES
Torsten> #undef HAVE_WAITPID
Torsten> #undef HAVE_VSNPRINTF
Torsten> +/* UNIX98 PTY support functions
Torsten> + getpt is a glibc addition which emulates the master device on
Torsten> + systems without kernel support. */
Torsten> +#undef HAVE_GRANTPT
Torsten> +#undef HAVE_UNLOCKPT
Torsten> +#undef HAVE_GETPT
Torsten> +#undef HAVE_PTSNAME
Torsten> +
Torsten> #undef HAVE_SOCKETS
Torsten> #undef HAVE_SOCKADDR_SUN_LEN
Torsten> #undef HAVE_MULTICAST
Torsten> diff -ur xemacs-21.1.orig/src/s/gnu.h xemacs-21.1/src/s/gnu.h
Torsten> --- xemacs-21.1.orig/src/s/gnu.h Sun Jul 20 00:12:29 1997
Torsten> +++ xemacs-21.1/src/s/gnu.h Tue Apr 25 16:11:54 2000
Torsten> ＠＠ -50,3 +50,49 ＠＠
Torsten> #ifndef NOT_C_CODE
Torsten> #include <fcntl.h>
Torsten> #endif
Torsten> +
Torsten> +#if defined(HAVE_GRANTPT) && defined(HAVE_UNLOCKPT) && defined(HAVE_PTSNAME)
Torsten> +/* UNIX98 PTYs are available.
Torsten> + Added by Florian Weimer <Florian.Weimer(a)RUS.Uni-Stuttgart.DE>,
Torsten> + RUS-CERT, University of Stuttgart. Based on Emacs code for DGUX. */
Torsten> +
Torsten> +#define PTY_ITERATION for (i = 0; i < 1; i++)
Torsten> +/* no iteration at all */
Torsten> +
Torsten> +/* Use getpt() if it's available, because it provides Unix98 PTY
Torsten> + emulation for kernels which doesn't support it natively. */
Torsten> +
Torsten> +#ifdef HAVE_GETPT
Torsten> +#define PTY_OPEN \
Torsten> + do { \
Torsten> + fd = getpt(); \
Torsten> + if (fcntl (fd, F_SETFL, O_NDELAY) == -1) \
Torsten> + fatal ("could not set master PTY to non-block mode"); \
Torsten> + } while (0)
Torsten> +
Torsten> +#else
Torsten> +/* the master PTY device */
Torsten> +#define PTY_NAME_SPRINTF strcpy (pty_name, "/dev/ptmx");
Torsten> +#endif
Torsten> +
Torsten> +/* This sets the name of the slave side of the PTY. grantpt(3) and
Torsten> + unlockpt(3) may fork a subprocess, so keep sigchld_handler() from
Torsten> + intercepting that death. */
Torsten> +
Torsten> +#define PTY_TTY_NAME_SPRINTF \
Torsten> + { \
Torsten> + char *ptsname(), *ptyname; \
Torsten> + \
Torsten> + sigblock(sigmask(SIGCHLD)); \
Torsten> + if (grantpt(fd) == -1) \
Torsten> + fatal("could not grant slave pty"); \
Torsten> + if (unlockpt(fd) == -1) \
Torsten> + fatal("could not unlock slave pty"); \
Torsten> + if (!(ptyname = ptsname(fd))) \
Torsten> + fatal ("could not enable slave pty"); \
Torsten> + strncpy(pty_name, ptyname, sizeof(pty_name)); \
Torsten> + pty_name[sizeof(pty_name) - 1] = 0; \
Torsten> + sigsetmask(siggetmask() & ~sigmask(SIGCHLD)); \
Torsten> + }
Torsten> +
Torsten> +#endif
Torsten> diff -ur xemacs-21.1.orig/src/s/hpux.h xemacs-21.1/src/s/hpux.h
Torsten> --- xemacs-21.1.orig/src/s/hpux.h Mon Jul 14 00:43:10 1997
Torsten> +++ xemacs-21.1/src/s/hpux.h Tue Apr 25 16:11:54 2000
Torsten> ＠＠ -157,6 +157,59 ＠＠
Torsten> /* This is needed for HPUX version 6.2; it may not be needed for 6.2.1. */
Torsten> #define SHORT_CAST_BUG
Torsten> +#if defined(HAVE_GRANTPT) && defined(HAVE_UNLOCKPT) && defined(HAVE_PTSNAME)
Torsten> +/* UNIX98 PTYs are available.
Torsten> + Added by Florian Weimer <Florian.Weimer(a)RUS.Uni-Stuttgart.DE>,
Torsten> + RUS-CERT, University of Stuttgart. Based on Emacs code for DGUX. */
Torsten> +
Torsten> +#ifdef emacs
Torsten> +#include <grp.h>
Torsten> +#include <sys/stropts.h>
Torsten> +#endif
Torsten> +
Torsten> +#define PTY_ITERATION for (i = 0; i < 1; i++)
Torsten> +/* no iteration at all */
Torsten> +
Torsten> +/* the master PTY device */
Torsten> +#define PTY_NAME_SPRINTF strcpy (pty_name, "/dev/ptmx");
Torsten> +
Torsten> +/* This sets the name of the slave side of the PTY. grantpt(3) and
Torsten> + unlockpt(3) may fork a subprocess, so keep sigchld_handler() from
Torsten> + intercepting that death. grantpt() behavior on HP-UX differs from
Torsten> + what's specified in the man page: the group of the slave PTY is set
Torsten> + to the user's primary group, and we fix that. */
Torsten> +
Torsten> +#define PTY_TTY_NAME_SPRINTF \
Torsten> + { \
Torsten> + char *ptsname(), *ptyname; \
Torsten> + struct group *getgrnam (), *tty_group = getgrnam ("tty"); \
Torsten> + if (tty_group == NULL) \
Torsten> + fatal ("group tty not found"); \
Torsten> + \
Torsten> + sigblock(sigmask(SIGCHLD)); \
Torsten> + if (grantpt(fd) == -1) \
Torsten> + fatal("could not grant slave pty"); \
Torsten> + if (!(ptyname = ptsname(fd))) \
Torsten> + fatal ("could not enable slave pty"); \
Torsten> + strncpy(pty_name, ptyname, sizeof(pty_name)); \
Torsten> + pty_name[sizeof(pty_name) - 1] = 0; \
Torsten> + if (chown (pty_name, (uid_t) -1, tty_group->gr_gid) == -1) \
Torsten> + fatal ("could not chown slave pty"); \
Torsten> + if (unlockpt(fd) == -1) \
Torsten> + fatal("could not unlock slave pty"); \
Torsten> + sigunblock(sigmask(SIGCHLD)); \
Torsten> + }
Torsten> +
Torsten> +/* Push various streams modules onto a PTY channel. */
Torsten> +
Torsten> +#define SETUP_SLAVE_PTY \
Torsten> + if (ioctl (xforkin, I_PUSH, "ptem") == -1) \
Torsten> + fatal ("ioctl I_PUSH ptem", errno); \
Torsten> + if (ioctl (xforkin, I_PUSH, "ldterm") == -1) \
Torsten> + fatal ("ioctl I_PUSH ldterm", errno);
Torsten> +
Torsten> +#else /* no UNIX98 PTYs */
Torsten> +
Torsten> /* This is how to get the device name of the tty end of a pty. */
Torsten> #define PTY_TTY_NAME_SPRINTF \
Torsten> sprintf (pty_name, "/dev/pty/tty%c%x", c, i);
Torsten> ＠＠ -164,6 +217,8 ＠＠
Torsten> /* This is how to get the device name of the control end of a pty. */
Torsten> #define PTY_NAME_SPRINTF \
Torsten> sprintf (pty_name, "/dev/ptym/pty%c%x", c, i);
Torsten> +
Torsten> +#endif /* UNIX 98 PTYs */
Torsten> /* This triggers a conditional in xfaces.c. */
Torsten> #define XOS_NEEDS_TIME_H
Torsten> diff -ur xemacs-21.1.orig/src/s/linux.h xemacs-21.1/src/s/linux.h
Torsten> --- xemacs-21.1.orig/src/s/linux.h Tue Apr 25 12:56:09 2000
Torsten> +++ xemacs-21.1/src/s/linux.h Tue Apr 25 16:12:54 2000
Torsten> ＠＠ -194,3 +194,49 ＠＠
Torsten> /* XEmacs: removed setpgrp() definition because we use setpgid() when
Torsten> it's available, and autodetect it. */
Torsten> +
Torsten> +#if defined(HAVE_GRANTPT) && defined(HAVE_UNLOCKPT) && defined(HAVE_PTSNAME)
Torsten> +/* UNIX98 PTYs are available.
Torsten> + Added by Florian Weimer <Florian.Weimer(a)RUS.Uni-Stuttgart.DE>,
Torsten> + RUS-CERT, University of Stuttgart. Based on Emacs code for DGUX. */
Torsten> +
Torsten> +#define PTY_ITERATION for (i = 0; i < 1; i++)
Torsten> +/* no iteration at all */
Torsten> +
Torsten> +/* Use getpt() if it's available, because it provides Unix98 PTY
Torsten> + emulation for kernels which doesn't support it natively. */
Torsten> +
Torsten> +#ifdef HAVE_GETPT
Torsten> +#define PTY_OPEN \
Torsten> + do { \
Torsten> + fd = getpt(); \
Torsten> + if (fcntl (fd, F_SETFL, O_NDELAY) == -1) \
Torsten> + fatal ("could not set master PTY to non-block mode"); \
Torsten> + } while (0)
Torsten> +
Torsten> +#else
Torsten> +/* the master PTY device */
Torsten> +#define PTY_NAME_SPRINTF strcpy (pty_name, "/dev/ptmx");
Torsten> +#endif
Torsten> +
Torsten> +/* This sets the name of the slave side of the PTY. grantpt(3) and
Torsten> + unlockpt(3) may fork a subprocess, so keep sigchld_handler() from
Torsten> + intercepting that death. */
Torsten> +
Torsten> +#define PTY_TTY_NAME_SPRINTF \
Torsten> + { \
Torsten> + char *ptsname(), *ptyname; \
Torsten> + \
Torsten> + sigblock(sigmask(SIGCHLD)); \
Torsten> + if (grantpt(fd) == -1) \
Torsten> + fatal("could not grant slave pty"); \
Torsten> + if (unlockpt(fd) == -1) \
Torsten> + fatal("could not unlock slave pty"); \
Torsten> + if (!(ptyname = ptsname(fd))) \
Torsten> + fatal ("could not enable slave pty"); \
Torsten> + strncpy(pty_name, ptyname, sizeof(pty_name)); \
Torsten> + pty_name[sizeof(pty_name) - 1] = 0; \
Torsten> + sigsetmask(siggetmask() & ~sigmask(SIGCHLD)); \
Torsten> + }
Torsten> +
Torsten> +#endif
Torsten> --- xemacs-21.1.10/src/editfns.c~ Wed Jul 7 05:10:33 1999
Torsten> +++ xemacs-21.1.10/src/editfns.c Wed May 10 12:05:10 2000
Torsten> ＠＠ -602,6 +602,12 ＠＠
Torsten> return make_char (BUF_FETCH_CHAR (b, n));
Torsten> }
Torsten> +#if !defined(WINDOWSNT) && !defined(MSDOS)
Torsten> +#include <sys/stat.h>
Torsten> +#include <fcntl.h>
Torsten> +#include <errno.h>
Torsten> +#include <limits.h>
Torsten> +#endif
Torsten>
Torsten> DEFUN ("temp-directory", Ftemp_directory, 0, 0, 0, /*
Torsten> Return the pathname to the directory to use for temporary files.
Torsten> ＠＠ -621,7 +627,47 ＠＠
Torsten> #else /* WINDOWSNT || MSDOS */
Torsten> tmpdir = getenv ("TMPDIR");
Torsten> if (!tmpdir)
Torsten> + {
Torsten> + struct stat st;
Torsten> + char * logname = user_login_name(NULL);
Torsten> + int myuid = getuid();
Torsten> + static char path[1+_POSIX_PATH_MAX];
Torsten> +
Torsten> + strcpy(path, "/tmp/"); strncat(path, logname, _POSIX_PATH_MAX);
Torsten> + if (lstat(path, &st) < 0 && errno == ENOENT)
Torsten> + {
Torsten> + mkdir(path, 0700); /* ignore retval -- checked next anyway. */
Torsten> + }
Torsten> + if (lstat(path, &st) == 0 && st.st_uid == myuid && S_ISDIR(st.st_mode))
Torsten> + {
Torsten> + tmpdir = path;
Torsten> + }
Torsten> + else
Torsten> + {
Torsten> + strcpy(path, getenv("HOME")); strncat(path, "/tmp/", _POSIX_PATH_MAX);
Torsten> + if (stat(path, &st) < 0 && errno == ENOENT)
Torsten> + {
Torsten> + int fd;
Torsten> + char warnpath[1+_POSIX_PATH_MAX];
Torsten> + mkdir(path, 0700); /* ignore retvals */
Torsten> + strcpy(warnpath, path);
Torsten> + strncat(warnpath, ".created_by_xemacs", _POSIX_PATH_MAX);
Torsten> + if ((fd = open(warnpath, O_WRONLY|O_CREAT, 0644)) > 0)
Torsten> + {
Torsten> + write(fd, "XEmacs created this directory because /tmp/<yourname> was unavailable -- \nPlease check !\n", 89);
Torsten> + close(fd);
Torsten> + }
Torsten> + }
Torsten> + if (stat(path, &st) == 0 && S_ISDIR(st.st_mode))
Torsten> + {
Torsten> + tmpdir = path;
Torsten> + }
Torsten> + else
Torsten> + {
Torsten> tmpdir = "/tmp";
Torsten> + }
Torsten> + }
Torsten> + }
Torsten> #endif
Torsten> return build_ext_string (tmpdir, FORMAT_FILENAME);
Torsten> --- xemacs-packages/lisp/xemacs-base/comint.el~ Wed Jan 19 10:36:46 2000
Torsten> +++ xemacs-packages/lisp/xemacs-base/comint.el Tue May 9 14:18:35 2000
Torsten> ＠＠ -1670,6 +1670,7 ＠＠
Torsten> (echo-keystrokes 0)
Torsten> (cursor-in-echo-area t)
Torsten> (message-log-max nil) ;turn of logging in GNU Emacs
Torsten> + (inhibit-input-event-recording t) ; and XEmacs
Torsten> (done nil))
Torsten> (while (not done)
Torsten> (if stars
Torsten> --- xemacs-packages/lisp/xemacs-base/passwd.el~ Thu Jan 20 11:29:37 2000
Torsten> +++ xemacs-packages/lisp/xemacs-base/passwd.el Tue May 9 15:47:38 2000
Torsten> ＠＠ -156,12 +156,7 ＠＠
Torsten> - do not pass it as an argument to a shell command - anyone will be
Torsten> able to see it if they run `ps' at the right time.
Torsten> -Note that the password will be temporarily recoverable with the `view-lossage'
Torsten> -command. This data will not be overwritten until another hundred or so
Torsten> -characters are typed. You can temporarily disable recording key strokes
Torsten> -by binding `inhibit-input-event-recording' to t."
Torsten> -
Torsten> -
Torsten> +"
Torsten> (save-excursion
Torsten> (let ((input (get-buffer-create " *password*"))
Torsten> (passwd-history-posn 0)
Torsten> ＠＠ -213,6 +207,7 ＠＠
Torsten> (let* ((minibuffer-completion-table nil)
Torsten> (cursor-in-echo-area t)
Torsten> (echo-keystrokes 0)
Torsten> + (inhibit-input-event-recording t)
Torsten> (key (passwd-read-key-sequence
Torsten> (concat (if (listp prompt)
Torsten> (car (nth passwd-history-posn passwd-history))
- 1 participants
- 0 comments
Re: [PATCH] pcl-cvs to work with "xemacs -nw"
25 years
Adrian Aichner
Adrian Aichner
>>>>> "Andreas" == Andreas Jaeger <aj(a)suse.de> writes:
Hi Andreas, hello Michael!
Michael is the pcl-cvs maintainer for XEmacs.
pcl-cvs(a)cyclic.com are the maintainers of pcl-cvs.
I sent my patch there also.
I have no problem with keeping this patch in my private workspace if
you don't want it in xemacs-packages.
I had to do this before for the dired package where my patch to make
! runs the command dired-do-shell-command
work on system-type 'windows-nt has not been applied for a year now.
Best regards,
Adrian
>>>>> Adrian Aichner writes:
APA> Greetings, All!
APA> I have a little patch for pcl-cvs to make it work with
APA> xemacs -nw
APA> pcl-cvs-version
APA> "v2_9_9"
APA> Andreas (Jaeger), I would like to commit this to xemacs-packages
APA> shortly, unless there are objections.
Andreas> Isn't pcl-cvs maintained externally? I don't have at the moment
Andreas> access to the sources - please check this yourself and only apply the
Andreas> patch if *we* (xemacs team) maintain pcl-cvs. Otherwise send it to
Andreas> the maintainers - and then integrate the maintainers' version.
Andreas> Andreas
Andreas> --
Andreas> Andreas Jaeger
Andreas> SuSE Labs aj(a)suse.de
Andreas> private aj(a)arthur.inka.de
- 1 participants
- 0 comments
Dump latest CVS, using -nw
25 years
Norbert Koch
Norbert Koch
Hi!
I see a couple of core dumps, when running xemacs (-nw) in an xterm.
This one happened while opening an article in gnus:
Generating summary...doneFatal error: assertion failed, file glyphs.c,
line 3810, !UNBOUNDP (image_instance)
Only C-Trace which reads
#0 0x1846b3d0 in kill () from /usr/lib/libc.so.3
#1 0x80b38b3 in fatal_error_signal ()
#2 0xbfbfdfdc in ?? ()
#3 0x80b58aa in assert_failed ()
#4 0x812953c in glyph_image_instance ()
#5 0x812b419 in update_glyph_cachel_data ()
#6 0x812b551 in add_glyph_cachel ()
#7 0x812b625 in get_glyph_cachel_index ()
#8 0x816c959 in add_glyph_rune ()
#9 0x816d07b in add_glyph_runes ()
#10 0x816d967 in create_text_block ()
#11 0x816b41b in generate_display_line ()
#12 0x817278d in regenerate_window ()
#13 0x8173aec in redisplay_window ()
#14 0x8174320 in redisplay_windows ()
#15 0x8173765 in redisplay_window ()
#16 0x8174320 in redisplay_windows ()
#17 0x8174656 in redisplay_frame ()
#18 0x81747e0 in redisplay_device ()
#19 0x8174b79 in redisplay_without_hooks ()
#20 0x8174e22 in redisplay ()
#21 0x80ebaab in Fnext_event ()
#22 0x809d5f9 in Fcommand_loop_1 ()
#23 0x809d3be in command_loop_1 ()
#24 0x80b8c9a in condition_case_1 ()
#25 0x809cae3 in command_loop_3 ()
#26 0x809cb07 in command_loop_2 ()
#27 0x80b885f in internal_catch ()
#28 0x809cdd6 in initial_command_loop ()
#29 0x80b4941 in xemacs_21_2_b35_i386_unknown_freebsd3_5 ()
#30 0x80b505f in main ()
#31 0x8083b2d in _start ()
I've already cut out displaying x-faces, but it seems every glyph
triggers this feature, eg I encounter the same behaviour, when trying
to open a buffer that has outline-mouse-minor-mode activated.
norbert
--
..Are we having FUN yet...?
- 1 participants
- 0 comments