* PATCH zsh-3.1.5-pws-7: cygwin make fixes
@ 1999-02-08 6:49 Matt Armstrong
1999-02-08 7:26 ` Bart Schaefer
0 siblings, 1 reply; 7+ messages in thread
From: Matt Armstrong @ 1999-02-08 6:49 UTC (permalink / raw)
To: zsh-workers
[-- Attachment #1: Type: text/plain, Size: 2013 bytes --]
There are two problems with the make system in zsh-3.1.5-pws-7 (and
surely previous versions) when running under cygwin.
1) The makefiles assume they are creating "zsh" and not "zsh.exe." Not
a problem unless you do a "make install" under cygwin.
2) The Src/signames.awk gets confused by cygwin's sys/signals.h since it
defines signal names for multiple platforms (many of which #define all
the signal names to 1).
I fixed #1 by adding AC_EXEEXT to configure.in and using the
corresponding $(EXEEXT) variable where needed in the other makefiles.
WARNING: I imagine you need a fairly recent autoconf to get the
AC_EXEEXT (I tested with version 2.13). You can get it at
http://sourceware.cygnus.com/autoconf/.
I fixed #2 by re-doing the way signames.c is created. Instead of a AWK
script that gets run on a signal.h file, a C program is compiled and
run. This gets it right even if the signal.h file is hard to parse.
My fix for #1 may be controversial because it may break for people who
need to keep using an older autoconf. Is upgrading a problem?
My fix for #2, I think, is a good thing. ;-) I'd do the same thing for
the RLIMIT stuff if cygwin had RLIMIT stuff.
The patch for both is attached (they intertwine).
Please let me know if sending patches as MIME attachments is an
inconvenience (though, pasting a patch file into this little hotmail
edit window may ruin it).
The patch file is generated with cvs rdiff. If you apply it with
'patch', do it as "patch -p1 -E < zsh-patch.txt" in the root of your zsh
tree. "-p1" (p-one, not "el") will strip off the leading "zsh" path
component that cvs rdiff likes to add. "-E" will delete
Src/signames.awk instead of leaving it as a zero length file.
REMEMBER: Src/mksignames.c is a new file, Src/signames.awk gets deleted.
Testing: You might check that "kill -l" prints the same thing before and
after this patch.
______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com
[-- Attachment #2: zsh-patch.txt --]
[-- Type: text/plain, Size: 20568 bytes --]
Index: zsh/configure.in
diff -c zsh/configure.in:1.1.1.3 zsh/configure.in:1.3
*** zsh/configure.in:1.1.1.3 Sat Jan 30 10:14:56 1999
--- zsh/configure.in Tue Feb 02 23:11:27 1999
***************
*** 287,292 ****
--- 287,294 ----
fi
AC_SUBST(U)
+ AC_EXEEXT dnl Determine the executable suffix.
+
AC_FUNC_ALLOCA dnl Check how to get `alloca'.
dnl If the compiler supports union initialisation
***************
*** 661,684 ****
fi
AC_DEFINE_UNQUOTED($signals_style)
AC_MSG_RESULT($signals_style)
-
- dnl Where is <signal.h> located? Needed as input for signals.awk
- AC_CACHE_CHECK(where signal.h is located, zsh_cv_path_signal_h,
- [for SIGNAL_H in /usr/include/bsd/sys/signal.h dnl Next
- /usr/include/asm/signum.h dnl alpha-Linux
- /usr/include/asm/signal.h dnl Linux 1.3.0 and above
- /usr/include/linux/signal.h dnl Linux up to 1.2.11
- /usr/include/sys/signal.h dnl Almost everybody else
- /dev/null; dnl Just in case we fall through
- do
- test -f $SIGNAL_H && \
- grep '#[ ]*define[ ][ ]*SIG[0-9A-Z]*[ ]*[0-9][0-9]*' $SIGNAL_H > /dev/null && \
- break
- done
- zsh_cv_path_signal_h=$SIGNAL_H
- ])
- SIGNAL_H=$zsh_cv_path_signal_h
- AC_SUBST(SIGNAL_H)dnl
dnl -----------------------------------------------------
dnl Look for the file containing the RLIMIT_* definitions
--- 663,668 ----
Index: zsh/Config/defs.mk
diff -c zsh/Config/defs.mk:1.1.1.2 zsh/Config/defs.mk:1.2
*** zsh/Config/defs.mk:1.1.1.2 Sat Jan 23 20:48:28 1999
--- zsh/Config/defs.mk Tue Feb 02 23:11:27 1999
***************
*** 58,63 ****
--- 58,64 ----
DLLD = @DLLD@
EXPOPT = @EXPOPT@
IMPOPT = @IMPOPT@
+ EXEEXT = @EXEEXT@
# utilities
AWK = @AWK@
Index: zsh/Src/.cvsignore
diff -c zsh/Src/.cvsignore:1.1.1.1 zsh/Src/.cvsignore:1.2
*** zsh/Src/.cvsignore:1.1.1.1 Sat Jan 23 18:10:11 1999
--- zsh/Src/.cvsignore Sat Jan 30 14:06:22 1999
***************
*** 23,25 ****
--- 23,26 ----
zshxmods.h
bltinmods.list
tags TAGS
+ mksignames mksignames.exe
\ No newline at end of file
Index: zsh/Src/Makefile.in
diff -c zsh/Src/Makefile.in:1.1.1.2 zsh/Src/Makefile.in:1.3
*** zsh/Src/Makefile.in:1.1.1.2 Sat Jan 23 20:48:38 1999
--- zsh/Src/Makefile.in Tue Feb 02 23:17:23 1999
***************
*** 41,47 ****
all: bin modules
! bin: zsh
modules: headers
--- 41,47 ----
all: bin modules
! bin: zsh$(EXEEXT)
modules: headers
***************
*** 63,69 ****
LDRUNPATH = LD_RUN_PATH=$(libdir)/zsh
NDRUNPATH =
! zsh: $(@L@IBZSH) $(@L@STMP) $(MAIN_OBJS)
rm -f $@
$(@L@DRUNPATH) $(LINK) $(MAIN_OBJS) $(@L@LIST) $(ZSH_@E@XPORT) $(@L@IBZSH) $(LIBS)
--- 63,69 ----
LDRUNPATH = LD_RUN_PATH=$(libdir)/zsh
NDRUNPATH =
! zsh$(EXEEXT): $(@L@IBZSH) $(@L@STMP) $(MAIN_OBJS)
rm -f $@
$(@L@DRUNPATH) $(LINK) $(MAIN_OBJS) $(@L@LIST) $(ZSH_@E@XPORT) $(@L@IBZSH) $(LIBS)
***************
*** 147,162 ****
uninstall.bin: uninstall.bin-here
# install binary, creating install directory if necessary
! install.bin-here: zsh install.bin-@L@
$(sdir_top)/mkinstalldirs $(bindir)
! $(INSTALL_PROGRAM) $(STRIPFLAGS) zsh $(bindir)/zsh-$(VERSION)
! if test -f $(bindir)/zsh; then \
! rm -f $(bindir)/zsh.old; \
! ln $(bindir)/zsh $(bindir)/zsh.old; \
else :; fi
! rm -f $(bindir)/zsh.new
! ln $(bindir)/zsh-$(VERSION) $(bindir)/zsh.new
! mv $(bindir)/zsh.new $(bindir)/zsh
install.bin-N:
install.bin-L: $(LIBZSH)
--- 147,162 ----
uninstall.bin: uninstall.bin-here
# install binary, creating install directory if necessary
! install.bin-here: zsh$(EXEEXT) install.bin-@L@
$(sdir_top)/mkinstalldirs $(bindir)
! $(INSTALL_PROGRAM) $(STRIPFLAGS) zsh$(EXEEXT) $(bindir)/zsh-$(VERSION)$(EXEEXT)
! if test -f $(bindir)/zsh$(EXEEXT); then \
! rm -f $(bindir)/zsh.old$(EXEEXT); \
! ln $(bindir)/zsh$(EXEEXT) $(bindir)/zsh.old$(EXEEXT); \
else :; fi
! rm -f $(bindir)/zsh.new$(EXEEXT)
! ln $(bindir)/zsh-$(VERSION)$(EXEEXT) $(bindir)/zsh.new$(EXEEXT)
! mv $(bindir)/zsh.new$(EXEEXT) $(bindir)/zsh$(EXEEXT)
install.bin-N:
install.bin-L: $(LIBZSH)
***************
*** 165,171 ****
# uninstall binary
uninstall.bin-here: uninstall.bin-@L@
! rm -f $(bindir)/zsh-$(VERSION) $(bindir)/zsh
uninstall.bin-N:
uninstall.bin-L:
--- 165,171 ----
# uninstall binary
uninstall.bin-here: uninstall.bin-@L@
! rm -f $(bindir)/zsh-$(VERSION)$(EXEEXT) $(bindir)/zsh$(EXEEXT)
uninstall.bin-N:
uninstall.bin-L:
***************
*** 179,185 ****
rm -f stamp-modobjs stamp-modobjs.tmp
clean-here:
! rm -f modules.index.tmp modules.stamp zsh ansi2knr.o ansi2knr
rm -f libzsh-*.$(DL_EXT)
distclean-here:
--- 179,185 ----
rm -f stamp-modobjs stamp-modobjs.tmp
clean-here:
! rm -f modules.index.tmp modules.stamp zsh$(EXEEXT) ansi2knr.o ansi2knr
rm -f libzsh-*.$(DL_EXT)
distclean-here:
Index: zsh/Src/mksignames.c
diff -c nul zsh/Src/mksignames.c:1.2
*** nul Sun Feb 07 22:08:50 1999
--- zsh/Src/mksignames.c Sun Jan 31 21:57:56 1999
***************
*** 0 ****
--- 1,320 ----
+ /*
+ * mksignames.c - create signames.c
+ *
+ * This file is part of zsh, the Z shell.
+ *
+ * Copyright (c) 1999 Matt Armstrong
+ * All rights reserved
+ *
+ * Permission is hereby granted, without written agreement and without
+ * license or royalty fees, to use, copy, modify, and distribute this
+ * software and to distribute modified versions of this software for
+ * any purpose, provided that the above copyright notice and the
+ * following two paragraphs appear in all copies of this software.
+ *
+ * In no event shall Matt Armstrong or the Zsh Development Group be
+ * liable to any party for direct, indirect, special, incidental, or
+ * consequential damages arising out of the use of this software and
+ * its documentation, even if Matt Armstrong and the Zsh Development
+ * Group have been advised of the possibility of such damage.
+ *
+ * Matt Armstrong and the Zsh Development Group specifically disclaim
+ * any warranties, including, but not limited to, the implied
+ * warranties of merchantability and fitness for a particular purpose.
+ * The software provided hereunder is on an "as is" basis, and Matt
+ * Armstrong and the Zsh Development Group have no obligation to
+ * provide maintenance, support, updates, enhancements, or
+ * modifications.
+ *
+ */
+
+ /*
+ * This code replaces the previous signames.awk that tried to deduce
+ * the number and kind of signals available on the system by examining
+ * the header files. The idea is that the C version will get the
+ * answer right more often.
+ */
+ #include <stdio.h>
+ #include <signal.h>
+
+ #ifdef NSIG
+ #define MAX_SIGNAL NSIG
+ #else
+ #define MAX_SIGNAL 60
+ #endif
+
+ #define MAX(x,y) ((x)>(y)?(x):(y))
+
+ char *sigmsg[MAX_SIGNAL];
+ char *sigs[MAX_SIGNAL];
+ int max_signal;
+ #ifdef USE_SUSPENDED
+ int suspended = 1;
+ #else
+ int suspended = 0;
+ #endif
+
+ /* For each signal in the system, describe it in the sigmsg[] and sigs[]
+ arrays. Bump max_signal to the largest possible signal value. */
+ void
+ init(void)
+ {
+ int i, missing;
+
+ sigs[0] = "EXIT";
+ sigmsg[0] = "done";
+
+ #ifdef SIGALRM
+ sigs[SIGALRM] = "ALRM";
+ sigmsg[SIGALRM] = "alarm";
+ max_signal = MAX(max_signal, SIGALRM);
+ #endif
+ #ifdef SIGBUS
+ sigs[SIGBUS] = "BUS";
+ sigmsg[SIGBUS] = "bus error";
+ max_signal = MAX(max_signal, SIGBUS);
+ #endif
+ #ifdef SIGCHLD
+ sigs[SIGCHLD] = "CHLD";
+ sigmsg[SIGCHLD] = "death of child";
+ max_signal = MAX(max_signal, SIGCHLD);
+ #endif
+ #ifdef SIGCLD
+ sigs[SIGCLD] = "CLD";
+ sigmsg[SIGCLD] = "death of child";
+ max_signal = MAX(max_signal, SIGCLD);
+ #endif
+ #ifdef SIGCONT
+ sigs[SIGCONT] = "CONT";
+ sigmsg[SIGCONT] = "continued";
+ max_signal = MAX(max_signal, SIGCONT);
+ #endif
+ #ifdef SIGEMT
+ sigs[SIGEMT] = "EMT";
+ sigmsg[SIGEMT] = "EMT instruction";
+ max_signal = MAX(max_signal, SIGEMT);
+ #endif
+ #ifdef SIGFPE
+ sigs[SIGFPE] = "FPE";
+ sigmsg[SIGFPE] = "floating point exception";
+ max_signal = MAX(max_signal, SIGFPE);
+ #endif
+ #ifdef SIGHUP
+ sigs[SIGHUP] = "HUP";
+ sigmsg[SIGHUP] = "hangup";
+ max_signal = MAX(max_signal, SIGHUP);
+ #endif
+ #ifdef SIGILL
+ sigs[SIGILL] = "ILL";
+ sigmsg[SIGILL] = "illegal hardware instruction";
+ max_signal = MAX(max_signal, SIGILL);
+ #endif
+ #ifdef SIGINFO
+ sigs[SIGINFO] = "INFO";
+ sigmsg[SIGINFO] = "status request from keyboard";
+ max_signal = MAX(max_signal, SIGINFO);
+ #endif
+ #ifdef SIGINT
+ sigs[SIGINT] = "INT";
+ sigmsg[SIGINT] = "interrupt";
+ max_signal = MAX(max_signal, SIGINT);
+ #endif
+ #ifdef SIGIO
+ sigs[SIGIO] = "IO";
+ sigmsg[SIGIO] = "i/o ready";
+ max_signal = MAX(max_signal, SIGIO);
+ #endif
+ #ifdef SIGIOT
+ sigs[SIGIOT] = "IOT";
+ sigmsg[SIGIOT] = "IOT instruction";
+ max_signal = MAX(max_signal, SIGIOT);
+ #endif
+ #ifdef SIGABRT
+ sigs[SIGABRT] = "ABRT";
+ sigmsg[SIGABRT] = "abort";
+ max_signal = MAX(max_signal, SIGABRT);
+ #endif
+ #ifdef SIGKILL
+ sigs[SIGKILL] = "KILL";
+ sigmsg[SIGKILL] = "killed";
+ max_signal = MAX(max_signal, SIGKILL);
+ #endif
+ #ifdef SIGLOST
+ sigs[SIGLOST] = "LOST";
+ sigmsg[SIGLOST] = "resource lost";
+ max_signal = MAX(max_signal, SIGLOST);
+ #endif
+ #ifdef SIGPIPE
+ sigs[SIGPIPE] = "PIPE";
+ sigmsg[SIGPIPE] = "broken pipe";
+ max_signal = MAX(max_signal, SIGPIPE);
+ #endif
+ #ifdef SIGPOLL
+ sigs[SIGPOLL] = "POLL";
+ sigmsg[SIGPOLL] = "pollable event occurred";
+ max_signal = MAX(max_signal, SIGPOLL);
+ #endif
+ #ifdef SIGPROF
+ sigs[SIGPROF] = "PROF";
+ sigmsg[SIGPROF] = "profile signal";
+ max_signal = MAX(max_signal, SIGPROF);
+ #endif
+ #ifdef SIGPWR
+ sigs[SIGPWR] = "PWR";
+ sigmsg[SIGPWR] = "power fail";
+ max_signal = MAX(max_signal, SIGPWR);
+ #endif
+ #ifdef SIGQUIT
+ sigs[SIGQUIT] = "QUIT";
+ sigmsg[SIGQUIT] = "quit";
+ max_signal = MAX(max_signal, SIGQUIT);
+ #endif
+ #ifdef SIGSEGV
+ sigs[SIGSEGV] = "SEGV";
+ sigmsg[SIGSEGV] = "segmentation fault";
+ max_signal = MAX(max_signal, SIGSEGV);
+ #endif
+ #ifdef SIGSYS
+ sigs[SIGSYS] = "SYS";
+ sigmsg[SIGSYS] = "invalid system call";
+ max_signal = MAX(max_signal, SIGSYS);
+ #endif
+ #ifdef SIGTERM
+ sigs[SIGTERM] = "TERM";
+ sigmsg[SIGTERM] = "terminated";
+ max_signal = MAX(max_signal, SIGTERM);
+ #endif
+ #ifdef SIGTRAP
+ sigs[SIGTRAP] = "TRAP";
+ sigmsg[SIGTRAP] = "trace trap";
+ max_signal = MAX(max_signal, SIGTRAP);
+ #endif
+ #ifdef SIGURG
+ sigs[SIGURG] = "URG";
+ sigmsg[SIGURG] = "urgent condition";
+ max_signal = MAX(max_signal, SIGURG);
+ #endif
+ #ifdef SIGUSR1
+ sigs[SIGUSR1] = "USR1";
+ sigmsg[SIGUSR1] = "user-defined signal 1";
+ max_signal = MAX(max_signal, SIGUSR1);
+ #endif
+ #ifdef SIGUSR2
+ sigs[SIGUSR2] = "USR2";
+ sigmsg[SIGUSR2] = "user-defined signal 2";
+ max_signal = MAX(max_signal, SIGUSR2);
+ #endif
+ #ifdef SIGVTALRM
+ sigs[SIGVTALRM] = "VTALRM";
+ sigmsg[SIGVTALRM] = "virtual time alarm";
+ max_signal = MAX(max_signal, SIGVTALRM);
+ #endif
+ #ifdef SIGWINCH
+ sigs[SIGWINCH] = "WINCH";
+ sigmsg[SIGWINCH] = "window size changed";
+ max_signal = MAX(max_signal, SIGWINCH);
+ #endif
+ #ifdef SIGXCPU
+ sigs[SIGXCPU] = "XCPU";
+ sigmsg[SIGXCPU] = "cpu limit exceeded";
+ max_signal = MAX(max_signal, SIGXCPU);
+ #endif
+ #ifdef SIGXFSZ
+ sigs[SIGXFSZ] = "XFSZ";
+ sigmsg[SIGXFSZ] = "file size limit exceeded";
+ max_signal = MAX(max_signal, SIGXFSZ);
+ #endif
+ #ifdef SIGSTOP
+ sigs[SIGSTOP] = "STOP";
+ sigmsg[SIGSTOP] = suspended ? "suspended (signal)" : "stopped (signal)";
+ max_signal = MAX(max_signal, SIGSTOP);
+ #endif
+ #ifdef SIGTSTP
+ sigs[SIGTSTP] = "TSTP";
+ sigmsg[SIGTSTP] = "";
+ max_signal = MAX(max_signal, SIGTSTP);
+ #endif
+ #ifdef SIGTTIN
+ sigs[SIGTTIN] = "TTIN";
+ sigmsg[SIGTTIN] = " (tty input)";
+ max_signal = MAX(max_signal, SIGTTIN);
+ #endif
+ #ifdef SIGTTOU
+ sigs[SIGTTOU] = "TTOU";
+ sigmsg[SIGTTOU] = " (tty output)";
+ max_signal = MAX(max_signal, SIGTTOU);
+ #endif
+
+ /* Now make sure we have them all. NSIG is a POSIX thing, I
+ think. */
+ #ifdef NSIG
+ if (max_signal != NSIG - 1) {
+ printf("Found %d signals, while signal.h's NSIG #define is %d.\n",
+ max_signal, NSIG);
+ printf("Modify mksignames.c to include the missing signal(s).\n");
+ exit(1);
+ }
+ #endif
+ }
+
+ /* Print signames.c to stdout */
+ void
+ printit(void)
+ {
+ int i;
+
+ printf("/** signames.c **/\n");
+ printf("/** architecture-customized signames.c for zsh **/\n");
+ printf("\n");
+ printf("#define SIGCOUNT\t%d\n", max_signal);
+ printf("\n");
+ printf("#include \"zsh.mdh\"\n");
+ printf("\n");
+
+ printf("/**/\n");
+ printf("char *sigmsg[SIGCOUNT+2] = {\n");
+ for (i = 0; i <= max_signal; i++) {
+ if (!sigmsg[i]) {
+ printf("\t\"\",\n");
+ } else {
+ if (i == SIGSTOP
+ || i == SIGTSTP
+ || i == SIGTTIN
+ || i == SIGTTOU) {
+ printf("# ifdef USE_SUSPENDED\n\t\"suspended%s\",\n",
+ sigmsg[i]);
+ printf("# else\n\t\"stopped%s\",\n# endif\n",
+ sigmsg[i]);
+ } else {
+ printf("\t\"%s\",\n", sigmsg[i]);
+ }
+ }
+ }
+ puts("\tNULL");
+ puts("};");
+ puts("");
+
+ printf("/**/\n");
+ printf("char *sigs[SIGCOUNT+4] = {\n");
+ for (i = 0; i <= max_signal; i++) {
+ if (sigs[i] == "") {
+ printf("\t\"%d\",\n", i);
+ } else {
+ printf("\t\"%s\",\n", sigs[i]);
+ }
+ }
+ printf("\t\"ZERR\",\n");
+ printf("\t\"DEBUG\",\n");
+ puts("\tNULL");
+ puts("};");
+ }
+
+ int
+ main()
+ {
+ init();
+ printit();
+ exit(0);
+ }
+
+
Index: zsh/Src/signames.awk
diff -c zsh/Src/signames.awk:1.1.1.1 zsh/Src/signames.awk:removed
*** zsh/Src/signames.awk:1.1.1.1 Sat Jan 23 18:10:11 1999
--- zsh/Src/signames.awk Sun Feb 07 22:08:50 1999
***************
*** 1,98 ****
- #
- # {g,n}awk script to generate signames.c
- #
- # NB: On SunOS 4.1.3 - user-functions don't work properly, also \" problems
- # Without 0 + hacks some nawks compare numbers as strings
- #
- /^[\t ]*#[\t ]*define[\t _]*SIG[A-Z][A-Z0-9]*[\t ]*[1-9][0-9]*/ {
- sigindex = index($0, "SIG")
- sigtail = substr($0, sigindex, 80)
- split(sigtail, tmp)
- signam = substr(tmp[1], 4, 20)
- signum = tmp[2]
- if (sig[signum] == "") {
- sig[signum] = signam
- if (0 + max < 0 + signum && signum < 60)
- max = signum
- if (signam == "ABRT") { msg[signum] = "abort" }
- if (signam == "ALRM") { msg[signum] = "alarm" }
- if (signam == "BUS") { msg[signum] = "bus error" }
- if (signam == "CHLD") { msg[signum] = "death of child" }
- if (signam == "CLD") { msg[signum] = "death of child" }
- if (signam == "CONT") { msg[signum] = "continued" }
- if (signam == "EMT") { msg[signum] = "EMT instruction" }
- if (signam == "FPE") { msg[signum] = "floating point exception" }
- if (signam == "HUP") { msg[signum] = "hangup" }
- if (signam == "ILL") { msg[signum] = "illegal hardware instruction" }
- if (signam == "INFO") { msg[signum] = "status request from keyboard" }
- if (signam == "INT") { msg[signum] = "interrupt" }
- if (signam == "IO") { msg[signum] = "i/o ready" }
- if (signam == "IOT") { msg[signum] = "IOT instruction" }
- if (signam == "KILL") { msg[signum] = "killed" }
- if (signam == "LOST") { msg[signum] = "resource lost" }
- if (signam == "PIPE") { msg[signum] = "broken pipe" }
- if (signam == "POLL") { msg[signum] = "pollable event occurred" }
- if (signam == "PROF") { msg[signum] = "profile signal" }
- if (signam == "PWR") { msg[signum] = "power fail" }
- if (signam == "QUIT") { msg[signum] = "quit" }
- if (signam == "SEGV") { msg[signum] = "segmentation fault" }
- if (signam == "SYS") { msg[signum] = "invalid system call" }
- if (signam == "TERM") { msg[signum] = "terminated" }
- if (signam == "TRAP") { msg[signum] = "trace trap" }
- if (signam == "URG") { msg[signum] = "urgent condition" }
- if (signam == "USR1") { msg[signum] = "user-defined signal 1" }
- if (signam == "USR2") { msg[signum] = "user-defined signal 2" }
- if (signam == "VTALRM") { msg[signum] = "virtual time alarm" }
- if (signam == "WINCH") { msg[signum] = "window size changed" }
- if (signam == "XCPU") { msg[signum] = "cpu limit exceeded" }
- if (signam == "XFSZ") { msg[signum] = "file size limit exceeded" }
- }
- }
-
- END {
- ps = "%s"
- ifdstr = sprintf("# ifdef USE_SUSPENDED\n\t%csuspended%s%c,\n%s else\n\t%cstopped%s%c,\n# endif\n", 34, ps, 34, "#", 34, ps, 34)
-
- printf "/** signames.c **/\n"
- printf "/** architecture-customized signames.c for zsh **/\n"
- printf "\n"
- printf "#define SIGCOUNT\t%d\n", max
- printf "\n"
- printf "#include %czsh.mdh%c\n", 34, 34
- printf "\n"
- printf "/**/\n"
- printf "char *sigmsg[SIGCOUNT+2] = {\n"
- printf "\t%c%s%c,\n", 34, "done", 34
-
- for (i = 1; i <= 0 + max; i++)
- if (msg[i] == "") {
- if (sig[i] == "")
- printf("\t%c%c,\n", 34, 34)
- else if (sig[i] == "STOP")
- printf ifdstr, " (signal)", " (signal)"
- else if (sig[i] == "TSTP")
- printf ifdstr, "", ""
- else if (sig[i] == "TTIN")
- printf ifdstr, " (tty input)", " (tty input)"
- else if (sig[i] == "TTOU")
- printf ifdstr, " (tty output)", " (tty output)"
- else
- printf("\t%cSIG%s%c,\n", 34, sig[i], 34)
- } else
- printf("\t%c%s%c,\n", 34, msg[i], 34)
- print "\tNULL"
- print "};"
- print ""
- print "/**/"
- printf "char *sigs[SIGCOUNT+4] = {\n"
- printf("\t%cEXIT%c,\n", 34, 34)
- for (i = 1; i <= 0 + max; i++)
- if (sig[i] == "")
- printf("\t%c%d%c,\n", 34, i, 34)
- else
- printf("\t%c%s%c,\n", 34, sig[i], 34)
- printf("\t%cZERR%c,\n", 34, 34)
- printf("\t%cDEBUG%c,\n", 34, 34)
- print "\tNULL"
- print "};"
- }
--- 0 ----
Index: zsh/Src/zsh.mdd
diff -c zsh/Src/zsh.mdd:1.1.1.2 zsh/Src/zsh.mdd:1.4
*** zsh/Src/zsh.mdd:1.1.1.2 Sat Jan 23 20:48:42 1999
--- zsh/Src/zsh.mdd Sun Feb 07 22:08:28 1999
***************
*** 12,20 ****
prototypes.h hashtable.h ztype.h"
:<<\Make
! signames.c: signames.awk ../config.h @SIGNAL_H@
! $(AWK) -f $(sdir)/signames.awk @SIGNAL_H@ > $@
sigcount.h: signames.c
grep 'define.*SIGCOUNT' signames.c > $@
--- 12,27 ----
prototypes.h hashtable.h ztype.h"
:<<\Make
! mksignames.o: mksignames.c
! $(CC) -c $(CPPFLAGS) $(CFLAGS) $(sdir)/mksignames.c
+ mksignames$(EXEEXT): mksignames.o
+ $(CC) $(LDFLAGS) $(EXELDFLAGS) -o $@ ansi2knr.o
+
+ signames.c: ../config.h mksignames.c
+ $(CC) $(sdir)/mksignames.c -o mksignames$(EXEEXT)
+ ./mksignames$(EXEEXT) > $@
+
sigcount.h: signames.c
grep 'define.*SIGCOUNT' signames.c > $@
***************
*** 60,66 ****
clean-here: clean.zsh
clean.zsh:
! rm -f sigcount.h signames.c bltinmods.list version.h zshpaths.h zshxmods.h
# This is not properly part of this module, but it is built as if it were.
main.o: main.c zsh.mdh main.pro
--- 67,73 ----
clean-here: clean.zsh
clean.zsh:
! rm -f sigcount.h signames.c bltinmods.list version.h zshpaths.h zshxmods.h mksignames.o mksignames$(EXEEXT)
# This is not properly part of this module, but it is built as if it were.
main.o: main.c zsh.mdh main.pro
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: PATCH zsh-3.1.5-pws-7: cygwin make fixes
1999-02-08 6:49 PATCH zsh-3.1.5-pws-7: cygwin make fixes Matt Armstrong
@ 1999-02-08 7:26 ` Bart Schaefer
1999-02-08 9:55 ` Peter Stephenson
0 siblings, 1 reply; 7+ messages in thread
From: Bart Schaefer @ 1999-02-08 7:26 UTC (permalink / raw)
To: Matt Armstrong, zsh-workers
On Feb 7, 10:49pm, Matt Armstrong wrote:
} Subject: PATCH zsh-3.1.5-pws-7: cygwin make fixes
}
} I fixed #2 by re-doing the way signames.c is created. Instead of a AWK
} script that gets run on a signal.h file, a C program is compiled and
} run. This gets it right even if the signal.h file is hard to parse.
I haven't examined the patch in detail to check this, but did you make
sure that the compilation of this program is performed in such a way that
it executes correctly on the current hardware even when cross-compiling
for another architecture?
} My fix for #1 may be controversial because it may break for people who
} need to keep using an older autoconf. Is upgrading a problem?
It'd be better if it weren't necessary. Copy it into aczsh.m4 and give
it a different name, e.g zsh_EXEEXT ?
} My fix for #2, I think, is a good thing. ;-) I'd do the same thing for
} the RLIMIT stuff if cygwin had RLIMIT stuff.
It's not a good thing if it won't cross-compile. I didn't use to pay much
attention to this, but I've just been through this particular plane of hell
with procmail's build system (which even compiles and runs C programs to do
sed-like substition on its man page templates!) so I'm feeling sensitive.
--
Bart Schaefer Brass Lantern Enterprises
http://www.well.com/user/barts http://www.brasslantern.com
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: PATCH zsh-3.1.5-pws-7: cygwin make fixes
1999-02-08 7:26 ` Bart Schaefer
@ 1999-02-08 9:55 ` Peter Stephenson
1999-02-08 10:29 ` Peter Stephenson
0 siblings, 1 reply; 7+ messages in thread
From: Peter Stephenson @ 1999-02-08 9:55 UTC (permalink / raw)
To: zsh-workers
"Bart Schaefer" wrote:
> On Feb 7, 10:49pm, Matt Armstrong wrote:
> } Subject: PATCH zsh-3.1.5-pws-7: cygwin make fixes
> }
> } I fixed #2 by re-doing the way signames.c is created. Instead of a AWK
> } script that gets run on a signal.h file, a C program is compiled and
> } run. This gets it right even if the signal.h file is hard to parse.
>
> I haven't examined the patch in detail to check this, but did you make
> sure that the compilation of this program is performed in such a way that
> it executes correctly on the current hardware even when cross-compiling
> for another architecture?
The trouble is mksignames actually has to run on the target
architecture to generate the appropriate signal names. I don't know
if there's any way of doing that at all with autoconf. Any more
ideas? We could maybe make it conditional on not cross-compiling,
since it does solve the #include problem fairly neatly.
> } My fix for #1 may be controversial because it may break for people who
> } need to keep using an older autoconf. Is upgrading a problem?
>
> It'd be better if it weren't necessary. Copy it into aczsh.m4 and give
> it a different name, e.g zsh_EXEEXT ?
In principle this shouldn't be a problem, except that autoconf 2.13
doesn't seemed to have appeared on any mirror yet. Even then I would
hope $(ZSH_EXEEXT) would just be empty for anyone who needed to run a
previous autoconf (which is supposed just people adding new patches to
the configuration behaviour), which wouldn't be so bad. However, the
patch won't be in pws-8 because of the cross-compilation problem.
(Does anyone know this actually works at the moment?)
--
Peter Stephenson <pws@ibmth.df.unipi.it> Tel: +39 050 844536
WWW: http://www.ifh.de/~pws/
Dipartimento di Fisica, Via Buonarroti 2, 56127 Pisa, Italy
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: PATCH zsh-3.1.5-pws-7: cygwin make fixes
1999-02-08 9:55 ` Peter Stephenson
@ 1999-02-08 10:29 ` Peter Stephenson
1999-02-08 16:54 ` Bart Schaefer
0 siblings, 1 reply; 7+ messages in thread
From: Peter Stephenson @ 1999-02-08 10:29 UTC (permalink / raw)
To: zsh-workers
I wrote:
> Bart wrote:
> > I haven't examined the patch in detail to check this, but did you make
> > sure that the compilation of this program is performed in such a way that
> > it executes correctly on the current hardware even when cross-compiling
> > for another architecture?
>
> The trouble is mksignames actually has to run on the target
> architecture to generate the appropriate signal names. I don't know
> if there's any way of doing that at all with autoconf. Any more
> ideas? We could maybe make it conditional on not cross-compiling,
> since it does solve the #include problem fairly neatly.
As configure.in contains AC_TRY_RUN statements, it looks to me
unlikely that the zsh build system supports cross-compilation anyway
--- this is specifically warned about in the autoconf manual ---
in which case there's no objection to Matt's patch. Does anyone know
for a fact that zsh has ever been successfully cross-compiled?
FYI, the tests that currently have to be run, rather than just
compiled, are the following. All but the first five are to do with
dynamically linked libraries.
- signed to unsigned casting
- if tgetent accepts NULL
- if rlim_t is quad_t
- if rlim_t is unsigned
- if named FIFOs work
- for ELF binaries
- if dlsym() needs a leading underscore
- if static/shared linking is broken
- if name clashes in shared objects are ok
- whether cross-library links are ok
- for working RTLD_GLOBAL
- if libraries can find symbols in exec
- whether executables can be stripped
- if dynamically loaded libs can be stripped
--
Peter Stephenson <pws@ibmth.df.unipi.it> Tel: +39 050 844536
WWW: http://www.ifh.de/~pws/
Dipartimento di Fisica, Via Buonarroti 2, 56127 Pisa, Italy
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: PATCH zsh-3.1.5-pws-7: cygwin make fixes
1999-02-08 10:29 ` Peter Stephenson
@ 1999-02-08 16:54 ` Bart Schaefer
1999-02-08 17:40 ` PATCH: zsh-3.1.5-pws-7: another go at signames Peter Stephenson
0 siblings, 1 reply; 7+ messages in thread
From: Bart Schaefer @ 1999-02-08 16:54 UTC (permalink / raw)
To: Peter Stephenson, zsh-workers
On Feb 8, 11:29am, Peter Stephenson wrote:
} Subject: Re: PATCH zsh-3.1.5-pws-7: cygwin make fixes
}
} As configure.in contains AC_TRY_RUN statements, it looks to me
} unlikely that the zsh build system supports cross-compilation anyway
I thought that might be it, but:
} - signed to unsigned casting
} - if tgetent accepts NULL
} - if rlim_t is quad_t
} - if rlim_t is unsigned
These four are just testing compiler capability, and can be done without
using the cross-compilation options (though I don't know if they ARE done
that way); and as you point out all but
} - if named FIFOs work
can be turned off (or should be) by specifying no dynamic module loading.
Signals are pretty basic to the shell with or without dynamic modules.
Seems to me it should be possible to do this particular thing with just
the preprocessor, rather than needing to compile and execute a program.
use the awk script to generate a bunch of #defines for every possible
signal in the header, feed that through "cc -E ...." and then awk or sed
the output again to extract the stuff zsh needs.
--
Bart Schaefer Brass Lantern Enterprises
http://www.well.com/user/barts http://www.brasslantern.com
^ permalink raw reply [flat|nested] 7+ messages in thread
* PATCH: zsh-3.1.5-pws-7: another go at signames
1999-02-08 16:54 ` Bart Schaefer
@ 1999-02-08 17:40 ` Peter Stephenson
0 siblings, 0 replies; 7+ messages in thread
From: Peter Stephenson @ 1999-02-08 17:40 UTC (permalink / raw)
To: Zsh hackers list
"Bart Schaefer" wrote:
> Seems to me it should be possible to do this particular thing with just
> the preprocessor, rather than needing to compile and execute a program.
> use the awk script to generate a bunch of #defines for every possible
> signal in the header, feed that through "cc -E ...." and then awk or sed
> the output again to extract the stuff zsh needs.
This is a good idea. There was another, more drastic problem with
what Matt had: it required you to know all the possible signals, since
you never got to see the names if you didn't know them already.
That's pretty hairy, given the stuff that gets shoved into OSes --- it
would have needed major additions just for AIX here.
This is pretty much Bart's idea above. (It still needs the right
signal.h, but so does every other idea.) signames1.awk does a first
pass through the signal header, extracting everything in sight, and
dumping lines like
XXNAMES XXSIGHUP SIGHUP
to sigtmp.c. This gets preprocessed; it's up to cpp to turn the last
word, SIGHUP here, into a number, and these get dumped to sigtmp.out.
Then signames2.awk runs on that; it's an only slightly modified
version of signames.awk which matches only those lines where the
substitution of the signal number has taken place. I managed to get
an identical signames.c to the one before. Matt, does this fix your
problems?
You ought to be able to get it to work by running config.status in the
top directory, then removing Makemod and Makemod.in in Src and running
make there.
signames.awk and now signames2.awk seem deliberately to stop before
60. Is this known to be right? I have extra definitions for signals
60..63, separated from the rest. I take it that means somebody's
thought about this.
The only problem I know of over and above the existing ones is that
now $(CPP) has to work, which zsh didn't need before. But configure
should make a good job of finding that. I can't think of a good
reason why anything more than `#include <signal.h>' should be
necessary in sigtmp.c.
Note this patch doesn't delete the old signames.awk, which nonetheless
becomes redundant. I hope.
--- Config/defs.mk.sn2 Mon Feb 8 10:43:29 1999
+++ Config/defs.mk Mon Feb 8 18:12:08 1999
@@ -44,6 +44,7 @@
# compilation
CC = @CC@
+CPP = @CPP@
CPPFLAGS = @CPPFLAGS@
DEFS = @DEFS@
CFLAGS = @CFLAGS@
--- Src/zsh.mdd.sn2 Mon Feb 8 10:44:28 1999
+++ Src/zsh.mdd Mon Feb 8 18:14:39 1999
@@ -12,8 +12,11 @@
prototypes.h hashtable.h ztype.h"
:<<\Make
-signames.c: signames.awk ../config.h @SIGNAL_H@
- $(AWK) -f $(sdir)/signames.awk @SIGNAL_H@ > $@
+signames.c: signames1.awk signames2.awk ../config.h @SIGNAL_H@
+ $(AWK) -f $(sdir)/signames1.awk @SIGNAL_H@ >sigtmp.c
+ $(CPP) sigtmp.c >sigtmp.out
+ $(AWK) -f $(sdir)/signames2.awk sigtmp.out > $@
+ rm -f sigtmp.c sigtmp.out
sigcount.h: signames.c
grep 'define.*SIGCOUNT' signames.c > $@
--- /dev/null Mon Feb 8 18:16:14 1999
+++ Src/signames1.awk Mon Feb 8 18:07:02 1999
@@ -0,0 +1,9 @@
+BEGIN { printf "#include <signal.h>\n\n" }
+
+/^[\t ]*#[\t ]*define[\t _]*SIG[A-Z][A-Z0-9]*[\t ]*[1-9][0-9]*/ {
+ sigindex = index($0, "SIG")
+ sigtail = substr($0, sigindex, 80)
+ split(sigtail, tmp)
+ signam = substr(tmp[1], 4, 20)
+ printf("XXNAMES XXSIG%s SIG%s\n", signam, signam)
+}
--- /dev/null Mon Feb 8 18:16:14 1999
+++ Src/signames2.awk Mon Feb 8 18:07:52 1999
@@ -0,0 +1,100 @@
+#
+# {g,n}awk script to generate signames.c
+# This version relies on the previous output of the preprocessor
+# on sigtmp.c, sigtmp.out, which is in turn generated by signames1.awk.
+#
+# NB: On SunOS 4.1.3 - user-functions don't work properly, also \" problems
+# Without 0 + hacks some nawks compare numbers as strings
+#
+/^XXNAMES XXSIG[A-Z][A-Z0-9]* [1-9][0-9]*/ {
+ sigindex = index($0, "SIG")
+ sigtail = substr($0, sigindex, 80)
+ split(sigtail, tmp)
+ signam = substr(tmp[1], 4, 20)
+ signum = tmp[2]
+ if (sig[signum] == "") {
+ sig[signum] = signam
+ if (0 + max < 0 + signum && signum < 60)
+ max = signum
+ if (signam == "ABRT") { msg[signum] = "abort" }
+ if (signam == "ALRM") { msg[signum] = "alarm" }
+ if (signam == "BUS") { msg[signum] = "bus error" }
+ if (signam == "CHLD") { msg[signum] = "death of child" }
+ if (signam == "CLD") { msg[signum] = "death of child" }
+ if (signam == "CONT") { msg[signum] = "continued" }
+ if (signam == "EMT") { msg[signum] = "EMT instruction" }
+ if (signam == "FPE") { msg[signum] = "floating point exception" }
+ if (signam == "HUP") { msg[signum] = "hangup" }
+ if (signam == "ILL") { msg[signum] = "illegal hardware instruction" }
+ if (signam == "INFO") { msg[signum] = "status request from keyboard" }
+ if (signam == "INT") { msg[signum] = "interrupt" }
+ if (signam == "IO") { msg[signum] = "i/o ready" }
+ if (signam == "IOT") { msg[signum] = "IOT instruction" }
+ if (signam == "KILL") { msg[signum] = "killed" }
+ if (signam == "LOST") { msg[signum] = "resource lost" }
+ if (signam == "PIPE") { msg[signum] = "broken pipe" }
+ if (signam == "POLL") { msg[signum] = "pollable event occurred" }
+ if (signam == "PROF") { msg[signum] = "profile signal" }
+ if (signam == "PWR") { msg[signum] = "power fail" }
+ if (signam == "QUIT") { msg[signum] = "quit" }
+ if (signam == "SEGV") { msg[signum] = "segmentation fault" }
+ if (signam == "SYS") { msg[signum] = "invalid system call" }
+ if (signam == "TERM") { msg[signum] = "terminated" }
+ if (signam == "TRAP") { msg[signum] = "trace trap" }
+ if (signam == "URG") { msg[signum] = "urgent condition" }
+ if (signam == "USR1") { msg[signum] = "user-defined signal 1" }
+ if (signam == "USR2") { msg[signum] = "user-defined signal 2" }
+ if (signam == "VTALRM") { msg[signum] = "virtual time alarm" }
+ if (signam == "WINCH") { msg[signum] = "window size changed" }
+ if (signam == "XCPU") { msg[signum] = "cpu limit exceeded" }
+ if (signam == "XFSZ") { msg[signum] = "file size limit exceeded" }
+ }
+}
+
+END {
+ ps = "%s"
+ ifdstr = sprintf("# ifdef USE_SUSPENDED\n\t%csuspended%s%c,\n%s else\n\t%cstopped%s%c,\n# endif\n", 34, ps, 34, "#", 34, ps, 34)
+
+ printf "/** signames.c **/\n"
+ printf "/** architecture-customized signames.c for zsh **/\n"
+ printf "\n"
+ printf "#define SIGCOUNT\t%d\n", max
+ printf "\n"
+ printf "#include %czsh.mdh%c\n", 34, 34
+ printf "\n"
+ printf "/**/\n"
+ printf "char *sigmsg[SIGCOUNT+2] = {\n"
+ printf "\t%c%s%c,\n", 34, "done", 34
+
+ for (i = 1; i <= 0 + max; i++)
+ if (msg[i] == "") {
+ if (sig[i] == "")
+ printf("\t%c%c,\n", 34, 34)
+ else if (sig[i] == "STOP")
+ printf ifdstr, " (signal)", " (signal)"
+ else if (sig[i] == "TSTP")
+ printf ifdstr, "", ""
+ else if (sig[i] == "TTIN")
+ printf ifdstr, " (tty input)", " (tty input)"
+ else if (sig[i] == "TTOU")
+ printf ifdstr, " (tty output)", " (tty output)"
+ else
+ printf("\t%cSIG%s%c,\n", 34, sig[i], 34)
+ } else
+ printf("\t%c%s%c,\n", 34, msg[i], 34)
+ print "\tNULL"
+ print "};"
+ print ""
+ print "/**/"
+ printf "char *sigs[SIGCOUNT+4] = {\n"
+ printf("\t%cEXIT%c,\n", 34, 34)
+ for (i = 1; i <= 0 + max; i++)
+ if (sig[i] == "")
+ printf("\t%c%d%c,\n", 34, i, 34)
+ else
+ printf("\t%c%s%c,\n", 34, sig[i], 34)
+ printf("\t%cZERR%c,\n", 34, 34)
+ printf("\t%cDEBUG%c,\n", 34, 34)
+ print "\tNULL"
+ print "};"
+}
--
Peter Stephenson <pws@ibmth.df.unipi.it> Tel: +39 050 844536
WWW: http://www.ifh.de/~pws/
Dipartimento di Fisica, Via Buonarroti 2, 56127 Pisa, Italy
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: PATCH: zsh-3.1.5-pws-7: another go at signames
@ 1999-02-09 7:01 Matt Armstrong
0 siblings, 0 replies; 7+ messages in thread
From: Matt Armstrong @ 1999-02-09 7:01 UTC (permalink / raw)
To: zsh-workers
Peter Stephenson <pws@ibmth.df.unipi.it> writes:
> There was another, more drastic problem with what Matt had: it
> required you to know all the possible signals, since you never got
> to see the names if you didn't know them already. That's pretty
> hairy, given the stuff that gets shoved into OSes --- it would have
> needed major additions just for AIX here.
Yes. I considered this as part of porting zsh to a platform.
> This is pretty much Bart's idea above. (It still needs the right
> signal.h, but so does every other idea.)
Not my original scheme. It always gets the right one because it
relies on the compiler to find it, not configure.
The signal.h on a cygwin system can be pretty much anywhere. On my
machine it is at
/cygnus/cygwin-b20/H-i586-cygwin32/i586-cygwin32/include/sys/signal.h.
Also, if the point is to support cross compiling, I don't think the
fixed list of header locations in configure.in will cut it. It will
likely pick up the signal.h for the host system, not the target
(though, I have to admit I've never cross compiled from one unix to
another -- maybe the same set of headers is typically used?).
It took me a while to figure out that zsh's "kill -l" listed only ZERR
because configure wasn't finding my signal.h. (maybe I should
consider this part of porting zsh to a new platform! :-)
> I managed to get an identical signames.c to the one before. Matt,
> does this fix your problems?
Fortunately, cygwin lets me mount the path mentioned above to
/usr/include. This lets configure find the right signal.h.
And, what really counts, it creates a correct signames.c!
If we stick with the improved AWK based scheme, I'll add a Cygwin
section to Etc/MACHINES describing how people should create their
/usr/include and that they should do it before running configure.
______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~1999-02-09 7:02 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
1999-02-08 6:49 PATCH zsh-3.1.5-pws-7: cygwin make fixes Matt Armstrong
1999-02-08 7:26 ` Bart Schaefer
1999-02-08 9:55 ` Peter Stephenson
1999-02-08 10:29 ` Peter Stephenson
1999-02-08 16:54 ` Bart Schaefer
1999-02-08 17:40 ` PATCH: zsh-3.1.5-pws-7: another go at signames Peter Stephenson
1999-02-09 7:01 Matt Armstrong
Code repositories for project(s) associated with this public inbox
https://git.vuxu.org/mirror/zsh/
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).