Re: Static pattern usage
I used this trick with rather complex define/eval constructs It really works but sometimes gives me 'virtual memory exhaused' If you (Sam) are interested, I can give you precise example. If only I figured out how to work out this (endless loop?) my build system would be perfect =) Paul D. Smith wrote: %% Sam Ravnborg <[EMAIL PROTECTED]> writes: sr> Is there any other way to get the name of the target sr> to be used in the prerequisite list? sr> Other make implementation expands $* to the name of sr> the target when listed in the prerequisites, but not gnu make. No make expands $* in the prerequisites list to the name of the target. Some makes support a feature where $$@ (note two $'s!) in the prerequisites list will expand to the name of the target. GNU make also supports this syntax (see the GNU make manual). However, this won't help you because again, the $$@ is not expanded until _after_ all the other variables are expanded. This is true even for other versions of make which support this syntax. The only way to do what you want is to use the $(eval ...) function to declare extra dependencies. Something like: OBJS = foo.o deps_foo.o := foo.h $(foreach target,$(OBJS),$(eval $(target): $$(deps_$(target (note, this is untested). -- Best Regards, Peter A. Kerzum ___ Bug-make mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/bug-make
make 3.80 subscript error if environ has small string
GNU make 3.80 assumes that every string in the environment has at least
MAKELEVEL_LENGTH bytes in it, and it executes a subscript error otherwise.
Normally this is harmless, but
Here is a patch.
2003-10-05 Paul Eggert <[EMAIL PROTECTED]>
* main.c (main):
Avoid potential subscript error if environ has short strings.
===
RCS file: RCS/main.c,v
retrieving revision 3.80
retrieving revision 3.80.0.1
diff -pu -r3.80 -r3.80.0.1
--- main.c 2002/08/10 01:27:17 3.80
+++ main.c 2003/10/06 05:46:41 3.80.0.1
@@ -1845,8 +1845,8 @@ int main (int argc, char ** argv)
#ifndef _AMIGA
for (p = environ; *p != 0; ++p)
- if ((*p)[MAKELEVEL_LENGTH] == '='
- && strneq (*p, MAKELEVEL_NAME, MAKELEVEL_LENGTH))
+ if (strneq (*p, MAKELEVEL_NAME, MAKELEVEL_LENGTH)
+ && (*p)[MAKELEVEL_LENGTH] == '=')
{
/* The SGI compiler apparently can't understand
the concept of storing the result of a function
___
Bug-make mailing list
[EMAIL PROTECTED]
http://mail.gnu.org/mailman/listinfo/bug-make
virtual memory exhausted
Excuse me for another silly question, but can you tell when could this error occure: virtual memory exhausted -- Best Regards, Peter A. Kerzum ___ Bug-make mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/bug-make
make 3.80 sprintf potential buffer overruns
I audited GNU make 3.80 for potential buffer overruns due to sprintf
calls, and came up with the following patch. Most of the buffer
overruns are fairly theoretical: they can occur only on hosts where
integers are fairly wide. Some of them are more practical, though:
they can occur on hosts that have 5-digit file descriptor numbers.
2003-10-06 Paul Eggert <[EMAIL PROTECTED]>
* arscan.c (ar_member_touch): Don't overrun buffer if a time stamp
is so large that it doesn't fit into ar_hdr.ar_date.
Report overflow instead.
* function.c (func_words):
Don't assume int can print in less than 20 bytes.
(func_shell): Don't assume that line number can print in 11
bytes or less.
(func_call): Don't assume that int can print in 10 bytes or less.
* main.c (main): Don't assume that file descriptors can be
printed as integers in 4 bytes or less. Don't assume that
make level can be printed in less than about 30 bytes.
(define_makeflags): Don't assume that unsigned can be printed
in less than 30 bytes. Don't assume that double %g prints in
less than 100 bytes.
* make.h (EOVERFLOW): Define if errno.h doesn't.
(INT_STRLEN_BOUND, INT_BUFSIZE_BOUND): New macros, taken from gnulib.
* misc.c (strerror): Don't refer to errno; that is bogus.
Don't translate "Unknown error %d", as the translation can
overflow the static buffer. Don't assume that ints can be
printed in 20 bytes or less.
* signame.c (strsignal):
Don't assume that ints can be printed in 20 bytes or less.
* variable.c (define_automatic_variables): Don't assume that
version number + "Customs" fits in 200 bytes. Don't assume
that int can be printed in less than 200 bytes.
(target_environment): Don't assume that double %g prints in less than
100 bytes.
===
RCS file: arscan.c,v
retrieving revision 3.80
retrieving revision 3.80.0.1
diff -pu -r3.80 -r3.80.0.1
--- arscan.c2001/06/01 03:56:50 3.80
+++ arscan.c2003/10/06 06:53:34 3.80.0.1
@@ -777,6 +777,7 @@ ar_member_touch (arname, memname)
struct ar_hdr ar_hdr;
register int i;
struct stat statbuf;
+ char mtime_buf[INT_BUFSIZE_BOUND (statbuf.st_mtime)];
if (pos < 0)
return (int) pos;
@@ -803,7 +804,16 @@ ar_member_touch (arname, memname)
/* Advance member's time to that time */
for (i = 0; i < sizeof ar_hdr.ar_date; i++)
ar_hdr.ar_date[i] = ' ';
- sprintf (ar_hdr.ar_date, "%ld", (long int) statbuf.st_mtime);
+ if (statbuf.st_mtime < 0)
+sprintf (mtime_buf, "%ld", (long int) statbuf.st_mtime);
+ else
+sprintf (mtime_buf, "%lu", (unsigned long int) statbuf.st_mtime);
+ if (sizeof ar_hdr.ar_date <= strlen (mtime_buf))
+{
+ errno = EOVERFLOW;
+ goto lose;
+}
+ strcpy (ar_hdr.ar_date, mtime_buf);
#ifdef AIAMAG
ar_hdr.ar_date[strlen(ar_hdr.ar_date)] = ' ';
#endif
===
RCS file: function.c,v
retrieving revision 3.80
retrieving revision 3.80.0.1
diff -pu -r3.80 -r3.80.0.1
--- function.c 2002/10/04 02:13:42 3.80
+++ function.c 2003/10/06 06:53:34 3.80.0.1
@@ -704,7 +704,7 @@ func_words (o, argv, funcname)
{
int i = 0;
char *word_iterator = argv[0];
- char buf[20];
+ char buf[INT_BUFSIZE_BOUND (i)];
while (find_next_token (&word_iterator, (unsigned int *) 0) != 0)
++i;
@@ -1531,7 +1531,9 @@ func_shell (o, argv, funcname)
/* For error messages. */
if (reading_file != 0)
{
- error_prefix = (char *) alloca (strlen (reading_file->filenm)+11+4);
+ error_prefix = (char *) alloca (strlen (reading_file->filenm) + 1
+ + INT_STRLEN_BOUND (reading_file->lineno)
+ + 3);
sprintf (error_prefix,
"%s:%lu: ", reading_file->filenm, reading_file->lineno);
}
@@ -2045,7 +2047,7 @@ func_call (o, argv, funcname)
for (i=0; *argv; ++i, ++argv)
{
- char num[11];
+ char num[INT_BUFSIZE_BOUND (i)];
sprintf (num, "%d", i);
define_variable (num, strlen (num), *argv, o_automatic, 0);
===
RCS file: main.c,v
retrieving revision 3.80.0.1
retrieving revision 3.80.0.2
diff -pu -r3.80.0.1 -r3.80.0.2
--- main.c 2003/10/06 05:46:41 3.80.0.1
+++ main.c 2003/10/06 07:02:01 3.80.0.2
@@ -1573,7 +1573,8 @@ int main (int argc, char ** argv)
jobserver_fds = (struct stringlist *)
xmalloc (sizeof (struct stringlist));
jobserver_fds->list = (char **) xmalloc (sizeof (char *));
- jobserver_fds->list[0] = xmalloc ((sizeof ("1024")*2)+1);
+ jobserver_fds->list[0] = xmalloc (INT_STRLEN_BOUND (job_fds[0]) + 1
+
Re: Static pattern usage
%% "Peter A. Kerzum" <[EMAIL PROTECTED]> writes: pak> I used this trick with rather complex define/eval constructs It pak> really works but sometimes gives me 'virtual memory exhaused' If pak> you (Sam) are interested, I can give you precise example. If pak> only I figured out how to work out this (endless loop?) my build pak> system would be perfect =) Apply the patch attached to bug #1517 at http://savannah.gnu.org/projects/make -- --- Paul D. Smith <[EMAIL PROTECTED]> Find some GNU make tips at: http://www.gnu.org http://make.paulandlesley.org "Please remain calm...I may be mad, but I am a professional." --Mad Scientist ___ Bug-make mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/bug-make
Re: virtual memory exhausted
On Monday 06 October 2003 11:49, Peter A. Kerzum wrote: > Excuse me for another silly question, but can you tell when could this > error occure: > > virtual memory exhausted When calling malloc(0) on some systems. See http://savannah.gnu.org/bugs/?func=detailbug&bug_id=2888&group_id=71 for details. ___ Bug-make mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/bug-make
Re: Static pattern usage
On Mon, Oct 06, 2003 at 01:53:48PM +0400, Peter A. Kerzum wrote: > I used this trick with rather complex define/eval constructs > It really works but sometimes gives me 'virtual memory exhaused' > If you (Sam) are interested, I can give you precise example. I would be glad to see that - if not for anything else then just to learn one more way to utilise make. TIA, Sam ___ Bug-make mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/bug-make
Unable to build win32 GNU make-3.80
Hello, Just tried to build win32 version of make-3.80. I had some problems with that though: 1. The build_w32.bat has a typo, + before the first if. This should not be there. 2. In the "NMakefile" $(MAKE) is used, which is in my environment to be set to gmake.exe, perhaps this NMakefile should use something like: NMAKE=nmake.exe Then use $(NMAKE) to avoid other make conflicts in the NMakefile in the archive. 3. This version of GNU make would not build, using MSVC6, or MSVC .Net C compiler, the batch file, or the "nmake NMakefile" methods. Attached is a log, does anyone have any ideas what could be the problem? I did this in cmd.exe, setting the appropriate environment variables before trying, using vcvars32.bat etc. Best regards J. Grant C:\Documents and Settings\jgrant\Desktop\make-3.80>set make=gnumake C:\Documents and Settings\jgrant\Desktop\make-3.80>if not exist config.h copy config.h.W32 config.h C:\Documents and Settings\jgrant\Desktop\make-3.80>cd w32\subproc C:\Documents and Settings\jgrant\Desktop\make-3.80\w32\subproc>echo "Creating the subproc library" "Creating the subproc library" C:\Documents and Settings\jgrant\Desktop\make-3.80\w32\subproc>C:\WINNT\system32\cmd.exe /c build.bat C:\Documents and Settings\jgrant\Desktop\make-3.80\w32\subproc>if not exist .\WinDebug\nul mkdir .\WinDebug C:\Documents and Settings\jgrant\Desktop\make-3.80\w32\subproc>cl.exe /nologo /MT /W3 /GX /Z7 /YX /Od /I .. /I . /I ../include /D WIN32 /D WINDOWS32 /D _DEBUG /D _WINDOWS /FR.\WinDebug/ /Fp.\WinDebug/subproc.pch /Fo.\WinDebug/ /c misc.c misc.c C:\Documents and Settings\jgrant\Desktop\make-3.80\w32\subproc>cl.exe /nologo /MT /W3 /GX /Z7 /YX /Od /I .. /I . /I ../include /I ../.. /D WIN32 /D WINDOWS32 /D _DEBUG /D _WINDOWS /FR.\WinDebug/ /Fp.\WinDebug/subproc.pch /Fo.\WinDebug/ /c sub_proc.c sub_proc.c C:\Documents and Settings\jgrant\Desktop\make-3.80\w32\subproc>cl.exe /nologo /MT /W3 /GX /Z7 /YX /Od /I .. /I . /I ../include /D WIN32 /D WINDOWS32 /D _DEBUG /D _WINDOWS /FR.\WinDebug/ /Fp.\WinDebug/subproc.pch /Fo.\WinDebug/ /c w32err.c w32err.c C:\Documents and Settings\jgrant\Desktop\make-3.80\w32\subproc>lib.exe /NOLOGO /OUT:.\WinDebug\subproc.lib .\WinDebug/misc.obj .\WinDebug/sub_proc.obj .\WinDebug/w32err.obj C:\Documents and Settings\jgrant\Desktop\make-3.80\w32\subproc>if not exist .\WinRel\nul mkdir .\WinRel C:\Documents and Settings\jgrant\Desktop\make-3.80\w32\subproc>cl.exe /nologo /MT /W3 /GX /YX /O2 /I ../include /D WIN32 /D WINDOWS32 /D NDEBUG /D _WINDOWS /FR.\WinRel/ /Fp.\WinRel/subproc.pch /Fo.\WinRel/ /c misc.c misc.c C:\Documents and Settings\jgrant\Desktop\make-3.80\w32\subproc>cl.exe /nologo /MT /W3 /GX /YX /O2 /I ../include /I ../.. /D WIN32 /D WINDOWS32 /D NDEBUG /D _WINDOWS /FR.\WinRel/ /Fp.\WinRel/subproc.pch /Fo.\WinRel/ /c sub_proc.c sub_proc.c C:\Documents and Settings\jgrant\Desktop\make-3.80\w32\subproc>cl.exe /nologo /MT /W3 /GX /YX /O2 /I ../include /D WIN32 /D WINDOWS32 /D NDEBUG /D _WINDOWS /FR.\WinRel/ /Fp.\WinRel/subproc.pch /Fo.\WinRel/ /c w32err.c w32err.c C:\Documents and Settings\jgrant\Desktop\make-3.80\w32\subproc>lib.exe /NOLOGO /OUT:.\WinRel\subproc.lib .\WinRel/misc.obj .\WinRel/sub_proc.obj .\WinRel/w32err.obj C:\Documents and Settings\jgrant\Desktop\make-3.80\w32\subproc>cd ..\.. C:\Documents and Settings\jgrant\Desktop\make-3.80>del link.dbg link.rel C:\Documents and Settings\jgrant\Desktop\make-3.80>del config.h C:\Documents and Settings\jgrant\Desktop\make-3.80>copy config.h.W32 config.h 1 file(s) copied. C:\Documents and Settings\jgrant\Desktop\make-3.80>echo off "Creating GNU make for Windows 95/NT" C:\Documents and Settings\jgrant\Desktop\make-3.80>if not exist .\WinDebug\nul mkdir .\WinDebug C:\Documents and Settings\jgrant\Desktop\make-3.80>cl.exe /nologo /MT /W3 /GX /Zi /YX /Od /I . /I glob /I w32/include /D TIVOLI /D _DEBUG /D WINDOWS32 /D WIN32 /D _CONSOLE /D HAVE_CONFIG_H /FR.\WinDebug/ /Fp.\WinDebug/gnumake.pch /Fo.\WinDebug/ /Fd.\WinDebug/gnumake.pdb /c variable.c variable.c variable.c(85) : warning C4113: 'unsigned long (__cdecl *)()' differs in parameter lists from 'unsigned long (__cdecl *)(const void *)' variable.c(85) : warning C4113: 'unsigned long (__cdecl *)()' differs in parameter lists from 'unsigned long (__cdecl *)(const void *)' variable.c(85) : warning C4113: 'int (__cdecl *)()' differs in parameter lists from 'int (__cdecl *)(const void *,const void *)' variable.c(399) : warning C4113: 'unsigned long (__cdecl *)()' differs in parameter lists from 'unsigned long (__cdecl *)(const void *)' variable.c(399) : warning C4113: 'unsigned long (__cdecl *)()' differs in parameter lists from 'unsigned long (__cdecl *)(const void *)' variable.c(399) : warning C4113: 'int (__cdecl *)()' differs in parameter lists from 'int (__cdecl *)(const void *,const void *)' variable.c(470) : warning C4113: 'void (__cdecl *)()' differs in parameter lists from '
Re: Unable to build win32 GNU make-3.80
%% [EMAIL PROTECTED] writes: j> 3. This version of GNU make would not build, using MSVC6, or MSVC j>.Net C compiler, the batch file, or the "nmake NMakefile" methods. See bug # 1687 on Savannah (http://savannah.gnu.org/projects/make) -- --- Paul D. Smith <[EMAIL PROTECTED]> Find some GNU make tips at: http://www.gnu.org http://make.paulandlesley.org "Please remain calm...I may be mad, but I am a professional." --Mad Scientist ___ Bug-make mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/bug-make
RE: Unable to build win32 GNU make-3.80
Hello, Thanks for the pointer [http://savannah.gnu.org/bugs/?func=detailbug&bug_id=1687&group_id=71]. Is there any plan to release a fixed version of 3.80? It seems like it is quite a while it has not been working. Is there a release Errata for GNU make? I think this would be useful if there is not one already, I could not find one. Regarding the warnings during compile, are they all still in CVS version? if they are, then I can create a patch that resolves some of these if you would like (I will test on several platforms etc). Please let me know if you would like me to do this. Best regards J. Grant > -Original Message- > From: Paul D. Smith [mailto:[EMAIL PROTECTED] > Sent: 06 October 2003 17:08 > To: [EMAIL PROTECTED] > Cc: [EMAIL PROTECTED]; [EMAIL PROTECTED] > Subject: Re: Unable to build win32 GNU make-3.80 > > > %% [EMAIL PROTECTED] writes: > > j> 3. This version of GNU make would not build, using MSVC6, or MSVC > j>.Net C compiler, the batch file, or the "nmake > NMakefile" methods. > > See bug # 1687 on Savannah (http://savannah.gnu.org/projects/make) > > -- > -- > - > Paul D. Smith <[EMAIL PROTECTED]> Find some GNU make tips at: > http://www.gnu.org http://make.paulandlesley.org > "Please remain calm...I may be mad, but I am a > professional." --Mad Scientist > ___ Bug-make mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/bug-make
