[patch] Use MAXPATHLEN, fix sunos build problem

[patch] Use MAXPATHLEN, fix sunos build problem

[Andrew Cagney]

[-- Attachment #1: Type: text/plain, Size: 81 bytes --]

FYI,

I've checked in the attached to fix a build problem on SunOS 4.x.

	Andrew

[-- Attachment #2: diffs --]
[-- Type: text/plain, Size: 1007 bytes --]

2002-01-19  Andrew Cagney  <ac131313@redhat.com>

	* utils.c: Include <sys/param.h> for MAXPATHLEN.
	(gdb_realpath): Use MAXPATHLEN when PATH_MAX is not defined.

Index: utils.c
===================================================================
RCS file: /cvs/src/src/gdb/utils.c,v
retrieving revision 1.55
diff -p -r1.55 utils.c
*** utils.c	2002/01/17 23:33:38	1.55
--- utils.c	2002/01/20 00:41:53
***************
*** 54,59 ****
--- 54,61 ----
  
  #include "inferior.h" /* for signed_pointer_to_address */
  
+ #include <sys/param.h>		/* For MAXPATHLEN */
+ 
  #include <readline/readline.h>
  
  #ifdef USE_MMALLOC
*************** char *
*** 2538,2544 ****
--- 2540,2552 ----
  gdb_realpath (const char *filename)
  {
  #ifdef HAVE_REALPATH
+ #if defined (PATH_MAX)
    char buf[PATH_MAX];
+ #elif defined (MAXPATHLEN)
+   char buf[MAXPATHLEN];
+ #else
+ #error "Neither PATH_MAX nor MAXPATHLEN defined"
+ #endif
    char *rp = realpath (filename, buf);
    return xstrdup (rp ? rp : filename);
  #else

Re: [patch] Use MAXPATHLEN, fix sunos build problem

[Jason R Thorpe]

On Sat, Jan 19, 2002 at 07:43:35PM -0500, Andrew Cagney wrote:

 > 2002-01-19  Andrew Cagney  <ac131313@redhat.com>
 > 
 > 	* utils.c: Include <sys/param.h> for MAXPATHLEN.
 > 	(gdb_realpath): Use MAXPATHLEN when PATH_MAX is not defined.

Seems like you really want PATH_MAX to be an autoconf test, and if
not, then test for MAXPATHLEN in <sys/param.h> and define PATH_MAX
in terms of MAXPATHLEN.

Basically, <machine/param.h> is by no means a standardized header, and
pulling it in causes all sorts of namespace pollution, so you wanna
use it only if you have to.

-- 
        -- Jason R. Thorpe <thorpej@wasabisystems.com>

Re: [patch] Use MAXPATHLEN, fix sunos build problem

[Andrew Cagney]

> On Sat, Jan 19, 2002 at 07:43:35PM -0500, Andrew Cagney wrote:
> 
>  > 2002-01-19  Andrew Cagney  <ac131313@redhat.com>
>> > * utils.c: Include <sys/param.h> for MAXPATHLEN.
>  > (gdb_realpath): Use MAXPATHLEN when PATH_MAX is not defined.
> 
> Seems like you really want PATH_MAX to be an autoconf test, and if
> not, then test for MAXPATHLEN in <sys/param.h> and define PATH_MAX
> in terms of MAXPATHLEN.
> 
> Basically, <machine/param.h> is by no means a standardized header, and
> pulling it in causes all sorts of namespace pollution, so you wanna
> use it only if you have to.


Did you mean machine/param.h or sys/param.h?  sys/param.h is currently 
pulled in all over the place un-guarded (which is probably a bad thing).

Andrew

Re: [patch] Use MAXPATHLEN, fix sunos build problem

[Jason R Thorpe]

On Sat, Jan 19, 2002 at 09:56:43PM -0500, Andrew Cagney wrote:

 > Did you mean machine/param.h or sys/param.h?  sys/param.h is currently 
 > pulled in all over the place un-guarded (which is probably a bad thing).

Err, sorry, yes, I meant <sys/param.h>.

-- 
        -- Jason R. Thorpe <thorpej@wasabisystems.com>

Re: [patch] Use MAXPATHLEN, fix sunos build problem

[Andrew Cagney]

> On Sat, Jan 19, 2002 at 07:43:35PM -0500, Andrew Cagney wrote:
> 
>  > 2002-01-19  Andrew Cagney  <ac131313@redhat.com>
>> > * utils.c: Include <sys/param.h> for MAXPATHLEN.
>  > (gdb_realpath): Use MAXPATHLEN when PATH_MAX is not defined.
> 
> Seems like you really want PATH_MAX to be an autoconf test, and if
> not, then test for MAXPATHLEN in <sys/param.h> and define PATH_MAX
> in terms of MAXPATHLEN.
> 
> Basically, <machine/param.h> is by no means a standardized header, and
> pulling it in causes all sorts of namespace pollution, so you wanna
> use it only if you have to.


This really begs a question.  Should we put a bomb under utils.c 
breaking it up into smaller chunks (gdb/utils/*.c) so that people can be 
more confident about this.

At present utils.c contains so much that someone trying to change it is 
never really sure what the consequences of an include file tweek are.

Andrew