Re: [RFC] Move/Rename include/wait.h ("wait.h")

Re: [RFC] Move/Rename include/wait.h ("wait.h")

[J.T. Conklin]

>>>>> "Andrew" == Andrew Cagney <ac131313@cygnus.com> writes:
Andrew> I'd like to put forward the proposal that the file
Andrew> ``include/wait.h'' be either:
Andrew> 	o	renamed to include/gnu-wait.h (?)
Andrew> 	o	moved to gdb/gdb-wait.h

If it's not used by other programs, I'd recommend that it be moved to
gdb_wait.h (with an underscore, not a dash).  I created gdb_string.h
and gdb_stat.h long ago for the same sort of reasons.

We might also want to consider creating a gdb_fcntl.h header to wrap
fcntl.h.  It would would define O_BINARY to 0 on hosts that don't
distinguish between text and binary files; instead of having that
definition scattered throughout various *.c files.

        --jtc

-- 
J.T. Conklin
RedBack Networks
From shebs@cygnus.com Thu Dec 09 12:15:00 1999
From: Stan Shebs <shebs@cygnus.com>
To: jtc@redback.com
Cc: gcc-patches@gcc.gnu.org, gdb-patches@sourceware.cygnus.com
Subject: Re: [RFC] Move/Rename include/wait.h ("wait.h")
Date: Thu, 09 Dec 1999 12:15:00 -0000
Message-id: <199912092015.MAA13565@andros.cygnus.com>
References: <5mn1rk6j43.fsf@jtc.redbacknetworks.com>
X-SW-Source: 1999-q4/msg00360.html
Content-length: 1153

   From: jtc@redback.com (J.T. Conklin)
   Date: 09 Dec 1999 10:53:48 -0800

   >>>>> "Andrew" == Andrew Cagney <ac131313@cygnus.com> writes:
   Andrew> I'd like to put forward the proposal that the file
   Andrew> ``include/wait.h'' be either:
   Andrew> 	o	renamed to include/gnu-wait.h (?)
   Andrew> 	o	moved to gdb/gdb-wait.h

   If it's not used by other programs, I'd recommend that it be moved to
   gdb_wait.h (with an underscore, not a dash).  I created gdb_string.h
   and gdb_stat.h long ago for the same sort of reasons.

I agree, this sounds like the right idea.

   We might also want to consider creating a gdb_fcntl.h header to wrap
   fcntl.h.  It would would define O_BINARY to 0 on hosts that don't
   distinguish between text and binary files; instead of having that
   definition scattered throughout various *.c files.

This would be good too, we have picked up a number of redundant
definitions...  Might be a good idea to consider doing the include
of fcntl.h in defs.h, it's mentioned explicitly in nearly half the
source files already, and the compiletime overhead of including it
always would be pretty minor.

								Stan

Re: [RFC] Move/Rename include/wait.h ("wait.h")

[Andrew Cagney]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 4225 bytes --]

"J.T. Conklin" wrote:
> 
> >>>>> "Andrew" == Andrew Cagney <ac131313@cygnus.com> writes:
> Andrew> I'd like to put forward the proposal that the file
> Andrew> ``include/wait.h'' be either:
> Andrew>         o       renamed to include/gnu-wait.h (?)
> Andrew>         o       moved to gdb/gdb-wait.h
> 
> If it's not used by other programs, I'd recommend that it be moved to
> gdb_wait.h (with an underscore, not a dash).  I created gdb_string.h
> and gdb_stat.h long ago for the same sort of reasons.

Ah, that explains the rationale.  We're going to have to put that one
into the GDB programing guidelines! :-)

	Andrew
From rth@cygnus.com Thu Dec 09 23:43:00 1999
From: Richard Henderson <rth@cygnus.com>
To: Andrew Cagney <ac131313@cygnus.com>
Cc: GCC Patches <gcc-patches@gcc.gnu.org>, GDB Patches <gdb-patches@sourceware.cygnus.com>
Subject: Re: [RFC] Move/Rename include/wait.h ("wait.h")
Date: Thu, 09 Dec 1999 23:43:00 -0000
Message-id: <19991209234347.B19791@cygnus.com>
References: <384F481D.A77D2CEF@cygnus.com>
X-SW-Source: 1999-q4/msg00362.html
Content-length: 287

On Thu, Dec 09, 1999 at 05:11:41PM +1100, Andrew Cagney wrote:
> I'd like to rename/move the file so that this problem goes away :-)  My
> preferred option is to move it to the GDB directory as gdb-wait.h since,
> to the best of my knowledge, nothing else is using it.

Have at it.


r~
From ac131313@cygnus.com Fri Dec 10 06:10:00 1999
From: Andrew Cagney <ac131313@cygnus.com>
To: "Brown, Rodney" <rodneybrown@pmsc.com>
Cc: "'gdb-patches@sourceware.cygnus.com'" <gdb-patches@sourceware.cygnus.com>
Subject: Re: Patches to 19991116 snapshot for vendor compilers on AIX and OSF14.0
Date: Fri, 10 Dec 1999 06:10:00 -0000
Message-id: <3851099E.66043C5C@cygnus.com>
References: <9150F3E779F0D211BD370008C733141C38AA0E@aus-msg-02.au.pmsc.com>
X-SW-Source: 1999-q4/msg00363.html
Content-length: 515

> "Brown, Rodney" wrote:
> 
> Digital (Compaq) errors comparing a function pointer with another with

Can you expand this one a bit more.  Are the consequence ``bad'' or just
that DEC is being pedantic? :-)

I ask as GDB currently contains many casts using (make_cleanup_func). 
The code has been identified as something that violates the ANSI spec
but, fortunatly, works.  If Alphas really do have a technical problem
with that cast/compare then there could easily be the same problem in
make_cleanups().

	Andrew
From rodneybrown@pmsc.com Fri Dec 10 13:10:00 1999
From: "Brown, Rodney" <rodneybrown@pmsc.com>
To: "'Andrew Cagney'" <ac131313@cygnus.com>
Cc: "'gdb-patches@sourceware.cygnus.com'" <gdb-patches@sourceware.cygnus.com>
Subject: RE: Patches to 19991116 snapshot for vendor compilers on AIX and OSF14.0
Date: Fri, 10 Dec 1999 13:10:00 -0000
Message-id: <9150F3E779F0D211BD370008C733141C38AA2B@aus-msg-02.au.pmsc.com>
X-SW-Source: 1999-q4/msg00364.html
Content-length: 1340

Title: RE: Patches to 19991116 snapshot for vendor compilers on AIX and OSF14.0





 
> > Digital (Compaq) errors comparing a function pointer with 
> another with
> 
> Can you expand this one a bit more.  Are the consequence 
> ``bad'' or just
> that DEC is being pedantic? :-)


I think it is only pedantry - it does fail to compile


cc: Error: ../../gdb-19991116/gdb/target.c, line 2840: In this statement, "current_target.to_rcmd" and "(void ...)tcomplain" cannot be compared for equality or inequality. (noequality)

  if ((current_target.to_rcmd == (void*) tcomplain)
-------^
cc: Error: ../../gdb-19991116/gdb/target.c, line 2842: In this statement, "debug_target.to_rcmd" and "(void ...)tcomplain" cannot be compared for equality or inequality. (noequality)

          && (debug_target.to_rcmd == (void*) tcomplain)))
--------------^ 
> I ask as GDB currently contains many casts using (make_cleanup_func). 
> The code has been identified as something that violates the ANSI spec
> but, fortunatly, works.  If Alphas really do have a technical problem
> with that cast/compare then there could easily be the same problem in
> make_cleanups().


If make_cleanups was already present in 19991116, then it has already compiled
without erroring out, otherwise I'll find out when I build the new snapshot
on the alpha.