From: Doug Evans <dje@google.com>
To: Sergio Durigan Junior <sergiodj@redhat.com>
Cc: gdb-patches <gdb-patches@sourceware.org>
Subject: Re: [patch] Create cleanups.[ch]
Date: Tue, 17 Apr 2012 22:54:00 -0000 [thread overview]
Message-ID: <CADPb22R_TTgwA2qhHLjApRT1rF1dmmu0ZxbFzqgz7BCymNxnVA@mail.gmail.com> (raw)
In-Reply-To: <m3mx6cjmo9.fsf@redhat.com>
Hi. I didn't want it to seem like I ignored you. :-)
Comments inline.
On Sun, Apr 15, 2012 at 10:00 PM, Sergio Durigan Junior
<sergiodj@redhat.com> wrote:
> On Sunday, April 15 2012, Doug Evans wrote:
>
>> Hi.
>
> Hi Doug.
>
>> Since cleanups are a big enough source of issues, I want to separate them out.
>> This patch moves the core cleanup API into its own file.
>> It makes no other changes.
>> I have at least one more patch to go, but I want to get this done
>> first.
>
> Thanks, I'm always in favor of such API separations. Sorry for
> nitpicking, I know you are just moving the code around, but since you
> touched it I felt I should take a look even if it's old code (maybe,
> *especially* because of that!).
>
>> 2012-04-15 Doug Evans <dje@sebabeach.org>
>>
>> * cleanups.h: New file.
>> * cleanups.c: New file.
>> * Makefile.in (SFILES): Add cleanups.c
>> (HFILES_NO_SRCDIR): Add cleanups.h
>> (COMMON_OBS): Add cleanups.o
>> * defs.h (struct cleanup): Moved to cleanups.h
>> (do_cleanups,do_final_cleanups): Ditto.
>> (discard_cleanups,discard_final_cleanups): Ditto
>> (make_cleanup,make_cleanup_dtor,make_final_cleanup): Ditto.
>> (save_cleanups,save_final_cleanups): Ditto.
>> (restore_cleanups,restore_final_cleanups): Ditto.
>> (null_cleanup): Ditto.
>> (make_my_cleanup,make_my_cleanup2): Delete.
>> (discard_my_cleanups,save_my_cleanups,restore_my_cleanups): Delete.
>> * utils.c (cleanup_chain,final_cleanup_chain): Moved to cleanups.c.
>> (do_cleanups,do_final_cleanups): Ditto.
>> (discard_cleanups,discard_final_cleanups): Ditto
>> (make_cleanup,make_cleanup_dtor,make_final_cleanup): Ditto.
>> (save_cleanups,save_final_cleanups): Ditto.
>> (restore_cleanups,restore_final_cleanups): Ditto.
>> (null_cleanup): Ditto.
>> (make_my_cleanup,make_my_cleanup2): Ditto, and make static.
>> All uses rewritten to use proper interface.
>> (discard_my_cleanups,save_my_cleanups,restore_my_cleanups): Ditto.
>
> I remember using `Ditto' once, and being told that I should use
> `Likewise' instead. Anyway, I'm just bringing this because I never know
> what rule to follow :-).
I'm not aware of a specific preference.
I probably use either depending on the lunar phase or whatever :-),
and there's precedent for both.
"Go with the flow." works well on so many levels.
>> Index: cleanups.c
>> ===================================================================
>> RCS file: cleanups.c
>> diff -N cleanups.c
>> --- /dev/null 1 Jan 1970 00:00:00 -0000
>> +++ cleanups.c 15 Apr 2012 19:23:28 -0000
>> @@ -0,0 +1,210 @@
>> +/* Cleanup routines for GDB, the GNU debugger.
>> +
>> + Copyright (C) 1986, 1988-2012 Free Software Foundation, Inc.
>
> It should be:
>
> Copyright (C) 2012 Free Software Foundation, Inc.
>
> AFAIK, since it's a new file.
I'm of two minds on this, but here I just "went with the flow."
>> +#include "defs.h"
>> +
>> +/* Chain of cleanup actions established with make_cleanup,
>> + to be executed if an error happens. */
>> +
>> +/* Cleaned up after a failed command. */
>> +static struct cleanup *cleanup_chain;
>> +
>> +/* Cleaned up when gdb exits. */
>> +static struct cleanup *final_cleanup_chain;
>> +
>> +static struct cleanup *
>> +make_my_cleanup2 (struct cleanup **pmy_chain, make_cleanup_ftype *function,
>> + void *arg, void (*free_arg) (void *))
>> +{
>> + struct cleanup *new
>> + = (struct cleanup *) xmalloc (sizeof (struct cleanup));
>> + struct cleanup *old_chain = *pmy_chain;
>> +
>> + new->next = *pmy_chain;
>> + new->function = function;
>> + new->free_arg = free_arg;
>> + new->arg = arg;
>> + *pmy_chain = new;
>> +
>> + return old_chain;
>> +}
>
> Maybe this function should be named `make_my_cleanup_1', just like other
> cases in GDB? I think it should also have a comment here, since it's
> static and its declaration/definition is here.
Such changes run afoul of the code movement rule. :-)
For comments, I went back and added the missing ones though.
>> +
>> +static struct cleanup *
>> +make_my_cleanup (struct cleanup **pmy_chain, make_cleanup_ftype *function,
>> + void *arg)
>> +{
>> + return make_my_cleanup2 (pmy_chain, function, arg, NULL);
>> +}
>
> Comment for this as well. Same thing for all static functions.
>
>> +
>> +/* Add a new cleanup to the cleanup_chain,
>> + and return the previous chain pointer
>> + to be passed later to do_cleanups or discard_cleanups.
>> + Args are FUNCTION to clean up with, and ARG to pass to it. */
>
> Maybe these comments can be made to fill more than 42 characters in the
> line?
>
> I also never know what's the best/recommended practice: to put the
> comments above the function's declaration (in this case, in the
> cleanups.h file), or to put comments above the function definition (as
> you did). Maybe someone more experienced can clarify.
I'm not sure there's a documented preference.
> I prefer comments in the declaration, FWIW.
Me too, though not enough to actually follow it religiously.
>> +
>> +struct cleanup *
>> +make_cleanup (make_cleanup_ftype *function, void *arg)
>> +{
>> + return make_my_cleanup (&cleanup_chain, function, arg);
>> +}
>> +
>> +/* Same as make_cleanup except also includes TDOR, a destructor to free ARG.
>> + DTOR is invoked when the cleanup is performed or when it is discarded. */
>> +
>> +struct cleanup *
>> +make_cleanup_dtor (make_cleanup_ftype *function, void *arg,
>> + void (*dtor) (void *))
>> +{
>> + return make_my_cleanup2 (&cleanup_chain,
>> + function, arg, dtor);
>
> No need to break the line, I think.
Code movement violation. :-)
It's easy enough to do a later pass to clean such things up.
>> +}
>> +
>> +static void
>> +do_my_cleanups (struct cleanup **pmy_chain,
>> + struct cleanup *old_chain)
>> +{
>> + struct cleanup *ptr;
>> +
>> + while ((ptr = *pmy_chain) != old_chain)
>> + {
>> + *pmy_chain = ptr->next; /* Do this first in case of recursion. */
>> + (*ptr->function) (ptr->arg);
>> + if (ptr->free_arg)
>> + (*ptr->free_arg) (ptr->arg);
>> + xfree (ptr);
>> + }
>> +}
>
> Comment above the function. Same thing for static functions below.
>
>> Index: cleanups.h
>> ===================================================================
>> RCS file: cleanups.h
>> diff -N cleanups.h
>> --- /dev/null 1 Jan 1970 00:00:00 -0000
>> +++ cleanups.h 15 Apr 2012 19:23:28 -0000
>> @@ -0,0 +1,86 @@
>> +/* Cleanups.
>> + Copyright (C) 1986, 1988-2005, 2007-2012 Free Software Foundation, Inc.
>
> Copyright should be 2012.
>
> --
> Sergio
Cheers.
next prev parent reply other threads:[~2012-04-17 22:41 UTC|newest]
Thread overview: 20+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-04-15 20:27 Doug Evans
2012-04-16 7:27 ` Sergio Durigan Junior
2012-04-16 10:52 ` Pedro Alves
2012-04-16 19:11 ` Sergio Durigan Junior
2012-04-18 9:43 ` Pedro Alves
2012-04-19 17:37 ` Sergio Durigan Junior
2012-04-20 15:09 ` Pedro Alves
2012-04-16 20:06 ` Tom Tromey
2012-04-16 20:42 ` Pedro Alves
2012-04-16 20:59 ` Tom Tromey
2012-04-17 22:41 ` Doug Evans
2012-04-16 20:10 ` Tom Tromey
2012-04-17 22:54 ` Doug Evans [this message]
2012-04-17 22:59 ` Joel Brobecker
2012-04-18 4:26 ` Sergio Durigan Junior
2012-04-18 6:22 ` Joel Brobecker
2012-04-16 10:40 ` Pedro Alves
2012-04-18 14:02 ` Yao Qi
2012-04-18 14:32 ` Tom Tromey
2012-04-16 2:07 Doug Evans
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=CADPb22R_TTgwA2qhHLjApRT1rF1dmmu0ZxbFzqgz7BCymNxnVA@mail.gmail.com \
--to=dje@google.com \
--cc=gdb-patches@sourceware.org \
--cc=sergiodj@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox