From: Simon Marchi <simark@simark.ca>
To: Tom Tromey <tom@tromey.com>, gdb-patches@sourceware.org
Subject: Re: [PATCH 00/47] Use methods for struct value
Date: Thu, 9 Feb 2023 21:54:18 -0500 [thread overview]
Message-ID: <43ad228f-55d0-d1b4-8dd3-20fec653967e@simark.ca> (raw)
In-Reply-To: <20230209-submit-value-fixups-2023-v1-0-55dc2794dbb9@tromey.com>
On 2/9/23 16:38, Tom Tromey wrote:
> This series changes many value-related functions to be methods. It
> also moves the definition of struct value to value.h, so that callers
> will benefit by inlining the smaller methods. The members of value
> are all made private at the end of the series.
>
> It's difficult to decide exactly which value functions ought to be
> converted, because there are so many. In the end I chose a relatively
> minimal approach -- only those that require direct access to the inner
> workings of value.
>
> It's possible, even easy, to convert more functions. If there are
> some in particular that seem worthwhile, let me know.
>
> I tacked on one small cleanup at the end of this series, but as it was
> already fairly long, I didn't try to do all the possible cleanups.
> For example, many things could use bool, some of the setter functions
> could be removed in favor of specialized constructors, etc.
>
> Regression tested on x86-64 Fedora 36.
>
> Tom
>
> PS - I'm testing the "b4 send" command, so my apologies if something
> goes wrong.
I was able to fetch it fine using "b4 am". However, it told me the DKIM
signature for "tromey.com" was incorrect :P.
I went relatively quickly over the whole series. I concentrated on
changes in value.h, since the rest is fairly mechanical. It looks
really nice, I had started to do that a little bit of these changes, but
I was sooo far away from what you have here.
If you want to go over the series again quickly, there are a few
comments in value.h that refer to now non-existent parameters (which
because "this"). They could be reworded to talk about "this value"
instead.
I agree with the instances where you removed "deprecated" from some
functions. I guess some refactoring work was started many moons ago to
get rid of them, but wasn't completed, and now the context is completely
lost. It's pointless to call something deprecated but also leave it
there forever.
Maybe others will want to take a look, but from my side:
Approved-By: Simon Marchi <simon.marchi@efficios.com>
Simon
next prev parent reply other threads:[~2023-02-10 2:54 UTC|newest]
Thread overview: 60+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-02-09 21:38 Tom Tromey
2023-02-09 21:38 ` [PATCH 01/47] Rename all fields of " Tom Tromey
2023-02-10 21:05 ` Tom Tromey
2023-02-09 21:38 ` [PATCH 02/47] Move ~value body out-of-line Tom Tromey
2023-02-09 21:38 ` [PATCH 03/47] Move struct value to value.h Tom Tromey
2023-02-09 21:38 ` [PATCH 04/47] Turn value_type into method Tom Tromey
2023-02-09 21:38 ` [PATCH 05/47] Turn deprecated_set_value_type into a method Tom Tromey
2023-02-09 21:38 ` [PATCH 06/47] Turn value_arch into method Tom Tromey
2023-02-10 2:08 ` Simon Marchi
2023-02-10 16:36 ` Tom Tromey
2023-02-09 21:38 ` [PATCH 07/47] Turn value_bitsize " Tom Tromey
2023-02-09 21:38 ` [PATCH 08/47] Turn value_bitpos " Tom Tromey
2023-02-09 21:38 ` [PATCH 09/47] Turn value_parent " Tom Tromey
2023-02-09 21:38 ` [PATCH 10/47] Turn value_offset " Tom Tromey
2023-02-09 21:38 ` [PATCH 11/47] Turn deprecated_value_modifiable " Tom Tromey
2023-02-09 21:38 ` [PATCH 12/47] Turn value_enclosing_type " Tom Tromey
2023-02-09 21:38 ` [PATCH 13/47] Turn some value offset functions " Tom Tromey
2023-02-09 21:38 ` [PATCH 14/47] Turn value_lazy and set_value_lazy functions into methods Tom Tromey
2023-02-09 21:38 ` [PATCH 15/47] Turn value_stack and set_value_stack " Tom Tromey
2023-02-09 21:38 ` [PATCH 16/47] Turn value_computed_closure and value_computed_funcs " Tom Tromey
2023-02-09 21:38 ` [PATCH 17/47] Convert value_lval_const and deprecated_lval_hack to methods Tom Tromey
2023-02-09 21:38 ` [PATCH 18/47] Turn value_initialized and set_value_initialized functions into methods Tom Tromey
2023-02-09 21:38 ` [PATCH 19/47] Turn value_address and set_value_address " Tom Tromey
2023-02-09 21:38 ` [PATCH 20/47] Turn more deprecated_* " Tom Tromey
2023-02-09 21:38 ` [PATCH 21/47] Turn allocate_value_lazy into a static "constructor" Tom Tromey
2023-02-09 21:38 ` [PATCH 22/47] Turn allocate_value " Tom Tromey
2023-02-09 21:38 ` [PATCH 23/47] Turn allocate_computed_value into " Tom Tromey
2023-02-09 21:38 ` [PATCH 24/47] Turn allocate_optimized_out_value " Tom Tromey
2023-02-09 21:38 ` [PATCH 25/47] Turn value_zero " Tom Tromey
2023-02-09 21:38 ` [PATCH 26/47] Turn some value_contents functions into methods Tom Tromey
2023-02-09 21:38 ` [PATCH 27/47] Turn value_fetch_lazy into a method Tom Tromey
2023-02-09 21:38 ` [PATCH 28/47] Turn allocate_value_contents " Tom Tromey
2023-02-09 21:38 ` [PATCH 29/47] Turn value_contents_eq " Tom Tromey
2023-02-10 2:18 ` Simon Marchi
2023-02-10 17:46 ` Tom Tromey
2023-02-09 21:38 ` [PATCH 30/47] Turn value_bits_synthetic_pointer " Tom Tromey
2023-02-09 21:38 ` [PATCH 31/47] Move value_ref_policy methods out-of-line Tom Tromey
2023-02-09 21:38 ` [PATCH 32/47] Turn value_incref and value_decref into methods Tom Tromey
2023-02-09 21:38 ` [PATCH 33/47] Turn remaining value_contents functions " Tom Tromey
2023-02-10 2:24 ` Simon Marchi
2023-02-10 17:46 ` Tom Tromey
2023-02-09 21:38 ` [PATCH 34/47] Fully qualify calls to copy in value.c Tom Tromey
2023-02-09 21:38 ` [PATCH 35/47] Turn value_copy into a method Tom Tromey
2023-02-10 2:42 ` Simon Marchi
2023-02-10 18:03 ` Tom Tromey
2023-02-09 21:38 ` [PATCH 36/47] Turn many optimized-out value functions into methods Tom Tromey
2023-02-09 21:38 ` [PATCH 37/47] Turn value_non_lval and value_force_lval " Tom Tromey
2023-02-09 21:38 ` [PATCH 38/47] Turn set_value_component_location into method Tom Tromey
2023-02-09 21:39 ` [PATCH 39/47] Change some code to use value methods Tom Tromey
2023-02-09 21:39 ` [PATCH 40/47] Turn some xmethod functions into methods Tom Tromey
2023-02-09 21:39 ` [PATCH 41/47] Turn preserve_one_value into method Tom Tromey
2023-02-09 21:39 ` [PATCH 42/47] Turn various value copying-related functions into methods Tom Tromey
2023-02-10 20:20 ` Tom Tromey
2023-02-09 21:39 ` [PATCH 43/47] Add value::set_modifiable Tom Tromey
2023-02-09 21:39 ` [PATCH 44/47] Make struct value data members private Tom Tromey
2023-02-09 21:39 ` [PATCH 45/47] Make ~value private Tom Tromey
2023-02-09 21:39 ` [PATCH 46/47] Introduce set_lval method on value Tom Tromey
2023-02-09 21:39 ` [PATCH 47/47] Remove deprecated_lval_hack Tom Tromey
2023-02-10 2:54 ` Simon Marchi [this message]
2023-02-10 18:08 ` [PATCH 00/47] Use methods for struct value Tom Tromey
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=43ad228f-55d0-d1b4-8dd3-20fec653967e@simark.ca \
--to=simark@simark.ca \
--cc=gdb-patches@sourceware.org \
--cc=tom@tromey.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;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).