From: Tom Tromey <tom@tromey.com>
To: gdb-patches@sourceware.org
Subject: [PATCH 00/36] C++-ify breakpoints
Date: Tue, 18 Jan 2022 12:39:31 -0700 [thread overview]
Message-ID: <20220118194007.2853108-1-tom@tromey.com> (raw)
This series somewhat C++-ifies breakpoints. It turns most
breakpoint_ops function pointers into virtual methods, and introduces
a class hierarchy for breakpoints.
The series starts with a few minor cleanups and clarifications. Then
it introduces a new vtable-based breakpoint_ops structure. This is
used to convert breakpoint types one-by-one, to avoid having a monster
patch. Then, near the end, this is removed again and some final
cleanups applied.
I built and regression tested each patch in this series on x86-64
Fedora 34.
There is still more work that could be done to improve this code in
the future:
* Right now, if you want to add a new breakpoint type, you may need a
new bptype enumerator constant. However, these can be a pain to
add. I think it would be better to remove this enum entirely, in
favor of ordinary object-oriented techniques like methods.
Similarly, perhaps create_breakpoint could be replaced with
constructors.
* breakpoint_ops isn't fully removed, as it has a couple of "static
virtual" methods -- methods that vary by breakpoint type, but which
can't be methods on the breakpoint itself. This could probably be
fixed somehow.
* I left the base class name as "breakpoint" to avoid rewriting a lot
of code. However, this isn't really the best name, so this could be
changed.
* Some methods on the base class could probably be made pure virtual.
This is currently faked, following the approach already used in the
source: the methods abort when called. Changing this to use the
real C++ feature would be an improvement.
* Relatedly, some members of 'breakpoint' could be pushed into the
relevant subclasses. For example, breakpoint::location_range_end is
only used by ranged breakpoints, and so could be moved to struct
ranged_breakpoint. This same idea probably applies to some methods
as well.
Tom
next reply other threads:[~2022-01-18 19:40 UTC|newest]
Thread overview: 49+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-01-18 19:39 Tom Tromey [this message]
2022-01-18 19:39 ` [PATCH 01/36] Move "catch load" to a new file Tom Tromey
2022-04-23 2:34 ` Simon Marchi
2022-04-24 19:35 ` Tom Tromey
2022-01-18 19:39 ` [PATCH 02/36] Boolify print_solib_event Tom Tromey
2022-01-18 19:39 ` [PATCH 03/36] Add an assertion to clone_momentary_breakpoint Tom Tromey
2022-04-25 13:25 ` Thiago Jung Bauermann
2022-04-25 19:03 ` Tom Tromey
2022-01-18 19:39 ` [PATCH 04/36] Delete some unnecessary wrapper functions Tom Tromey
2022-04-25 14:08 ` Thiago Jung Bauermann
2022-04-25 19:06 ` Tom Tromey
2022-04-25 20:06 ` Thiago Jung Bauermann
2022-01-18 19:39 ` [PATCH 05/36] Return bool from breakpoint_ops::print_one Tom Tromey
2022-01-18 19:39 ` [PATCH 06/36] Add a vtable-based breakpoint ops Tom Tromey
2022-01-18 19:39 ` [PATCH 07/36] Convert break-catch-sig to use vtable ops Tom Tromey
2022-01-18 19:39 ` [PATCH 08/36] Convert break-catch-syscall to " Tom Tromey
2022-01-18 19:39 ` [PATCH 09/36] Convert break-catch-exec " Tom Tromey
2022-01-18 19:39 ` [PATCH 10/36] Convert break-catch-fork " Tom Tromey
2022-01-18 19:39 ` [PATCH 11/36] Convert break-catch-load " Tom Tromey
2022-01-18 19:39 ` [PATCH 12/36] Convert watchpoints " Tom Tromey
2022-01-18 19:39 ` [PATCH 13/36] Convert tracepoints " Tom Tromey
2022-01-18 19:39 ` [PATCH 14/36] Add some new subclasses of breakpoint Tom Tromey
2022-01-18 19:39 ` [PATCH 15/36] Convert base breakpoints to vtable ops Tom Tromey
2022-05-02 10:27 ` Tom de Vries
2022-05-02 16:40 ` Tom Tromey
2022-01-18 19:39 ` [PATCH 16/36] Convert break-catch-throw " Tom Tromey
2022-01-18 19:39 ` [PATCH 17/36] Convert internal breakpoints " Tom Tromey
2022-01-18 19:39 ` [PATCH 18/36] Convert momentary " Tom Tromey
2022-01-18 19:39 ` [PATCH 19/36] Change inheritance of dprintf Tom Tromey
2022-01-18 19:39 ` [PATCH 20/36] Convert ordinary breakpoints to vtable ops Tom Tromey
2022-01-18 19:39 ` [PATCH 21/36] Convert Ada catchpoints " Tom Tromey
2022-01-18 19:39 ` [PATCH 22/36] Convert dprintf " Tom Tromey
2022-01-18 19:39 ` [PATCH 23/36] Convert ranged breakpoints " Tom Tromey
2022-01-18 19:39 ` [PATCH 24/36] Add bp_static_marker_tracepoint Tom Tromey
2022-01-18 19:39 ` [PATCH 25/36] Convert static marker tracepoints to vtable ops Tom Tromey
2022-01-18 19:39 ` [PATCH 26/36] Remove bkpt_base_breakpoint_ops Tom Tromey
2022-01-18 19:39 ` [PATCH 27/36] Merge probe and ordinary breakpoints Tom Tromey
2022-01-18 19:39 ` [PATCH 28/36] Merge probe and ordinary tracepoints Tom Tromey
2022-01-18 19:40 ` [PATCH 29/36] Remove breakpoint_ops from init_ada_exception_breakpoint Tom Tromey
2022-01-18 19:40 ` [PATCH 30/36] Remove breakpoint_ops from init_catchpoint Tom Tromey
2022-01-18 19:40 ` [PATCH 31/36] Remove most fields from breakpoint_ops Tom Tromey
2022-01-18 19:40 ` [PATCH 32/36] Remove vtable_breakpoint_ops Tom Tromey
2022-01-18 19:40 ` [PATCH 33/36] Remove breakpoint ops initialization Tom Tromey
2022-01-18 19:40 ` [PATCH 34/36] Constify breakpoint_ops Tom Tromey
2022-01-18 19:40 ` [PATCH 35/36] Remove allocate_bp_location Tom Tromey
2022-01-18 19:40 ` [PATCH 36/36] Remove create_breakpoints_sal_default Tom Tromey
2022-04-22 20:21 ` [PATCH 00/36] C++-ify breakpoints Tom Tromey
2022-04-23 2:59 ` Simon Marchi
2022-04-29 22:15 ` 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=20220118194007.2853108-1-tom@tromey.com \
--to=tom@tromey.com \
--cc=gdb-patches@sourceware.org \
/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).