From: Lancelot SIX <lsix@lancelotsix.com>
To: Tom Tromey <tromey@adacore.com>
Cc: gdb-patches@sourceware.org
Subject: Re: [PATCH 1/7] Introduce type-safe variant of gdb_bfd_openr_iovec
Date: Mon, 18 Sep 2023 11:29:13 +0100 [thread overview]
Message-ID: <20230918102849.ksgb4c33dwsqpcjc@octopus> (raw)
In-Reply-To: <20230824-gdb-bfd-vec-v1-1-850e4e907ed1@adacore.com>
Hi Tom,
On Thu, Aug 24, 2023 at 11:12:18AM -0600, Tom Tromey via Gdb-patches wrote:
> This patch adds a new, type-safe variant of gdb_bfd_openr_iovec. In
> this approach, the underlying user data is simply an object, the
> callbacks are methods, and the "open" function is a function view.
> Nothing uses this new code yet.
> ---
> gdb/gdb_bfd.c | 42 ++++++++++++++++++++++++++++++++++++++++++
> gdb/gdb_bfd.h | 31 +++++++++++++++++++++++++++++++
> 2 files changed, 73 insertions(+)
>
> diff --git a/gdb/gdb_bfd.c b/gdb/gdb_bfd.c
> index 3765561cbe9..0bb97460898 100644
> --- a/gdb/gdb_bfd.c
> +++ b/gdb/gdb_bfd.c
> @@ -907,6 +907,48 @@ gdb_bfd_openw (const char *filename, const char *target)
>
> /* See gdb_bfd.h. */
>
> +gdb_bfd_ref_ptr
> +gdb_bfd_openr_iovec (const char *filename, const char *target,
> + gdb_iovec_opener_ftype open_fn)
> +{
> + auto do_open = [] (bfd *nbfd, void *closure) -> void *
> + {
> + auto real_opener = (gdb_iovec_opener_ftype *) closure;
I think I would have a personal preference toward using static_cast
here, but in this case the C-style cast and static cast are equivalent.
Other than this remark, this looks good to me.
Reviewed-By: Lancelot Six <lancelot.six@amd.com>
Best,
Lancelot.
> + return (*real_opener) (nbfd);
> + };
> +
> + auto read_trampoline = [] (bfd *nbfd, void *stream, void *buf,
> + file_ptr nbytes, file_ptr offset) -> file_ptr
> + {
> + gdb_bfd_iovec_base *obj = (gdb_bfd_iovec_base *) stream;
> + return obj->read (nbfd, buf, nbytes, offset);
> + };
> +
> + auto stat_trampoline = [] (struct bfd *abfd, void *stream,
> + struct stat *sb) -> int
> + {
> + gdb_bfd_iovec_base *obj = (gdb_bfd_iovec_base *) stream;
> + return obj->stat (abfd, sb);
> + };
> +
> + auto close_trampoline = [] (struct bfd *nbfd, void *stream) -> int
> + {
> + gdb_bfd_iovec_base *obj = (gdb_bfd_iovec_base *) stream;
> + delete obj;
> + /* Success. */
> + return 0;
> + };
> +
> + bfd *result = bfd_openr_iovec (filename, target,
> + do_open, &open_fn,
> + read_trampoline, close_trampoline,
> + stat_trampoline);
> +
> + return gdb_bfd_ref_ptr::new_reference (result);
> +}
> +
> +/* See gdb_bfd.h. */
> +
> gdb_bfd_ref_ptr
> gdb_bfd_openr_iovec (const char *filename, const char *target,
> void *(*open_func) (struct bfd *nbfd,
> diff --git a/gdb/gdb_bfd.h b/gdb/gdb_bfd.h
> index d15b1106d9a..ae374f5d7ae 100644
> --- a/gdb/gdb_bfd.h
> +++ b/gdb/gdb_bfd.h
> @@ -22,6 +22,7 @@
>
> #include "registry.h"
> #include "gdbsupport/byte-vector.h"
> +#include "gdbsupport/function-view.h"
> #include "gdbsupport/gdb_ref_ptr.h"
> #include "gdbsupport/iterator-range.h"
> #include "gdbsupport/next-iterator.h"
> @@ -150,6 +151,36 @@ gdb_bfd_ref_ptr gdb_bfd_openr (const char *, const char *);
>
> gdb_bfd_ref_ptr gdb_bfd_openw (const char *, const char *);
>
> +/* The base class for BFD "iovec" implementations. This is used by
> + gdb_bfd_openr_iovec and enables better type safety. */
> +
> +class gdb_bfd_iovec_base
> +{
> +protected:
> +
> + gdb_bfd_iovec_base () = default;
> +
> +public:
> +
> + virtual ~gdb_bfd_iovec_base () = default;
> +
> + /* The "read" callback. */
> + virtual file_ptr read (bfd *abfd, void *buffer, file_ptr nbytes,
> + file_ptr offset) = 0;
> +
> + /* The "stat" callback. */
> + virtual int stat (struct bfd *abfd, struct stat *sb) = 0;
> +};
> +
> +/* The type of the function used to open a new iovec-based BFD. */
> +using gdb_iovec_opener_ftype
> + = gdb::function_view<gdb_bfd_iovec_base * (bfd *)>;
> +
> +/* A type-safe wrapper for bfd_openr_iovec. */
> +
> +gdb_bfd_ref_ptr gdb_bfd_openr_iovec (const char *filename, const char *target,
> + gdb_iovec_opener_ftype open_fn);
> +
> /* A wrapper for bfd_openr_iovec that initializes the gdb-specific
> reference count. */
>
>
> --
> 2.40.1
>
next prev parent reply other threads:[~2023-09-18 10:29 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-08-24 17:12 [PATCH 0/7] Rewrite gdb_bfd_openr_iovec to be type-safe Tom Tromey
2023-08-24 17:12 ` [PATCH 1/7] Introduce type-safe variant of gdb_bfd_openr_iovec Tom Tromey
2023-09-18 10:29 ` Lancelot SIX [this message]
2023-09-18 14:44 ` Tom Tromey
2023-08-24 17:12 ` [PATCH 2/7] Small constructor change to target_buffer Tom Tromey
2023-08-24 17:12 ` [PATCH 3/7] Convert mem_bfd_iovec to new type-safe gdb_bfd_openr_iovec Tom Tromey
2023-08-24 17:12 ` [PATCH 4/7] Convert target fileio " Tom Tromey
2023-08-24 17:12 ` [PATCH 5/7] Convert minidebug " Tom Tromey
2023-08-24 17:12 ` [PATCH 6/7] Convert solib-rocm " Tom Tromey
2023-09-18 9:41 ` Lancelot SIX
2023-09-18 14:49 ` Tom Tromey
2023-08-24 17:12 ` [PATCH 7/7] Remove old gdb_bfd_openr_iovec Tom Tromey
2023-09-18 12:34 ` [PATCH 0/7] Rewrite gdb_bfd_openr_iovec to be type-safe Lancelot SIX
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=20230918102849.ksgb4c33dwsqpcjc@octopus \
--to=lsix@lancelotsix.com \
--cc=gdb-patches@sourceware.org \
--cc=tromey@adacore.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).