From: Simon Marchi <simark@simark.ca>
To: Tom Tromey <tom@tromey.com>, gdb-patches@sourceware.org
Subject: Re: [RFA 01/10] Remove some cleanups from search_minsyms_for_name
Date: Mon, 02 Apr 2018 01:39:00 -0000 [thread overview]
Message-ID: <8f38f8d0-fe2e-7023-dd57-7a32af834e4a@simark.ca> (raw)
In-Reply-To: <20180401163539.15314-2-tom@tromey.com>
On 2018-04-01 12:35 PM, Tom Tromey wrote:
> This changes struct collect_minsyms to use a std::vector, which
> enables the removal of a cleanup from search_minsyms_for_name. This
> also changes iterate_over_minimal_symbols to take a
> gdb::function_view, which makes a function in linespec.c more
> type-safe.
>
> gdb/ChangeLog
> 2018-03-31 Tom Tromey <tom@tromey.com>
>
> * minsyms.h (iterate_over_minimal_symbols): Update.
> * minsyms.c (iterate_over_minimal_symbols): Take a
> gdb::function_view.
> * linespec.c (struct collect_minsyms): Add constructor and
> initializers.
> <msyms>: Now a std::vector.
> (compare_msyms): Now a std::sort comparator.
> (add_minsym): Change type of second parameter.
> (search_minsyms_for_name): Update.
> ---
> gdb/ChangeLog | 12 +++++++
> gdb/linespec.c | 112 ++++++++++++++++++++++++++++-----------------------------
> gdb/minsyms.c | 9 ++---
> gdb/minsyms.h | 8 ++---
> 4 files changed, 72 insertions(+), 69 deletions(-)
>
> diff --git a/gdb/linespec.c b/gdb/linespec.c
> index 1236b3f475..bd09f57b05 100644
> --- a/gdb/linespec.c
> +++ b/gdb/linespec.c
> @@ -46,6 +46,7 @@
> #include "location.h"
> #include "common/function-view.h"
> #include "common/def-vector.h"
> +#include <algorithm>
>
> /* An enumeration of the various things a user might attempt to
> complete for a linespec location. */
> @@ -4375,8 +4376,15 @@ minsym_found (struct linespec_state *self, struct objfile *objfile,
>
> struct collect_minsyms
> {
> + collect_minsyms (int funfirstline_, int list_mode_, struct symtab *symtab_)
> + : symtab (symtab_),
> + funfirstline (funfirstline_),
Note, funfirstline is unused, so you might as well remove it.
> + list_mode (list_mode_)
> + {
> + }
> +
> /* The objfile we're examining. */
> - struct objfile *objfile;
> + struct objfile *objfile = nullptr;
>
> /* Only search the given symtab, or NULL to search for all symbols. */
> struct symtab *symtab;
> @@ -4388,7 +4396,7 @@ struct collect_minsyms
> int list_mode;
>
> /* The resulting symbols. */
> - VEC (bound_minimal_symbol_d) *msyms;
> + std::vector<struct bound_minimal_symbol> msyms;
> };
I would lean towards removing the collect_minsyms structure and just pass anything
that's required by parameter to add_minsym. It's easier to follow the flow of
execution when parameters are passed explicitly than through a structure.
>
> /* A helper function to classify a minimal_symbol_type according to
> @@ -4415,47 +4423,43 @@ classify_mtype (enum minimal_symbol_type t)
> }
> }
>
> -/* Callback for qsort that sorts symbols by priority. */
> +/* Callback for std::sort that sorts symbols by priority. */
>
> -static int
> -compare_msyms (const void *a, const void *b)
> +static bool
> +compare_msyms (const bound_minimal_symbol &a, const bound_minimal_symbol &b)
> {
> - const bound_minimal_symbol_d *moa = (const bound_minimal_symbol_d *) a;
> - const bound_minimal_symbol_d *mob = (const bound_minimal_symbol_d *) b;
> - enum minimal_symbol_type ta = MSYMBOL_TYPE (moa->minsym);
> - enum minimal_symbol_type tb = MSYMBOL_TYPE (mob->minsym);
> + enum minimal_symbol_type ta = MSYMBOL_TYPE (a.minsym);
> + enum minimal_symbol_type tb = MSYMBOL_TYPE (b.minsym);
>
> - return classify_mtype (ta) - classify_mtype (tb);
> + return classify_mtype (ta) < classify_mtype (tb);
> }
>
> /* Callback for iterate_over_minimal_symbols that adds the symbol to
> the result. */
This comment should probably updated. It is not really a callback for
iterate_over_minimal_symbols anymore (at least not directly).
>
> static void
> -add_minsym (struct minimal_symbol *minsym, void *d)
> +add_minsym (struct minimal_symbol *minsym, struct collect_minsyms &info)
...
> diff --git a/gdb/minsyms.c b/gdb/minsyms.c
> index 72969b7778..08efa1dc7e 100644
> --- a/gdb/minsyms.c
> +++ b/gdb/minsyms.c
> @@ -471,11 +471,8 @@ linkage_name_str (const lookup_name_info &lookup_name)
> void
> iterate_over_minimal_symbols (struct objfile *objf,
> const lookup_name_info &lookup_name,
> - void (*callback) (struct minimal_symbol *,
> - void *),
> - void *user_data)
> + gdb::function_view<void (struct minimal_symbol *)> callback)
This line is too long, you should probably wrap the whole parameter list.
> {
> -
> /* The first pass is over the ordinary hash table. */
> {
> const char *name = linkage_name_str (lookup_name);
> @@ -490,7 +487,7 @@ iterate_over_minimal_symbols (struct objfile *objf,
> iter = iter->hash_next)
> {
> if (mangled_cmp (MSYMBOL_LINKAGE_NAME (iter), name) == 0)
> - (*callback) (iter, user_data);
> + callback (iter);
> }
> }
>
> @@ -509,7 +506,7 @@ iterate_over_minimal_symbols (struct objfile *objf,
> iter != NULL;
> iter = iter->demangled_hash_next)
> if (name_match (MSYMBOL_SEARCH_NAME (iter), lookup_name, NULL))
> - (*callback) (iter, user_data);
> + callback (iter);
> }
> }
>
> diff --git a/gdb/minsyms.h b/gdb/minsyms.h
> index 11a202025d..b05f717575 100644
> --- a/gdb/minsyms.h
> +++ b/gdb/minsyms.h
> @@ -268,11 +268,9 @@ struct bound_minimal_symbol lookup_minimal_symbol_by_pc (CORE_ADDR);
> For each matching symbol, CALLBACK is called with the symbol and
> USER_DATA as arguments. */
This comment should be updated.
>
> -void iterate_over_minimal_symbols (struct objfile *objf,
> - const lookup_name_info &name,
> - void (*callback) (struct minimal_symbol *,
> - void *),
> - void *user_data);
> +void iterate_over_minimal_symbols
> + (struct objfile *objf, const lookup_name_info &name,
> + gdb::function_view<void (struct minimal_symbol *)> callback);
>
> /* Compute the upper bound of MINSYM. The upper bound is the last
> address thought to be part of the symbol. If the symbol has a
>
Thanks,
Simon
next prev parent reply other threads:[~2018-04-02 1:39 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-04-01 16:35 [RFA 00/10] Remove some cleanups from linespec.c Tom Tromey
2018-04-01 16:35 ` [RFA 02/10] Fix some indentation in linespec.c Tom Tromey
2018-04-02 1:49 ` Simon Marchi
2018-04-01 16:35 ` [RFA 06/10] Remove a string copy from event_location_to_sals Tom Tromey
2018-04-01 16:35 ` [RFA 09/10] Remove typep and VEC(typep) from linespec.c Tom Tromey
2018-04-01 16:35 ` [RFA 01/10] Remove some cleanups from search_minsyms_for_name Tom Tromey
2018-04-02 1:39 ` Simon Marchi [this message]
2018-04-01 16:35 ` [RFA 05/10] Have filter_results take a std::vector Tom Tromey
2018-04-01 16:35 ` [RFA 04/10] Return std::string from canonical_to_fullform Tom Tromey
2018-04-02 2:15 ` Simon Marchi
2018-04-01 16:35 ` [RFA 10/10] Remove unnecessary include from linespec.h Tom Tromey
2018-04-01 16:35 ` [RFA 08/10] More use of std::vector in linespec.c Tom Tromey
2018-04-02 2:35 ` Simon Marchi
2018-04-01 16:35 ` [RFA 03/10] Make copy_token_string return unique_xmalloc_ptr Tom Tromey
2018-04-01 16:35 ` [RFA 07/10] Change streq to return bool Tom Tromey
2018-04-02 2:28 ` Simon Marchi
2018-04-02 2:44 ` [RFA 00/10] Remove some cleanups from linespec.c Simon Marchi
2018-04-03 22:31 ` 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=8f38f8d0-fe2e-7023-dd57-7a32af834e4a@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).