From: Simon Marchi <simon.marchi@polymtl.ca>
To: Tom Tromey <tom@tromey.com>,
Pedro Alves via Gdb-patches <gdb-patches@sourceware.org>
Cc: Pedro Alves <palves@redhat.com>,
Simon Marchi <simon.marchi@efficios.com>,
Jon Turney <jon.turney@dronecode.org.uk>
Subject: [PATCH] gdb: use bfd_get_section_contents to read section contents in, is_linked_with_cygwin_dll (was: Re: [PATCH 6/7] gdb: select "Cygwin" OS ABI for Cygwin binaries)
Date: Thu, 2 Apr 2020 10:55:06 -0400 [thread overview]
Message-ID: <6b6b7467-2db4-56c9-dd98-3082b7b68abe@polymtl.ca> (raw)
In-Reply-To: <87ftdmvx85.fsf@tromey.com>
On 2020-04-02 9:22 a.m., Tom Tromey wrote:
> There's no way to ever release this data. Since it's only used once, it
> may be better to use bfd_get_full_section_contents, so the memory can be
> freed when done.
Good point, I hadn't thought of that.
Here's a patch. I only tested it on GNU/Linux, making sure GDB is still able
to recognize a mingw binary and a cygwin binary.
From b7f4f941018c7b4a7431837ccf5eee5c992f4111 Mon Sep 17 00:00:00 2001
From: Simon Marchi <simon.marchi@polymtl.ca>
Date: Thu, 2 Apr 2020 10:34:39 -0400
Subject: [PATCH] gdb: use bfd_get_section_contents to read section contents in
is_linked_with_cygwin_dll
The function is_linked_with_cygwin_dll currently uses
gdb_bfd_map_section to get some section contents. This is not ideal
because that memory, which is only used in this function, can't be
released. Instead, it was suggested to use bfd_get_section_contents,
and this is what this patch does.
I decided to make a small bfd_get_section_contents wrapper in gdb_bfd.c,
which returns the contents in a gdb::byte_vector. The wrapper also
makes it easy to get the full section contents into a gdb::byte_vector.
Note that BFD provides the bfd_get_full_section_contents function, but
this one returns a newly-allocated buffer. gdb_bfd_get_section_contents
offers a consistent interface, regardless of whether you need to read
part of a section or the full section.
gdb_bfd_get_section_contents could be used at many places that already
allocate a vector of the size of the section and then call
bfd_get_section_contents. I think these call sites can be updated over
time.
gdb/ChangeLog:
* gdb_bfd.h: Include gdbsupport/byte-vector.h and
gdbsupport/gdb_optional.h.
(BFD_SIZE_TYPE_MAX): New macro.
(gdb_bfd_get_section_contents): New declaration.
* gdb_bfd.c (gdb_bfd_get_section_contents): New function.
* windows-tdep.c (is_linked_with_cygwin_dll): Use
gdb_bfd_get_section_contents.
---
gdb/gdb_bfd.c | 25 ++++++++++++++++++++++++-
gdb/gdb_bfd.h | 15 +++++++++++++++
gdb/windows-tdep.c | 13 ++++++-------
3 files changed, 45 insertions(+), 8 deletions(-)
diff --git a/gdb/gdb_bfd.c b/gdb/gdb_bfd.c
index 5a6dee2d51a8..2b34e0f3e17c 100644
--- a/gdb/gdb_bfd.c
+++ b/gdb/gdb_bfd.c
@@ -926,7 +926,30 @@ gdb_bfd_requires_relocations (bfd *abfd)
return gdata->needs_relocations;
}
-\f
+/* See gdb_bfd.h. */
+
+bool
+gdb_bfd_get_section_contents (bfd *abfd, asection *section,
+ gdb::byte_vector *contents, file_ptr offset,
+ bfd_size_type count)
+{
+ bfd_size_type section_size = bfd_section_size (section);
+
+ /* Allow `offset == section_size`, to allow the corner case of
+ `offset == section_size`, `count = 0`. */
+ gdb_assert (offset <= section_size);
+
+ /* If COUNT is unspecified, get the contents from OFFSET until the end of the
+ section. */
+ if (count == BFD_SIZE_TYPE_MAX)
+ count = section_size - offset;
+
+ gdb_assert ((offset + count) <= section_size);
+
+ contents->resize (count);
+
+ return bfd_get_section_contents (abfd, section, contents->data (), offset, count);
+}
/* A callback for htab_traverse that prints a single BFD. */
diff --git a/gdb/gdb_bfd.h b/gdb/gdb_bfd.h
index 9b1e292bf18f..ffea4f6b715e 100644
--- a/gdb/gdb_bfd.h
+++ b/gdb/gdb_bfd.h
@@ -21,6 +21,8 @@
#define GDB_BFD_H
#include "registry.h"
+#include "gdbsupport/byte-vector.h"
+#include "gdbsupport/gdb_optional.h"
#include "gdbsupport/gdb_ref_ptr.h"
DECLARE_REGISTRY (bfd);
@@ -30,6 +32,8 @@ DECLARE_REGISTRY (bfd);
#define TARGET_SYSROOT_PREFIX "target:"
+#define BFD_SIZE_TYPE_MAX ((bfd_size_type) -1)
+
/* Returns nonzero if NAME starts with TARGET_SYSROOT_PREFIX, zero
otherwise. */
@@ -181,4 +185,15 @@ int gdb_bfd_count_sections (bfd *abfd);
int gdb_bfd_requires_relocations (bfd *abfd);
+/* Wrapper around bfd_get_section_contents, returning the requested section
+ contents in *CONTENTS. Return true on success, false otherwise.
+
+ If COUNT is not specified, read from OFFSET until the end of the
+ section. */
+
+bool
+gdb_bfd_get_section_contents (bfd *abfd, asection *section,
+ gdb::byte_vector *contents, file_ptr offset = 0,
+ bfd_size_type count = BFD_SIZE_TYPE_MAX);
+
#endif /* GDB_BFD_H */
diff --git a/gdb/windows-tdep.c b/gdb/windows-tdep.c
index 9c5dfd183bfa..9b7dc1d12b26 100644
--- a/gdb/windows-tdep.c
+++ b/gdb/windows-tdep.c
@@ -1005,18 +1005,17 @@ is_linked_with_cygwin_dll (bfd *abfd)
bfd_vma idata_addr
= pe_data (abfd)->pe_opthdr.DataDirectory[PE_IMPORT_TABLE].VirtualAddress;
- /* Map the section's data. */
- bfd_size_type idata_size;
- const gdb_byte *const idata_contents
- = gdb_bfd_map_section (idata_section, &idata_size);
- if (idata_contents == nullptr)
+ /* Get the section's data. */
+ gdb::byte_vector idata_contents;
+ if (!gdb_bfd_get_section_contents(abfd, idata_section, &idata_contents))
{
warning (_("Failed to get content of .idata section."));
return false;
}
- const gdb_byte *iter = idata_contents;
- const gdb_byte *end = idata_contents + idata_size;
+ size_t idata_size = idata_contents.size ();
+ const gdb_byte *iter = idata_contents.data ();
+ const gdb_byte *end = idata_contents.data () + idata_size;
const pe_import_directory_entry null_dir_entry = { 0 };
/* Iterate through all directory entries. */
--
2.26.0
next prev parent reply other threads:[~2020-04-02 14:55 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-03-16 17:08 [PATCH 0/7] Add "Windows" OS ABI Simon Marchi
2020-03-16 17:08 ` [PATCH 1/7] gdb: recognize 64 bits Windows executables as Cygwin osabi Simon Marchi
2020-03-16 17:08 ` [PATCH 2/7] gdb: move enum gdb_osabi to osabi.h Simon Marchi
2020-03-16 17:08 ` [PATCH 3/7] gdb: add Windows OS ABI Simon Marchi
2020-03-16 17:08 ` [PATCH 4/7] gdb: rename i386-cygwin-tdep.c to i386-windows-tdep.c Simon Marchi
2020-03-16 17:08 ` [PATCH 5/7] gdb: rename content of i386-windows-tdep.c, cygwin to windows Simon Marchi
2020-03-16 17:08 ` [PATCH 6/7] gdb: select "Cygwin" OS ABI for Cygwin binaries Simon Marchi
2020-03-16 18:16 ` Christian Biesinger
2020-03-16 18:18 ` Simon Marchi
2020-03-16 19:03 ` Jon Turney
2020-03-16 21:00 ` Simon Marchi
2020-04-01 19:05 ` Tom Tromey
2020-04-01 19:25 ` Simon Marchi
2020-04-01 21:36 ` Pedro Alves
2020-04-01 21:53 ` Simon Marchi
2020-04-02 13:56 ` Pedro Alves
2020-04-02 14:01 ` Simon Marchi
2020-04-02 14:03 ` Pedro Alves
2020-04-02 14:08 ` Simon Marchi
2020-04-02 14:17 ` Simon Marchi
2020-04-02 13:22 ` Tom Tromey
2020-04-02 14:55 ` Simon Marchi [this message]
2020-04-02 14:57 ` [PATCH] gdb: use bfd_get_section_contents to read section contents in, is_linked_with_cygwin_dll Simon Marchi
2020-04-02 19:01 ` Tom Tromey
2020-04-02 19:42 ` Simon Marchi
2020-04-02 19:45 ` Tom Tromey
2020-04-02 19:47 ` Simon Marchi
2020-03-16 17:08 ` [PATCH 7/7] gdb: define builtin long type to be 64 bits on amd64 Cygwin Simon Marchi
2020-03-16 17:46 ` [PATCH 0/7] Add "Windows" OS ABI Eli Zaretskii
2020-03-16 17:48 ` Simon Marchi
2020-03-16 19:04 ` Jon Turney
2020-04-01 21:42 ` Pedro Alves
2020-04-01 21:56 ` Simon Marchi
2020-04-02 3:06 ` [PATCH] gdb: stop using host-dependent signal numbers in, windows-tdep.c (was: Re: [PATCH 0/7] Add "Windows" OS ABI) Simon Marchi
2020-04-02 14:00 ` [PATCH] gdb: stop using host-dependent signal numbers in, windows-tdep.c Pedro Alves
2020-04-02 14:02 ` Simon Marchi
2020-04-02 15:12 ` Eli Zaretskii
2020-04-08 12:45 ` Jon Turney
2020-04-08 18:16 ` Simon Marchi
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=6b6b7467-2db4-56c9-dd98-3082b7b68abe@polymtl.ca \
--to=simon.marchi@polymtl.ca \
--cc=gdb-patches@sourceware.org \
--cc=jon.turney@dronecode.org.uk \
--cc=palves@redhat.com \
--cc=simon.marchi@efficios.com \
--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).