From: Ilya Leoshkevich <iii@linux.ibm.com>
To: Tom Tromey <tom@tromey.com>
Cc: gdb-patches@sourceware.org, Andreas Arnez <arnez@linux.ibm.com>,
Ilya Leoshkevich <iii@linux.ibm.com>
Subject: [PATCH] gdb: Fix derive_heap_segment for PIE
Date: Tue, 9 May 2023 13:24:01 +0200 [thread overview]
Message-ID: <20230509112401.1043723-1-iii@linux.ibm.com> (raw)
Using gcore with qemu-user gdbstub produces huge core dumps:
$ qemu-s390x-static -g 1234 /bin/true
$ gdb -ex 'target remote :1234' /bin/true
(gdb) generate-core-file
Saved corefile core.1
$ wc -c <core.1
274878128632
The reason is that derive_heap_segment thinks that heap starts quite
close to 0, because it works with bfd objects, which are not aware of
actual PIE load addresses.
Fix by using objfiles instead.
---
gdb/gcore.c | 31 ++++++++++++++-----------------
1 file changed, 14 insertions(+), 17 deletions(-)
diff --git a/gdb/gcore.c b/gdb/gcore.c
index 05cad94526e..d044c887c22 100644
--- a/gdb/gcore.c
+++ b/gdb/gcore.c
@@ -283,18 +283,15 @@ call_target_sbrk (int sbrk_arg)
return top_of_heap;
}
-/* Derive a reasonable heap segment for ABFD by looking at sbrk and
- the static data sections. Store its limits in *BOTTOM and *TOP.
+/* Derive a reasonable heap segment by looking at sbrk and the executable's
+ static data sections. Store its limits in *BOTTOM and *TOP.
Return non-zero if successful. */
static int
-derive_heap_segment (bfd *abfd, bfd_vma *bottom, bfd_vma *top)
+derive_heap_segment (bfd_vma *bottom, bfd_vma *top)
{
bfd_vma top_of_data_memory = 0;
bfd_vma top_of_heap = 0;
- bfd_size_type sec_size;
- bfd_vma sec_vaddr;
- asection *sec;
gdb_assert (bottom);
gdb_assert (top);
@@ -315,17 +312,18 @@ derive_heap_segment (bfd *abfd, bfd_vma *bottom, bfd_vma *top)
| heap |
--------------------------------- */
- for (sec = abfd->sections; sec; sec = sec->next)
- {
- if (bfd_section_flags (sec) & SEC_DATA
- || strcmp (".bss", bfd_section_name (sec)) == 0)
+ for (objfile *objfile : current_program_space->objfiles ())
+ if (objfile->obfd == current_program_space->exec_bfd ())
+ for (obj_section *objsec : objfile->sections ())
{
- sec_vaddr = bfd_section_vma (sec);
- sec_size = bfd_section_size (sec);
- if (sec_vaddr + sec_size > top_of_data_memory)
- top_of_data_memory = sec_vaddr + sec_size;
+ asection *sec = objsec->the_bfd_section;
+ if (bfd_section_flags (sec) & SEC_DATA
+ || strcmp (".bss", bfd_section_name (sec)) == 0)
+ {
+ if (objsec->endaddr () > top_of_data_memory)
+ top_of_data_memory = objsec->endaddr ();
+ }
}
- }
top_of_heap = call_target_sbrk (0);
if (top_of_heap == (bfd_vma) 0)
@@ -553,8 +551,7 @@ objfile_find_memory_regions (struct target_ops *self,
obfd);
/* Make a heap segment. */
- if (derive_heap_segment (current_program_space->exec_bfd (), &temp_bottom,
- &temp_top))
+ if (derive_heap_segment (&temp_bottom, &temp_top))
(*func) (temp_bottom, temp_top - temp_bottom,
1, /* Heap section will be readable. */
1, /* Heap section will be writable. */
--
2.40.1
next reply other threads:[~2023-05-09 11:24 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-05-09 11:24 Ilya Leoshkevich [this message]
2023-06-21 22:13 ` [PING] " Ilya Leoshkevich
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=20230509112401.1043723-1-iii@linux.ibm.com \
--to=iii@linux.ibm.com \
--cc=arnez@linux.ibm.com \
--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).