From: Pedro Alves <palves@redhat.com>
To: gdb-patches@sourceware.org
Cc: Binutils <binutils@sourceware.org>
Subject: Re: [PATCH v2 13/15] PPC64: always make synthetic .text symbols for GNU ifunc symbols
Date: Sun, 25 Mar 2018 19:33:00 -0000 [thread overview]
Message-ID: <6844045b-c400-ea4e-430e-37fe7973ebcf@redhat.com> (raw)
In-Reply-To: <20180325191943.8246-14-palves@redhat.com>
Dear binutils friends,
This is a bfd patch that is part of a larger gdb patch series,
whose cover letter is found here:
https://sourceware.org/ml/gdb-patches/2018-03/msg00504.html
Thanks,
Pedro Alves
On 03/25/2018 08:19 PM, Pedro Alves wrote:
> If you create an ifunc using GCC's __attribute__ ifunc, like:
>
> extern int gnu_ifunc (int arg);
> static int gnu_ifunc_target (int arg) { return 0; }
> __typeof (gnu_ifunc) *gnu_ifunc_resolver (unsigned long hwcap) { return gnu_ifunc_target; }
> __typeof (gnu_ifunc) gnu_ifunc __attribute__ ((ifunc ("gnu_ifunc_resolver")));
>
> then you end up with two (function descriptor) symbols, one for the
> ifunc itself, and another for the resolver:
>
> (...)
> 12: 0000000000020060 104 FUNC GLOBAL DEFAULT 18 gnu_ifunc_resolver
> (...)
> 16: 0000000000020060 104 GNU_IFUNC GLOBAL DEFAULT 18 gnu_ifunc
> (...)
>
> Both ifunc and resolver symbols have the same address/value, so
> ppc64_elf_get_synthetic_symtab only creates a synthetic text symbol
> for one of them. In the case above, it ends up being created for the
> resolver, only:
>
> (gdb) maint print msymbols
> (...)
> [ 7] t 0x980 .frame_dummy section .text
> [ 8] T 0x9e4 .gnu_ifunc_resolver section .text
> [ 9] T 0xa58 __glink_PLTresolve section .text
> (...)
>
> GDB needs to know when a program stepped into an ifunc resolver, so
> that it can know whether to step past the resolver into the target
> function without the user noticing. The way GDB does it is my
> checking whether the current PC points to an ifunc symbol (since
> resolver and ifunc have the same address by design).
>
> The problem is then that ppc64_elf_get_synthetic_symtab never creates
> the synchetic symbol for the ifunc, so GDB stops stepping at the
> resolver (in a test added by the following patch):
>
> (gdb) step
> gnu_ifunc_resolver (hwcap=21) at gdb/testsuite/gdb.base/gnu-ifunc-lib.c:33
> 33 {
> (gdb) FAIL: gdb.base/gnu-ifunc.exp: resolver_attr=1: resolver_debug=1: final_debug=0: step
>
> After this commit, we get:
>
> [ 8] i 0x9e4 .gnu_ifunc section .text
> [ 9] T 0x9e4 .gnu_ifunc_resolver section .text
>
> And stepping an ifunc call takes to the final function:
> (gdb) step
> 0x00000000100009e8 in .final ()
> (gdb) PASS: gdb.base/gnu-ifunc.exp: resolver_attr=1: resolver_debug=1: final_debug=0: step
>
> An alternative to touching bfd I considered was for GDB to check
> whether there's an ifunc data symbol / function descriptor that points
> to the current PC, whenever the program stops, but discarded it
> because we'd have to do a linear scan over .opd over an over to find a
> matching function descriptor for the current PC. At that point I
> considered caching that info, but quickly dismissed it as then that
> has no advantage (memory or performance) over just creating the
> synthetic ifunc text symbol in the first place.
>
> I ran the binutils and ld testsuites on PPC64 ELFv1 (machine gcc110 on
> the GCC compile farm), and saw no regressions. This commit is part of
> a GDB patch series that includes GDB tests that fail without this fix.
>
> OK to apply?
>
> bfd/ChangeLog:
> yyyy-mm-dd Pedro Alves <palves@redhat.com>
>
> * elf64-ppc.c (ppc64_elf_get_synthetic_symtab): Don't consider
> ifunc and non-ifunc symbols duplicates.
> ---
> bfd/elf64-ppc.c | 22 ++++++++++++++++------
> 1 file changed, 16 insertions(+), 6 deletions(-)
>
> diff --git a/bfd/elf64-ppc.c b/bfd/elf64-ppc.c
> index 751ad778b26..791ec49b73d 100644
> --- a/bfd/elf64-ppc.c
> +++ b/bfd/elf64-ppc.c
> @@ -3330,13 +3330,23 @@ ppc64_elf_get_synthetic_symtab (bfd *abfd,
>
> if (!relocatable && symcount > 1)
> {
> - /* Trim duplicate syms, since we may have merged the normal and
> - dynamic symbols. Actually, we only care about syms that have
> - different values, so trim any with the same value. */
> + /* Trim duplicate syms, since we may have merged the normal
> + and dynamic symbols. Actually, we only care about syms
> + that have different values, so trim any with the same
> + value. Don't consider ifunc and ifunc resolver symbols
> + duplicates however, because GDB wants to know whether a
> + text symbol is an ifunc resolver. */
> for (i = 1, j = 1; i < symcount; ++i)
> - if (syms[i - 1]->value + syms[i - 1]->section->vma
> - != syms[i]->value + syms[i]->section->vma)
> - syms[j++] = syms[i];
> + {
> + const asymbol *s0 = syms[i - 1];
> + const asymbol *s1 = syms[i];
> +
> + if ((s0->value + s0->section->vma
> + != s1->value + s1->section->vma)
> + || ((s0->flags & BSF_GNU_INDIRECT_FUNCTION)
> + != (s1->flags & BSF_GNU_INDIRECT_FUNCTION)))
> + syms[j++] = syms[i];
> + }
> symcount = j;
> }
>
>
next prev parent reply other threads:[~2018-03-25 19:33 UTC|newest]
Thread overview: 33+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-03-25 19:19 [PATCH v2 00/15] Fixing GNU ifunc support Pedro Alves
2018-03-25 19:19 ` [PATCH v2 12/15] For PPC64/ELFv1: Introduce mst_data_gnu_ifunc Pedro Alves
2018-03-25 19:19 ` [PATCH v2 07/15] Breakpoints, don't skip prologue of ifunc resolvers with debug info Pedro Alves
2018-03-25 19:19 ` [PATCH v2 11/15] Fix stepping past GNU ifunc resolvers (introduce lookup_msym_prefer) Pedro Alves
2018-06-18 20:26 ` [PATCH] Silence GCC "uninitialized" warning on minsyms.c:lookup_minimal_symbol_by_pc_section Sergio Durigan Junior
2018-06-19 15:22 ` Pedro Alves
2018-06-19 16:55 ` Sergio Durigan Junior
2018-06-19 18:47 ` Tom Tromey
2018-03-25 19:19 ` [PATCH v2 05/15] Fix elf_gnu_ifunc_resolve_by_got buglet Pedro Alves
2018-04-01 4:32 ` Simon Marchi
2018-04-10 21:52 ` Pedro Alves
2018-03-25 19:19 ` [PATCH v2 01/15] Fix breakpoints in ifunc after inferior resolved it (@got.plt symbol creation) Pedro Alves
2018-04-01 3:35 ` Simon Marchi
2018-04-10 21:20 ` Pedro Alves
2018-04-14 16:36 ` Simon Marchi
2018-03-25 19:19 ` [PATCH v2 02/15] Fix calling ifunc functions when resolver has debug info and different name Pedro Alves
2018-04-01 3:44 ` Simon Marchi
2018-04-10 21:20 ` Pedro Alves
2018-03-25 19:19 ` [PATCH v2 03/15] Calling ifunc functions when target has no debug info but resolver has Pedro Alves
2018-04-01 4:22 ` Simon Marchi
2018-04-10 21:48 ` Pedro Alves
2018-04-10 21:54 ` Pedro Alves
2018-03-25 19:19 ` [PATCH v2 15/15] Fix resolving GNU ifunc bp locations when inferior runs resolver Pedro Alves
2018-03-25 19:19 ` [PATCH v2 08/15] Eliminate find_pc_partial_function_gnu_ifunc Pedro Alves
2018-03-25 19:25 ` [PATCH v2 09/15] Factor out minsym_found/find_function_start_sal overload Pedro Alves
2018-03-25 19:25 ` [PATCH v2 04/15] Calling ifunc functions when resolver has debug info, user symbol same name Pedro Alves
2018-03-25 19:28 ` [PATCH v2 14/15] Extend GNU ifunc testcases Pedro Alves
2018-03-25 19:29 ` [PATCH v2 13/15] PPC64: always make synthetic .text symbols for GNU ifunc symbols Pedro Alves
2018-03-25 19:33 ` Pedro Alves [this message]
2018-03-26 7:54 ` Alan Modra
2018-03-25 19:29 ` [PATCH v2 10/15] For PPC64: elf_gnu_ifunc_record_cache: handle plt symbols in .text section Pedro Alves
2018-03-25 19:29 ` [PATCH v2 06/15] Fix setting breakpoints on ifunc functions after they're already resolved Pedro Alves
2018-04-26 12:23 ` [PATCH v2 00/15] Fixing GNU ifunc support Pedro Alves
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=6844045b-c400-ea4e-430e-37fe7973ebcf@redhat.com \
--to=palves@redhat.com \
--cc=binutils@sourceware.org \
--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).