public inbox for libc-alpha@sourceware.org
 help / color / mirror / Atom feed
From: "H.J. Lu" <hjl.tools@gmail.com>
To: Alan Modra <amodra@gmail.com>
Cc: GNU C Library <libc-alpha@sourceware.org>,
	 Tulio Magno Quites Machado Filho <tuliom@linux.ibm.com>
Subject: Re: [PATCH 3/5] [PowerPC] Relocate stinfo->main
Date: Sun, 23 Jan 2022 20:48:44 -0800	[thread overview]
Message-ID: <CAMe9rOpa2WVZuawbNTJw4XkeGe9FTLM-wgm5dEJF-fvrQNKUkA@mail.gmail.com> (raw)
In-Reply-To: <Ye4iHPb/sTm517LH@squeak.grove.modra.org>

On Sun, Jan 23, 2022 at 7:51 PM Alan Modra via Libc-alpha
<libc-alpha@sourceware.org> wrote:
>
> The previous 2 patches got us to the point of segfaulting when trying
> to transfer to main.
>
> start_addresses in sysdeps/powerpc/powerpc64/start.S is historical
> baggage that should disappear.  Until someone does that, relocating
> stinfo->main by hand is one solution to the fact that the field may be
> unrelocated at the time it is accessed.  This is similar to what is
> done for dynamic tags via the D_PTR macro.  The makefile change is
> needed to pick up elf/dl-static-tls.h from dl-machine.h.
>
> diff --git a/sysdeps/unix/sysv/linux/powerpc/Makefile b/sysdeps/unix/sysv/linux/powerpc/Makefile
> index fc7c29c695..93783cae00 100644
> --- a/sysdeps/unix/sysv/linux/powerpc/Makefile
> +++ b/sysdeps/unix/sysv/linux/powerpc/Makefile
> @@ -27,3 +27,9 @@ tests += $(tests-static)
>  tests += test-gettimebasefreq
>  tests += test-powerpc-linux-sysconf
>  endif
> +
> +ifeq ($(subdir),csu)
> +# to relocate stinfo->main
> +CPPFLAGS-libc-start.o += -I../elf
> +CPPFLAGS-libc-start.op += -I../elf
> +endif
> diff --git a/sysdeps/unix/sysv/linux/powerpc/libc-start.c b/sysdeps/unix/sysv/linux/powerpc/libc-start.c
> index bb97d16145..8f0d3a0eac 100644
> --- a/sysdeps/unix/sysv/linux/powerpc/libc-start.c
> +++ b/sysdeps/unix/sysv/linux/powerpc/libc-start.c
> @@ -23,6 +23,10 @@
>  #ifndef SHARED
>  #include <hwcapinfo.h>
>  #endif
> +#if ENABLE_STATIC_PIE && !defined SHARED
> +/* For elf_machine_load_address.  */
> +#include <dl-machine.h>
> +#endif
>
>  /* The main work is done in the generic function.  */
>  #define LIBC_START_MAIN generic_start_main
> @@ -95,8 +99,15 @@ __libc_start_main_impl (int argc, char **argv,
>    __tcb_parse_hwcap_and_convert_at_platform ();
>  #endif
>
> -  return generic_start_main (stinfo->main, argc, argv, auxvec,
> -                            stinfo->init, stinfo->fini, rtld_fini,
> +  void *main = stinfo->main;
> +#if ENABLE_STATIC_PIE && !defined SHARED
> +  struct link_map *map = _dl_get_dl_main_map ();
> +  if (!map->l_relocated)
> +    main = (char *) main + elf_machine_load_address ();
> +#endif

Can you use __ehdr_start t here, similar to

https://gitlab.com/x86-glibc/glibc/-/commit/1df93c500d0f81d30d1b7e323153512a8768bb72#ffb950808032d5085fdf882753d7bf8889e6a3d7


> +  return generic_start_main (main, argc, argv, auxvec,
> +                            NULL, NULL, rtld_fini,
>                              stack_on_entry);
>  }
>  DEFINE_LIBC_START_MAIN_VERSION
>
> --
> Alan Modra
> Australia Development Lab, IBM



--
H.J.

  reply	other threads:[~2022-01-24  4:49 UTC|newest]

Thread overview: 35+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-01-23 12:42 [PowerPC64] Use medium model toc accesses throughout Alan Modra
2022-01-24  3:47 ` [PATCH 2/5] [PowerPC64] Set up thread register for _dl_relocate_static_pie Alan Modra
2022-02-16 23:02   ` Paul E Murphy
2022-02-19  0:49     ` Alan Modra
2022-01-24  3:50 ` [PATCH 3/5] [PowerPC] Relocate stinfo->main Alan Modra
2022-01-24  4:48   ` H.J. Lu [this message]
2022-01-24  6:51     ` Alan Modra
2022-01-24  3:52 ` [PATCH 4/5] Constify a variable in dl_vdso_vsym Alan Modra
2022-01-24  4:06 ` [PATCH 5/5] Enable static-pie on powerpc64 Alan Modra
2022-01-27 18:39   ` Florian Weimer
2022-01-28  8:45     ` Alan Modra
2022-01-28 17:48 ` [PowerPC64] Use medium model toc accesses throughout Paul E Murphy
2022-01-29  1:24   ` Alan Modra
2022-02-28  6:40 ` [PATCH v2 0/4] PowerPC64 static-pie Alan Modra
2022-02-28  6:40   ` [PATCH v2 1/4] powerpc64: Use medium model toc accesses throughout Alan Modra
2022-04-08 22:28     ` Tulio Magno Quites Machado Filho
2022-02-28  6:40   ` [PATCH v2 2/4] powerpc64: Set up thread register for _dl_relocate_static_pie Alan Modra
2022-04-08 22:28     ` Tulio Magno Quites Machado Filho
2022-02-28  6:40   ` [PATCH v2 3/4] powerpc: Relocate stinfo->main Alan Modra
2022-04-08 22:32     ` Tulio Magno Quites Machado Filho
2022-02-28  6:40   ` [PATCH v2 4/4] powerpc64: Enable static-pie Alan Modra
2022-04-08 22:49     ` Tulio Magno Quites Machado Filho
2022-04-14  1:16       ` Alan Modra
2022-04-14  3:42         ` Fangrui Song
2022-04-20  7:21         ` [PATCH v3] " Alan Modra
2022-03-04 12:48   ` [PATCH v2 0/4] PowerPC64 static-pie Alan Modra
2022-04-08  8:06     ` Alan Modra
2022-04-09  0:14       ` Fangrui Song
2022-04-14  0:33         ` Alan Modra
2022-04-14  1:54           ` DT_RELR without libc.so dependency H.J. Lu
2022-04-14  3:43             ` Fangrui Song
2022-04-14  5:18             ` Alan Modra
2022-04-14 17:55               ` H.J. Lu
2022-04-08 22:27   ` [PATCH v2 0/4] PowerPC64 static-pie Tulio Magno Quites Machado Filho
2022-04-11  1:38     ` Alan Modra

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=CAMe9rOpa2WVZuawbNTJw4XkeGe9FTLM-wgm5dEJF-fvrQNKUkA@mail.gmail.com \
    --to=hjl.tools@gmail.com \
    --cc=amodra@gmail.com \
    --cc=libc-alpha@sourceware.org \
    --cc=tuliom@linux.ibm.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).