From: Jason Merrill <jason@redhat.com>
To: "Martin Liška" <mliska@suse.cz>, gcc-patches@gcc.gnu.org
Cc: Richard Biener <richard.guenther@gmail.com>
Subject: Re: [PATCH] use proper DECL_INITIAL for VTV
Date: Thu, 13 Oct 2022 10:46:40 -0400 [thread overview]
Message-ID: <902f2778-ac8d-7b8a-0aeb-eb609b90729a@redhat.com> (raw)
In-Reply-To: <79ffd1f4-684e-dead-9d77-f1567acbc1d8@suse.cz>
On 10/13/22 10:25, Martin Liška wrote:
> Hi.
>
> I am working on the early debug info emission that would benefit from a late
> use of asm_put_file. This is last blocker where C++ emits early a section directive
> in assemble_vtv_preinit_initializer. We can use a proper DECL_INITIAL for that.
>
> Patch can bootstrap on x86_64-linux-gnu and survives regression tests.
> with --enable-vtable-verify.
>
> Ready to be installed?
OK.
> gcc/cp/ChangeLog:
>
> * vtable-class-hierarchy.cc (vtv_generate_init_routine): Emit
> an artificial variable that would be put into .preinit_array
> section.
>
> gcc/ChangeLog:
>
> * output.h (assemble_vtv_preinit_initializer): Remove.
> * varasm.cc (assemble_vtv_preinit_initializer): Remove.
> ---
> gcc/cp/vtable-class-hierarchy.cc | 14 ++++++++++++--
> gcc/output.h | 4 ----
> gcc/varasm.cc | 17 -----------------
> 3 files changed, 12 insertions(+), 23 deletions(-)
>
> diff --git a/gcc/cp/vtable-class-hierarchy.cc b/gcc/cp/vtable-class-hierarchy.cc
> index 79cb5f8de02..cc1df1ebdb2 100644
> --- a/gcc/cp/vtable-class-hierarchy.cc
> +++ b/gcc/cp/vtable-class-hierarchy.cc
> @@ -1192,8 +1192,18 @@ vtv_generate_init_routine (void)
> cgraph_node::add_new_function (vtv_fndecl, false);
>
> if (flag_vtable_verify == VTV_PREINIT_PRIORITY && !TARGET_PECOFF)
> - assemble_vtv_preinit_initializer (vtv_fndecl);
> -
> + {
> + tree vtv_var
> + = build_decl (BUILTINS_LOCATION, VAR_DECL,
> + get_identifier ("__vtv_preinit"),
> + build_pointer_type (TREE_TYPE (vtv_fndecl)));
> + TREE_STATIC (vtv_var) = 1;
> + DECL_ARTIFICIAL (vtv_var) = 1;
> + DECL_INITIAL (vtv_var) = build_fold_addr_expr (vtv_fndecl);
> + set_decl_section_name (vtv_var, ".preinit_array");
> +
> + varpool_node::add (vtv_var);
> + }
> }
> pop_lang_context ();
> }
> diff --git a/gcc/output.h b/gcc/output.h
> index 6dea630913a..6936bdeeb6c 100644
> --- a/gcc/output.h
> +++ b/gcc/output.h
> @@ -199,10 +199,6 @@ extern void assemble_end_function (tree, const char *);
> initial value (that will be done by the caller). */
> extern void assemble_variable (tree, int, int, int);
>
> -/* Put the vtable verification constructor initialization function
> - into the preinit array. */
> -extern void assemble_vtv_preinit_initializer (tree);
> -
> /* Assemble everything that is needed for a variable declaration that has
> no definition in the current translation unit. */
> extern void assemble_undefined_decl (tree);
> diff --git a/gcc/varasm.cc b/gcc/varasm.cc
> index 423f3f91af8..a11184584a2 100644
> --- a/gcc/varasm.cc
> +++ b/gcc/varasm.cc
> @@ -2419,23 +2419,6 @@ assemble_variable (tree decl, int top_level ATTRIBUTE_UNUSED,
> }
> }
>
> -
> -/* Given a function declaration (FN_DECL), this function assembles the
> - function into the .preinit_array section. */
> -
> -void
> -assemble_vtv_preinit_initializer (tree fn_decl)
> -{
> - section *sect;
> - unsigned flags = SECTION_WRITE;
> - rtx symbol = XEXP (DECL_RTL (fn_decl), 0);
> -
> - flags |= SECTION_NOTYPE;
> - sect = get_section (".preinit_array", flags, fn_decl);
> - switch_to_section (sect);
> - assemble_addr_to_section (symbol, sect);
> -}
> -
> /* Return 1 if type TYPE contains any pointers. */
>
> static int
prev parent reply other threads:[~2022-10-13 14:46 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-10-13 14:25 Martin Liška
2022-10-13 14:46 ` Jason Merrill [this message]
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=902f2778-ac8d-7b8a-0aeb-eb609b90729a@redhat.com \
--to=jason@redhat.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=mliska@suse.cz \
--cc=richard.guenther@gmail.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).