From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <sourceware-bugzilla@sourceware.org>
Received: by sourceware.org (Postfix, from userid 48)
	id 22BDF3858CDB; Tue,  1 Nov 2022 20:42:07 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 22BDF3858CDB
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=sourceware.org;
	s=default; t=1667335341;
	bh=0zcqTyAh5/p8MMj9yUeFRaPcHx4b8iCms6eIW8RRgb8=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=SVBIRPR7ee/z23zYNz+av3PTGgPazD756NzTGD+pAukdm5z9pIktYg98OlL5E5GLe
	 hIL12pn7qVfqUXQ8grRhb9u0V4ct3+YMeoTLomnB5vxQSy4fzkjYcyMSMRG7ZhwjET
	 XZIzgXNEpNjDeAHdLn70c8bftoXs6upcym8zO5lA=
From: "tromey at sourceware dot org" <sourceware-bugzilla@sourceware.org>
To: gdb-prs@sourceware.org
Subject: [Bug symtab/29105] new DWARF reader still slow
Date: Tue, 01 Nov 2022 20:42:05 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gdb
X-Bugzilla-Component: symtab
X-Bugzilla-Version: HEAD
X-Bugzilla-Keywords: 
X-Bugzilla-Severity: normal
X-Bugzilla-Who: tromey at sourceware dot org
X-Bugzilla-Status: NEW
X-Bugzilla-Resolution: 
X-Bugzilla-Priority: P2
X-Bugzilla-Assigned-To: tromey at sourceware dot org
X-Bugzilla-Target-Milestone: 13.1
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-29105-4717-xoIIP3AX2f@http.sourceware.org/bugzilla/>
In-Reply-To: <bug-29105-4717@http.sourceware.org/bugzilla/>
References: <bug-29105-4717@http.sourceware.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
X-Bugzilla-URL: http://sourceware.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
List-Id: <gdb-prs.sourceware.org>

https://sourceware.org/bugzilla/show_bug.cgi?id=3D29105

--- Comment #15 from Tom Tromey <tromey at sourceware dot org> ---
I've been debugging this a bit.

For the sizeof test, gdb makes an index entry for "short int".
This comes directly from the DWARF:

 <1><bc>: Abbrev Number: 2 (DW_TAG_base_type)
    <bd>   DW_AT_byte_size   : 2
    <be>   DW_AT_encoding    : 5        (signed)
    <bf>   DW_AT_name        : (indirect string, offset: 0x120): short int

lookup_signed_typename first looks for "signed short" though.

(top-gdb) up
#15 0x00000000007bb408 in lookup_typename (language=3D0x24c0da0
<c_language_defn>, name=3D0x7fffffffab50 "signed short", block=3D0x0,=20
    noerr=3D1) at ../../binutils-gdb/gdb/gdbtypes.c:1700
(top-gdb) up
#16 0x00000000007bb5ac in lookup_signed_typename (language=3D0x24c0da0
<c_language_defn>, name=3D0x12528f8 "short")
    at ../../binutils-gdb/gdb/gdbtypes.c:1729


Now, trying this with an older gdb shows that psyms also make these
same symbol names:

(gdb) maint print psymbols=20
[...]
    `short int', type, 0x0

I'm not sure why it works in this case and not in the new code.

I added a 'maint canonicalize' command and it shows:

(gdb) maint canonicalize short
No change.
(gdb) maint canonicalize short int
canonical =3D short
(gdb) maint canonicalize signed short int
canonical =3D short
(gdb) maint canonicalize signed short=20
canonical =3D short

That is, one oddity here is that the C++ canonical form for these names
differs from what's in the DWARF and what things like
lookup_signed_typename expect.

Next I'm going to look at canonicalization for C, at least for those
types that have multiple spellings.

--=20
You are receiving this mail because:
You are on the CC list for the bug.=