From: Florian Weimer <fweimer@redhat.com>
To: Alistair Francis <alistair23@gmail.com>
Cc: GNU C Library <libc-alpha@sourceware.org>
Subject: Re: Generate VDSO_HASH
Date: Tue, 12 Nov 2019 10:47:12 +0100 [thread overview]
Message-ID: <87v9rpqvdr.fsf@oldenburg2.str.redhat.com> (raw)
In-Reply-To: <CAKmqyKPQ5sQHM67ykotRFTsRpG0p2cMgQox7-r+hot22icNXBA@mail.gmail.com> (Alistair Francis's message of "Mon, 11 Nov 2019 15:57:03 -0800")
* Alistair Francis:
> How is VDSO_HASH generated? I would like to generate one for the 5.4
> kernel.
It's the ELF hash. Not to be confused with the GNU hash, which is used
for symbol name lookup these days. See <sysdeps/generic/dl-hash.h> and
the program below.
LINUX_5.4: ELF: 61765876 (0x3ae78f4), GNU: 650524491 (0x26c6374b)
The ELF hash is still used for symbol versions, only the symbol names
switched to the GNU hash.
Thanks,
Florian
/* Compute hash value for given string according to ELF standard.
Copyright (C) 1995-2019 Free Software Foundation, Inc.
This file is part of the GNU C Library.
The GNU C Library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2.1 of the License, or (at your option) any later version.
The GNU C Library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with the GNU C Library; if not, see
<https://www.gnu.org/licenses/>. */
#ifndef _DL_HASH_H
#define _DL_HASH_H 1
/* This is the hashing function specified by the ELF ABI. In the
first five operations no overflow is possible so we optimized it a
bit. */
static unsigned int
__attribute__ ((unused))
_dl_elf_hash (const char *name_arg)
{
const unsigned char *name = (const unsigned char *) name_arg;
unsigned long int hash = *name;
if (hash != 0 && name[1] != '\0')
{
hash = (hash << 4) + name[1];
if (name[2] != '\0')
{
hash = (hash << 4) + name[2];
if (name[3] != '\0')
{
hash = (hash << 4) + name[3];
if (name[4] != '\0')
{
hash = (hash << 4) + name[4];
name += 5;
while (*name != '\0')
{
unsigned long int hi;
hash = (hash << 4) + *name++;
hi = hash & 0xf0000000;
/* The algorithm specified in the ELF ABI is as
follows:
if (hi != 0)
hash ^= hi >> 24;
hash &= ~hi;
But the following is equivalent and a lot
faster, especially on modern processors. */
hash ^= hi >> 24;
}
/* Second part of the modified formula. This
operation can be lifted outside the loop. */
hash &= 0x0fffffff;
}
}
}
}
return hash;
}
#endif /* dl-hash.h */
next prev parent reply other threads:[~2019-11-12 9:47 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-11-11 23:57 Generate VDSO_HASH Alistair Francis
2019-11-12 9:47 ` Florian Weimer [this message]
2019-11-12 17:34 ` Alistair Francis
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
List information: https://www.gnu.org/software/libc/involved.html
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=87v9rpqvdr.fsf@oldenburg2.str.redhat.com \
--to=fweimer@redhat.com \
--cc=alistair23@gmail.com \
--cc=libc-alpha@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).