From: Palmer Dabbelt <palmer@dabbelt.com>
To: vincent.chen@sifive.com
Cc: libc-alpha@sourceware.org, Darius Rad <darius@bluespec.com>,
Andrew Waterman <andrew@sifive.com>,
dj@redhat.com, kito.cheng@sifive.com, greentime.hu@sifive.com,
vincent.chen@sifive.com
Subject: Re: [PATCH v2 1/2] RISC-V: remove riscv-specific sigcontext.h
Date: Wed, 19 Jan 2022 18:36:29 -0800 (PST) [thread overview]
Message-ID: <mhng-28d9eb84-d94a-4f2c-8863-9dd6512cb387@palmer-ri-x1c9> (raw)
In-Reply-To: <20220118043159.27521-2-vincent.chen@sifive.com>
On Mon, 17 Jan 2022 20:31:58 PST (-0800), vincent.chen@sifive.com wrote:
> Remove riscv-specific sigcontext.h so that Glibc can directly use
> sigcontext.h provided by the kernel to reduce synchronization work
> when new extension support is introduced.
> ---
> .../unix/sysv/linux/riscv/bits/sigcontext.h | 31 -------------------
> 1 file changed, 31 deletions(-)
> delete mode 100644 sysdeps/unix/sysv/linux/riscv/bits/sigcontext.h
>
> diff --git a/sysdeps/unix/sysv/linux/riscv/bits/sigcontext.h b/sysdeps/unix/sysv/linux/riscv/bits/sigcontext.h
> deleted file mode 100644
> index b6e15b5f62..0000000000
> --- a/sysdeps/unix/sysv/linux/riscv/bits/sigcontext.h
> +++ /dev/null
> @@ -1,31 +0,0 @@
> -/* Machine-dependent signal context structure for Linux. RISC-V version.
> - Copyright (C) 1996-2022 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 _BITS_SIGCONTEXT_H
> -#define _BITS_SIGCONTEXT_H 1
> -
> -#if !defined _SIGNAL_H && !defined _SYS_UCONTEXT_H
> -# error "Never use <bits/sigcontext.h> directly; include <signal.h> instead."
> -#endif
> -
> -struct sigcontext {
> - /* gregs[0] holds the program counter. */
> - unsigned long int gregs[32];
> - unsigned long long int fpregs[66] __attribute__ ((__aligned__ (16)));
> -};
> -
> -#endif
This will definitely break API compatibility (the fields have
different names) but should be fine for ABI compatibility. IIUC that's
within the rules, but I'm not sure it's a desirable outcome. Probably
would have been better to get this right the first time around, but I'm
not sure it's worth fixing -- essentially we're making a bunch of users
change things so we don't have to. That said, it's pretty ugly to have
two different definitions of a structure with the same name and layout.
Maybe there's some sort of macro-related trick we can use? ie, provide
the current definition unless users opt into the Linux one (presumably
so they can talk about the V state). There's going to be some hoops to
jump through there to maintain ABI compatibility either way, so it's
possible we could tie these two together?
next prev parent reply other threads:[~2022-01-20 2:36 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-01-18 4:31 [PATCH v2 0/2] RISC-V: Add vector ISA support Vincent Chen
2022-01-18 4:31 ` [PATCH v2 1/2] RISC-V: remove riscv-specific sigcontext.h Vincent Chen
2022-01-20 2:36 ` Palmer Dabbelt [this message]
2022-01-20 2:47 ` Kito Cheng
2022-01-21 1:29 ` Vincent Chen
2022-01-24 9:42 ` Vincent Chen
2022-02-24 20:56 ` Palmer Dabbelt
2022-02-25 0:32 ` Vincent Chen
2022-01-18 4:31 ` [PATCH v2 2/2] riscv: Resolve symbols directly for symbols with STO_RISCV_VARIANT_CC Vincent Chen
2022-01-20 2:21 ` Palmer Dabbelt
2022-01-20 2:38 ` H.J. Lu
2022-01-20 2:43 ` Palmer Dabbelt
2022-01-21 1:43 ` Vincent Chen
2022-02-24 20:56 ` Palmer Dabbelt
2022-12-09 4:11 ` Vineet Gupta
2022-12-09 4:22 ` Kito Cheng
2022-12-09 4:26 ` Vineet Gupta
2022-12-09 4:35 ` Kito Cheng
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=mhng-28d9eb84-d94a-4f2c-8863-9dd6512cb387@palmer-ri-x1c9 \
--to=palmer@dabbelt.com \
--cc=andrew@sifive.com \
--cc=darius@bluespec.com \
--cc=dj@redhat.com \
--cc=greentime.hu@sifive.com \
--cc=kito.cheng@sifive.com \
--cc=libc-alpha@sourceware.org \
--cc=vincent.chen@sifive.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).