From: Stefan Liebler <stli@linux.ibm.com>
To: libc-alpha@sourceware.org
Subject: Re: [PATCH v2 1/4] s390x: Add the clone3 wrapper
Date: Wed, 12 Jul 2023 16:51:54 +0200 [thread overview]
Message-ID: <72a387fa-b08e-221d-6073-c623dd8c569e@linux.ibm.com> (raw)
In-Reply-To: <20230710184022.3701346-2-adhemerval.zanella@linaro.org>
Hi Adhemerval,
I have applied your patches and tested it on s390x. I've ensured that
the wrapper is now used for pthread_create/posix_spawn and the testsuite
runs fine. Nevertheless I have some minor comments: see below.
And one more question:
Is there a reason, why the wrapper is only added to s390x (64bit), but
not for s390 (31bit)? Should there be an #undef __ASSUME_CLONE3?
According to the kernel commit it is also available for compat-mode:
"s390: wire up clone3 system call"
https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git/commit/?id=5518aed82d2abd97f8d3ec91d8ba455d939e8cd1
On 10.07.23 20:40, Adhemerval Zanella via Libc-alpha wrote:
> It follows the internal signature:
>
> extern int clone3 (struct clone_args *__cl_args, size_t __size,
> int (*__func) (void *__arg), void *__arg);
>
> Checked on s390x-linux-gnu.
> ---
> sysdeps/unix/sysv/linux/s390/s390-64/clone3.S | 81 +++++++++++++++++++
> sysdeps/unix/sysv/linux/s390/sysdep.h | 1 +
> 2 files changed, 82 insertions(+)
> create mode 100644 sysdeps/unix/sysv/linux/s390/s390-64/clone3.S
>
> diff --git a/sysdeps/unix/sysv/linux/s390/s390-64/clone3.S b/sysdeps/unix/sysv/linux/s390/s390-64/clone3.S
> new file mode 100644
> index 0000000000..417fdcb785
> --- /dev/null
> +++ b/sysdeps/unix/sysv/linux/s390/s390-64/clone3.S
> @@ -0,0 +1,81 @@
> +/* The clone3 syscall wrapper. Linux/s390x version.
> + Copyright (C) 2023 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/>. */
> +
> +#include <sysdep.h>
> +#define _ERRNO_H 1
> +#include <bits/errno.h>
> +
> +/* The userland implementation is:
> + int clone3 (struct clone_args *cl_args, size_t size,
> + int (*func)(void *arg), void *arg);
> +
> + the kernel entry is:
> + int clone3 (struct clone_args *cl_args, size_t size);
> +
> + The parameters are passed in registers from userland:
> + r2: cl_args
> + r3: size
> + r4: func
> + r5: arg */
> +
> + .text
> +ENTRY(__clone3)
> + stg %r6, 48(%r15)
Add cfi_offset (%r6,-112)
> +
> + /* Sanity check args. */
> + ltgr %r2, %r2
> + je error
> + ltgr %r6, %r4
> + je error
> +
> + /* Do the system call, the kernel expects:
> + r1: system call number
> + r2: cl_args
> + r3: size */
> + lghi %r1, SYS_ify(clone3)
> + svc 0
> + ltgr %r2,%r2 /* check return code */
> + jz thread_start
> + lg %r6, 48(%r15)
> + jgm SYSCALL_ERROR_LABEL
> + br %r14
> +error:
> + lghi %r2,-EINVAL
> + jg SYSCALL_ERROR_LABEL
> +PSEUDO_END (__clone3)
> +
> + .align 4
The ENTRY-macro from sysdep.h is using an alignment of 16 bytes. Can you
please also use it here?
> + .type thread_start, %function
> +thread_start:
> + cfi_startproc
> + /* Mark r14 as undefined in order to stop unwinding here. */
> + cfi_undefined (r14)
> +
> + /* func is in gpr 6, arg in gpr 5. */
> + lgr %r2, %r5
> + aghi %r15, -160
> + xc 0(8,%r15),0(%r15)
> + basr %r14, %r6
> +
> + DO_CALL (exit, 1)
> + cfi_endproc
> + .size thread_start, .-thread_start
If you want, you can use ASM_SIZE_DIRECTIVE (thread_start) from sysdep.h
here, which is also used in END-macros.
> +
> +libc_hidden_def (__clone3)
> +weak_alias (__clone3, clone3)
> diff --git a/sysdeps/unix/sysv/linux/s390/sysdep.h b/sysdeps/unix/sysv/linux/s390/sysdep.h
> index 3bde8ba19c..b099fea9c6 100644
> --- a/sysdeps/unix/sysv/linux/s390/sysdep.h
> +++ b/sysdeps/unix/sysv/linux/s390/sysdep.h
> @@ -72,6 +72,7 @@
> #ifdef __s390x__
> #define HAVE_CLOCK_GETRES64_VSYSCALL "__kernel_clock_getres"
> #define HAVE_CLOCK_GETTIME64_VSYSCALL "__kernel_clock_gettime"
> +#define HAVE_CLONE3_WRAPPER 1
> #else
> #define HAVE_CLOCK_GETRES_VSYSCALL "__kernel_clock_getres"
> #define HAVE_CLOCK_GETTIME_VSYSCALL "__kernel_clock_gettime"
Thanks,
Stefan
next prev parent reply other threads:[~2023-07-12 14:52 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-07-10 18:40 [PATCH v2 0/4] Add clone3 support for multiple architectures Adhemerval Zanella
2023-07-10 18:40 ` [PATCH v2 1/4] s390x: Add the clone3 wrapper Adhemerval Zanella
2023-07-12 14:51 ` Stefan Liebler [this message]
2023-07-12 18:08 ` Adhemerval Zanella Netto
2023-07-13 9:30 ` Stefan Liebler
2023-07-13 11:52 ` Adhemerval Zanella Netto
2023-07-10 18:40 ` [PATCH v2 2/4] arm: " Adhemerval Zanella
2023-07-10 18:40 ` [PATCH v2 3/4] mips: " Adhemerval Zanella
2023-07-10 18:40 ` [PATCH v2 4/4] linux: Undef __ASSUME_CLONE3 for alpha, ia64, nios2, sh, and sparc Adhemerval Zanella
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=72a387fa-b08e-221d-6073-c623dd8c569e@linux.ibm.com \
--to=stli@linux.ibm.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).