unofficial mirror of libc-alpha@sourceware.org
 help / color / mirror / Atom feed
From: Szabolcs Nagy via Libc-alpha <libc-alpha@sourceware.org>
To: Naohiro Tamura <naohirot@fujitsu.com>
Cc: Naohiro Tamura <naohirot@jp.fujitsu.com>, libc-alpha@sourceware.org
Subject: Re: [PATCH 4/5] scripts: Added Vector Length Set test helper script
Date: Mon, 29 Mar 2021 14:20:17 +0100	[thread overview]
Message-ID: <20210329132017.GF23289@arm.com> (raw)
In-Reply-To: <20210317023510.323258-1-naohirot@fujitsu.com>

The 03/17/2021 02:35, Naohiro Tamura wrote:
> +"""Set Scalable Vector Length test helper.
> +
> +Set Scalable Vector Length for child process.
> +
> +examples:
> +
> +ubuntu@bionic:~/build$ make check subdirs=string \
> +test-wrapper='~/glibc/scripts/vltest.py 16'
> +
> +ubuntu@bionic:~/build$ ~/glibc/scripts/vltest.py 16 make test \
> +t=string/test-memcpy
> +
> +ubuntu@bionic:~/build$ ~/glibc/scripts/vltest.py 32 ./debugglibc.sh \
> +string/test-memmove
> +
> +ubuntu@bionic:~/build$ ~/glibc/scripts/vltest.py 64 ./testrun.sh \
> +string/test-memset
> +"""
> +import argparse
> +from ctypes import cdll, CDLL
> +import os
> +import sys
> +
> +EXIT_SUCCESS = 0
> +EXIT_FAILURE = 1
> +EXIT_UNSUPPORTED = 77
> +
> +AT_HWCAP = 16
> +HWCAP_SVE = (1 << 22)
> +
> +PR_SVE_GET_VL = 51
> +PR_SVE_SET_VL = 50
> +PR_SVE_SET_VL_ONEXEC = (1 << 18)
> +PR_SVE_VL_INHERIT = (1 << 17)
> +PR_SVE_VL_LEN_MASK = 0xffff
> +
> +def main(args):
> +    libc = CDLL("libc.so.6")
> +    if not libc.getauxval(AT_HWCAP) & HWCAP_SVE:
> +        print("CPU doesn't support SVE")
> +        sys.exit(EXIT_UNSUPPORTED)
> +
> +    libc.prctl(PR_SVE_SET_VL,
> +               args.vl[0] | PR_SVE_SET_VL_ONEXEC | PR_SVE_VL_INHERIT)
> +    os.execvp(args.args[0], args.args)
> +    print("exec system call failure")
> +    sys.exit(EXIT_FAILURE)


this only works on a (new enough) glibc based system and python's
CDLL path lookup can fail too (it does not follow the host system
configuration).

but i think there is no simple solution without compiling c code and
this seems useful, so i'm happy to have this script.

  reply	other threads:[~2021-03-29 13:20 UTC|newest]

Thread overview: 36+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-03-17  2:28 [PATCH 0/5] Added optimized memcpy/memmove/memset for A64FX Naohiro Tamura
2021-03-17  2:33 ` [PATCH 1/5] config: Added HAVE_SVE_ASM_SUPPORT for aarch64 Naohiro Tamura
2021-03-29 12:11   ` Szabolcs Nagy via Libc-alpha
2021-03-30  6:19     ` naohirot
2021-03-17  2:34 ` [PATCH 2/5] aarch64: Added optimized memcpy and memmove for A64FX Naohiro Tamura
2021-03-29 12:44   ` Szabolcs Nagy via Libc-alpha
2021-03-30  7:17     ` naohirot
2021-03-17  2:34 ` [PATCH 3/5] aarch64: Added optimized memset " Naohiro Tamura
2021-03-17  2:35 ` [PATCH 4/5] scripts: Added Vector Length Set test helper script Naohiro Tamura
2021-03-29 13:20   ` Szabolcs Nagy via Libc-alpha [this message]
2021-03-30  7:25     ` naohirot
2021-03-17  2:35 ` [PATCH 5/5] benchtests: Added generic_memcpy and generic_memmove to large benchtests Naohiro Tamura
2021-03-29 12:03 ` [PATCH 0/5] Added optimized memcpy/memmove/memset for A64FX Szabolcs Nagy via Libc-alpha
2021-05-10  1:45 ` naohirot
2021-05-14 13:35   ` Szabolcs Nagy via Libc-alpha
2021-05-19  0:11     ` naohirot
2021-05-12  9:23 ` [PATCH v2 0/6] aarch64: " Naohiro Tamura
2021-05-12  9:26   ` [PATCH v2 1/6] config: Added HAVE_AARCH64_SVE_ASM for aarch64 Naohiro Tamura
2021-05-26 10:05     ` Szabolcs Nagy via Libc-alpha
2021-05-12  9:27   ` [PATCH v2 2/6] aarch64: define BTI_C and BTI_J macros as NOP unless HAVE_AARCH64_BTI Naohiro Tamura
2021-05-26 10:06     ` Szabolcs Nagy via Libc-alpha
2021-05-12  9:28   ` [PATCH v2 3/6] aarch64: Added optimized memcpy and memmove for A64FX Naohiro Tamura
2021-05-26 10:19     ` Szabolcs Nagy via Libc-alpha
2021-05-12  9:28   ` [PATCH v2 4/6] aarch64: Added optimized memset " Naohiro Tamura
2021-05-26 10:22     ` Szabolcs Nagy via Libc-alpha
2021-05-12  9:29   ` [PATCH v2 5/6] scripts: Added Vector Length Set test helper script Naohiro Tamura
2021-05-12 16:58     ` Joseph Myers
2021-05-13  9:53       ` naohirot
2021-05-20  7:34     ` Naohiro Tamura
2021-05-26 10:24       ` Szabolcs Nagy via Libc-alpha
2021-05-12  9:29   ` [PATCH v2 6/6] benchtests: Fixed bench-memcpy-random: buf1: mprotect failed Naohiro Tamura
2021-05-26 10:25     ` Szabolcs Nagy via Libc-alpha
2021-05-27  0:22   ` [PATCH v2 0/6] aarch64: Added optimized memcpy/memmove/memset for A64FX naohirot
2021-05-27 23:50     ` naohirot
2021-05-27  7:42   ` [PATCH v3 1/2] aarch64: Added optimized memcpy and memmove " Naohiro Tamura
2021-05-27  7:44   ` [PATCH v3 2/2] aarch64: Added optimized memset " Naohiro Tamura

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=20210329132017.GF23289@arm.com \
    --to=libc-alpha@sourceware.org \
    --cc=naohirot@fujitsu.com \
    --cc=naohirot@jp.fujitsu.com \
    --cc=szabolcs.nagy@arm.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).