lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:   Fri, 21 Apr 2023 13:03:31 +0200
From:   John Paul Adrian Glaubitz <glaubitz@...sik.fu-berlin.de>
To:     Geert Uytterhoeven <geert+renesas@...der.be>,
        Yoshinori Sato <ysato@...rs.sourceforge.jp>,
        Rich Felker <dalias@...c.org>
Cc:     Palmer Dabbelt <palmer@...osinc.com>, linux-sh@...r.kernel.org,
        linux-kbuild@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3] sh: Use generic GCC library routines

Hi Geert!

On Thu, 2023-03-23 at 11:18 +0100, Geert Uytterhoeven wrote:
> The C implementations of __ashldi3(), __ashrdi3__(), and __lshrdi3() in
> arch/sh/lib/ are identical to the generic C implementations in lib/.
> Reduce duplication by switching SH to the generic versions.
> 
> Update the include path in arch/sh/boot/compressed accordingly.
> 
> Signed-off-by: Geert Uytterhoeven <geert+renesas@...der.be>
> Acked-by: Palmer Dabbelt <palmer@...osinc.com>
> ---
> v3:
>   - Add Acked-by,
> 
> v2:
>   - Fix silly typo in subject.
> 
> Tested on landisk and qemu/rts7751r2d.
> 
> Note that it also works without the change to arch/sh/boot/compressed/,
> as lib/ashldi3.c can be reached via both include/uapi/../../lib/ashldi3.c
> and arch/sh/boot/compressed/../../../../lib/ashldi3.c.
> 
> Palmer tried a similar thing before:
> https://lore.kernel.org/linux-arch/20170523220546.16758-1-palmer@dabbelt.com/
> but initially it broke the SH build due to a missing change to
> arch/sh/boot/compressed/, and the later update never got picked up.
> In the mean time, arch/sh/boot/compressed/ was changed, so his patch no
> longer applies.
> 
> Similar for the other architectures, I guess?
> ---
>  arch/sh/Kconfig                   |  3 +++
>  arch/sh/boot/compressed/ashldi3.c |  4 ++--
>  arch/sh/lib/Makefile              |  4 +---
>  arch/sh/lib/ashldi3.c             | 30 -----------------------------
>  arch/sh/lib/ashrdi3.c             | 32 -------------------------------
>  arch/sh/lib/lshrdi3.c             | 30 -----------------------------
>  6 files changed, 6 insertions(+), 97 deletions(-)
>  delete mode 100644 arch/sh/lib/ashldi3.c
>  delete mode 100644 arch/sh/lib/ashrdi3.c
>  delete mode 100644 arch/sh/lib/lshrdi3.c
> 
> diff --git a/arch/sh/Kconfig b/arch/sh/Kconfig
> index ccb866750a884643..892903dc74990c36 100644
> --- a/arch/sh/Kconfig
> +++ b/arch/sh/Kconfig
> @@ -20,6 +20,9 @@ config SUPERH
>  	select GENERIC_CMOS_UPDATE if SH_SH03 || SH_DREAMCAST
>  	select GENERIC_IDLE_POLL_SETUP
>  	select GENERIC_IRQ_SHOW
> +	select GENERIC_LIB_ASHLDI3
> +	select GENERIC_LIB_ASHRDI3
> +	select GENERIC_LIB_LSHRDI3
>  	select GENERIC_PCI_IOMAP if PCI
>  	select GENERIC_SCHED_CLOCK
>  	select GENERIC_SMP_IDLE_THREAD
> diff --git a/arch/sh/boot/compressed/ashldi3.c b/arch/sh/boot/compressed/ashldi3.c
> index 7cebd646df839b48..7c12121702309e8c 100644
> --- a/arch/sh/boot/compressed/ashldi3.c
> +++ b/arch/sh/boot/compressed/ashldi3.c
> @@ -1,2 +1,2 @@
> -// SPDX-License-Identifier: GPL-2.0-only
> -#include "../../lib/ashldi3.c"
> +// SPDX-License-Identifier: GPL-2.0-or-later
> +#include "../../../../lib/ashldi3.c"
> diff --git a/arch/sh/lib/Makefile b/arch/sh/lib/Makefile
> index eb473d373ca43a4b..d20a0768b31fa2b6 100644
> --- a/arch/sh/lib/Makefile
> +++ b/arch/sh/lib/Makefile
> @@ -7,9 +7,7 @@ lib-y  = delay.o memmove.o memchr.o \
>  	 checksum.o strlen.o div64.o div64-generic.o
>  
>  # Extracted from libgcc
> -obj-y += movmem.o ashldi3.o ashrdi3.o lshrdi3.o \
> -	 ashlsi3.o ashrsi3.o ashiftrt.o lshrsi3.o \
> -	 udiv_qrnnd.o
> +obj-y += movmem.o ashlsi3.o ashrsi3.o ashiftrt.o lshrsi3.o udiv_qrnnd.o

Why are the single-precision (denoted as "si") variants not being replaced?

Don't we have generic versions for these?

Adrian

-- 
 .''`.  John Paul Adrian Glaubitz
: :' :  Debian Developer
`. `'   Physicist
  `-    GPG: 62FF 8A75 84E0 2956 9546  0006 7426 3B37 F5B5 F913

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ