linux-kernel - Re: [PATCH v2 2/4] MIPS: Introduce config options for LLSC availability

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: Thu, 20 Jun 2024 18:06:28 +0200
From: Thomas Bogendoerfer <tsbogend@...ha.franken.de>
To: Jiaxun Yang <jiaxun.yang@...goat.com>
Cc: Jonas Gorski <jonas.gorski@...il.com>,
	"Maciej W. Rozycki" <macro@...am.me.uk>, linux-mips@...r.kernel.org,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 2/4] MIPS: Introduce config options for LLSC
 availability

On Wed, Jun 12, 2024 at 10:53:30AM +0100, Jiaxun Yang wrote:
> Introduce CPU_HAS_LLSC and CPU_MAY_HAVE_LLSC to determine availability
> of LLSC and Kconfig level.
> 
> They are both true for almost all supported CPUs besides:
> 
> R3000: Doesn't have LLSC, so both false.
> R5000 series: LLSC is unusable for 64bit kernel, so both false.
> R10000: Some platforms decided to opt-out LLSC due to errata, so only
> 	select CPU_MAY_HAVE_LLSC.
> WAR_4KC_LLSC: LLSC is buggy on certain reversion, which can be detected
> 	at cpu-probe or platform override, so only select CPU_MAY_HAVE_LLSC.
> 
> Signed-off-by: Jiaxun Yang <jiaxun.yang@...goat.com>
> ---
> v2: Make cpu_has_llsc logic clear
> ---
>  arch/mips/Kconfig                    | 20 ++++++++++++++++++++
>  arch/mips/include/asm/cpu-features.h |  9 ++++++++-
>  2 files changed, 28 insertions(+), 1 deletion(-)
> 
> diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig
> index 8ac467c1f9c8..50260a7e9b54 100644
> --- a/arch/mips/Kconfig
> +++ b/arch/mips/Kconfig
> @@ -1548,6 +1548,7 @@ config CPU_R3000
>  config CPU_R4300
>  	bool "R4300"
>  	depends on SYS_HAS_CPU_R4300
> +	select CPU_HAS_LLSC
>  	select CPU_SUPPORTS_32BIT_KERNEL
>  	select CPU_SUPPORTS_64BIT_KERNEL
>  	help
> @@ -1556,6 +1557,7 @@ config CPU_R4300
>  config CPU_R4X00
>  	bool "R4x00"
>  	depends on SYS_HAS_CPU_R4X00
> +	select CPU_HAS_LLSC
>  	select CPU_SUPPORTS_32BIT_KERNEL
>  	select CPU_SUPPORTS_64BIT_KERNEL
>  	select CPU_SUPPORTS_HUGEPAGES
> @@ -1566,6 +1568,7 @@ config CPU_R4X00
>  config CPU_TX49XX
>  	bool "R49XX"
>  	depends on SYS_HAS_CPU_TX49XX
> +	select CPU_HAS_LLSC
>  	select CPU_HAS_PREFETCH
>  	select CPU_SUPPORTS_32BIT_KERNEL
>  	select CPU_SUPPORTS_64BIT_KERNEL
> @@ -1574,6 +1577,7 @@ config CPU_TX49XX
>  config CPU_R5000
>  	bool "R5000"
>  	depends on SYS_HAS_CPU_R5000
> +	select CPU_HAS_LLSC if !64BIT
>  	select CPU_SUPPORTS_32BIT_KERNEL
>  	select CPU_SUPPORTS_64BIT_KERNEL
>  	select CPU_SUPPORTS_HUGEPAGES
> @@ -1583,6 +1587,7 @@ config CPU_R5000
>  config CPU_R5500
>  	bool "R5500"
>  	depends on SYS_HAS_CPU_R5500
> +	select CPU_HAS_LLSC

do you have an user manual stating that the R5k bug is fixed on this CPUs ?

Thomas.

-- 
Crap can work. Given enough thrust pigs will fly, but it's not necessarily a
good idea.                                                [ RFC1925, 2.3 ]