linux-kernel - Re: [PATCH v2] arm64/fpsimd: Only provide the length to cpufeature for xCR registers

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]

Message-ID: <ZMvYSmpCfFQ2+m7q@arm.com>
Date:   Thu, 3 Aug 2023 17:39:38 +0100
From:   Catalin Marinas <catalin.marinas@....com>
To:     Mark Brown <broonie@...nel.org>
Cc:     Will Deacon <will@...nel.org>,
        Suzuki K Poulose <suzuki.poulose@....com>,
        linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] arm64/fpsimd: Only provide the length to cpufeature
 for xCR registers

On Mon, Jul 31, 2023 at 02:58:48PM +0100, Mark Brown wrote:
> For both SVE and SME we abuse the generic register field comparison
> support in the cpufeature code as part of our detection of unsupported
> variations in the vector lengths available to PEs, reporting the maximum
> vector lengths via ZCR_EL1.LEN and SMCR_EL1.LEN.  Since these are
> configuration registers rather than identification registers the
> assumptions the cpufeature code makes about how unknown bitfields behave
> are invalid, leading to warnings when SME features like FA64 are enabled
> and we hotplug a CPU:
> 
>   CPU features: SANITY CHECK: Unexpected variation in SYS_SMCR_EL1. Boot CPU: 0x0000000000000f, CPU3: 0x0000008000000f
>   CPU features: Unsupported CPU feature variation detected.
> 
> SVE has no controls other than the vector length so is not yet impacted
> but the same issue will apply there if any are defined.
> 
> Since the only field we are interested in having the cpufeature code
> handle is the length field and we use a custom read function to obtain
> the value we can avoid these warnings by filtering out all other bits
> when we return the register value, if we're doing that we don't need to
> bother reading the register at all and can simply use the RDVL/RDSVL
> value we were filling in instead.

Maybe that's the simplest fix, especially if you want it in stable, but
I wonder why we even bother with with treating ZCR_EL1 and SMCR_EL1 as
feature registers. We already have verify_sme_features() to check for
the mismatch. BTW, is vec_verify_vq_map() sufficient so that we can skip
the maximum vector length check?

-- 
Catalin