Re: [PATCH] arm64/sve: Lower the maximum allocation for the SVE ptrace regset

From: Mark Brown <broonie@kernel.org>
To: Will Deacon <will@kernel.org>
Cc: Doug Anderson <dianders@chromium.org>,
	Catalin Marinas <catalin.marinas@arm.com>,
	Dave Martin <Dave.Martin@arm.com>,
	Oleg Nesterov <oleg@redhat.com>,
	Al Viro <viro@zeniv.linux.org.uk>,
	linux-arm-kernel@lists.infradead.org,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH] arm64/sve: Lower the maximum allocation for the SVE ptrace regset
Date: Fri, 9 Feb 2024 17:40:06 +0000	[thread overview]
Message-ID: <ZcZjdnchVblj4iP+@finisterre.sirena.org.uk> (raw)
In-Reply-To: <20240209171155.GB25069@willie-the-truck>

[-- Attachment #1: Type: text/plain, Size: 889 bytes --]

On Fri, Feb 09, 2024 at 05:11:55PM +0000, Will Deacon wrote:

> I can pick this up as a short-term hack if it solves the problem for you,
> but I also saw that you posted:

> https://lore.kernel.org/r/20240205092626.v2.1.Id9ad163b60d21c9e56c2d686b0cc9083a8ba7924@changeid

> to fallback onto vmalloc() for large allocations.

> What's your preference for a fix?

We need the change kvzalloc() regardless since the ZA regset is just
over 64K which is on the big side, I do think that reducing the reported
regset size is useful as a fix independently since even with kvzalloc()
it's still pointless and rude to ask for such a big allocation and we
might reasonably be generating core files or dumping crashing processes
under memory pressure.

Dave was looking at sizing the regsets dynamically but there's enough
going on there that it's not great if people want to pick it up for
stable.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]