From mboxrd@z Thu Jan 1 00:00:00 1970 From: alex.bennee@linaro.org (Alex =?utf-8?Q?Benn=C3=A9e?=) Date: Mon, 30 Oct 2017 16:12:13 +0000 Subject: [PATCH v4 20/28] arm64/sve: Add prctl controls for userspace vector length management In-Reply-To: <20171028160506.GI19485@e103592.cambridge.arm.com> References: <1509101470-7881-1-git-send-email-Dave.Martin@arm.com> <1509101470-7881-21-git-send-email-Dave.Martin@arm.com> <87r2toiikt.fsf@linaro.org> <20171028160506.GI19485@e103592.cambridge.arm.com> Message-ID: <87fua07gyq.fsf@linaro.org> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org Dave Martin writes: > On Fri, Oct 27, 2017 at 06:52:50PM +0100, Alex Benn?e wrote: >> >> Dave Martin writes: >> >> > This patch adds two arm64-specific prctls, to permit userspace to >> > control its vector length: >> > >> > * PR_SVE_SET_VL: set the thread's SVE vector length and vector >> > length inheritance mode. >> > >> > * PR_SVE_GET_VL: get the same information. >> > >> > Although these prctls resemble instruction set features in the SVE >> > architecture, they provide additional control: the vector length >> > inheritance mode is Linux-specific and nothing to do with the >> > architecture, and the architecture does not permit EL0 to set its >> > own vector length directly. Both can be used in portable tools >> > without requiring the use of SVE instructions. >> > >> > Signed-off-by: Dave Martin >> > Reviewed-by: Catalin Marinas >> > Cc: Alex Benn?e >> >> FYI there is a minor conflict applying this on current master. > > There are some trivial conflicts with one or two patches that already > went into arm64/for-next/core, so I based on that for this posting, not > torvalds/master. > > There's a note in the cover letter giving the precise commit I based > on, though the branch doesn't seem to have moved yet since I posted. Ahh probably those. It was only a few header file shuffles. I was rushing to apply before my flight back I failed to note the change in base commit. > > Otherwise, I don't see any conflict -- can you give details? Only that #endif _ASM_ had been dropped from the header file. > > Cheers > ---Dave -- Alex Benn?e