This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH 10/27] arm64/sve: Low-level CPU setup
- From: Dave Martin <Dave dot Martin at arm dot com>
- To: Alex Bennée <alex dot bennee at linaro dot org>
- Cc: linux-arch at vger dot kernel dot org, libc-alpha at sourceware dot org, Ard Biesheuvel <ard dot biesheuvel at linaro dot org>, Szabolcs Nagy <szabolcs dot nagy at arm dot com>, Catalin Marinas <catalin dot marinas at arm dot com>, Will Deacon <will dot deacon at arm dot com>, Richard Sandiford <richard dot sandiford at arm dot com>, kvmarm at lists dot cs dot columbia dot edu, linux-arm-kernel at lists dot infradead dot org
- Date: Tue, 22 Aug 2017 16:33:03 +0100
- Subject: Re: [PATCH 10/27] arm64/sve: Low-level CPU setup
- Authentication-results: sourceware.org; auth=none
- References: <1502280338-23002-1-git-send-email-Dave.Martin@arm.com> <1502280338-23002-11-git-send-email-Dave.Martin@arm.com> <87r2w34pab.fsf@linaro.org>
On Tue, Aug 22, 2017 at 04:04:28PM +0100, Alex Bennée wrote:
>
> Dave Martin <Dave.Martin@arm.com> writes:
>
> > To enable the kernel to use SVE, all SVE traps from EL1 must be
> > disabled. To take maximum advantage of the hardware, the full
> > available vector length also needs to be enabled for EL1 by
> > programming ZCR_EL2.LEN. (The kernel will program ZCR_EL1.LEN as
> > required, but this cannot override the limit set by ZCR_EL2.)
> >
> > In advance of full SVE support being implemented for userspace, it
> > also necessary to ensure that SVE traps from EL0 are enabled.
> >
> > This patch makes the appropriate changes to the primary and
> > secondary CPU initialisation code.
> >
> > Signed-off-by: Dave Martin <Dave.Martin@arm.com>
> > ---
> > arch/arm64/kernel/head.S | 13 ++++++++++++-
> > arch/arm64/mm/proc.S | 14 ++++++++++++--
> > 2 files changed, 24 insertions(+), 3 deletions(-)
> >
> > diff --git a/arch/arm64/kernel/head.S b/arch/arm64/kernel/head.S
> > index 973df7d..0ae1713 100644
> > --- a/arch/arm64/kernel/head.S
> > +++ b/arch/arm64/kernel/head.S
> > @@ -514,8 +514,19 @@ CPU_LE( movk x0, #0x30d0, lsl #16 ) // Clear EE and E0E on LE systems
> > mov x0, #0x33ff
> > msr cptr_el2, x0 // Disable copro. traps to EL2
> >
> > + /* SVE register access */
> > + mrs x1, id_aa64pfr0_el1
> > + ubfx x1, x1, #ID_AA64PFR0_SVE_SHIFT, #4
> > + cbz x1, 7f
> > +
> > + bic x0, x0, #CPTR_EL2_TZ // Also disable SVE traps
> > + msr cptr_el2, x0 // Disable copro. traps
> > to EL2
>
> It seems a shame to write to cptr_el2 twice rather than compute and
> write.
>
> > + isb
>
> Especially as the second one needs an isb :-/
>
> But I don't see a much neater way of doing it so:
>
> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Thanks
Originally I think I did merge the CPTR writes here, but since this is
slow-path anyway, I figured it was better to keep the code simple rather
than introducing new interdependencies.
I could revisit it if someone has a strong view.
Cheers
---Dave
>
> > + mov x1, #ZCR_ELx_LEN_MASK // SVE: Enable full vector
> > + msr_s SYS_ZCR_EL2, x1 // length for EL1.
> > +
> > /* Hypervisor stub */
> > - adr_l x0, __hyp_stub_vectors
> > +7: adr_l x0, __hyp_stub_vectors
> > msr vbar_el2, x0
[...]