Dave Martin <Dave.Martin@xxxxxxx> writes:
> On Tue, Aug 22, 2017 at 11:22:44AM +0100, Alex Bennée wrote:
>>
>> Dave Martin <Dave.Martin@xxxxxxx> writes:
>>
>> > This patch defines the representation that will be used for the SVE
>> > register state in the signal frame, and implements support for
>> > saving and restoring the SVE registers around signals.
>> >
>> > The same layout will also be used for the in-kernel task state.
>> >
>> > Due to the variability of the SVE vector length, it is not possible
>> > to define a fixed C struct to describe all the registers. Instead,
>> > Macros are defined in sigcontext.h to facilitate access to the
>> > parts of the structure.
>> >
>> > Signed-off-by: Dave Martin <Dave.Martin@xxxxxxx>
>> > ---
>> > arch/arm64/include/uapi/asm/sigcontext.h | 113 ++++++++++++++++++++++++++++++-
>> > 1 file changed, 112 insertions(+), 1 deletion(-)
>> >
>> > diff --git a/arch/arm64/include/uapi/asm/sigcontext.h b/arch/arm64/include/uapi/asm/sigcontext.h
>> > index f0a76b9..0533bdf 100644
>> > --- a/arch/arm64/include/uapi/asm/sigcontext.h
>> > +++ b/arch/arm64/include/uapi/asm/sigcontext.h
>> > @@ -16,6 +16,8 @@
>> > #ifndef _UAPI__ASM_SIGCONTEXT_H
>> > #define _UAPI__ASM_SIGCONTEXT_H
>> >
>> > +#ifndef __ASSEMBLY__
>> > +
>> > #include <linux/types.h>
>> >
>> > /*
>> > @@ -41,10 +43,11 @@ struct sigcontext {
>> > *
>> > * 0x210 fpsimd_context
>> > * 0x10 esr_context
>> > + * 0x8a0 sve_context (vl <= 64) (optional)
>> > * 0x20 extra_context (optional)
>> > * 0x10 terminator (null _aarch64_ctx)
>> > *
>> > - * 0xdb0 (reserved for future allocation)
>> > + * 0x510 (reserved for future allocation)
>> > *
>> > * New records that can exceed this space need to be opt-in for userspace, so
>> > * that an expanded signal frame is not generated unexpectedly. The mechanism
>> > @@ -116,4 +119,112 @@ struct extra_context {
>> > __u32 __reserved[3];
>> > };
>> >
>> > +#define SVE_MAGIC 0x53564501
>> > +
>> > +struct sve_context {
>> > + struct _aarch64_ctx head;
>> > + __u16 vl;
>> > + __u16 __reserved[3];
>> > +};
>> > +
>> > +#endif /* !__ASSEMBLY__ */
>> > +
>> > +/*
>> > + * The SVE architecture leaves space for future expansion of the
>> > + * vector length beyond its initial architectural limit of 2048 bits
>> > + * (16 quadwords).
>> > + */
>> > +#define SVE_VQ_MIN 1
>> > +#define SVE_VQ_MAX 0x200
>> > +
>> > +#define SVE_VL_MIN (SVE_VQ_MIN * 0x10)
>> > +#define SVE_VL_MAX (SVE_VQ_MAX * 0x10)
>> > +
>> > +#define SVE_NUM_ZREGS 32
>> > +#define SVE_NUM_PREGS 16
>> > +
>> > +#define sve_vl_valid(vl) \
>> > + ((vl) % 0x10 == 0 && (vl) >= SVE_VL_MIN && (vl) <= SVE_VL_MAX)
>> > +#define sve_vq_from_vl(vl) ((vl) / 0x10)
>> > +#define sve_vl_from_vq(vq) ((vq) * 0x10)
>>
>> I got a little confused first time through over what VQ and VL where.
>> Maybe it would make sense to expand a little more from first principles?
>>
>> /*
>> * The SVE architecture defines vector registers as a multiple of 128
>> * bit quadwords. The current architectural limit is 2048 bits (16
>> * quadwords) but there is room for future expansion beyond that.
>> */
>
> This comes up in several places and so I didn't want to comment it
> repeatedly everywhere.
>
> Instead, I wrote up something in section 2 (Vector length terminology)
> of Documentation/arm64/sve.txt -- see patch 25. Can you take a look and
> see whether that's adequate?
Ahh, I hadn't got to that yet. I'm unsure to the order the kernel likes
to put things but I like to put design documents at the front of the
patch queue as they are useful primers and saves you having to patch a:
modified arch/arm64/include/uapi/asm/sigcontext.h
@@ -132,19 +132,24 @@ struct sve_context {
/*
* The SVE architecture leaves space for future expansion of the
* vector length beyond its initial architectural limit of 2048 bits
- * (16 quadwords).
+ * (16 quadwords). See Documentation/arm64/sve.txt for a summary of
+ * the terminology of Vector Quads (VQ) and Vector Lengths (VL).
*/
+
+#define SVE_VQ_BITS 128 /* 128 bits in one quadword */
+#define SVE_VQ_BYTES (SVE_VQ_BITS / 8)
+
#define SVE_VQ_MIN 1
#define SVE_VQ_MAX 0x200
-#define SVE_VL_MIN (SVE_VQ_MIN * 0x10)
-#define SVE_VL_MAX (SVE_VQ_MAX * 0x10)
+#define SVE_VL_MIN (SVE_VQ_MIN * SVE_VQ_BYTES)
+#define SVE_VL_MAX (SVE_VQ_MAX * SVE_VQ_BYTES)
#define SVE_NUM_ZREGS 32
#define SVE_NUM_PREGS 16
#define sve_vl_valid(vl) \
- ((vl) % 0x10 == 0 && (vl) >= SVE_VL_MIN && (vl) <= SVE_VL_MAX)
+ ((vl) % SVE_VQ_BYTES == 0 && (vl) >= SVE_VL_MIN && (vl) <= SVE_VL_MAX)
#define sve_vq_from_vl(vl) ((vl) / 0x10)
#define sve_vl_from_vq(vq) ((vq) * 0x10)
>
> [...]
>
> Cheers
> ---Dave
--
Alex Bennée
