From a1fcaab90ce21e0c63e7d3b34308faef59879a8b Mon Sep 17 00:00:00 2001 From: Luke Kenneth Casson Leighton Date: Wed, 19 Jun 2019 15:55:43 +0100 Subject: [PATCH] add SV VLIW idea --- simple_v_extension/specification.mdwn | 59 +++++++++++++++++++++++++++ 1 file changed, 59 insertions(+) diff --git a/simple_v_extension/specification.mdwn b/simple_v_extension/specification.mdwn index 9bf1bf5e4..44a6b8213 100644 --- a/simple_v_extension/specification.mdwn +++ b/simple_v_extension/specification.mdwn @@ -2208,6 +2208,65 @@ in as a *parameter* to the HINT operation. No specific hints are yet defined in Simple-V +# VLIW Format + +One issue with SV is the setup and teardown time of the CSRs. The cost +of the use of a full CSRRW (requiring LI) is quite high. A VLIW format +therefore makes sense. + +A suitable prefix, which fits the Expanded Instruction-Length encoding +for "(80 + 16 times instruction_length)", as defined in Section 1.5 +of the RISC-V ISA, is as follows: + +| 15 | 14:12 | 11:10 | 9:8 | 7 | 6:0 | +| - | ----- | ----- | ----- | --- | ------- | +| rsvd | 16xil | pplen | rplen | S/VL| 1111111 | + +Optional VL/MAXVL/SubVL Block: + +| 15 | 14:12 | 11:6 | 5:0 | +| - | ----- | ------ | ------- | +| rsvd | SubVL | MAXVL | VLEN | + +Reminder of the variable-length format from Section 1.5 of the RISC-V ISA: + +| - | ---- | ---------------- | ---------------- | -------------------------- | +| .. | xxxx | xxxxxxxxxxxxxxxx | xnnnxxxxx1111111 | (80+16\*nnn)-bit, nnn!=111 | +| - | ---- | ---------------- | ---------------- | -------------------------- | + +Notes: + +* Bit 7 does not specify the VL, it specifies if VL, MAXVL and Sub-Vector + Length are to be in an optional block immediately following the prefix. +* Bits 8 and 9 define how many RegCam entries (0 to 3) follow (after + the optional VL block) +* Bits 10 and 11 define how many PredCam entries (0 to 3) follow after + the (optional) RegCam entries +* Bits 14 to 12 (IL) define the actual length of the instruction: total + number of bits is 80 + 16 times IL. Standard RV32, RVC and also + SVPrefix (P48-\*-Type) instructions fit into this space, after the + (optional) RegCam / PredCam entries +* Anything - any registers - within the VLIW-prefixed format *MUST* have the + RegCam and PredCam CSRs applied to it. +* At the end of the VLIW Group, the RegCam and PredCam CSRs *no longer apply*. + +This would greatly reduce the amount of space utilised by Vectorised +instructions, given that 64-bit CSRRW requires 3, even 4 32-bit opcodes: the +CSR itself, a LI, and the setting up of the value into the RS register +of the CSR, which, again, requires a LI / LUI to get the full 64 bit +data into the CSR. + +Given that each 64-bit CSR can hold only 4x PredCAM entries (or 4 RegCAM +entries), that's potentially 6 to eight 32-bit instructions, just to +establish the Vector State! + +Not only that: even CSRRW on VL and MAXVL requires 64-bits (even more bits if +VL needs to be set to greater than 32). Bear in mind that in SV, both MAXVL +and VL need to be set. + +In this light, embedding the VL/MAXVL, PredCam and RegCam CSR entries into +a VLIW format makes a lot of sense. + # Subsets of RV functionality This section describes the differences when SV is implemented on top of -- 2.30.2