Optimize the MVE 1op-immediate insns (VORR, VBIC, VMOV) to
use TCG vector ops when possible.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20210913095440.13462-13-peter.maydell@linaro.org
Optimize the MVE shift-and-insert insns by using TCG
vector ops when possible.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20210913095440.13462-12-peter.maydell@linaro.org
Optimize the MVE VSHLL insns by using TCG vector ops when possible.
This includes the VMOVL insn, which we handle in mve.decode as "VSHLL
with zero shift count".
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20210913095440.13462-11-peter.maydell@linaro.org
Optimize the MVE VSHL and VSHR immediate forms by using TCG vector
ops when possible.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20210913095440.13462-10-peter.maydell@linaro.org
Optimize the MVE VMVN insn by using TCG vector ops when possible.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20210913095440.13462-9-peter.maydell@linaro.org
Optimize the MVE VDUP insns by using TCG vector ops when possible.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20210913095440.13462-8-peter.maydell@linaro.org
Optimize the MVE VNEG and VABS insns by using TCG
vector ops when possible.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20210913095440.13462-7-peter.maydell@linaro.org
Optimize MVE arithmetic ops when we have a TCG
vector operation we can use.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20210913095440.13462-6-peter.maydell@linaro.org
When not predicating, implement the MVE bitwise logical insns
directly using TCG vector operations.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20210913095440.13462-5-peter.maydell@linaro.org
Our current codegen for MVE always calls out to helper functions,
because some byte lanes might be predicated. The common case is that
in fact there is no predication active and all lanes should be
updated together, so we can produce better code by detecting that and
using the TCG generic vector infrastructure.
Add a TB flag that is set when we can guarantee that there is no
active MVE predication, and a bool in the DisasContext. Subsequent
patches will use this flag to generate improved code for some
instructions.
In most cases when the predication state changes we simply end the TB
after that instruction. For the code called from vfp_access_check()
that handles lazy state preservation and creating a new FP context,
we can usually avoid having to try to end the TB because luckily the
new value of the flag following the register changes in those
sequences doesn't depend on any runtime decisions. We do have to end
the TB if the guest has enabled lazy FP state preservation but not
automatic state preservation, but this is an odd corner case that is
not going to be common in real-world code.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20210913095440.13462-4-peter.maydell@linaro.org
Implement the MVE VRINT insns, which round floating point inputs
to integer values, leaving them in floating point format.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE VCVT instruction which converts between single
and half precision floating point.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE VCVT which converts from floating-point to integer
using a rounding mode specified by the instruction. We implement
this similarly to the Neon equivalents, by passing the required
rounding mode as an extra integer parameter to the helper functions.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE "VCVT (between floating-point and integer)" insn.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE VCVT insns which convert between floating and fixed
point. As with the Neon equivalents, these use essentially the same
constant encoding as right-shift-by-immediate.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE fp scalar comparisons VCMP and VPT.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE fp vector comparisons VCMP and VPT.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE VMAXNMV, VMINNMV, VMAXNMAV, VMINNMAV insns. These
calculate the maximum or minimum of floating point elements across a
vector, starting with a value in a general purpose register and
returning the result there.
The pseudocode silences a possible SNaN in the accumulating result
on every iteration (by calling FPConvertNaN), but we do it only
on the input ra, because if none of the inputs to float*_maxnum
or float*_minnum are SNaNs then the result can't be an SNaN.
Note that we can't use the float*_maxnuma() etc functions we defined
earlier for VMAXNMA and VMINNMA, because we mustn't take the absolute
value of the starting general-purpose register value, which could be
negative.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE fp-with-scalar VFMA and VFMAS insns.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE scalar floating point insns VADD, VSUB and VMUL.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE VMAXNMA and VMINNMA insns; these are 2-operand, but
the destination register must be the same as one of the source
registers.
We defer the decode of the size in bit 28 to the individual insn
patterns rather than doing it in the format, because otherwise we
would have a single insn pattern that overlapped with two groups (eg
VMAXNMA with the VMULH_S and VMULH_U groups). Having two insn
patterns per insn seems clearer than a complex multilevel nesting
of overlapping and non-overlapping groups.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Implement the MVE VCMUL and VCMLA insns.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Implement the MVE VFMA and VFMS insns.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Implement the MVE VCADD insn. Note that here the size bit is the
opposite sense to the other 2-operand fp insns.
We don't check for the sz == 1 && Qd == Qm UNPREDICTABLE case,
because that would mean we can't use the DO_2OP_FP macro in
translate-mve.c.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Implement more simple 2-operand floating point MVE insns.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Implement the MVE VADD (floating-point) insn. Handling of this is
similar to the 2-operand integer insns, except that we must take care
to only update the floating point exception status if the least
significant bit of the predicate mask for each element is active.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Implement the MVE interleaving load/store functions VLD2, VLD4, VST2
and VST4. VLD2 loads 16 bytes of data from memory and writes to 2
consecutive Qregs; VLD4 loads 16 bytes of data from memory and writes
to 4 consecutive Qregs. The 'pattern' field in the encoding
determines the offset into memory which is accessed and also which
elements in the Qregs are written to. (The intention is that a
sequence of four consecutive VLD4 with different pattern values
performs a complete de-interleaving load of 64 bytes into all
elements of the 4 Qregs.) VST2 and VST4 do the same, but for stores.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE VLDR/VSTR insns which do scatter-gather using base
addresses from Qm plus or minus an immediate offset (possibly with
writeback). Note that writeback is not predicated but it does have
to honour ECI state, so we have to add an eci_mask check to the
VSTR_SG macros (the VLDR_SG macros already needed this to be able
to distinguish "skip beat" from "set predicated element to 0").
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE gather-loads and scatter-stores which
form the address by adding a base value from a scalar
register to an offset in each element of a vector.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE VCTP insn, which sets the VPR.P0 predicate bits so
as to predicate any element at index Rn or greater is predicated. As
with VPNOT, this insn itself is predicable and subject to beatwise
execution.
The calculation of the mask is the same as is used to determine
ltpmask in mve_element_mask(), but we precalculate masklen in
generated code to avoid having to have 4 helpers specialized by size.
We put the decode line in with the low-overhead-loop insns in
t32.decode because it's logically part of that collection of insn
patterns, even though it is an MVE only insn.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE VPNOT insn, which inverts the bits in VPR.P0
(subject to both predication and to beatwise execution).
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE VMOV forms that move data between 2 general-purpose
registers and 2 32-bit lanes in a vector register.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE VMAXA and VMINA insns, which take the absolute
value of the signed elements in the input vector and then accumulate
the unsigned max or min into the destination vector.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE 1-operand saturating operations VQABS and VQNEG.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE saturating doubling multiply accumulate insns
VQDMLAH, VQRDMLAH, VQDMLASH and VQRDMLASH. These perform a multiply,
double, add the accumulator shifted by the element size, possibly
round, saturate to twice the element size, then take the high half of
the result. The *MLAH insns do vector * scalar + vector, and the
*MLASH insns do vector * vector + scalar.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE VMLA insn, which multiplies a vector by a scalar
and accumulates into another vector.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE VMLADAV and VMLSLDAV insns. Like the VMLALDAV and
VMLSLDAV insns already implemented, these accumulate multiplied
vector elements; but they accumulate a 32-bit result rather than a
64-bit one.
Note that these encodings overlap with what would be RdaHi=0b111 for
VMLALDAV, VMLSLDAV, VRMLALDAVH and VRMLSLDAVH.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
The MVEGenDualAccOpFn is a bit misnamed, since it is used for
the "long dual accumulate" operations that use a 64-bit
accumulator. Rename it to MVEGenLongDualAccOpFn so we can
use the former name for the 32-bit accumulator insns.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE narrowing move insns VMOVN, VQMOVN and VQMOVUN.
These take a double-width input, narrow it (possibly saturating) and
store the result to either the top or bottom half of the output
element.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE VABAV insn, which computes absolute differences
between elements of two vectors and accumulates the result into
a general purpose register.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE integer min/max across vector insns
VMAXV, VMINV, VMAXAV and VMINAV, which find the maximum
from the vector elements and a general purpose register,
and store the maximum back into the general purpose
register.
These insns overlap with VRMLALDAVH (they use what would
be RdaHi=0b110).
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE instructions which perform shifts by a scalar.
These are VSHL T2, VRSHL T2, VQSHL T1 and VQRSHL T2. They take the
shift amount in a general purpose register and shift every element in
the vector by that amount.
Mostly we can reuse the helper functions for shift-by-immediate; we
do need two new helpers for VQRSHL.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE VMLAS insn, which multiplies a vector by a vector
and adds a scalar.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE VPSEL insn, which sets each byte of the destination
vector Qd to the byte from either Qn or Qm depending on the value of
the corresponding bit in VPR.P0.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE integer vector comparison instructions that compare
each element against a scalar from a general purpose register. These
are "VCMP (vector)" encodings T4, T5 and T6 and "VPT (vector)"
encodings T4, T5 and T6.
We have to move the decodetree pattern for VPST, because it
overlaps with VCMP T4 with size = 0b11.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE integer vector comparison instructions. These are
"VCMP (vector)" encodings T1, T2 and T3, and "VPT (vector)" encodings
T1, T2 and T3.
These insns compare corresponding elements in each vector, and update
the VPR.P0 predicate bits with the results of the comparison. VPT
also sets the VPR.MASK01 and VPR.MASK23 fields -- it is effectively
"VCMP then VPST".
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Factor out the "generate code to update VPR.MASK01/MASK23" part of
trans_VPST(); we are going to want to reuse it for the VPT insns.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE incrementing/decrementing dup insns VIDUP, VDDUP,
VIWDUP and VDWDUP. These fill the elements of a vector with
successively incrementing values, starting at the offset specified in
a general purpose register. The final value of the offset is written
back to this register. The wrapping variants take a second general
purpose register which specifies the point where the count should
wrap back to 0.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE VMULL (polynomial) insn. Unlike Neon, this comes
in two flavours: 8x8->16 and a 16x16->32. Also unlike Neon, the
inputs are in either the low or the high half of each double-width
element.
The assembler for this insn indicates the size with "P8" or "P16",
encoded into bit 28 as size = 0 or 1. We choose to follow the
same encoding as VQDMULL and decode this into a->size as MO_16
or MO_32 indicating the size of the result elements. This then
carries through to the helper function names where it then
matches up with the existing pmull_h() which does an 8x8->16
operation and a new pmull_w() which does the 16x16->32.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Implement the MVE VADDLV insn; this is similar to VADDV, except
that it accumulates 32-bit elements into a 64-bit accumulator
stored in a pair of general-purpose registers.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20210628135835.6690-15-peter.maydell@linaro.org
