motorhead/qemu
1
0
Fork 0
mirror of https://github.com/Motorhead1991/qemu.git synced 2025-08-04 08:13:54 -06:00
Code Issues Projects Releases Packages Wiki Activity Actions

target/arm: Implement MVE VDUP

Browse source 
Implement the MVE VDUP insn, which duplicates a value from
a general-purpose register into every lane of a vector
register (subject to predication).

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20210617121628.20116-11-peter.maydell@linaro.org
This commit is contained in:
Peter Maydell 2021-06-17 13:15:54 +01:00
parent 614dd4f3ba
commit ab59362fca
4 changed files with 55 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

16
target/arm/mve_helper.c
View file

@ -246,6 +246,22 @@ static void mergemask_sq(int64_t *d, int64_t r, uint16_t mask)
uint64_t *: mergemask_uq, \
int64_t *: mergemask_sq)(D, R, M)
void HELPER(mve_vdup)(CPUARMState *env, void *vd, uint32_t val)
{
/*
* The generated code already replicated an 8 or 16 bit constant
* into the 32-bit value, so we only need to write the 32-bit
* value to all elements of the Qreg, allowing for predication.
*/
uint32_t *d = vd;
uint16_t mask = mve_element_mask(env);
unsigned e;
for (e = 0; e < 16 / 4; e++, mask >>= 4) {
mergemask(&d[H4(e)], val, mask);
}
mve_advance_vpt(env);
}
#define DO_1OP(OP, ESIZE, TYPE, FN) \
void HELPER(mve_##OP)(CPUARMState *env, void *vd, void *vm) \
{ \