gh-mirror/mikepaul-LuaJIT
1
0
Fork 0
mirror of https://github.com/LuaJIT/LuaJIT.git synced 2025-02-08 07:34:07 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

ARM64: Improve generation of immediates.

Browse Source
This commit is contained in:
Mike Pall 2021-03-23 02:08:44 +01:00
parent a4c9fc3d6c
commit ebc4919aff
1 changed files with 33 additions and 31 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

64
src/lj_emit_arm64.h
View File

@ -194,39 +194,41 @@ static int emit_kdelta(ASMState *as, Reg rd, uint64_t k, int lim)
static void emit_loadk(ASMState *as, Reg rd, uint64_t u64, int is64) static void emit_loadk(ASMState *as, Reg rd, uint64_t u64, int is64)
{ {
uint32_t k13 = emit_isk13(u64, is64); int i, zeros = 0, ones = 0, neg;
if (k13) { /* Can the constant be represented as a bitmask immediate? */ if (!is64) u64 = (int64_t)(int32_t)u64; /* Sign-extend. */
emit_dn(as, (is64|A64I_ORRw)^k13, rd, RID_ZERO); /* Count homogeneous 16 bit fragments. */
} else { for (i = 0; i < 4; i++) {
int i, zeros = 0, ones = 0, neg; uint64_t frag = (u64 >> i*16) & 0xffff;
if (!is64) u64 = (int64_t)(int32_t)u64; /* Sign-extend. */ zeros += (frag == 0);
/* Count homogeneous 16 bit fragments. */ ones += (frag == 0xffff);
for (i = 0; i < 4; i++) { }
uint64_t frag = (u64 >> i*16) & 0xffff; neg = ones > zeros; /* Use MOVN if it pays off. */
zeros += (frag == 0); if ((neg ? ones : zeros) < 3) { /* Need 2+ ins. Try shorter K13 encoding. */
ones += (frag == 0xffff); uint32_t k13 = emit_isk13(u64, is64);
if (k13) {
emit_dn(as, (is64|A64I_ORRw)^k13, rd, RID_ZERO);
return;
} }
neg = ones > zeros; /* Use MOVN if it pays off. */ }
if (!emit_kdelta(as, rd, u64, 4 - (neg ? ones : zeros))) { if (!emit_kdelta(as, rd, u64, 4 - (neg ? ones : zeros))) {
int shift = 0, lshift = 0; int shift = 0, lshift = 0;
uint64_t n64 = neg ? ~u64 : u64; uint64_t n64 = neg ? ~u64 : u64;
if (n64 != 0) { if (n64 != 0) {
/* Find first/last fragment to be filled. */ /* Find first/last fragment to be filled. */
shift = (63-emit_clz64(n64)) & ~15; shift = (63-emit_clz64(n64)) & ~15;
lshift = emit_ctz64(n64) & ~15; lshift = emit_ctz64(n64) & ~15;
}
/* MOVK requires the original value (u64). */
while (shift > lshift) {
uint32_t u16 = (u64 >> shift) & 0xffff;
/* Skip fragments that are correctly filled by MOVN/MOVZ. */
if (u16 != (neg ? 0xffff : 0))
emit_d(as, is64 | A64I_MOVKw | A64F_U16(u16) | A64F_LSL16(shift), rd);
shift -= 16;
}
/* But MOVN needs an inverted value (n64). */
emit_d(as, (neg ? A64I_MOVNx : A64I_MOVZx) |
A64F_U16((n64 >> lshift) & 0xffff) | A64F_LSL16(lshift), rd);
} }
/* MOVK requires the original value (u64). */
while (shift > lshift) {
uint32_t u16 = (u64 >> shift) & 0xffff;
/* Skip fragments that are correctly filled by MOVN/MOVZ. */
if (u16 != (neg ? 0xffff : 0))
emit_d(as, is64 | A64I_MOVKw | A64F_U16(u16) | A64F_LSL16(shift), rd);
shift -= 16;
}
/* But MOVN needs an inverted value (n64). */
emit_d(as, (neg ? A64I_MOVNx : A64I_MOVZx) |
A64F_U16((n64 >> lshift) & 0xffff) | A64F_LSL16(lshift), rd);
} }
} }