VFPU: vsqrt, vrsq, vrcp and vnrcp are exact in every backend

These went through the host's sqrt and division everywhere except the
interpreter's vrcp and vnrcp (vsqrt and vrsq there only behind
USE_VFPU_SQRT, now gone). They now always give the PSP's bits: the IR
gets FVSqrt (FSqrt stays the FPU's IEEE sqrt.s), and FRSqrt and FRecip,
which only the VFPU emits, become vfpu_rsqrt and vfpu_rcp; the IR
interpreter and the x64, arm64, RISC-V and LoongArch backends call them.
The old JITs call them directly, the ARM ones keeping the lanes in
callee-saved registers across the calls. cpu/vfpu/exact now passes on every core.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
This commit is contained in:
Henrik RydgårdandClaude Opus 5.5 committed 2026-09-24 12:24:42 -06:00
1 parent 87ad32d578
commit a2b10778be
17 files changed
+225 -132

No files matched your search

+1 -1
View File
@@ -1128,7 +1128,7 @@ namespace MIPSComp {
DISABLE;
break;
case 22: // d[i] = sqrtf(s[i]); break; //vsqrt
ir.Write(IROp::FSqrt, tempregs[i], sregs[i]);
ir.Write(IROp::FVSqrt, tempregs[i], sregs[i]);
break;
case 23: // d[i] = asinf(s[i]) / M_PI_2; break; //vasin
ir.Write(IROp::FAsin, tempregs[i], sregs[i]);