A float inline constant (240-248) read by a 64-bit operand is the f64 bit
pattern for integer B64 opcodes too: AMD's compilers emit
v_lshrrev_b64 v[4:5], v0, 1.0 for 0x3ff0000000000000 >> lane. ReadU32Pair
returned {f32 bits, 0}, so only F64-typed reads had the right value.
Based on TheCruZ's 284deab28 (closed#1219).
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
V_PK_MIN_F16 and V_PK_MAX_F16 return the quieted NaN for a signaling NaN
operand and the other operand for a quiet one (per half, as V_MIN_F16 and
V_MAX_F16 do). The signaling test masked the half with 0x7c00 and checked
for a nonzero payload, so a quiet NaN with payload bits such as 0x7e01 was
taken as signaling and returned. Mask with 0x7e00 so the quiet bit must be
clear.
Two more vectors in PackedMinMaxF16NanAndSignedZeroEdges cover it.
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* shader: support DS_CMPST_RTN_B32
Assassin's Creed Valhalla (PPSA01490, #281) stops on DS opcode 0x30 in a
compute shader. DS compares DATA and stores DATA2, the opposite of
BUFFER_ATOMIC_CMPSWAP, so the 32-bit atomic path swaps the two for this
opcode, as the 64-bit form does.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* shader: reject unsupported GDS compare-store variant
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: nmzik <Nmzik@mail.ru>
A patched TLS load calls TlsRegStub and SafeCall, which moved RSP past the
red zone with sub/add rsp,0x80. ADD and SUB write the arithmetic flags, so
the sub ran before SafeCall's pushfq saved them and the add ran after its
popfq restored them: guest code that compares, loads fs:[0] and then
branches on the flags could branch the wrong way. lea moves RSP without
touching the flags. The entry lea plus a two-byte nop keeps every patched
offset in both stubs; the exit lea is one byte longer and moves only ret.
TlsStubsPreserveFlags calls the stubs with two flag patterns and checks
the status flags, the output register and RAX; it fails with the old
bytes.
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Performance changes from #1495:
- Sync changed BDA ranges instead of scanning all mappings.
- Let VMA suballocate BDA buffers.
- Remove the renderer lock from presentation.
- Keep compatible pipeline caches across builds.
- Fold masked shifts that always produce zero.
nmzik