The hand-written parts of the manual still described how the retired xenia-rs interpreter behaved: its snapshots, Rust casts and helpers. Each of those 490 statements is now either restated as what Canary's emitters and x64 backend actually do (at the pinned canary_experimental commit), or dropped where it only made sense for xenia-rs. Checking them turned up claims that were wrong, not just outdated: - VSCR[SAT] is never modelled in Canary (DID_SATURATE is a stub and mfvscr cannot see it); the pages said saturating ops set it stickily. - Canary does not implement lswi/lswx/stswi/stswx, dcbi, mtfsb0/mtfsb1, vmsum*, vmhaddshs, vupkhpx/vupklpx, and most SPRs; pages described them as working. - Traps evaluate TO in Canary; stvebx/stvehx/stvewx store one element, not 16 bytes; mtmsrd writes only EE; fres/frsqrte/vrsqrtefp precision claims and the stfs "rounds under RN / sets FPSCR" claim contradicted the spec. - Reservations are a 64 KiB block bitmap plus a value compare, not per-address tracking. Claims that neither Canary's source nor a public spec settles are marked unverified (NI at boot, vmaddcfp128 operand order, estimate bit-exactness). Generated regions are untouched; re-running the generator changes nothing. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
11 KiB
11 KiB
lha — Load Half Word Algebraic
Assembler Mnemonics
| Mnemonic | XML entry | Flags | Description |
|---|---|---|---|
lha |
lha |
— | Load Half Word Algebraic |
lhau |
lhau |
— | Load Half Word Algebraic with Update |
lhaux |
lhaux |
— | Load Half Word Algebraic with Update Indexed |
lhax |
lhax |
— | Load Half Word Algebraic Indexed |
Syntax
lha [RD], [d]([RA0])
lhau [RD], [d]([RA])
lhaux [RD], [RA], [RB]
lhax [RD], [RA0], [RB]
Encoding
lha — form D
- Opcode word:
0xa8000000 - Primary opcode (bits 0–5):
42 - Extended opcode: —
- Synchronising: no
| Bits | Field | Meaning |
|---|---|---|
| 0–5 | OPCD |
primary opcode |
| 6–10 | RT |
destination GPR (or RS when storing) |
| 11–15 | RA |
source GPR (0 ⇒ literal 0 for RA0 forms) |
| 16–31 | D/SI/UI |
16-bit signed or unsigned immediate |
lhau — form D
- Opcode word:
0xac000000 - Primary opcode (bits 0–5):
43 - Extended opcode: —
- Synchronising: no
| Bits | Field | Meaning |
|---|---|---|
| 0–5 | OPCD |
primary opcode |
| 6–10 | RT |
destination GPR (or RS when storing) |
| 11–15 | RA |
source GPR (0 ⇒ literal 0 for RA0 forms) |
| 16–31 | D/SI/UI |
16-bit signed or unsigned immediate |
lhaux — form X
- Opcode word:
0x7c0002ee - Primary opcode (bits 0–5):
31 - Extended opcode:
375 - Synchronising: no
| Bits | Field | Meaning |
|---|---|---|
| 0–5 | OPCD |
primary opcode |
| 6–10 | RT/FRT/VRT |
destination |
| 11–15 | RA/FRA/VRA |
source A |
| 16–20 | RB/FRB/VRB |
source B |
| 21–30 | XO |
extended opcode (10 bits) |
| 31 | Rc |
record-form flag |
lhax — form X
- Opcode word:
0x7c0002ae - Primary opcode (bits 0–5):
31 - Extended opcode:
343 - Synchronising: no
| Bits | Field | Meaning |
|---|---|---|
| 0–5 | OPCD |
primary opcode |
| 6–10 | RT/FRT/VRT |
destination |
| 11–15 | RA/FRA/VRA |
source A |
| 16–20 | RB/FRB/VRB |
source B |
| 21–30 | XO |
extended opcode (10 bits) |
| 31 | Rc |
record-form flag |
Operands
| Field | Role | Description |
|---|---|---|
RA0 |
lha: read; lhax: read | Source GPR; when the encoded register number is 0 the operand is the literal 64-bit zero, not r0. |
d |
lha: read; lhau: read | 16-bit signed displacement (d) added to the base address register. |
RD |
lha: write; lhau: write; lhaux: write; lhax: write | Destination GPR. |
RA |
lhau: read; lhau: write; lhaux: read; lhaux: write | Source GPR (r0–r31). |
RB |
lhaux: read; lhax: read | Source GPR. |
Register Effects
lha
- Reads (always):
RA0,d - Reads (conditional): none
- Writes (always):
RD - Writes (conditional): none
lhau
- Reads (always):
RA,d - Reads (conditional): none
- Writes (always):
RD,RA - Writes (conditional): none
lhaux
- Reads (always):
RA,RB - Reads (conditional): none
- Writes (always):
RD,RA - Writes (conditional): none
lhax
- Reads (always):
RA0,RB - Reads (conditional): none
- Writes (always):
RD - Writes (conditional): none
Status-Register Effects
No condition-register or status-register effects.
Operation (pseudocode)
EA <- (RA|0) + EXTS(d)
RT <- SEXT16_to_64(MEM(EA, 2))
C Translation Example
/* No hand-written C yet. Translate the Canary emitter snapshot */
/* under Implementation References; its HIR maps directly: */
/* f.LoadGPR(n) / f.StoreGPR(n, v) -> r[n] / r[n] = v */
/* f.LoadFPR / StoreFPR, f.LoadVR / StoreVR -> f[n], v[n] */
/* f.Load(ea, T), f.Store(ea, v) -> raw read / write; emitters */
/* wrap them in f.ByteSwap for the big-endian guest value */
/* f.UpdateCR(n, v) -> CR field n from v's LOW 32 BITS vs 0 */
/* f.LoadCA / f.StoreCA -> xer.CA; f.StoreSAT -> vscr.SAT */
/* i.XO.RA, i.D.DS, ... -> the bit-fields listed under Operands */
/* The Register Effects and Status-Register Effects tables above */
/* enumerate every side effect a faithful translation must emit. */
Implementation References
lha
- Canary XML:
tools/ppc-instructions.xml— search formnem="lha" - Canary emitter:
src/xenia/cpu/ppc/ppc_emit_memory.cc:128 - Sylpheed opcode:
crates/sylpheed-ppc/src/opcode.rs:122 - Sylpheed decoder:
crates/sylpheed-ppc/src/decoder.rs:480
Canary emitter (frozen snapshot @ f21ebd49e9)
int InstrEmit_lha(PPCHIRBuilder& f, const InstrData& i) {
// if RA = 0 then
// b <- 0
// else
// b <- (RA)
// EA <- b + EXTS(D)
// RT <- EXTS(MEM(EA, 2))
Value* b;
if (i.D.RA == 0) {
b = f.LoadZeroInt64();
} else {
b = f.LoadGPR(i.D.RA);
}
Value* offset = f.LoadConstantInt64(XEEXTS16(i.D.DS));
Value* rt =
f.SignExtend(f.ByteSwap(f.LoadOffset(b, offset, INT16_TYPE)), INT64_TYPE);
f.StoreGPR(i.D.RT, rt);
return 0;
}
lhau
- Canary XML:
tools/ppc-instructions.xml— search formnem="lhau" - Canary emitter:
src/xenia/cpu/ppc/ppc_emit_memory.cc:149 - Sylpheed opcode:
crates/sylpheed-ppc/src/opcode.rs:123 - Sylpheed decoder:
crates/sylpheed-ppc/src/decoder.rs:481
Canary emitter (frozen snapshot @ f21ebd49e9)
int InstrEmit_lhau(PPCHIRBuilder& f, const InstrData& i) {
// EA <- (RA) + EXTS(D)
// RT <- EXTS(MEM(EA, 2))
// RA <- EA
Value* ra = f.LoadGPR(i.D.RA);
Value* offset = f.LoadConstantInt64(XEEXTS16(i.D.DS));
Value* rt = f.SignExtend(f.ByteSwap(f.LoadOffset(ra, offset, INT16_TYPE)),
INT64_TYPE);
f.StoreGPR(i.D.RT, rt);
StoreEA(f, i.D.RA, f.Add(ra, offset));
return 0;
}
lhaux
- Canary XML:
tools/ppc-instructions.xml— search formnem="lhaux" - Canary emitter:
src/xenia/cpu/ppc/ppc_emit_memory.cc:162 - Sylpheed opcode:
crates/sylpheed-ppc/src/opcode.rs:124 - Sylpheed decoder:
crates/sylpheed-ppc/src/decoder.rs:920
Canary emitter (frozen snapshot @ f21ebd49e9)
int InstrEmit_lhaux(PPCHIRBuilder& f, const InstrData& i) {
// EA <- (RA) + (RB)
// RT <- EXTS(MEM(EA, 2))
// RA <- EA
Value* ea = CalculateEA(f, i.X.RA, i.X.RB);
Value* rt = f.SignExtend(f.ByteSwap(f.Load(ea, INT16_TYPE)), INT64_TYPE);
f.StoreGPR(i.X.RT, rt);
StoreEA(f, i.X.RA, ea);
return 0;
}
lhax
- Canary XML:
tools/ppc-instructions.xml— search formnem="lhax" - Canary emitter:
src/xenia/cpu/ppc/ppc_emit_memory.cc:173 - Sylpheed opcode:
crates/sylpheed-ppc/src/opcode.rs:125 - Sylpheed decoder:
crates/sylpheed-ppc/src/decoder.rs:916
Canary emitter (frozen snapshot @ f21ebd49e9)
int InstrEmit_lhax(PPCHIRBuilder& f, const InstrData& i) {
// if RA = 0 then
// b <- 0
// else
// b <- (RA)
// EA <- b + (RB)
// RT <- EXTS(MEM(EA, 2))
Value* ea = CalculateEA_0(f, i.X.RA, i.X.RB);
Value* rt = f.SignExtend(f.ByteSwap(f.Load(ea, INT16_TYPE)), INT64_TYPE);
f.StoreGPR(i.X.RT, rt);
return 0;
}
Special Cases & Edge Conditions
- Sign-extending half-word load. Reads 2 bytes big-endian, treats them as a signed 16-bit integer, sign-extends to 64 bits. Compare with
lhz, which zero-extends. Canary emitsSignExtend(ByteSwap(LoadOffset(…, INT16)), INT64). - Big-endian read. Byte at
EAis the most-significant 8 bits of the half; byte atEA+1is the least-significant. On little-endian hostsmem.read_u16returns the big-endian word in host-native form already. RA0(non-update forms).RA = 0inlhaandlhaxselects literal zero — useful for absolute-address access patterns.- Update-form invalid forms.
lhau/lhauxinvokeRA = 0andRA = RTas invalid forms; Canary performs the load before writing backRA ← EA, so anRA = RTcollision silently destroys the loaded value. - No alignment requirement. Xenon executes unaligned half-word loads without a fault.
- Common in audio / graphics code.
lhais the standard load for signed 16-bit PCM samples and signed 16-bit packed vertex deltas. - Use
lharather thanlhz+extsh. Both produce the same result, butlhais one fused instruction and the compiler will pick it whenever the source type isint16_t/short.
Related Instructions
lhz,lhzu,lhzx,lhzux— zero-extending counterparts.lwa,lwax,lwaux— sign-extending word loads (32→64).lbz— byte load (no sign-extending byte load exists; uselbz+extsb).lhbrx— byte-reversed half-word load (zero-extending).sth,sthu,sthx,sthux— corresponding stores.