The type of the immediates should not matter as long as the encoding is
equivalent to the encoding of one of the legal inline constants.

Tested-by: Michel Dänzer <michel.daenzer@amd.com>
llvm-svn: 204056

Added checks for number of operands and operand register classes.

Tested-by: Michel Dänzer <michel.daenzer@amd.com>
llvm-svn: 204054

operator* on the by-operand iterators to return a MachineOperand& rather than
a MachineInstr&.  At this point they almost behave like normal iterators!

Again, this requires making some existing loops more verbose, but should pave
the way for the big range-based for-loop cleanups in the future.

llvm-svn: 203865

llvm-svn: 203517

DS instructions that access local memory can only uses addresses that
are less than or equal to the value of M0.  When M0 is uninitialized,
then we experience undefined behavior.

This patch also changes the behavior to emit S_WQM_B64 on pixel shaders
no matter what kind of DS instruction is used.

llvm-svn: 201097

Allow MachineCSE to coalesce trivial subregister copies the same way that it coalesces normal copies.

Without this, MachineCSE is powerless to handle redundant operations with truncated source operands.

This required fixing the 2-addr pass to handle tied subregisters. It isn't clear what combinations of subregisters can legally be tied, but the simple case of truncated source operands is now safely handled:

     %vreg11<def> = COPY %vreg1:sub_32bit; GR32:%vreg11 GR64:%vreg1
     %vreg12<def> = COPY %vreg2:sub_32bit; GR32:%vreg12 GR64:%vreg2
     %vreg13<def,tied1> = ADD32rr %vreg11<tied0>, %vreg12<kill>, %EFLAGS<imp-def>

Test case: cse-add-with-overflow.ll.

This exposed an existing bug in
PPCInstrInfo::commuteInstruction. Thanks to Rafael for the test case:
PowerPC/crash.ll.

llvm-svn: 197465

SGPRs are spilled into VGPRs using the {READ,WRITE}LANE_B32 instructions.

v2:
  - Fix encoding of Lane Mask
  - Use correct register flags, so we don't overwrite the low dword
    when restoring multi-dword registers.

v3:
  - Register spilling seems to hang the GPU, so replace all shaders
    that need spilling with a dummy shader.

v4:
  - Fix *LANE definitions
  - Change destination reg class for 32-bit SMRD instructions

v5:
  - Remove small optimization that was crashing Serious Sam 3.

https://bugs.freedesktop.org/show_bug.cgi?id=68224
https://bugs.freedesktop.org/show_bug.cgi?id=71285

NOTE: This is a candidate for the 3.4 branch.
llvm-svn: 195880

Moving into a VSrc doesn't always work, since it could be
replaced with an SGPR later.

llvm-svn: 195042

No other SGPR operands are allowed, so if VCC is
used, move the other to a VGPR.

llvm-svn: 195041

llvm-svn: 195034

When replacing scalar operations with vector,
the wrong implicit output register was used.

llvm-svn: 195033

llvm-svn: 194858

llvm-svn: 194688

llvm-svn: 194684

llvm-svn: 194632

Private address space is emulated using the register file with
MOVRELS and MOVRELD instructions.

llvm-svn: 194626

All shift operations will be selected as SALU instructions and then
if necessary lowered to VALU instructions in the SIFixSGPRCopies pass.

This allows us to do more operations on the SALU which will improve
performance and is also required for implementing private memory
using indirect addressing, since the private memory pointers must stay
in the scalar registers.

This patch includes some fixes from Matt Arsenault.

llvm-svn: 194625

llvm-svn: 193510

llvm-svn: 193183

llvm-svn: 193180

The AMDGPUIndirectAddressing pass was previously responsible for
lowering private loads and stores to indirect addressing instructions.
However, this pass was buggy and way too complicated.  The only
advantage it had over the new simplified code was that it saved one
instruction per direct write to private memory.  This optimization
likely has a minimal impact on performance, and we may be able
to duplicate it using some other transformation.

For the private address space, we now:
1. Lower private loads/store to Register(Load|Store) instructions
2. Reserve part of the register file as 'private memory'
3. After regalloc lower the Register(Load|Store) instructions to
   MOV instructions that use indirect addressing.

llvm-svn: 193179

llvm-svn: 193178

llvm-svn: 192743

The function is used by the machine verifier and checks that VOP*
instructions have legal operands.

llvm-svn: 192367

llvm-svn: 188626

The logic in SIInsertWaits::getHwCounts() only really made sense for SMRD
instructions, and trying to shoehorn it into handling DS_WRITE_B32 caused
it to corrupt the encoding of that by clobbering the first operand with
the second one.

Undo that damage and only apply the SMRD logic to that.

Fixes some derivates related piglit regressions with radeonsi.

Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 188558

The previous code declared the operand as unknown:$vaddr, which made
it possible for scalar registers to be used instead of vector registers.

llvm-svn: 188425

llvm-svn: 186307

the internals of TargetMachine could change.

No functionality change intended.

llvm-svn: 183561

Depending on the number of bits set in the writemask.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
llvm-svn: 179166

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Tested-by: Michel Dänzer <michel.daenzer@amd.com>
llvm-svn: 178127

Prevent loading M0 multiple times.

Signed-off-by: Christian König <christian.koenig@amd.com>
llvm-svn: 178023

v2: based on Michels patch, but now allows copying of all registers sizes.

Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Christian König <christian.koenig@amd.com>
llvm-svn: 176346

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 176102

Seems to be allot simpler, and also paves the
way for further improvements.

v2: rebased on master, use 0 in BUFFER_LOAD_FORMAT_XYZW,
    use VGPR0 in dummy EXP, avoid compiler warning, break
    after encoding the first literal.
v3: correctly use V_ADD_F32_e64

This is a candidate for the stable branch.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 175354

Allows nexuiz to run with radeonsi.

Patch by: Michel Dänzer

Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 174655

Only implemented for R600 so far.  SI is missing implementations of a
few callbacks used by the Indirect Addressing pass and needs code to
handle frame indices.

At the moment R600 only supports array sizes of 16 dwords or less.
Register packing of vector types is currently disabled, which means that a
vec4 is stored in T0_X, T1_X, T2_X, T3_X, rather than T0_XYZW. In order
to correctly pack registers in all cases, we will need to implement an
analysis pass for R600 that determines the correct vector width for each
array.

v2:
  - Add support for i8 zext load from stack.
  - Coding style fixes

v3:
  - Don't reserve registers for indirect addressing when it isn't
    being used.
  - Fix bug caused by LLVM limiting the number of SubRegIndex
    declarations.

v4:
  - Fix 64-bit defines

llvm-svn: 174525

utils/sort_includes.py script.

Most of these are updating the new R600 target and fixing up a few
regressions that have creeped in since the last time I sorted the
includes.

llvm-svn: 171362

llvm-svn: 170620

A new backend supporting AMD GPUs: Radeon HD2XXX - HD7XXX

llvm-svn: 169915