in favour of teaching CCAssignToStack that size 0 and/or align
0 means to use the ABI values. This seems a neater solution.
It is safe since no legal value type has size 0.
llvm-svn: 44107
This makes DwarfRegNum to accept list of numbers instead.
Added three different "flavours", but only slightly tested on x86-32/linux.
Please check another subtargets if possible,
llvm-svn: 43997
take a deleted nodes vector, instead of requiring it.
One more significant change: Implement the start of a legalizer that
just works on types. This legalizer is designed to run before the
operation legalizer and ensure just that the input dag is transformed
into an output dag whose operand and result types are all legal, even
if the operations on those types are not.
This design/impl has the following advantages:
1. When finished, this will *significantly* reduce the amount of code in
LegalizeDAG.cpp. It will remove all the code related to promotion and
expansion as well as splitting and scalarizing vectors.
2. The new code is very simple, idiomatic, and modular: unlike
LegalizeDAG.cpp, it has no 3000 line long functions. :)
3. The implementation is completely iterative instead of recursive, good
for hacking on large dags without blowing out your stack.
4. The implementation updates nodes in place when possible instead of
deallocating and reallocating the entire graph that points to some
mutated node.
5. The code nicely separates out handling of operations with invalid
results from operations with invalid operands, making some cases
simpler and easier to understand.
6. The new -debug-only=legalize-types option is very very handy :),
allowing you to easily understand what legalize types is doing.
This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is
enabled, this does nothing. However, this code is sufficient to legalize
all of the code in 186.crafty, olden and freebench on an x86 machine. The
biggest issues are:
1. Vectors aren't implemented at all yet
2. SoftFP is a mess, I need to talk to Evan about it.
3. No lowering to libcalls is implemented yet.
4. Various operations are missing etc.
5. There are FIXME's for stuff I hax0r'd out, like softfp.
Hey, at least it is a step in the right direction :). If you'd like to help,
just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If
this explodes it will tell you what needs to be implemented. Help is
certainly appreciated.
Once this goes in, we can do three things:
1. Add a new pass of dag combine between the "type legalizer" and "operation
legalizer" passes. This will let us catch some long-standing isel issues
that we miss because operation legalization often obfuscates the dag with
target-specific nodes.
2. We can rip out all of the type legalization code from LegalizeDAG.cpp,
making it much smaller and simpler. When that happens we can then
reimplement the core functionality left in it in a much more efficient and
non-recursive way.
3. Once the whole legalizer is non-recursive, we can implement whole-function
selectiondags maybe...
llvm-svn: 42981
1.
[(set GR32:$dst, (add GR32:$src1, GR32:$src2)),
(modify EFLAGS)]
This indicates the source pattern expects the instruction would produce 2 values. The first is the result of the addition. The second is an implicit definition in register EFLAGS.
2.
def : Pat<(parallel (addc GR32:$src1, GR32:$src2), (modify EFLAGS)), ()>
Similar to #1 except this is used for def : Pat patterns.
llvm-svn: 41897
This also changes the syntax for llvm.bswap, llvm.part.set, llvm.part.select, and llvm.ct* intrinsics. They are automatically upgraded by both the LLVM ASM reader and the bitcode reader. The test cases have been updated, with special tests added to ensure the automatic upgrading is supported.
llvm-svn: 40807
This commit fixes two things. One is a pair of VStudio compiler errors stemming from variables
which defined within the for loop statement and also within the body of the for loop. I fixed these
by renaming one of the two variables. Additionally, I've made the Function*->ExFunc map in
ExternalFunctions.cpp a ManagedStatic object, so that cleanup will be done on llvm_shutdown. In repeated
uses of the interpreter, where the same Function* address may get used for completely differnet functions,
this was causing a crash.
llvm-svn: 40558
InOperandList. This gives one piece of important information: # of results
produced by an instruction.
An example of the change:
def ADD32rr : I<0x01, MRMDestReg, (ops GR32:$dst, GR32:$src1, GR32:$src2),
"add{l} {$src2, $dst|$dst, $src2}",
[(set GR32:$dst, (add GR32:$src1, GR32:$src2))]>;
=>
def ADD32rr : I<0x01, MRMDestReg, (outs GR32:$dst), (ins GR32:$src1, GR32:$src2),
"add{l} {$src2, $dst|$dst, $src2}",
[(set GR32:$dst, (add GR32:$src1, GR32:$src2))]>;
llvm-svn: 40033
instruction flag, and use the flag along with a virtual member function
hook for targets to override if there are instructions that are only
trivially rematerializable with specific operands (i.e. constant pool
loads).
llvm-svn: 37728
with a general target hook to identify rematerializable instructions. Some
instructions are only rematerializable with specific operands, such as loads
from constant pools, while others are always rematerializable. This hook
allows both to be identified as being rematerializable with the same
mechanism.
llvm-svn: 37644
types of the iAny types involved in the overloaded intrinsic. Thus, we
can't use the argument number as the index but have to count them separately
in order to index Tys correctly. This patch rectifies this situation.
llvm-svn: 37296
feature is set, then the features in the implied list should be set also.
The opposite is also enforced: if a feature in the implied list isn't set,
then the feature that owns that implies list shouldn't be set either.
llvm-svn: 36756
Don't assert everytime an intrinsic name isn't recognized. Instead, make
the assert optional when callin getIntrinsicID(). This allows the assembler
to handle invalid intrinsic names gracefully.
llvm-svn: 36120
Implement code generation for overloaded intrinsic functions. The basic
difference is that "actual" argument types must be provided when
constructing intrinsic names and types. Also, for recognition, only the
prefix is examined. If it matches, the suffix is assumed to match. The
suffix is checked by the Verifier, however.
llvm-svn: 35539
Implement the arbitrary bit-width integer feature. The feature allows
integers of any bitwidth (up to 64) to be defined instead of just 1, 8,
16, 32, and 64 bit integers.
This change does several things:
1. Introduces a new Derived Type, IntegerType, to represent the number of
bits in an integer. The Type classes SubclassData field is used to
store the number of bits. This allows 2^23 bits in an integer type.
2. Removes the five integer Type::TypeID values for the 1, 8, 16, 32 and
64-bit integers. These are replaced with just IntegerType which is not
a primitive any more.
3. Adjust the rest of LLVM to account for this change.
Note that while this incremental change lays the foundation for arbitrary
bit-width integers, LLVM has not yet been converted to actually deal with
them in any significant way. Most optimization passes, for example, will
still only deal with the byte-width integer types. Future increments
will rectify this situation.
llvm-svn: 33113
not be used for anything other than backwards compat constraint handling.
Add support for a new DisableEncoding property which contains a list of
registers that should not be encoded by the generated code emitter. Convert
the codeemitter generator to use this, fixing some PPC JIT regressions.
llvm-svn: 31769
that there were two input operands before the variable operand portion. This
*happened* to be true for all call instructions, which took a chain and a
destination, but was not true for the PPC BCTRL instruction, whose destination
is implicit.
Making this code more general allows elimination of the custom selection logic
for BCTRL.
llvm-svn: 31732
X86ISD::CMP, etc.) instead of SDNode names (add, x86cmp, etc). We now allow
multiple SDNodes to map to the same SelectionDAG node (e.g. store, indexed
store).
llvm-svn: 31575
chain operand to point to the load being folded. Now we relax this, traversing
up the chain, if it doesn't reach the load, then it's ok. We will create a
TokenFactor (of all the chain operands and the load's chain) to capture all
the control flow dependencies.
llvm-svn: 30897
The dag/inst combiners often 'simplify' the masked value based on whether
or not the bits are live or known zero/one. This is good and dandy, but
often causes special case patterns to fail, such as alpha's CMPBGE pattern,
which looks like "(set GPRC:$RC, (setuge (and GPRC:$RA, 255), (and GPRC:$RB, 255)))".
Here the pattern for (and X, 255) should match actual dags like (and X, 254) if
the dag combiner proved that the missing bits are already zero (one for 'or').
For CodeGen/Alpha/cmpbge.ll:test2 for example, this results in:
sll $16,1,$0
cmpbge $0,$17,$0
ret $31,($26),1
instead of:
sll $16,1,$0
and $0,254,$0
and $17,255,$1
cmpule $1,$0,$0
ret $31,($26),1
... and requires no target-specific code.
llvm-svn: 30871
AggregateString += "\0\0";
Doesn't add two nuls to the AggregateString (for obvious reasons), which
broke the asmprinter when the first character of an asm string was not
literal text.
llvm-svn: 30625
the branch's chain is also produced by cmp.
[ch, r : ld]
^ ^
| |
[XX]--/ \- [flag : cmp]
^ ^
| |
\---[br flag]-
Remove an isel check which prevents loads from being folded into cmp / test
instructions.
2) Whenever possible, delete a selected node to allow more load folding
opportunities. Note not all nodes can be deleted after it has been
selected. Some may have simply morphed; some have not changed at all (e.g.
EntryToken).
llvm-svn: 30242
actually *removes* one of the operands, instead of just assigning both operands
the same register. This make reasoning about instructions unnecessarily complex,
because you need to know if you are before or after register allocation to match
up operand #'s with the target description file.
Changing this also gets rid of a bunch of hacky code in various places.
This patch also includes changes to fold loads into cmp/test instructions in
the X86 backend, along with a significant simplification to the X86 spill
folding code.
llvm-svn: 30108
- Clean up the code generated by tablegen:
* AddToISelQueue now takes one argument.
* ComplexPattern matching condition can now be shared.
* Eliminate passing unnecessary arguments to emit routines.
* Eliminate some unneeded SDOperand declarations in select routines.
* Other minor clean ups.
- This reduces foot print slightly: X86ISelDAGToDAG.o is reduced from 971k
to 823k.
llvm-svn: 29892
introduced by previous commit.
- SelectCode now returns a SDNode*. If it is not null, the selected node
produces the same number of results as the input node. The seletion loop
is responsible for calling ReplaceAllUsesWith() to replace the input node
with the output target node. For other cases, e.g. when load is folded,
the selection code is responsible for calling ReplaceAllUsesOfValueWith()
and SelectCode returns NULL.
- Other clean ups.
llvm-svn: 29602
in the start of an array and a count of operands where applicable. In many
cases, the number of operands is known, so this static array can be allocated
on the stack, avoiding the heap. In many other cases, a SmallVector can be
used, which has the same benefit in the common cases.
I updated a lot of code calling getNode that takes a vector, but ran out of
time. The rest of the code should be updated, and these methods should be
removed.
We should also do the same thing to eliminate the methods that take a
vector of MVT::ValueTypes.
It would be extra nice to convert the dagiselemitter to avoid creating vectors
for operands when calling getTargetNode.
llvm-svn: 29566
per possible ValueType of the node. e.g. Select_add is split into Select_add_i8,
Select_add_i16, etc.
For opcodes which do not produce a non-chain result, it is split on the
ValueType of its first non-chain operand. e.g. Select_store.
On X86 / Mac OS X, Select_store used to be the largest function. It had a stack
frame size of 8.5k. Now the largest one is Store_i32 with a frame size of 3.1k.
llvm-svn: 29404
instructions not handled would have a case value of #0, throwing things off.
This marginally shrinks the X86 asmprinter, but shrinks the sparc asmwriter
by 25 lines.
llvm-svn: 29187
series of identical commands, handle them all with one switch. In the case
of the x86 at&t asm printer, only 3 switches are needed for all instructions.
llvm-svn: 29184
based and less switch-statements-with-hundreds-of-cases based. This shrinks
the x86 asmprinters to about 1/3 their previous size.
Other improvements coming.
llvm-svn: 29177
code that emit target specific nodes into emit functions that are uniquified
and shared among selection routines.
e.g. This reduces X86ISelDAGToDAG.o (release) from ~2M to ~1.5M. Stack frame
size of Select_store from ~13k down to ~8k.
This is the first step. Further work to enable more sharing will follow.
llvm-svn: 29158
and index into it, instead of emitting it like this:
static const char * const OpStrs[] = {
"PHINODE\n", // PHI
0, // INLINEASM
"adc ", // ADC32mi
"adc ", // ADC32mi8
...
The old way required thousands of relocations that slows down link time and
dynamic load times.
This also cuts about 10K off each of the X86 asmprinters, and should shrink
the others as well.
llvm-svn: 29152
because information about one can help refine the other. This allows us to
write:
def : Pat<(i32 (extload xaddr:$src, i8)),
(LBZX xaddr:$src)>;
as:
def : Pat<(extload xaddr:$src, i8),
(LBZX xaddr:$src)>;
because tblgen knows LBZX returns i32.
llvm-svn: 28865
instruction, and the result type of the instruction to refine the pattern.
This allows us to write things like this:
def : Pat<(v2i64 (bitconvert (v16i8 VR128:$src))), (v2i64 VR128:$src)>;
as:
def : Pat<(v2i64 (bitconvert (v16i8 VR128:$src))), (VR128:$src)>
and fixes a ppc64 issue.
llvm-svn: 28863
operands. e.g.
def CALL32r : I<0xFF, MRM2r, (ops GR32:$dst, variable_ops),
"call {*}$dst", [(X86call GR32:$dst)]>;
TableGen should emit operand informations for the "required" operands.
Added a target instruction info flag M_VARIABLE_OPS to indicate the target
instruction may have more operands in addition to the minimum required
operands.
llvm-svn: 28791
a cycle. This increase the search space and will increase compile time (in
practice it appears to be small, e.g. 176.gcc goes from 62 sec to 65 sec)
that will be addressed later.
llvm-svn: 28476
