4c059f4962
Minor speedup for legalize by avoiding some malloc traffic
...
llvm-svn: 30477
2006-09-19 04:51:23 +00:00
1fc7c363e6
Fix a typo.
...
llvm-svn: 30474
2006-09-18 23:28:33 +00:00
4bfaf0bd2c
Allow i32 UDIV, SDIV, UREM, SREM to be expanded into libcalls.
...
llvm-svn: 30470
2006-09-18 21:49:04 +00:00
c50458fb90
absolute addresses must match pointer size
...
llvm-svn: 30461
2006-09-18 17:59:35 +00:00
e50f5d1fb1
Oh yeah, this is needed too
...
llvm-svn: 30407
2006-09-16 05:08:34 +00:00
1b63391fdf
simplify control flow, no functionality change
...
llvm-svn: 30403
2006-09-16 00:21:44 +00:00
fbadbda6ba
Allow custom expand of mul
...
llvm-svn: 30402
2006-09-16 00:09:24 +00:00
46d710e6ea
Fold (X & C1) | (Y & C2) -> (X|Y) & C3 when possible.
...
This implements CodeGen/X86/and-or-fold.ll
llvm-svn: 30379
2006-09-14 21:11:37 +00:00
97614c86ce
Split rotate matching code out to its own function. Make it stronger, by
...
matching things like ((x >> c1) & c2) | ((x << c3) & c4) to (rot x, c5) & c6
llvm-svn: 30376
2006-09-14 20:50:57 +00:00
84cc1f7cb8
If LSR went through a lot of trouble to put constants (e.g. the addr of a global
...
in a specific BB, don't undo this!). This allows us to compile
CodeGen/X86/loop-hoist.ll into:
_foo:
xorl %eax, %eax
*** movl L_Arr$non_lazy_ptr, %ecx
movl 4(%esp), %edx
LBB1_1: #cond_true
movl %eax, (%ecx,%eax,4)
incl %eax
cmpl %edx, %eax
jne LBB1_1 #cond_true
LBB1_2: #return
ret
instead of:
_foo:
xorl %eax, %eax
movl 4(%esp), %ecx
LBB1_1: #cond_true
*** movl L_Arr$non_lazy_ptr, %edx
movl %eax, (%edx,%eax,4)
incl %eax
cmpl %ecx, %eax
jne LBB1_1 #cond_true
LBB1_2: #return
ret
This was noticed in 464.h264ref. This doesn't usually affect PPC,
but strikes X86 all the time.
llvm-svn: 30290
2006-09-13 06:02:42 +00:00
72b503bcad
Compile X << 1 (where X is a long-long) to:
...
addl %ecx, %ecx
adcl %eax, %eax
instead of:
movl %ecx, %edx
addl %edx, %edx
shrl $31, %ecx
addl %eax, %eax
orl %ecx, %eax
and to:
addc r5, r5, r5
adde r4, r4, r4
instead of:
slwi r2,r9,1
srwi r0,r11,31
slwi r3,r11,1
or r2,r0,r2
on PPC.
llvm-svn: 30284
2006-09-13 03:50:39 +00:00
45fe3bc72c
Added support for machine specific constantpool values. These are useful for
...
representing expressions that can only be resolved at link time, etc.
llvm-svn: 30278
2006-09-12 21:00:35 +00:00
2e0dfb0b16
This code was trying too hard. By eliminating redundant edges in the CFG
...
due to switch cases going to the same place, it make #pred != #phi entries,
breaking live interval analysis.
This fixes 458.sjeng on x86 with llc.
llvm-svn: 30236
2006-09-10 06:36:57 +00:00
f0359b343a
Implement the fpowi now by lowering to a libcall
...
llvm-svn: 30225
2006-09-09 06:03:30 +00:00
e4bbb6c341
Allow targets to custom lower expanded BIT_CONVERT's
...
llvm-svn: 30217
2006-09-09 00:20:27 +00:00
707339a57b
Fix CodeGen/Generic/2006-09-06-SwitchLowering.ll, a bug where SDIsel inserted
...
too many phi operands when lowering a switch to branches in some cases.
llvm-svn: 30142
2006-09-07 01:59:34 +00:00
0dce3311c4
Change the default to 0, which means 'default'.
...
llvm-svn: 30114
2006-09-05 17:39:15 +00:00
af23f9b5f6
Completely eliminate def&use operands. Now a register operand is EITHER a
...
def operand or a use operand.
llvm-svn: 30109
2006-09-05 02:31:13 +00:00
373be1d1a2
forgot this
...
llvm-svn: 30097
2006-09-04 07:44:11 +00:00
e93762d36e
Allow legalizer to expand ISD::MUL using only MULHS in the rare case that is
...
possible and the target only supports MULHS.
llvm-svn: 30022
2006-09-01 18:17:58 +00:00
31305c45da
DAG combiner fix for rotates. Previously the outer-most condition checks
...
for ROTL availability. This prevents it from forming ROTR for targets that
has ROTR only.
llvm-svn: 29997
2006-08-31 07:41:12 +00:00
e5570a4c3f
Move isCommutativeBinOp from SelectionDAG.cpp and DAGCombiner.cpp out. Make it a static method of SelectionDAG.
...
llvm-svn: 29951
2006-08-29 06:42:35 +00:00
3d27be1333
s|llvm/Support/Visibility.h|llvm/Support/Compiler.h|
...
llvm-svn: 29911
2006-08-27 12:54:02 +00:00
849f4bf8dd
Eliminate SelectNodeTo() and getTargetNode() variants which take more than
...
3 SDOperand operands. They are replaced by versions which take an array
of SDOperand and the number of operands.
llvm-svn: 29905
2006-08-27 08:08:54 +00:00
34b70eea5c
SelectNodeTo now returns a SDNode*.
...
llvm-svn: 29901
2006-08-26 08:00:10 +00:00
451b099113
Fix PR861
...
llvm-svn: 29796
2006-08-21 20:24:53 +00:00
d86418ab20
switch the SUnit pred/succ sets from being std::sets to being smallvectors.
...
This reduces selectiondag time on kc++ from 5.43s to 4.98s (9%). More
significantly, this speeds up the default ppc scheduler from ~1571ms to 1063ms,
a 33% speedup.
llvm-svn: 29743
2006-08-17 00:09:56 +00:00
65879caf07
minor changes.
...
llvm-svn: 29740
2006-08-16 22:57:46 +00:00
a4f3625c23
Use the appropriate typedef
...
llvm-svn: 29730
2006-08-16 20:59:32 +00:00
a5a3eafbd0
Start using SDVTList more consistently
...
llvm-svn: 29711
2006-08-15 19:11:05 +00:00
f98411a220
add a new SDVTList type and new SelectionDAG::getVTList methods to streamline
...
the creation of canonical VTLists.
llvm-svn: 29709
2006-08-15 17:46:01 +00:00
bd8877744b
eliminate use of getNode that takes vector of valuetypes.
...
llvm-svn: 29687
2006-08-14 23:53:35 +00:00
3bf4be453f
Add a new getNode() method that takes a pointer to an already-intern'd list
...
of value-type nodes. This avoids having to do mallocs for std::vectors of
valuetypes when a node returns more than one type.
llvm-svn: 29685
2006-08-14 23:31:51 +00:00
e93a39f2d7
remove SelectionDAG::InsertISelMapEntry, it is dead
...
llvm-svn: 29677
2006-08-14 22:24:39 +00:00
63268f0672
Add code to resize the CSEMap hash table. This doesn't speedup codegen of
...
kimwitu, but seems like a good idea from a "avoid performance cliffs" standpoint :)
llvm-svn: 29675
2006-08-14 22:19:25 +00:00
8e37283d8b
Add the actual constant to the hash for ConstantPool nodes. Thanks to
...
Rafael Espindola for pointing this out.
llvm-svn: 29669
2006-08-14 20:12:44 +00:00
0a60294fa0
Switch to using SuperFastHash instead of adding all elements together. This
...
doesn't significantly improve performance but it helps a small amount.
llvm-svn: 29642
2006-08-12 01:07:10 +00:00
04aa034f38
Switch NodeID to track 32-bit chunks instead of 8-bit chunks, for a 2.5%
...
speedup in isel time.
llvm-svn: 29640
2006-08-11 23:55:53 +00:00
0c2e5412bb
Remove 8 more std::map's.
...
llvm-svn: 29631
2006-08-11 21:55:30 +00:00
3f16b201e2
Move the BBNodes, GlobalValues, TargetGlobalValues, Constants, TargetConstants,
...
RegNodes, and ValueNodes maps into the CSEMap.
llvm-svn: 29626
2006-08-11 21:01:22 +00:00
fcb16470ec
eliminate the NullaryOps map, use CSEMap instead.
...
llvm-svn: 29621
2006-08-11 18:38:11 +00:00
6f22ebd8be
change internal impl of dag combiner so that calls to CombineTo never have to
...
make a temporary vector.
llvm-svn: 29618
2006-08-11 17:56:38 +00:00
a2f4086828
Change one ReplaceAllUsesWith method to take an array of operands to replace
...
instead of a vector of operands.
llvm-svn: 29616
2006-08-11 17:46:28 +00:00
c24a1d3093
Start eliminating temporary vectors used to create DAG nodes. Instead, pass
...
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
2006-08-08 02:23:42 +00:00
97af9d5d3a
Eliminate some malloc traffic by allocating vectors on the stack. Change some
...
method that took std::vector<SDOperand> to take a pointer to a first operand
and #operands.
This speeds up isel on kc++ by about 3%.
llvm-svn: 29561
2006-08-08 01:09:31 +00:00
1ee75ce65d
Revamp the "CSEMap" datastructure used in the SelectionDAG class. This
...
eliminates a bunch of std::map's in the SelectionDAG, replacing them with a
home-grown hashtable.
This is still a work in progress: not all the maps have been moved over and the
hashtable never resizes. That said, this still speeds up llc 20% on kimwitu++
with -fast -regalloc=local using a release build.
llvm-svn: 29550
2006-08-07 23:03:03 +00:00
445b91a041
Clear TopOrder before assigning topological order. Some clean ups.
...
llvm-svn: 29546
2006-08-07 22:13:29 +00:00
1640ae5a84
Reverse the FlaggedNodes after scanning up for flagged preds or else the order would be reversed.
...
llvm-svn: 29545
2006-08-07 22:12:12 +00:00
8927c875bb
Make SelectionDAG::RemoveDeadNodes iterative instead of recursive, which
...
also make it simpler.
llvm-svn: 29524
2006-08-04 17:45:20 +00:00
a5b707e3ad
Copy the liveins for the first block. PR859
...
llvm-svn: 29511
2006-08-03 20:51:06 +00:00
524c1a21f2
Work around a GCC 3.3.5 bug noticed by a user.
...
llvm-svn: 29490
2006-08-03 00:18:59 +00:00
bba1ebda32
- Change AssignTopologicalOrder to return vector of SDNode* by reference.
...
- Tweak implementation to avoid using std::map.
llvm-svn: 29479
2006-08-02 22:00:34 +00:00
29e635d3c9
Final polish on machine pass registries.
...
llvm-svn: 29471
2006-08-02 12:30:23 +00:00
17c67efe8a
Now that the ISel is available, it's possible to create a default instruction
...
scheduler creator.
llvm-svn: 29452
2006-08-01 19:14:14 +00:00
03593f72db
1. Change use of "Cache" to "Default".
...
2. Added argument to instruction scheduler creators so the creators can do
special things.
3. Repaired target hazard code.
4. Misc.
More to follow.
llvm-svn: 29450
2006-08-01 18:29:48 +00:00
95eda5b1f3
Introducing plugable register allocators and instruction schedulers.
...
llvm-svn: 29434
2006-08-01 14:21:23 +00:00
9631a60020
Added AssignTopologicalOrder() to assign each node an unique id based on their topological order.
...
llvm-svn: 29431
2006-08-01 08:20:41 +00:00
6ae6ac1216
PIC jump table entries are always 32-bit even in 64-bit mode.
...
llvm-svn: 29422
2006-08-01 01:03:13 +00:00
b572401bea
Remove InFlightSet hack. No longer needed.
...
llvm-svn: 29373
2006-07-28 00:47:19 +00:00
efc312a5c7
Code cleanups, per review
...
llvm-svn: 29347
2006-07-27 16:46:58 +00:00
acb606ff33
AssignNodeIds should return unsigned.
...
llvm-svn: 29343
2006-07-27 07:36:47 +00:00
29eefc164c
AssignNodeIds assign each node in the DAG an unique id.
...
llvm-svn: 29337
2006-07-27 06:39:06 +00:00
85ea83e821
Add some advice
...
llvm-svn: 29324
2006-07-27 04:24:14 +00:00
787565024a
Support jump tables when in PIC relocation model
...
llvm-svn: 29318
2006-07-27 01:13:04 +00:00
4488f0c303
Fix a case where LegalizeAllNodesNotLeadingTo could take exponential time.
...
This manifested itself as really long time to compile
Regression/CodeGen/Generic/2003-05-28-ManyArgs.ll on ppc.
This is PR847.
llvm-svn: 29313
2006-07-26 23:55:56 +00:00
421475cd3b
For PR780:
...
1. Move IncludeFile.h to System library
2. Move IncludeFile.cpp to System library
3. #1 and #2 required to prevent cyclic library dependencies for libSystem
4. Convert all existing uses of Support/IncludeFile.h to System/IncludeFile.h
5. Add IncludeFile support to various lib/System classes.
6. Add new lib/System classes to LinkAllVMCore.h
All this in an attempt to pull in lib/System to what's required for VMCore
llvm-svn: 29287
2006-07-26 16:18:00 +00:00
658b9476f0
Initialize some variables the compiler warns about.
...
llvm-svn: 29277
2006-07-25 20:44:41 +00:00
4e153f1b91
Use an enumeration to eliminate data relocations.
...
llvm-svn: 29249
2006-07-21 20:57:35 +00:00
7c970b98d0
If a shuffle is a splat, check if the argument is a build_vector with all elements being the same. If so, return the argument.
...
llvm-svn: 29242
2006-07-21 08:25:53 +00:00
55782c6c41
Build more debugger/selectiondag libraries as archives instead of .o files.
...
This works around bugs in some versions of the cygwin linker.
Patch contributed by Anton Korobeynikov.
llvm-svn: 29239
2006-07-21 00:10:47 +00:00
8472e0c4af
If a shuffle is unary, i.e. one of the vector argument is not needed, turn the
...
operand into a undef and adjust mask accordingly.
llvm-svn: 29232
2006-07-20 22:44:41 +00:00
b030532910
Mems can be in the output list also. This is the second half of a fix for
...
PR833
llvm-svn: 29224
2006-07-20 19:02:21 +00:00
ec104a2b41
80 cols
...
llvm-svn: 29221
2006-07-20 17:43:27 +00:00
c496b418b5
Reduce number of exported symbols
...
llvm-svn: 29220
2006-07-20 17:28:38 +00:00
c0973edc69
Add an out-of-line virtual method for the sdnode class to give it a home.
...
llvm-svn: 29192
2006-07-19 00:00:37 +00:00
f7300b2706
It was pointed out that DEBUG() is only available with -debug.
...
llvm-svn: 29106
2006-07-11 18:25:13 +00:00
c3d341ea98
Ensure that dump calls that are associated with asserts are removed from
...
non-debug build.
llvm-svn: 29105
2006-07-11 17:58:07 +00:00
1b8ea1f5ba
Fix CodeGen/Alpha/2006-07-03-ASMFormalLowering.ll and PR818.
...
llvm-svn: 29099
2006-07-11 01:40:09 +00:00
d19938834b
Ugly hack! Add helper functions InsertInFlightSetEntry and
...
RemoveInFlightSetEntry. They are used in place of direct set operators to
reduce instruction selection function stack size.
llvm-svn: 28987
2006-06-29 23:57:05 +00:00
996795b0dd
Use hidden visibility to make symbols in an anonymous namespace get
...
dropped. This shrinks libllvmgcc.dylib another 67K
llvm-svn: 28975
2006-06-28 23:17:24 +00:00
e097e6f7c7
Shave another 27K off libllvmgcc.dylib with visibility hidden
...
llvm-svn: 28973
2006-06-28 22:17:39 +00:00
54a34cd20b
Mark these two classes as hidden, shrinking libllbmgcc.dylib by 25K
...
llvm-svn: 28970
2006-06-28 21:58:30 +00:00
710b3d5ea1
Fix CodeGen/Generic/2006-06-28-SimplifySetCCCrash.ll
...
llvm-svn: 28965
2006-06-28 18:29:47 +00:00
ee7eaa25cf
For PR801:
...
Refactor the Graph writing code to use a common implementation which is
now in lib/Support/GraphWriter.cpp. This completes the PR.
Patch by Anton Korobeynikov. Thanks, Anton!
llvm-svn: 28925
2006-06-27 16:49:46 +00:00
ef9e07d3f0
Consistency. EXTRACT_ELEMENT index operand should have ptr type.
...
llvm-svn: 28795
2006-06-15 08:11:54 +00:00
55772ccfd6
Instructions with variable operands (variable_ops) can have a number required
...
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
2006-06-15 07:22:16 +00:00
32d92e004d
Make sure to update the CFG correctly if a switch only has a default dest.
...
This fixes CodeGen/Generic/2006-06-12-LowerSwitchCrash.ll
llvm-svn: 28755
2006-06-12 18:25:29 +00:00
0e57b2cb92
Start on my todo list
...
llvm-svn: 28752
2006-06-12 16:07:18 +00:00
c03a9259c0
Fix X86/inline-asm.ll:test2, a case where an input value was implicitly
...
truncated.
llvm-svn: 28733
2006-06-08 18:27:11 +00:00
705948d742
Fix Regression/CodeGen/X86/inline-asm.ll, a case where inline asm causes
...
implement extension of a register.
llvm-svn: 28731
2006-06-08 18:22:48 +00:00
614cb2ff82
For PR798:
...
Provide GraphViz support for MingW32. Patch provided by Anton Korobeynikov
llvm-svn: 28688
2006-06-05 16:26:06 +00:00
a647c7ff42
Use archive libraries instead of object files for VMCore, BCReader,
...
BCWriter, and bzip2 libraries. Adjust the various makefiles to accommodate
these changes. This was done to speed up link times.
llvm-svn: 28610
2006-06-01 01:30:27 +00:00
0c0996a97b
commuteInstruction() does not always create a new MI!
...
llvm-svn: 28592
2006-05-31 18:03:39 +00:00
9d91caa053
Eliminate a memory leak.
...
llvm-svn: 28585
2006-05-31 07:13:03 +00:00
64d2846017
visitVBinOp: Can't fold divide by zero!
...
llvm-svn: 28584
2006-05-31 06:08:35 +00:00
d12c97d23a
Make sure the register pressure reduction schedulers work for non-uniform
...
latency targets, e.g. PPC32.
llvm-svn: 28561
2006-05-30 18:05:39 +00:00
61e9f0d680
When a priority_queue is empty, the behavior of top() operator is
...
non-deterministic. Returns NULL when it's empty!
llvm-svn: 28560
2006-05-30 18:04:34 +00:00
8f872d2091
Fix a nasty dag combiner bug that caused nondeterminstic crashes (MY FAVORITE!):
...
SimplifySelectOps would eliminate a Select, delete it, then return true.
The clients would see that it did something and return null.
The top level would see a null return, and decide that nothing happened,
proceeding to process the node in other ways: boom.
The fix is simple: clients of SimplifySelectOps should return the select
node itself.
In order to catch really obnoxious boogs like this in the future, add an
assert that nodes are not deleted. We do this by checking for a sentry node
type that the SDNode dtor sets when a node is destroyed.
llvm-svn: 28514
2006-05-27 00:43:02 +00:00
21dee4e0b2
Make CALL node consistent with RET node. Signness of value has type MVT::i32
...
instead of MVT::i1. Either is fine except MVT::i32 is probably a legal type
for most (if not all) platforms while MVT::i1 is not.
llvm-svn: 28511
2006-05-26 23:13:20 +00:00
a2e9953c54
Change RET node to include signness information of the return values. e.g.
...
RET chain, value1, sign1, value2, sign2
llvm-svn: 28509
2006-05-26 23:09:09 +00:00
009f5f55f7
Turn on -sched-commute-nodes by default.
...
llvm-svn: 28465
2006-05-25 08:37:31 +00:00
4582771f3f
CALL node change: now including signness of every argument.
...
llvm-svn: 28461
2006-05-25 00:55:32 +00:00
aa2372562e
Patches to make the LLVM sources more -pedantic clean. Patch provided
...
by Anton Korobeynikov! This is a step towards closing PR786.
llvm-svn: 28447
2006-05-24 17:04:05 +00:00
ac4f66ff24
-enable-unsafe-fp-math implies -enable-finite-only-fp-math
...
llvm-svn: 28437
2006-05-23 18:18:46 +00:00
df1d439849
Fix missing include
...
llvm-svn: 28435
2006-05-23 13:43:15 +00:00
1c5b7d12df
Incorrect SETCC CondCode used for FP comparisons.
...
llvm-svn: 28433
2006-05-23 06:40:47 +00:00
d8e2f6ebc1
lib/Target/Target.td
...
llvm-svn: 28386
2006-05-18 20:42:07 +00:00
7949c2e8b2
Fix the result of the call to use a correct vbitconvert. There is no need to
...
use getPackedTypeBreakdown at all here.
llvm-svn: 28365
2006-05-17 20:49:36 +00:00
938155ca57
Correct a previous patch which broke CodeGen/PowerPC/vec_call.ll
...
llvm-svn: 28364
2006-05-17 20:43:21 +00:00
751cd7653d
Fixed a LowerCallTo and LowerArguments bug. They were introducing illegal
...
VBIT_VECTOR nodes. There were some confusion about the semantics of
getPackedTypeBreakdown(). e.g. for <4 x f32> it returns 1 and v4f32, not 4,
and f32.
llvm-svn: 28352
2006-05-17 18:16:39 +00:00
62f1b83c0e
When we legalize target nodes, do not use getNode to create a new node,
...
use UpdateNodeOperands to just update the operands! This is important because
getNode will allocate a new node if the node returns a flag and this breaks
assumptions in the legalizer that you can legalize some things multiple times
and get exactly the same results.
This latent bug was exposed by my ppc patch last night, and this fixes
gsm/toast.
llvm-svn: 28348
2006-05-17 18:00:08 +00:00
a1cec0106a
Add an assertion, avoid some unneeded work for each call. No functionality
...
change.
llvm-svn: 28347
2006-05-17 17:55:45 +00:00
b77ba73a29
Add support for calls that pass and return legal vectors.
...
llvm-svn: 28340
2006-05-16 23:39:44 +00:00
aaa23d953f
Add a new ISD::CALL node, make the default impl of TargetLowering::LowerCallTo
...
produce it.
llvm-svn: 28338
2006-05-16 22:53:20 +00:00
1dc9ec5874
Move this code to a common place
...
llvm-svn: 28329
2006-05-16 17:42:15 +00:00
3d82699605
Add a chain to FORMAL_ARGUMENTS. This is a minimal port of the X86 backend,
...
it doesn't currently use/maintain the chain properly. Also, make the
X86ISelLowering.cpp file 80-col clean.
llvm-svn: 28320
2006-05-16 06:45:34 +00:00
957cb6733a
Move function-live-in-handling code from the sdisel code to the scheduler.
...
This code should be emitted after legalize, so it can't be in sdisel.
Note that the EmitFunctionEntryCode hook should be updated to operate on the
DAG. The X86 backend is the only one currently using this hook.
llvm-svn: 28315
2006-05-16 06:10:58 +00:00
5f0edfb849
Legalize FORMAL_ARGUMENTS nodes correctly, we don't want to legalize them once
...
for each argument.
llvm-svn: 28313
2006-05-16 05:49:56 +00:00
99f2f79e2f
Fixing 2006-05-01-SchedCausingSpills.ll; some clean up
...
llvm-svn: 28279
2006-05-13 08:22:24 +00:00
d1915cfa6f
Revert an un-intended change
...
llvm-svn: 28278
2006-05-13 05:53:47 +00:00
69a0ce6261
Merge identical code.
...
llvm-svn: 28274
2006-05-13 02:11:14 +00:00
53cdb2f2b0
Remove dead vars
...
llvm-svn: 28255
2006-05-12 18:06:45 +00:00
da076e41ab
remove dead vars
...
llvm-svn: 28254
2006-05-12 18:04:28 +00:00
afe72481f6
Comment out dead variables
...
llvm-svn: 28252
2006-05-12 17:57:54 +00:00
8c02c3f41a
Compile:
...
%tmp152 = setgt uint %tmp144, %tmp149 ; <bool> [#uses=1]
%tmp159 = setlt uint %tmp144, %tmp149 ; <bool> [#uses=1]
%bothcond2 = or bool %tmp152, %tmp159 ; <bool> [#uses=1]
To setne, not setune, which causes an assertion fault.
llvm-svn: 28244
2006-05-12 17:03:46 +00:00
8c2c1e90c4
Refactor a bunch of includes so that TargetMachine.h doesn't have to include
...
TargetData.h. This should make recompiles a bit faster with my current
TargetData tinkering.
llvm-svn: 28238
2006-05-12 06:33:49 +00:00
095c9d9b7f
Duh. That could take a long time.
...
llvm-svn: 28235
2006-05-12 06:05:18 +00:00
66adee93aa
Two simplifications for token factor nodes: simplify tf(x,x) -> x.
...
simplify tf(x,y,y,z) -> tf(x,y,z).
llvm-svn: 28233
2006-05-12 05:01:37 +00:00
afed73eebe
Add capability to scheduler to commute nodes for profit.
...
If a two-address code whose first operand has uses below, it should be commuted
when possible.
llvm-svn: 28230
2006-05-12 01:58:24 +00:00
d38c22bdd3
Refactor scheduler code. Move register-reduction list scheduler to a
...
separate file. Added an initial implementation of top-down register pressure
reduction list scheduler.
llvm-svn: 28226
2006-05-11 23:55:42 +00:00
9665ba053f
Templatify RegReductionPriorityQueue
...
llvm-svn: 28212
2006-05-10 06:16:44 +00:00
1a225d23ae
Fix PR773
...
llvm-svn: 28207
2006-05-09 18:20:51 +00:00
7d693898ee
Add pseudo dependency to force a def&use operand to be scheduled last (unless
...
the distance between the def and another use is much longer). This is under
option control for now "-sched-lower-defnuse".
llvm-svn: 28201
2006-05-09 07:13:34 +00:00
2c74848af1
Debugging info
...
llvm-svn: 28200
2006-05-09 06:55:15 +00:00
446e1ef26a
Make the case I just checked in stronger. Now we compile this:
...
short test2(short X, short x) {
int Y = (short)(X+x);
return Y >> 1;
}
to:
_test2:
add r2, r3, r4
extsh r2, r2
srawi r3, r2, 1
blr
instead of:
_test2:
add r2, r3, r4
extsh r2, r2
srwi r2, r2, 1
extsh r3, r2
blr
llvm-svn: 28175
2006-05-08 21:18:59 +00:00
29062da0ac
Implement and_sext.ll:test3, generating:
...
_test4:
srawi r3, r3, 16
blr
instead of:
_test4:
srwi r2, r3, 16
extsh r3, r2
blr
for:
short test4(unsigned X) {
return (X >> 16);
}
llvm-svn: 28174
2006-05-08 20:59:41 +00:00
2935d8190c
Compile this:
...
short test4(unsigned X) {
return (X >> 16);
}
to:
_test4:
movl 4(%esp), %eax
sarl $16, %eax
ret
instead of:
_test4:
movl $-65536, %eax
andl 4(%esp), %eax
sarl $16, %eax
ret
llvm-svn: 28171
2006-05-08 20:51:54 +00:00
78da6792e7
Fold shifts with undef operands.
...
llvm-svn: 28167
2006-05-08 17:29:49 +00:00
d7a19102d1
Make emission of jump tables a bit less conservative; they are now required
...
to be only 31.25% dense, rather than 75% dense.
llvm-svn: 28165
2006-05-08 16:51:36 +00:00
e5ce5bb6da
Fix PR772
...
llvm-svn: 28161
2006-05-08 01:35:01 +00:00
7e7bcf3a54
Simplify some code, add a couple minor missed folds
...
llvm-svn: 28152
2006-05-06 23:06:26 +00:00
751817c54f
constant fold sign_extend_inreg
...
llvm-svn: 28151
2006-05-06 23:05:41 +00:00
2a4d7b845b
remove cases handled elsewhere
...
llvm-svn: 28150
2006-05-06 22:43:44 +00:00
1ecb2a2dac
Use the new TargetLowering::ComputeNumSignBits method to eliminate
...
sign_extend_inreg operations. Though ComputeNumSignBits is still rudimentary,
this is enough to compile this:
short test(short X, short x) {
int Y = X+x;
return (Y >> 1);
}
short test2(short X, short x) {
int Y = (short)(X+x);
return Y >> 1;
}
into:
_test:
add r2, r3, r4
srawi r3, r2, 1
blr
_test2:
add r2, r3, r4
extsh r2, r2
srawi r3, r2, 1
blr
instead of:
_test:
add r2, r3, r4
srawi r2, r2, 1
extsh r3, r2
blr
_test2:
add r2, r3, r4
extsh r2, r2
srawi r2, r2, 1
extsh r3, r2
blr
llvm-svn: 28146
2006-05-06 09:30:03 +00:00
21cd99024a
When inserting casts, be careful of where we put them. We cannot insert
...
a cast immediately before a PHI node.
This fixes Regression/CodeGen/Generic/2006-05-06-GEP-Cast-Sink-Crash.ll
llvm-svn: 28143
2006-05-06 09:10:37 +00:00
907e392dba
Fold trunc(any_ext). This gives stuff like:
...
27,28c27
< movzwl %di, %edi
< movl %edi, %ebx
---
> movw %di, %bx
llvm-svn: 28137
2006-05-05 22:56:26 +00:00
57f8c5a387
Shrink shifts when possible.
...
llvm-svn: 28136
2006-05-05 22:53:17 +00:00
3d26577396
Fold (fpext (load x)) -> (extload x)
...
llvm-svn: 28130
2006-05-05 21:34:35 +00:00
3e3f2c63c3
More aggressively sink GEP offsets into loops. For example, before we
...
generated:
movl 8(%esp), %eax
movl %eax, %edx
addl $4316, %edx
cmpb $1, %cl
ja LBB1_2 #cond_false
LBB1_1: #cond_true
movl L_QuantizationTables720$non_lazy_ptr, %ecx
movl %ecx, (%edx)
movl L_QNOtoQuantTableShift720$non_lazy_ptr, %edx
movl %edx, 4460(%eax)
ret
...
Now we generate:
movl 8(%esp), %eax
cmpb $1, %cl
ja LBB1_2 #cond_false
LBB1_1: #cond_true
movl L_QuantizationTables720$non_lazy_ptr, %ecx
movl %ecx, 4316(%eax)
movl L_QNOtoQuantTableShift720$non_lazy_ptr, %ecx
movl %ecx, 4460(%eax)
ret
... which uses one fewer register.
llvm-svn: 28129
2006-05-05 21:17:49 +00:00
25a5283a86
Fold some common code.
...
llvm-svn: 28124
2006-05-05 06:32:04 +00:00
002ee91457
Implement:
...
// fold (and (sext x), (sext y)) -> (sext (and x, y))
// fold (or (sext x), (sext y)) -> (sext (or x, y))
// fold (xor (sext x), (sext y)) -> (sext (xor x, y))
// fold (and (aext x), (aext y)) -> (aext (and x, y))
// fold (or (aext x), (aext y)) -> (aext (or x, y))
// fold (xor (aext x), (aext y)) -> (aext (xor x, y))
llvm-svn: 28123
2006-05-05 06:31:05 +00:00
5ac4293606
Pull and through and/or/xor. This compiles some bitfield code to:
...
mov EAX, DWORD PTR [ESP + 4]
mov ECX, DWORD PTR [EAX]
mov EDX, ECX
add EDX, EDX
or EDX, ECX
and EDX, -2147483648
and ECX, 2147483647
or EDX, ECX
mov DWORD PTR [EAX], EDX
ret
instead of:
sub ESP, 4
mov DWORD PTR [ESP], ESI
mov EAX, DWORD PTR [ESP + 8]
mov ECX, DWORD PTR [EAX]
mov EDX, ECX
add EDX, EDX
mov ESI, ECX
and ESI, -2147483648
and EDX, -2147483648
or EDX, ESI
and ECX, 2147483647
or EDX, ECX
mov DWORD PTR [EAX], EDX
mov ESI, DWORD PTR [ESP]
add ESP, 4
ret
llvm-svn: 28122
2006-05-05 06:10:43 +00:00
812646aa0c
Implement a variety of simplifications for ANY_EXTEND.
...
llvm-svn: 28121
2006-05-05 05:58:59 +00:00
8d6fc20181
Factor some code, add these transformations:
...
// fold (and (trunc x), (trunc y)) -> (trunc (and x, y))
// fold (or (trunc x), (trunc y)) -> (trunc (or x, y))
// fold (xor (trunc x), (trunc y)) -> (trunc (xor x, y))
llvm-svn: 28120
2006-05-05 05:51:50 +00:00
78a7f0e05e
Fix VC++ compilation error.
...
llvm-svn: 28117
2006-05-05 01:47:05 +00:00
7a3ecf7993
Sink noop copies into the basic block that uses them. This reduces the number
...
of cross-block live ranges, and allows the bb-at-a-time selector to always
coallesce these away, at isel time.
This reduces the load on the coallescer and register allocator. For example
on a codec on X86, we went from:
1643 asm-printer - Number of machine instrs printed
419 liveintervals - Number of loads/stores folded into instructions
1144 liveintervals - Number of identity moves eliminated after coalescing
1022 liveintervals - Number of interval joins performed
282 liveintervals - Number of intervals after coalescing
1304 liveintervals - Number of original intervals
86 regalloc - Number of times we had to backtrack
1.90232 regalloc - Ratio of intervals processed over total intervals
40 spiller - Number of values reused
182 spiller - Number of loads added
121 spiller - Number of stores added
132 spiller - Number of register spills
6 twoaddressinstruction - Number of instructions commuted to coalesce
360 twoaddressinstruction - Number of two-address instructions
to:
1636 asm-printer - Number of machine instrs printed
403 liveintervals - Number of loads/stores folded into instructions
1155 liveintervals - Number of identity moves eliminated after coalescing
1033 liveintervals - Number of interval joins performed
279 liveintervals - Number of intervals after coalescing
1312 liveintervals - Number of original intervals
76 regalloc - Number of times we had to backtrack
1.88998 regalloc - Ratio of intervals processed over total intervals
1 spiller - Number of copies elided
41 spiller - Number of values reused
191 spiller - Number of loads added
114 spiller - Number of stores added
128 spiller - Number of register spills
4 twoaddressinstruction - Number of instructions commuted to coalesce
356 twoaddressinstruction - Number of two-address instructions
On this testcase, this change provides a modest reduction in spill code,
regalloc iterations, and total instructions emitted. It increases the number
of register coallesces.
llvm-svn: 28115
2006-05-05 01:04:50 +00:00
9add880566
Initial support for register pressure aware scheduling. The register reduction
...
scheduler can go into a "vertical mode" (i.e. traversing up the two-address
chain, etc.) when the register pressure is low.
This does seem to reduce the number of spills in the cases I've looked at. But
with x86, it's no guarantee the performance of the code improves.
It can be turned on with -sched-vertically option.
llvm-svn: 28108
2006-05-04 19:16:39 +00:00
469647bf38
Remove and simplify some more machineinstr/machineoperand stuff.
...
llvm-svn: 28105
2006-05-04 18:16:01 +00:00
10b71c0d08
Rename MO_VirtualRegister -> MO_Register. Clean up immediate handling.
...
llvm-svn: 28104
2006-05-04 18:05:43 +00:00
940cc978ef
Remove a bunch more SparcV9 specific stuff
...
llvm-svn: 28093
2006-05-04 01:15:02 +00:00
df4883971e
Finish up the initial jump table implementation by allowing jump tables to
...
not be 100% dense. Increase the minimum threshold for the number of cases
in a switch statement from 4 to 6 in order to create a jump table.
llvm-svn: 28079
2006-05-03 03:48:02 +00:00
ffef8b9412
Bottom up register pressure reduction work: clean up some hacks and enhanced
...
the heuristic to further reduce spills for several test cases. (Note, it may
not necessarily translate to runtime win!)
llvm-svn: 28076
2006-05-03 02:10:45 +00:00
20a631fde7
Refactor TargetMachine, pushing handling of TargetData into the target-specific subclasses. This has one caller-visible change: getTargetData() now returns a pointer instead of a reference.
...
This fixes PR 759.
llvm-svn: 28074
2006-05-03 01:29:57 +00:00
0d084fb9ca
Dis-favor stores more
...
llvm-svn: 28035
2006-05-01 09:20:44 +00:00
24e795496d
Bottom up register-pressure reduction scheduler now pushes store operations
...
up the schedule. This helps code that looks like this:
loads ...
computations (first set) ...
stores (first set) ...
loads
computations (seccond set) ...
stores (seccond set) ...
Without this change, the stores and computations are more likely to
interleave:
loads ...
loads ...
computations (first set) ...
computations (second set) ...
computations (first set) ...
stores (first set) ...
computations (second set) ...
stores (stores set) ...
This can increase the number of spills if we are unlucky.
llvm-svn: 28033
2006-05-01 09:14:40 +00:00
10ff7b27ce
Didn't mean ScheduleDAGList.cpp to make the last checkin.
...
llvm-svn: 28030
2006-05-01 08:56:34 +00:00
a656242690
Remove temp. option -spiller-check-liveout, it didn't cause any failure nor performance regressions.
...
llvm-svn: 28029
2006-05-01 08:54:57 +00:00
2b48a94413
Remove a bogus transformation. This fixes SingleSource/UnitTests/2006-01-23-InitializedBitField.c
...
with some changes I have to the new CFE.
llvm-svn: 28022
2006-04-28 23:33:20 +00:00
c5e8ce8b8c
Remove the temporary option: -no-isel-fold-inflight
...
llvm-svn: 28012
2006-04-28 18:54:11 +00:00
d43c5c6046
TargetLowering::LowerArguments should return a VBIT_CONVERT of
...
FORMAL_ARGUMENTS SDOperand in the return result vector.
llvm-svn: 28009
2006-04-28 05:25:15 +00:00
51ab4498e7
Added a temporary option -no-isel-fold-inflight to control whether a "inflight"
...
node can be folded.
llvm-svn: 28003
2006-04-28 02:09:19 +00:00
3784f3c57c
Insert a VBIT_CONVERT between a FORMAL_ARGUMENT node and its vector uses
...
(VAND, VADD, etc.). Legalizer will assert otherwise.
llvm-svn: 27991
2006-04-27 08:29:42 +00:00
393d96a56c
Fix Regression/CodeGen/Generic/2006-04-26-SetCCAnd.ll and
...
PR748.
llvm-svn: 27987
2006-04-27 05:01:07 +00:00
9618df1190
Don't forget return void.
...
llvm-svn: 27974
2006-04-25 23:03:35 +00:00
866b4b4d45
Fix the updating of the machine CFG when a PHI node was in a successor of
...
the jump table's range check block. This re-enables 100% dense jump tables
by default on PPC & x86
llvm-svn: 27952
2006-04-23 06:26:20 +00:00
ecb1dafd3d
Turn of jump tables for a bit, there are still some issues to work out with
...
updating the machine CFG.
llvm-svn: 27949
2006-04-22 23:51:56 +00:00
4ca2ea5b43
JumpTable support! What this represents is working asm and jit support for
...
x86 and ppc for 100% dense switch statements when relocations are non-PIC.
This support will be extended and enhanced in the coming days to support
PIC, and less dense forms of jump tables.
llvm-svn: 27947
2006-04-22 18:53:45 +00:00
b21d3bfd1f
The BFS scheduler is apparently nondeterminstic (causes many llvmgcc bootstrap
...
miscompares). Switch RISC targets to use the list-td scheduler, which isn't.
llvm-svn: 27933
2006-04-21 17:16:16 +00:00
662e940f73
Fix a couple more memory issues
...
llvm-svn: 27930
2006-04-21 15:32:26 +00:00
cc47ab3305
Fix a really subtle and obnoxious memory bug that caused issues with an
...
llvm-gcc4 boostrap. Whenever a node is deleted by the dag combiner, it
*must* be returned by the visit function, or the dag combiner will not
know that the node has been processed (and will, e.g., send it to the
target dag combine xforms).
llvm-svn: 27922
2006-04-20 23:55:59 +00:00
a320abc494
Turn a VAND into a VECTOR_SHUFFLE is applicable.
...
DAG combiner can turn a VAND V, <-1, 0, -1, -1>, i.e. vector clear elements,
into a vector shuffle with a zero vector. It only does so when TLI tells it
the xform is profitable.
llvm-svn: 27874
2006-04-20 08:56:16 +00:00
bc1b262725
Implement folding of a bunch of binops with undef
...
llvm-svn: 27863
2006-04-20 05:39:12 +00:00
73eb58e1a2
Simplify some code
...
llvm-svn: 27846
2006-04-19 23:17:50 +00:00
916ae0775e
Fix handling of calls in functions that use vectors. This fixes a crash on
...
the code in GCC PR26546.
llvm-svn: 27780
2006-04-17 22:10:08 +00:00
326870b40b
Codegen insertelement with constant insertion points as scalar_to_vector
...
and a shuffle. For this:
void %test2(<4 x float>* %F, float %f) {
%tmp = load <4 x float>* %F ; <<4 x float>> [#uses=2]
%tmp3 = add <4 x float> %tmp, %tmp ; <<4 x float>> [#uses=1]
%tmp2 = insertelement <4 x float> %tmp3, float %f, uint 2 ; <<4 x float>> [#uses=2]
%tmp6 = add <4 x float> %tmp2, %tmp2 ; <<4 x float>> [#uses=1]
store <4 x float> %tmp6, <4 x float>* %F
ret void
}
we now get this on X86 (which will get better):
_test2:
movl 4(%esp), %eax
movaps (%eax), %xmm0
addps %xmm0, %xmm0
movaps %xmm0, %xmm1
shufps $3, %xmm1, %xmm1
movaps %xmm0, %xmm2
shufps $1, %xmm2, %xmm2
unpcklps %xmm1, %xmm2
movss 8(%esp), %xmm1
unpcklps %xmm1, %xmm0
unpcklps %xmm2, %xmm0
addps %xmm0, %xmm0
movaps %xmm0, (%eax)
ret
instead of:
_test2:
subl $28, %esp
movl 32(%esp), %eax
movaps (%eax), %xmm0
addps %xmm0, %xmm0
movaps %xmm0, (%esp)
movss 36(%esp), %xmm0
movss %xmm0, 8(%esp)
movaps (%esp), %xmm0
addps %xmm0, %xmm0
movaps %xmm0, (%eax)
addl $28, %esp
ret
llvm-svn: 27765
2006-04-17 19:21:01 +00:00
91226e5799
Add support for promoting stores from one legal type to another, allowing us
...
to write one pattern for vector stores instead of 4.
llvm-svn: 27730
2006-04-16 01:36:45 +00:00
7e7ad593cc
Make these predicates return true for bit_convert(buildvector)'s as well as
...
buildvectors.
llvm-svn: 27723
2006-04-15 23:38:00 +00:00
086e986e94
Make this assertion better
...
llvm-svn: 27695
2006-04-14 06:08:35 +00:00
119266ea92
Promote vector AND, OR, and XOR
...
llvm-svn: 27632
2006-04-12 21:20:24 +00:00
be8a8933e6
Vector type promotion for ISD::LOAD and ISD::SELECT
...
llvm-svn: 27606
2006-04-12 16:33:18 +00:00
d3b504ae10
Implement support for the formal_arguments node. To get this, targets shouldcustom legalize it and remove their XXXTargetLowering::LowerArguments overload
...
llvm-svn: 27604
2006-04-12 16:20:43 +00:00
417b96b6dd
Don't memoize vloads in the load map! Don't memoize them anywhere here, let
...
getNode do it. This fixes CodeGen/Generic/2006-04-11-vecload.ll
llvm-svn: 27602
2006-04-12 03:25:41 +00:00
7256b0ae05
Only get Tmp2 for cases where number of operands is > 1. Fixed return void.
...
llvm-svn: 27586
2006-04-11 06:33:39 +00:00
6cf3bbbe17
add some todos
...
llvm-svn: 27580
2006-04-11 02:00:08 +00:00
2eb22eef7d
Add basic support for legalizing returns of vectors
...
llvm-svn: 27578
2006-04-11 01:31:51 +00:00
cb73b8d419
Missing break
...
llvm-svn: 27559
2006-04-10 18:54:36 +00:00
02274a5265
Add code generator support for VSELECT
...
llvm-svn: 27542
2006-04-08 22:22:57 +00:00
e1401e3610
Canonicalize vvector_shuffle(x,x) -> vvector_shuffle(x,undef) to enable patterns
...
to match again :)
llvm-svn: 27533
2006-04-08 05:34:25 +00:00
098c01e94e
Codegen shufflevector as VVECTOR_SHUFFLE
...
llvm-svn: 27529
2006-04-08 04:15:24 +00:00
101ea66813
add a sanity check: LegalizeOp should return a value that is the same type
...
as its input.
llvm-svn: 27528
2006-04-08 04:13:17 +00:00
Chris Lattner