Commit Graph

79 Commits

Author SHA1 Message Date
Chris Lattner
a16b7cb1d3 Two changes:
1) codegen a shift of a register as a shift, not an LEA.
2) teach the RA to convert a shift to an LEA instruction if it wants something
   in three-address form.

This gives us asm diffs like:

-       leal (,%eax,4), %eax
+       shll $2, %eax

which is faster on some processors and smaller on all of them.

and, more interestingly:

-       movl 24(%esi), %eax
-       leal (,%eax,4), %edi
+       movl 24(%esi), %edi
+       shll $2, %edi

Without #2, #1 was a significant pessimization in some cases.

This implements CodeGen/X86/shift-codegen.ll


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35204 91177308-0d34-0410-b5e6-96231b3b80d8
2007-03-20 06:08:29 +00:00
Bill Wendling
2f88dcdfb3 Added "padd*" support for MMX. Added MMX move stuff to X86InstrInfo so that
moves, loads, etc. are recognized.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@35031 91177308-0d34-0410-b5e6-96231b3b80d8
2007-03-08 22:09:11 +00:00
Jim Laskey
1ee2925742 Make LABEL a builtin opcode.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@33537 91177308-0d34-0410-b5e6-96231b3b80d8
2007-01-26 14:34:52 +00:00
Evan Cheng
258ff6726d convertToThreeAddress() is now responsible for updating live info as well as inserting the new MI's.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@32097 91177308-0d34-0410-b5e6-96231b3b80d8
2006-12-01 21:52:41 +00:00
Evan Cheng
c0f64ffab9 Change MachineInstr ctor's to take a TargetInstrDescriptor reference instead
of opcode and number of operands.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31947 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-27 23:37:22 +00:00
Evan Cheng
e23fc60854 Fix a potential bug: MOVPDI2DI, etc. are not copy instructions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31794 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-16 23:22:26 +00:00
Evan Cheng
6ce7dc2a97 Properly transfer kill / dead info.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31765 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-15 20:58:11 +00:00
Evan Cheng
7ce4578353 Matches MachineInstr changes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31712 91177308-0d34-0410-b5e6-96231b3b80d8
2006-11-13 23:36:35 +00:00
Chris Lattner
6ce644392e fix wonky indentation
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31298 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-30 22:27:23 +00:00
Chris Lattner
c24ff8ed12 add another target hook for branch folding.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31262 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-28 17:29:57 +00:00
Chris Lattner
9cd6875917 Implement support for branch condition reversal.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31099 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-21 05:52:40 +00:00
Chris Lattner
879d09cf13 Simplify code, no functionality change
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31097 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-21 05:42:09 +00:00
Chris Lattner
34a84ac81c allow insertion of a conditional branch with fall-through
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31095 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-21 05:34:23 +00:00
Chris Lattner
2a445add12 update assert message
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31093 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-21 04:42:29 +00:00
Chris Lattner
5860715e9d bugfix
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31074 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-20 20:44:34 +00:00
Chris Lattner
7fbe9723e3 Implement branch analysis/xform hooks required by the branch folding pass.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31065 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-20 17:42:20 +00:00
Chris Lattner
ae1dc40327 expose DWARF_LABEL opcode# so the branch folder can update debug info properly.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@31024 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-17 22:41:45 +00:00
Chris Lattner
d77ddbc0ba remove some dead code
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30938 91177308-0d34-0410-b5e6-96231b3b80d8
2006-10-13 20:40:42 +00:00
Chris Lattner
6458f1807d update comments
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30663 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-28 23:33:12 +00:00
Evan Cheng
25ab690a43 Committing X86-64 support.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30177 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-08 06:48:29 +00:00
Chris Lattner
2926869b4a Fix a long-standing wart in the code generator: two-address instruction lowering
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.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@30108 91177308-0d34-0410-b5e6-96231b3b80d8
2006-09-05 02:12:02 +00:00
Evan Cheng
55371739de Can't commute shufps. The high / low parts elements come from different vectors.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@29275 91177308-0d34-0410-b5e6-96231b3b80d8
2006-07-25 20:25:40 +00:00
Evan Cheng
efeaed8fb4 Commute shufps / shufpd.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28577 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-30 23:34:30 +00:00
Evan Cheng
a0eaf2d0c5 Somehow I lost a condition when I was shuffling some code around. Anyway,
only transform a shufps to pshufd when the first two operands are the same.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28575 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-30 22:13:36 +00:00
Evan Cheng
aa3c1410b4 Fix a build breaker.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28574 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-30 21:45:53 +00:00
Evan Cheng
51da42c28f Oops. PSHUFD is only available with SSE2.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28573 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-30 21:30:59 +00:00
Evan Cheng
ccba76bb25 Allow shufps x, x, mask to be converted to pshufd x, mask to save a move.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28565 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-30 20:26:50 +00:00
Evan Cheng
6de0163201 These can be transformed into lea as well. Not that we use this feature
currently...


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28393 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-19 18:43:41 +00:00
Evan Cheng
f4df680ad4 Add MOV16_rm / MOV32_rm and MOV16_mr / MOV32_mr to isLoadFromStackSlot and isStoreToStackSlot
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28223 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-11 07:33:49 +00:00
Evan Cheng
403be7eafc Fixing truncate. Previously we were emitting truncate from r16 to r8 as
movw. That is we promote the destination operand to r16. So
        %CH = TRUNC_R16_R8 %BP
is emitted as
        movw %bp, %cx.

This is incorrect. If %cl is live, it would be clobbered.
Ideally we want to do the opposite, that is emitted it as
        movb ??, %ch
But this is not possible since %bp does not have a r8 sub-register.

We are now defining a new register class R16_ which is a subclass of R16
containing only those 16-bit registers that have r8 sub-registers (i.e.
AX - DX). We isel the truncate to two instructions, a MOV16to16_ to copy the
value to the R16_ class, followed by a TRUNC_R16_R8.

Due to bug 770, the register colaescer is not going to coalesce between R16 and
R16_. That will be fixed later so we can eliminate the MOV16to16_. Right now, it
can only be eliminated if we are lucky that source and destination registers are
the same.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@28164 91177308-0d34-0410-b5e6-96231b3b80d8
2006-05-08 08:01:26 +00:00
Chris Lattner
993c897390 Teach the codegen about instructions used for SSE spill code, allowing it
to optimize cases where it has to spill a lot


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27801 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-18 16:44:51 +00:00
Evan Cheng
11e15b38e9 - More efficient extract_vector_elt with shuffle and movss, movsd, movd, etc.
- Some bug fixes and naming inconsistency fixes.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27377 91177308-0d34-0410-b5e6-96231b3b80d8
2006-04-03 20:53:28 +00:00
Evan Cheng
bc4832bc64 Support for scalar to vector with zero extension.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@27091 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-24 23:15:12 +00:00
Evan Cheng
82521dd838 - Remove scalar to vector pseudo ops. They are just wrong.
- Handle FR32 to VR128:v4f32 and FR64 to VR128:v2f64 with aliases of MOVAPS
and MOVAPD. Mark them as move instructions and *hope* they will be deleted.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26919 91177308-0d34-0410-b5e6-96231b3b80d8
2006-03-21 07:09:35 +00:00
Evan Cheng
fe5cb19405 1. Use pxor instead of xoraps / xorapd to clear FR32 / FR64 registers. This
proves to be worth 20% on Ptrdist/ks. Might be related to dependency
   breaking support.
2. Added FsMOVAPSrr and FsMOVAPDrr as aliases to MOVAPSrr and MOVAPDrr. These
   are used for FR32 / FR64 reg-to-reg copies.
3. Tell reg-allocator to generate MOVSSrm / MOVSDrm and MOVSSmr / MOVSDmr to
   spill / restore FsMOVAPSrr and FsMOVAPDrr.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26241 91177308-0d34-0410-b5e6-96231b3b80d8
2006-02-16 22:45:17 +00:00
Chris Lattner
1c07e7286d fix operand numbers
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@25915 91177308-0d34-0410-b5e6-96231b3b80d8
2006-02-02 20:38:12 +00:00
Chris Lattner
4083960147 Move isLoadFrom/StoreToStackSlot from MRegisterInfo to TargetInstrInfo,a far more logical place. Other methods should also be moved if anyoneis interested. :)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@25913 91177308-0d34-0410-b5e6-96231b3b80d8
2006-02-02 20:12:32 +00:00
Evan Cheng
bda54cdd47 Tell codegen MOVAPSrr and MOVAPDrr are copies.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@25889 91177308-0d34-0410-b5e6-96231b3b80d8
2006-02-01 23:03:16 +00:00
Nate Begeman
14e2cf62f4 Properly split f32 and f64 into separate register classes for scalar sse fp
fixing a bunch of nasty hackery


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@23735 91177308-0d34-0410-b5e6-96231b3b80d8
2005-10-14 22:06:00 +00:00
Nate Begeman
11cefd926a Teach the register allocator that movaps is also a move instruction
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@22451 91177308-0d34-0410-b5e6-96231b3b80d8
2005-07-16 02:00:20 +00:00
Nate Begeman
f63be7d395 First round of support for doing scalar FP using the SSE2 ISA extension and
XMM registers.  There are many known deficiencies and fixmes, which will be
addressed ASAP.  The major benefit of this work is that it will allow the
LLVM register allocator to allocate FP registers across basic blocks.

The x86 backend will still default to x87 style FP.  To enable this work,
you must pass -enable-sse-scalar-fp and either -sse2 or -sse3 to llc.

An example before and after would be for:
double foo(double *P) { double Sum = 0; int i; for (i = 0; i < 1000; ++i)
                        Sum += P[i]; return Sum; }

The inner loop looks like the following:
x87:
.LBB_foo_1:     # no_exit
        fldl (%esp)
        faddl (%eax,%ecx,8)
        fstpl (%esp)
        incl %ecx
        cmpl $1000, %ecx
        #FP_REG_KILL
        jne .LBB_foo_1  # no_exit

SSE2:
        addsd (%eax,%ecx,8), %xmm0
        incl %ecx
        cmpl $1000, %ecx
        #FP_REG_KILL
        jne .LBB_foo_1  # no_exit


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@22340 91177308-0d34-0410-b5e6-96231b3b80d8
2005-07-06 18:59:04 +00:00
Misha Brukman
0e0a7a45d3 * Remove trailing whitespace
* Convert tabs to spaces


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@21426 91177308-0d34-0410-b5e6-96231b3b80d8
2005-04-21 23:38:14 +00:00
Chris Lattner
a76f04828a When commuting these instructions, make sure to actually swap the operands too.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19694 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-19 16:55:52 +00:00
Chris Lattner
0df53d22c3 Improve coverage of the X86 instruction set by adding 16-bit shift doubles.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19687 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-19 07:31:24 +00:00
Chris Lattner
41e431ba04 Teach the code generator that shrd/shld is commutable if it has an immediate.
This allows us to generate this:

foo:
        mov %EAX, DWORD PTR [%ESP + 4]
        mov %EDX, DWORD PTR [%ESP + 8]
        shld %EDX, %EDX, 2
        shl %EAX, 2
        ret

instead of this:

foo:
        mov %EAX, DWORD PTR [%ESP + 4]
        mov %ECX, DWORD PTR [%ESP + 8]
        mov %EDX, %EAX
        shrd %EDX, %ECX, 30
        shl %EAX, 2
        ret

Note the magically transmogrifying immediate.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19686 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-19 07:11:01 +00:00
Chris Lattner
5aee0b97aa Disable 2->3 address promotion of add and inc instructions to LEA's. In
addition to being three address, LEA's don't set the flags.

This fixes 186.crafty.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19251 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-02 04:18:17 +00:00
Chris Lattner
bcea4d6f28 Implement the convertToThreeAddress method, add support for inverting JP/JNP
branches.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19247 91177308-0d34-0410-b5e6-96231b3b80d8
2005-01-02 02:37:07 +00:00
Chris Lattner
bcdda01210 Fix a warning
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@15409 91177308-0d34-0410-b5e6-96231b3b80d8
2004-08-01 19:31:30 +00:00
Alkis Evlogimenos
31e155e610 Align breaks.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@15371 91177308-0d34-0410-b5e6-96231b3b80d8
2004-07-31 10:05:44 +00:00
Chris Lattner
167cf33e1d Add breaks
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@15365 91177308-0d34-0410-b5e6-96231b3b80d8
2004-07-31 09:53:31 +00:00