Commit Graph

11709 Commits

Author SHA1 Message Date
Andrew Lenharth
701f5ac73c Random sampling (aka Arnold and Ryder) profiling. This is still preliminary, but it works on spec on x86 and alpha. The idea is to allow profiling passes to remember what profiling they inserted, then a random sampling framework is inserted which consists of duplicated basic blocks (without profiling), such that at each backedge in the program and entry into every function, the framework chooses whether to use the instrumented code or the instrumentation free code. The goal of such a framework is to make it reasonably cheap to do random sampling of very expensive profiling products (such as load-value profiling).
The code is organized into 3 parts (2 passes)
1) a linked set of profiling passes, which implement an analysis group (linked, like alias analysis are).  These insert profiling into the program, and remember what they inserted, so that at a later time they can be queried about any instruction.

2) a pass that handles inserting the random sampling framework.  This also has options to control how random samples are choosen.  Currently implemented are Global counters, register allocated global counters, and read cycle counter (see? there was a reason for it).

The profiling passes are almost identical to the existing ones (block, function, and null profiling is supported right now), and they are valid passes without the sampling framework (hence the existing passes can be unified with the new ones, not done yet).

Some things are a bit ugly still, but that should be fixed up soon enough.

Other todo? making the counter values not "magic 2^16 -1" values, but dynamically choosable.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24493 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-28 00:58:09 +00:00
Nate Begeman
01595c52b3 Small tweaks noticed while on the plane.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24492 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-26 22:39:34 +00:00
Andrew Lenharth
b0826529f8 since reg2mem requires it, might as well mention that it preserves it
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24491 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-25 16:04:54 +00:00
Duraid Madina
7b1e154c5a add support for dynamic_stackalloc to the dag isel (thanks andrew ;)
next up: support argument passing in memory, not just registers


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24490 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-25 07:49:25 +00:00
Nate Begeman
e4f17a5f9b Some first bits of AltiVec stuff: Instruction Formats, Encodings, and
Registers.  Apologies to Jim if the scheduling info so far isn't accurate.

There's a few more things like VRsave support that need to be finished up
in my local tree before I can commit code that Does The Right Thing for
turning 4 x float into the various altivec packed float instructions.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24489 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-23 05:29:52 +00:00
Andrew Lenharth
7c0c567058 Reg2Mem is something a pass may depend on, so allow that
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24488 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-22 22:14:23 +00:00
Andrew Lenharth
7045f6c56e turns out, demotion and invokes and critical edges don't mix
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24487 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-22 21:45:19 +00:00
Andrew Lenharth
4052f02cdb Fix warning, the better way. Really, this is what this instruction is for, so use it
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24486 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-22 20:59:00 +00:00
Andrew Lenharth
475d31729d Fix warning
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24485 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-22 20:56:05 +00:00
Nate Begeman
ab48be3772 Check in code to scalarize arbitrarily wide packed types for some simple
vector operations (load, add, sub, mul).

This allows us to codegen:
void %foo(<4 x float> * %a) {
entry:
  %tmp1 = load <4 x float> * %a;
  %tmp2 = add <4 x float> %tmp1, %tmp1
  store <4 x float> %tmp2, <4 x float> *%a
  ret void
}

on ppc as:
_foo:
        lfs f0, 12(r3)
        lfs f1, 8(r3)
        lfs f2, 4(r3)
        lfs f3, 0(r3)
        fadds f0, f0, f0
        fadds f1, f1, f1
        fadds f2, f2, f2
        fadds f3, f3, f3
        stfs f0, 12(r3)
        stfs f1, 8(r3)
        stfs f2, 4(r3)
        stfs f3, 0(r3)
        blr


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24484 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-22 18:16:00 +00:00
Andrew Lenharth
50b37845ef massive DAGISel patch. lots and lots more stuff compiles now
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24483 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-22 04:20:06 +00:00
Nate Begeman
4ef3b817fe Rather than attempting to legalize 1 x float, make sure the SD ISel never
generates it.  Make MVT::Vector expand-only, and remove the code in
Legalize that attempts to legalize it.

The plan for supporting N x Type is to continually epxand it in ExpandOp
until it gets down to 2 x Type, where it will be scalarized into a pair of
scalars.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24482 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-22 01:29:36 +00:00
Chris Lattner
ac2902bcb5 Use HasDotTypeDotSizeDirective instead of forELF
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24481 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 23:06:54 +00:00
Chris Lattner
9787c6443f Remove a level of indentation by using a continue.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24479 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 22:48:18 +00:00
Chris Lattner
d460f57d65 Simplify the subtarget info, allow the asmwriter to do some target sensing
based on TargetType.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24478 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 22:43:58 +00:00
Chris Lattner
a35a8e87fa Use subtarget information computed by X86Subtarget instead of rolling our own.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24477 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 22:39:40 +00:00
Chris Lattner
e5600e5509 Make the X86 subtarget compute the basic target type: ELF, Cygwin, Darwin,
or native Win32


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24476 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 22:31:58 +00:00
Chris Lattner
5df14ca0a1 Add a forELF flag, allowing the removal of forCygwin and simplification of
conditionals.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24475 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 22:19:48 +00:00
Chris Lattner
8fccc97b0b simplify and genericize this code
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24473 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 19:50:31 +00:00
Duraid Madina
b6f023ac21 add support for div/rem to the dag->dag isel. yay.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24472 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 14:14:54 +00:00
Duraid Madina
e0b632a5d8 I think I know what you meant here, but just to be safe I'll let you
do it. :)

    <_sabre_> excuses excuses


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24471 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 14:09:40 +00:00
Chris Lattner
42a76cdb04 Eliminate unneeded intermediate class. Move doFinalizeMethod to bottom of
file.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24470 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 08:40:17 +00:00
Chris Lattner
3dfbe2ed34 Start using shared asmprinter Constant Pool emitter, use shorter cpi names.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24469 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 08:38:26 +00:00
Chris Lattner
69191f8f05 prune #include
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24468 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 08:33:17 +00:00
Chris Lattner
d939f6c85d Switch to using the shared constant pool printer, along with using shorter
CPI ids


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24467 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 08:32:23 +00:00
Chris Lattner
6053010f9c Switch to using the generic constant pool emitter impl, use shorter
CPI names


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24466 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 08:29:17 +00:00
Chris Lattner
c569e6108a Use generic constant pool emission code in the AsmPrinter class.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24465 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 08:26:15 +00:00
Chris Lattner
3b4fd32a41 Allow target to customize directive used to switch to arbitrary section in SwitchSection,
add generic constant pool emitter


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24464 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 08:25:09 +00:00
Chris Lattner
0745536c00 Use the FunctionNumber provided by the AsmPrinter class
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24462 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 08:14:07 +00:00
Chris Lattner
77bc22866a increment the function number in SetupMachineFunction
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24461 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 08:13:27 +00:00
Chris Lattner
6d5a4f6a8e Use CommentString where possible, fix a bug where aix mode wouldn't assemble
due to basic blocks being misnamed.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24459 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 08:02:41 +00:00
Chris Lattner
ef65874151 unify the darwin and aix constant pool printers
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24458 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 07:57:37 +00:00
Chris Lattner
25045bdcda Adjust to capitalized asmprinter method names
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24457 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 07:51:36 +00:00
Chris Lattner
8b8b9515d6 Adjust to capitalized AsmPrinter method names
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24456 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 07:51:23 +00:00
Chris Lattner
64965ba8b0 Use PrivateGlobalPrefix for basic block labels. This allows the x86 darwin
port to properly use L for the bb prefix instead of .


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24454 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 07:43:59 +00:00
Chris Lattner
7f9ccde91e use PrivateGlobalPrefix for basic blocks
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24453 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 07:41:05 +00:00
Chris Lattner
2ec30b5e82 Use PrivateGlobalPrefix for basic block labels
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24452 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 07:39:22 +00:00
Chris Lattner
87744a2dbd Use PrivateGlobalPrefix for basic blocks
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24451 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 07:38:08 +00:00
Chris Lattner
0a70a49dd6 Switch to the new shared SwitchSection
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24450 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 07:30:28 +00:00
Chris Lattner
4bfa3a3eb0 Start using SwitchSection, allowing globals and functions to be emitted
to specific sections.  Delete some dead functions copied from the X86 backend.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24449 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 07:26:04 +00:00
Chris Lattner
7b6e53cde7 convert the rest of this over to use SwitchSection
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24448 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 07:16:34 +00:00
Chris Lattner
224cfc2eac Start using the AsmPrinter shared SwitchSection code. This allows the X86
backend to implement global variables in sections.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24447 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 07:11:11 +00:00
Chris Lattner
f6163a0079 This is now implemented in common codegen code
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24446 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 07:06:58 +00:00
Chris Lattner
ac28fbd043 Add section switching to common code generator code. Add a couple of
asserts.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24445 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 07:06:27 +00:00
Chris Lattner
62cbf2aab9 Rename SwitchSection -> switchSection to avoid conflicting with a future
change.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24443 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 06:55:27 +00:00
Chris Lattner
81a994e42e Start using PrivateGlobalPrefix correctly
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24442 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 06:51:52 +00:00
Chris Lattner
f55366e3c2 set PrivateGlobalPrefix on darwin, use it when printing out CP references
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24441 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 06:47:58 +00:00
Chris Lattner
c41cc83117 Naturally align doubles in the constant pool, set PrivateGlobalPrefix on
darwin, use it when printing the constant pool indices so the labels are
appropriately private, emit cp entries to .const instead of .data on darwin
and only emit a single .section for the constant pool, not one for each
entry.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24440 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-21 06:46:22 +00:00
Chris Lattner
81f803df80 Lower READCYCLECOUNTER correctly, preserving the chain result
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24438 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-20 22:57:19 +00:00
Chris Lattner
308575be66 Legalize MERGE_VALUES, expand READCYCLECOUNTER correctly, so it doesn't
break control dependence.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24437 91177308-0d34-0410-b5e6-96231b3b80d8
2005-11-20 22:56:56 +00:00