latency for certain models of the Intel Atom family, by converting
instructions into their equivalent LEA instructions, when it is both
useful and possible to do so.
llvm-svn: 180573
The current Intel Atom microarchitecture has a feature whereby
when a function returns early then it is slightly faster to execute
a sequence of NOP instructions to wait until the return address is ready,
as opposed to simply stalling on the ret instruction until
the return address is ready.
When compiling for X86 Atom only, this patch will run a pass,
called "X86PadShortFunction" which will add NOP instructions where less
than four cycles elapse between function entry and return.
It includes tests.
This patch has been updated to address Nadav's review comments
- Optimize only at >= O1 and don't do optimization if -Os is set
- Stores MachineBasicBlock* instead of BBNum
- Uses DenseMap instead of std::map
- Fixes placement of braces
Patch by Andy Zhang.
llvm-svn: 171879
a TargetMachine to construct (and thus isn't always available), to an
analysis group that supports layered implementations much like
AliasAnalysis does. This is a pretty massive change, with a few parts
that I was unable to easily separate (sorry), so I'll walk through it.
The first step of this conversion was to make TargetTransformInfo an
analysis group, and to sink the nonce implementations in
ScalarTargetTransformInfo and VectorTargetTranformInfo into
a NoTargetTransformInfo pass. This allows other passes to add a hard
requirement on TTI, and assume they will always get at least on
implementation.
The TargetTransformInfo analysis group leverages the delegation chaining
trick that AliasAnalysis uses, where the base class for the analysis
group delegates to the previous analysis *pass*, allowing all but tho
NoFoo analysis passes to only implement the parts of the interfaces they
support. It also introduces a new trick where each pass in the group
retains a pointer to the top-most pass that has been initialized. This
allows passes to implement one API in terms of another API and benefit
when some other pass above them in the stack has more precise results
for the second API.
The second step of this conversion is to create a pass that implements
the TargetTransformInfo analysis using the target-independent
abstractions in the code generator. This replaces the
ScalarTargetTransformImpl and VectorTargetTransformImpl classes in
lib/Target with a single pass in lib/CodeGen called
BasicTargetTransformInfo. This class actually provides most of the TTI
functionality, basing it upon the TargetLowering abstraction and other
information in the target independent code generator.
The third step of the conversion adds support to all TargetMachines to
register custom analysis passes. This allows building those passes with
access to TargetLowering or other target-specific classes, and it also
allows each target to customize the set of analysis passes desired in
the pass manager. The baseline LLVMTargetMachine implements this
interface to add the BasicTTI pass to the pass manager, and all of the
tools that want to support target-aware TTI passes call this routine on
whatever target machine they end up with to add the appropriate passes.
The fourth step of the conversion created target-specific TTI analysis
passes for the X86 and ARM backends. These passes contain the custom
logic that was previously in their extensions of the
ScalarTargetTransformInfo and VectorTargetTransformInfo interfaces.
I separated them into their own file, as now all of the interface bits
are private and they just expose a function to create the pass itself.
Then I extended these target machines to set up a custom set of analysis
passes, first adding BasicTTI as a fallback, and then adding their
customized TTI implementations.
The fourth step required logic that was shared between the target
independent layer and the specific targets to move to a different
interface, as they no longer derive from each other. As a consequence,
a helper functions were added to TargetLowering representing the common
logic needed both in the target implementation and the codegen
implementation of the TTI pass. While technically this is the only
change that could have been committed separately, it would have been
a nightmare to extract.
The final step of the conversion was just to delete all the old
boilerplate. This got rid of the ScalarTargetTransformInfo and
VectorTargetTransformInfo classes, all of the support in all of the
targets for producing instances of them, and all of the support in the
tools for manually constructing a pass based around them.
Now that TTI is a relatively normal analysis group, two things become
straightforward. First, we can sink it into lib/Analysis which is a more
natural layer for it to live. Second, clients of this interface can
depend on it *always* being available which will simplify their code and
behavior. These (and other) simplifications will follow in subsequent
commits, this one is clearly big enough.
Finally, I'm very aware that much of the comments and documentation
needs to be updated. As soon as I had this working, and plausibly well
commented, I wanted to get it committed and in front of the build bots.
I'll be doing a few passes over documentation later if it sticks.
Commits to update DragonEgg and Clang will be made presently.
llvm-svn: 171681
URL: http://llvm.org/viewvc/llvm-project?rev=171524&view=rev
Log:
The current Intel Atom microarchitecture has a feature whereby when a function
returns early then it is slightly faster to execute a sequence of NOP
instructions to wait until the return address is ready,
as opposed to simply stalling on the ret instruction
until the return address is ready.
When compiling for X86 Atom only, this patch will run a pass, called
"X86PadShortFunction" which will add NOP instructions where less than four
cycles elapse between function entry and return.
It includes tests.
Patch by Andy Zhang.
llvm-svn: 171603
returns early then it is slightly faster to execute a sequence of NOP
instructions to wait until the return address is ready,
as opposed to simply stalling on the ret instruction
until the return address is ready.
When compiling for X86 Atom only, this patch will run a pass, called
"X86PadShortFunction" which will add NOP instructions where less than four
cycles elapse between function entry and return.
It includes tests.
Patch by Andy Zhang.
llvm-svn: 171524
This pass was conservative in that it always reserved the FP to enable dynamic
stack realignment, which allowed the RA to use aligned spills for vector
registers. This happens even when spills were not necessary. The RA has
since been improved to use unaligned spills when necessary.
The new behavior is to realign the stack if the frame pointer was already
reserved for some other reason, but don't reserve the frame pointer just
because a function contains vector virtual registers.
Part of rdar://12719844
llvm-svn: 168627
This implements codegen support for accesses to thread-local variables
using the local-dynamic model, and adds a clean-up pass so that the base
address for the TLS block can be re-used between local-dynamic access on
an execution path.
llvm-svn: 157818
This also enables domain swizzling for AVX code which required a few
trivial test changes.
The pass will be moved to lib/CodeGen shortly.
llvm-svn: 140659
SSE transition penalty. The pass is enabled through the "x86-use-vzeroupper"
llc command line option. This is only the first step (very naive and
conservative one) to sketch out the idea, but proper DFA is coming next
to allow smarter decisions. Comments and ideas now and in further commits
will be very appreciated.
llvm-svn: 138317
and MCSubtargetInfo.
- Added methods to update subtarget features (used when targets automatically
detect subtarget features or switch modes).
- Teach X86Subtarget to update MCSubtargetInfo features bits since the
MCSubtargetInfo layer can be shared with other modules.
- These fixes .code 16 / .code 32 support since mode switch is updated in
MCSubtargetInfo so MC code emitter can do the right thing.
llvm-svn: 134884
target machine from those that are only needed by codegen. The goal is to
sink the essential target description into MC layer so we can start building
MC based tools without needing to link in the entire codegen.
First step is to refactor TargetRegisterInfo. This patch added a base class
MCRegisterInfo which TargetRegisterInfo is derived from. Changed TableGen to
separate register description from the rest of the stuff.
llvm-svn: 133782
- Check getBytesToPopOnReturn().
- Eschew ST0 and ST1 for return values.
- Fix the PIC base register initialization so that it doesn't ever
fail to end up the top of the entry block.
llvm-svn: 108039
When a frame pointer is not otherwise required, and dynamic stack alignment
is necessary solely due to the spilling of a register with larger alignment
requirements than the default stack alignment, the frame pointer can be both
used as a general purpose register and a frame pointer. That goes poorly, for
obvious reasons. This patch brings back a bit of old logic for identifying
the use of such registers and conservatively reserves the frame pointer
during register allocation in such cases.
For now, implement for X86 only since it's 32-bit linux which is hitting this,
and we want a targeted fix for 2.7. As a follow-on, this will be expanded
to handle other targets, as theoretically the problem could arise elsewhere
as well.
llvm-svn: 100559
On Nehalem and newer CPUs there is a 2 cycle latency penalty on using a register
in a different domain than where it was defined. Some instructions have
equvivalents for different domains, like por/orps/orpd.
The SSEDomainFix pass tries to minimize the number of domain crossings by
changing between equvivalent opcodes where possible.
This is a work in progress, in particular the pass doesn't do anything yet. SSE
instructions are tagged with their execution domain in TableGen using the last
two bits of TSFlags. Note that not all instructions are tagged correctly. Life
just isn't that simple.
The SSE execution domain issue is very similar to the ARM NEON/VFP pipeline
issue handled by NEONMoveFixPass. This pass may become target independent to
handle both.
llvm-svn: 99524
This is work in progress. So far, SSE execution domain tables are added to
X86InstrInfo, and a skeleton pass is enabled with -sse-domain-fix.
llvm-svn: 99345
- Note, this is a gigantic hack, with the sole purpose of unblocking further
work on the assembler (its also possible to test the mathcer more completely
now).
- Despite being a hack, its actually good enough to work over all of 403.gcc
(although some encodings are probably incorrect). This is a testament to the
beauty of X86's MachineInstr, no doubt! ;)
llvm-svn: 80234
--- Reverse-merging r75799 into '.':
U test/Analysis/PointerTracking
U include/llvm/Target/TargetMachineRegistry.h
U include/llvm/Target/TargetMachine.h
U include/llvm/Target/TargetRegistry.h
U include/llvm/Target/TargetSelect.h
U tools/lto/LTOCodeGenerator.cpp
U tools/lto/LTOModule.cpp
U tools/llc/llc.cpp
U lib/Target/PowerPC/PPCTargetMachine.h
U lib/Target/PowerPC/AsmPrinter/PPCAsmPrinter.cpp
U lib/Target/PowerPC/PPCTargetMachine.cpp
U lib/Target/PowerPC/PPC.h
U lib/Target/ARM/ARMTargetMachine.cpp
U lib/Target/ARM/AsmPrinter/ARMAsmPrinter.cpp
U lib/Target/ARM/ARMTargetMachine.h
U lib/Target/ARM/ARM.h
U lib/Target/XCore/XCoreTargetMachine.cpp
U lib/Target/XCore/XCoreTargetMachine.h
U lib/Target/PIC16/PIC16TargetMachine.cpp
U lib/Target/PIC16/PIC16TargetMachine.h
U lib/Target/Alpha/AsmPrinter/AlphaAsmPrinter.cpp
U lib/Target/Alpha/AlphaTargetMachine.cpp
U lib/Target/Alpha/AlphaTargetMachine.h
U lib/Target/X86/X86TargetMachine.h
U lib/Target/X86/X86.h
U lib/Target/X86/AsmPrinter/X86ATTAsmPrinter.h
U lib/Target/X86/AsmPrinter/X86AsmPrinter.cpp
U lib/Target/X86/AsmPrinter/X86IntelAsmPrinter.h
U lib/Target/X86/X86TargetMachine.cpp
U lib/Target/MSP430/MSP430TargetMachine.cpp
U lib/Target/MSP430/MSP430TargetMachine.h
U lib/Target/CppBackend/CPPTargetMachine.h
U lib/Target/CppBackend/CPPBackend.cpp
U lib/Target/CBackend/CTargetMachine.h
U lib/Target/CBackend/CBackend.cpp
U lib/Target/TargetMachine.cpp
U lib/Target/IA64/IA64TargetMachine.cpp
U lib/Target/IA64/AsmPrinter/IA64AsmPrinter.cpp
U lib/Target/IA64/IA64TargetMachine.h
U lib/Target/IA64/IA64.h
U lib/Target/MSIL/MSILWriter.cpp
U lib/Target/CellSPU/SPUTargetMachine.h
U lib/Target/CellSPU/SPU.h
U lib/Target/CellSPU/AsmPrinter/SPUAsmPrinter.cpp
U lib/Target/CellSPU/SPUTargetMachine.cpp
U lib/Target/Mips/AsmPrinter/MipsAsmPrinter.cpp
U lib/Target/Mips/MipsTargetMachine.cpp
U lib/Target/Mips/MipsTargetMachine.h
U lib/Target/Mips/Mips.h
U lib/Target/Sparc/AsmPrinter/SparcAsmPrinter.cpp
U lib/Target/Sparc/SparcTargetMachine.cpp
U lib/Target/Sparc/SparcTargetMachine.h
U lib/ExecutionEngine/JIT/TargetSelect.cpp
U lib/Support/TargetRegistry.cpp
llvm-svn: 75820
Massive check in. This changes the "-fast" flag to "-O#" in llc. If you want to
use the old behavior, the flag is -O0. This change allows for finer-grained
control over which optimizations are run at different -O levels.
Most of this work was pretty mechanical. The majority of the fixes came from
verifying that a "fast" variable wasn't used anymore. The JIT still uses a
"Fast" flag. I'll change the JIT with a follow-up patch.
llvm-svn: 70343
Preston Gurd