model, except for external calls; this makes
addressing modes PC-relative. Incomplete.
The assertion at the top of Emitter::runOnMachineFunction
was obviously bogus (always true) so I removed it.
If someone knows what the correct test should be to cover
all the various targets, please fix.
llvm-svn: 54656
are allocated in the same buffer as the code,
jump tables, etc.
The default JIT memory manager does not handle buffer
overflow well. I didn't introduce this and I'm not
attempting to fix it here, but it is more likely to
be hit now since we're putting more stuff in the
buffer. This affects one test that I know of so far,
MultiSource/Benchmarks/NPB-serial/is.
llvm-svn: 54442
the need for a flavor operand, and add a new SDNode subclass,
LabelSDNode, for use with them to eliminate the need for a label id
operand.
Change instruction selection to let these label nodes through
unmodified instead of creating copies of them. Teach the MachineInstr
emitter how to emit a MachineInstr directly from an ISD label node.
This avoids the need for allocating SDNodes for the label id and
flavor value, as well as SDNodes for each of the post-isel label,
label id, and label flavor.
llvm-svn: 52943
1. The "JITState" object creates a PassManager with the ModuleProvider that the
jit is created with. If the ModuleProvider is removed and deleted, the
PassManager is invalid.
2. The Global maps in the JIT were not invalidated with a ModuleProvider was
removed. This could lead to a case where the Module would be freed, and a
new Module with Globals at the same addresses could return invalid results.
llvm-svn: 51384
function has already been codegen'd. This is required by the Java class loading
mechanism which executes Java code when materializing a function.
llvm-svn: 49988
was actually passing a completely incorrect size to sys_icache_invalidate.
Instead of having the JITEmitter do this (which doesn't have the correct
size), just make the target sync its own stubs.
llvm-svn: 46354
The meaning of getTypeSize was not clear - clarifying it is important
now that we have x86 long double and arbitrary precision integers.
The issue with long double is that it requires 80 bits, and this is
not a multiple of its alignment. This gives a primitive type for
which getTypeSize differed from getABITypeSize. For arbitrary precision
integers it is even worse: there is the minimum number of bits needed to
hold the type (eg: 36 for an i36), the maximum number of bits that will
be overwriten when storing the type (40 bits for i36) and the ABI size
(i.e. the storage size rounded up to a multiple of the alignment; 64 bits
for i36).
This patch removes getTypeSize (not really - it is still there but
deprecated to allow for a gradual transition). Instead there is:
(1) getTypeSizeInBits - a number of bits that suffices to hold all
values of the type. For a primitive type, this is the minimum number
of bits. For an i36 this is 36 bits. For x86 long double it is 80.
This corresponds to gcc's TYPE_PRECISION.
(2) getTypeStoreSizeInBits - the maximum number of bits that is
written when storing the type (or read when reading it). For an
i36 this is 40 bits, for an x86 long double it is 80 bits. This
is the size alias analysis is interested in (getTypeStoreSize
returns the number of bytes). There doesn't seem to be anything
corresponding to this in gcc.
(3) getABITypeSizeInBits - this is getTypeStoreSizeInBits rounded
up to a multiple of the alignment. For an i36 this is 64, for an
x86 long double this is 96 or 128 depending on the OS. This is the
spacing between consecutive elements when you form an array out of
this type (getABITypeSize returns the number of bytes). This is
TYPE_SIZE in gcc.
Since successive elements in a SequentialType (arrays, pointers
and vectors) need to be aligned, the spacing between them will be
given by getABITypeSize. This means that the size of an array
is the length times the getABITypeSize. It also means that GEP
computations need to use getABITypeSize when computing offsets.
Furthermore, if an alloca allocates several elements at once then
these too need to be aligned, so the size of the alloca has to be
the number of elements multiplied by getABITypeSize. Logically
speaking this doesn't have to be the case when allocating just
one element, but it is simpler to also use getABITypeSize in this
case. So alloca's and mallocs should use getABITypeSize. Finally,
since gcc's only notion of size is that given by getABITypeSize, if
you want to output assembler etc the same as gcc then getABITypeSize
is the size you want.
Since a store will overwrite no more than getTypeStoreSize bytes,
and a read will read no more than that many bytes, this is the
notion of size appropriate for alias analysis calculations.
In this patch I have corrected all type size uses except some of
those in ScalarReplAggregates, lib/Codegen, lib/Target (the hard
cases). I will get around to auditing these too at some point,
but I could do with some help.
Finally, I made one change which I think wise but others might
consider pointless and suboptimal: in an unpacked struct the
amount of space allocated for a field is now given by the ABI
size rather than getTypeStoreSize. I did this because every
other place that reserves memory for a type (eg: alloca) now
uses getABITypeSize, and I didn't want to make an exception
for unpacked structs, i.e. I did it to make things more uniform.
This only effects structs containing long doubles and arbitrary
precision integers. If someone wants to pack these types more
tightly they can always use a packed struct.
llvm-svn: 43620
Use APFloat in UpgradeParser and AsmParser.
Change all references to ConstantFP to use the
APFloat interface rather than double. Remove
the ConstantFP double interfaces.
Use APFloat functions for constant folding arithmetic
and comparisons.
(There are still way too many places APFloat is
just a wrapper around host float/double, but we're
getting there.)
llvm-svn: 41747
JITer (short path is added for darwin). This is needed to properly JIT llvm-gcc-4.2-built
binaries, since cxa_atexit is enabled by default on much more targets.
llvm-svn: 40600
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
recommended that getBoolValue be replaced with getZExtValue and that
get(bool) be replaced by get(const Type*, uint64_t). This implements
those changes.
llvm-svn: 33110
Turn on -Wunused and -Wno-unused-parameter. Clean up most of the resulting
fall out by removing unused variables. Remaining warnings have to do with
unused functions (I didn't want to delete code without review) and unused
variables in generated code. Maintainers should clean up the remaining
issues when they see them. All changes pass DejaGnu tests and Olden.
llvm-svn: 31380
This patch implements the first increment for the Signless Types feature.
All changes pertain to removing the ConstantSInt and ConstantUInt classes
in favor of just using ConstantInt.
llvm-svn: 31063
DLL* linkages got full (I hope) codegeneration support in C & both x86
assembler backends.
External weak linkage added for future use, we don't provide any
codegeneration, etc. support for it.
llvm-svn: 30374
This pass:
1. Splits TargetMachine into TargetMachine (generic targets, can be implemented
any way, like the CBE) and LLVMTargetMachine (subclass of TM that is used by
things using libcodegen and other support).
2. Instead of having each target fully populate the passmgr for file or JIT
output, move all this to common code, and give targets hooks they can
implement.
3. Commonalize the target population stuff between file emission and JIT
emission.
4. All (native code) codegen stuff now happens in a FunctionPassManager, which
paves the way for "fast -O0" stuff in the CFE later, and now LLC could
lazily stream .bc files from disk to use less memory.
5. There are now many fewer #includes and the targets don't depend on the
scalar xforms or libanalysis anymore (but codegen does).
6. Changing common code generator pass ordering stuff no longer requires
touching all targets.
7. The JIT now has the option of "-fast" codegen or normal optimized codegen,
which is now orthogonal to the fact that JIT'ing is being done.
llvm-svn: 30081
I've been told apple gcc version number is not guaranteed to increase
monotonically. Change the preprocess condition to make it less risky.
The configuration change is done during the middle 10.4 life cycle so we have
to check __APPLE_CC. For future OS X release, we should be able to assume
-fenable-cxa-atexit is the default.
llvm-svn: 30024
method.
- Added synchronizeICache() to TargetJITInfo. It is called after each block
of code is emitted to flush the icache. This ensures correct execution
on targets that have separate dcache and icache.
- Added PPC / Mac OS X specific code to do icache flushing.
llvm-svn: 29276
1. Change several methods in the MachineCodeEmitter class to be pure virtual.
2. Suck emitConstantPool/initJumpTableInfo into startFunction, removing them
from the MachineCodeEmitter interface, and reducing the amount of target-
specific code.
3. Change the JITEmitter so that it allocates constantpools and jump tables
*right* next to the functions that they belong to, instead of in a separate
pool of memory. This makes all memory for a function be contiguous, and
means the JITEmitter only tracks one block of memory now.
llvm-svn: 28065
code emission location into the base class, instead of being in the derived classes.
This change means that low-level methods like emitByte/emitWord now are no longer
virtual (yaay for speed), and we now have a framework to support growable code
segments. This implements feature request #1 of PR469.
llvm-svn: 28059
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
near the GOT, which new doesn't do. So break out the allocate into a new function.
Also move GOT index handling into JITResolver. This lets it update the mapping when a Lazy
function is JITed. It doesn't managed the table, just the mapping. Note that this is
still non-ideal, as any function that takes a function address should also take a GOT
index, but that is a lot of changes. The relocation resolve process updates any GOT entry
it sees is out of date.
llvm-svn: 22537
This patch completes the changes for making lli thread-safe. Here's the list
of changes:
* The Support/ThreadSupport* files were removed and replaced with the
MutexGuard.h file since all ThreadSupport* declared was a Mutex Guard.
The implementation of MutexGuard.h is now based on sys::Mutex which hides
its implementation and makes it unnecessary to have the -NoSupport.h and
-PThreads.h versions of ThreadSupport.
* All places in ExecutionEngine that previously referred to "Mutex" now
refer to sys::Mutex
* All places in ExecutionEngine that previously referred to "MutexLocker"
now refer to MutexGuard (this is frivolous but I believe the technically
correct name for such a class is "Guard" not a "Locker").
These changes passed all of llvm-test. All we need now are some test cases
that actually use multiple threads.
llvm-svn: 22404
immediately instead of lazily.
In this program, for example:
int main() {
printf("hello world\n");
printf("hello world\n");
printf("hello world\n");
printf("hello world\n");
}
We used to have to go through compilation callback 4 times (once for each
call to printf), now we don't go to it at all.
Thanks to Misha for noticing this, and for adding the initial ghost linkage
patches.
llvm-svn: 17864
Move include/Config and include/Support into include/llvm/Config,
include/llvm/ADT and include/llvm/Support. From here on out, all LLVM
public header files must be under include/llvm/.
llvm-svn: 16137
by trying to get the compiler to generate an undefined reference for it
and related functions which live in libc_nonshared.a on Linux.
Linkers... sigh.
llvm-svn: 12256
VM.cpp and JIT.cpp files into JIT.cpp. This also splits some nasty code out
into TargetSelect.cpp so that people hopefully won't notice it. :)
llvm-svn: 10544
allow unaligned loads, that is probably the problem I've been seeing in numerous
SPARC test cases failing. X86, on the other hand, just slows down unaligned
accesses, since it must make 2 aligned accesses for each unaligned one.
llvm-svn: 10266
recompile and relink. This keeps it from failing an assertion when
it goes and tries to construct a new MachineFunction for that Function.
llvm-svn: 9459
the #define up there too
* Since we're including system headers, use the ones in include/llvm/Config
* While we're here, use the canonical LLVM header ordering algorithm
llvm-svn: 8463
Switch Interpreter and JIT's "run" methods to take a Function and a vector of
GenericValues.
Move (almost all of) the stuff that constructs a canonical call to main()
into lli (new methods "callAsMain", "makeStringVector").
Nuke getCurrentExecutablePath(), enableTracing(), getCurrentFunction(),
isStopped(), and many dead decls from interpreter.
Add linux strdup() support to interpreter.
Make interpreter's atexit handler runner and JIT's runAtExitHandlers() look
more alike, in preparation for refactoring.
atexit() is spelled "atexit", not "at_exit".
llvm-svn: 8366
static method here.
Remove some extra blank lines.
ExecutionEngine.h: Add its prototype.
lli.cpp: Call it.
Make creation method for each type of EE into a static method of its
own subclass.
Interpreter/Interpreter.cpp: ExecutionEngine::createInterpreter -->
Interpreter::create
Interpreter/Interpreter.h: Likewise.
JIT/JIT.cpp: ExecutionEngine::createJIT --> VM::create
JIT/VM.h: Likewise.
llvm-svn: 8343
now handle far calls (i.e., beyond the 30-bit limit in call instructions).
* As a side-effect, this allows us to unify and clean up the mmap() call and
code around it.
llvm-svn: 7381
This us used by bugpoint -- when code is compiled to a shared object to be
JITted, it must use the JIT's lazy resolution method to find function addresses,
because some functions will not be available at .so load time, as they are in
the bytecode file.
llvm-svn: 7363
* Setting ENABLE_X86_JIT or ENABLE_SPARC_JIT on the `make' command-line will
force the inclusion of that JIT on a different architecture
* If neither JIT is enabled (e.g., compiling on a different architecture), the
-march option will not be available to LLI.
* As a side effect of the $ARCH variable, the Sparc LLI can now link just a bit
faster by not including the x86 library.
llvm-svn: 7070
(We're already talking about autoconf'ing this, so I'm assuming this hack
will be short-lived...I just don't want it to get lost in my working files.)
llvm-svn: 6761
The JIT is designed to code-generate a function at-a-time. That means that any
pass can only make local changes to its function. Period.
Because the Sparc PreSelection pass claims to be a BasicBlock pass while adding
globals to the Module, it cannot be run with the other passes, because by this
time, the globals have been output already by the JIT, and the addresses of any
globals appearing AFTER this point are not recognized.
However, the PreSelection pass is a requirement for correctness in the Sparc
codegen path, so it MUST be run.
::: HACK ALERT ::: HACK ALERT ::: HACK ALERT ::: HACK ALERT ::: HACK ALERT :::
llvm-svn: 6650
laid out closer to the VM so that calls to library functions (e.g. puts()) and
callback (e.g. JITResolver::CompilationCallback) fit into 30 bits of the call
instruction.
* Abort if architecture is not yet supported (not X86 or Sparc) because it
likely requires a different set of parameters to mmap() .
* Stop using hard-coded values for page size; use sysconf(_SC_PAGESIZE) instead.
llvm-svn: 6610
* No more createX86Emitter() vs. createSparcEmitter() -- there can be only one
* As a result, the memory management semantics must be handled according to
platform -- the parameters to mmap() are particularly sensitive to the host
architecture.
llvm-svn: 6527
`lli -march=x86' or `lli -march=sparc' will forcefully select the JIT even on a
different platform. Running lli without the -march option will select the JIT
for the platform that it's currently running on.
Pro: can test Sparc JIT (debug printing mode) on X86 -- faster to compile/link
LLVM source base to test changes.
Con: Linking lli on x86 now pulls in all the Sparc libs -> longer link time
(but X86 can bear it, right?)
In the future, perhaps this should be a ./configure option to enable/disable
target JITting...
llvm-svn: 6360
Evan Cheng