We currently error in clang with:
"error: thread-local storage is unsupported for the current target", but we
can start to get the llvm level ready.
When compiling
template<typename T>
struct foo {
static __declspec(thread) int bar;
};
template<typename T>
__declspec(therad) int foo<T>::bar;
template struct foo<int>;
msvc produces
SECTION HEADER #3
.tls$ name
0 physical address
0 virtual address
4 size of raw data
12F file pointer to raw data (0000012F to 00000132)
0 file pointer to relocation table
0 file pointer to line numbers
0 number of relocations
0 number of line numbers
C0301040 flags
Initialized Data
COMDAT; sym= "public: static int foo<int>::bar" (?bar@?$foo@H@@2HA)
4 byte align
Read Write
gcc produces a ".data$__emutls_v.<symbol>" for the testcase with
__declspec(thread) replaced with thread_local.
llvm-svn: 195849
With this patch we use simple names for COMDAT sections (like .text or .bss).
This matches the MSVC behavior.
When merging it is the COMDAT symbol that is used to decide if two sections
should be merged, so there is no point in building a fancy name.
This survived a bootstrap on mingw32.
llvm-svn: 195798
A Direct stack map location records the address of frame index. This
address is itself the value that the runtime requested. This differs
from IndirectMemRefOp locations, which refer to a stack locations from
which the requested values must be loaded. Direct locations can
directly communicate the address if an alloca, while IndirectMemRefOp
handle register spills.
For example:
entry:
%a = alloca i64...
llvm.experimental.stackmap(i32 <ID>, i32 <shadowBytes>, i64* %a)
Since both the alloca and stackmap intrinsic are in the entry block,
and the intrinsic takes the address of the alloca, the runtime can
assume that LLVM will not substitute alloca with any intervening
value. This must be verified by the runtime by checking that the stack
map's location is a Direct location type. The runtime can then
determine the alloca's relative location on the stack immediately after
compilation, or at any time thereafter. This differs from Register and
Indirect locations, because the runtime can only read the values in
those locations when execution reaches the instruction address of the
stack map.
llvm-svn: 195712
r195698 moved the type unit checking up into getOrCreateTypeDIE so
remove the redundant check and fold the functions back together again.
llvm-svn: 195700
It might be possible to eventually use one data structure, but I haven't
looked at the exact criteria used for accelerator tables and pubtypes to
see if there's good reason for the differences between the two or not.
llvm-svn: 195696
Summary:
Moved the requirement for SelectionDAG::getConstant() to return legally
typed nodes slightly earlier. There were two optional DAGCombine passes
that were missed out and were required to produce type-legal DAGs.
Simplified a code-path in tryFoldToZero() to use SelectionDAG::getConstant().
This provides support for both promoted and expanded vector types whereas the
previous code only supported promoted vector types.
Fixes a "Type for zero vector elements is not legal" assertion detected by
an llvm-stress generated test.
Reviewers: resistor
CC: llvm-commits
Differential Revision: http://llvm-reviews.chandlerc.com/D2251
llvm-svn: 195635
This avoids the need for an extra list of SkeletonCUs and associated
cleanup while staging things to be cleaner for further type unit
improvements.
Also hopefully fixes a memory leak introduced in r195166.
llvm-svn: 195536
a non-relocatable number offset.
One fixme to make the ranges as discrete data structures and
have range lists explicitly represented rather than as a list of symbols.
llvm-svn: 195523
Improvements over r195317:
- Set/restore EnableFastISel flag instead of just running FastISel within
SelectAllBasicBlocks; the flag is checked in various places, and
FastISel won't run properly if those places don't do the right thing.
- Test looks for normal ISel versus FastISel behavior, and not
something more subtle that doesn't work everywhere.
Based on work by Andrea Di Biagio.
llvm-svn: 195491
The legalizer can now do this type of expansion for more
type combinations without loading and storing to and
from the stack.
NOTE: This is a candidate for the 3.4 branch.
llvm-svn: 195398
This patch is a rewrite of the original patch commited in r194542. Instead of
relying on the type legalizer to do the splitting for us, we now peform the
splitting ourselves in the DAG combiner. This is necessary for the case where
the vector mask is a legal type after promotion and still wouldn't require
splitting.
Patch by: Juergen Ributzka
NOTE: This is a candidate for the 3.4 branch.
llvm-svn: 195397
section use the form DW_FORM_data4 whilst in Dwarf 4 and later they
use the form DW_FORM_sec_offset.
This patch updates the places where such attributes are generated to
use the appropriate form depending on the Dwarf version. The DIE entries
affected have the following tags:
DW_AT_stmt_list, DW_AT_ranges, DW_AT_location, DW_AT_GNU_pubnames,
DW_AT_GNU_pubtypes, DW_AT_GNU_addr_base, DW_AT_GNU_ranges_base
It also adds a hidden command line option "--dwarf-version=<uint>"
to llc which allows the version of Dwarf to be generated to override
what is specified in the metadata; this makes it possible to update
existing tests to check the debugging information generated for both
Dwarf 4 (the default) and Dwarf 3 using the same metadata.
Patch (slightly modified) by Keith Walker!
llvm-svn: 195391
Summary:
LegalizeSetCCCondCode can now legalize SETEQ and SETNE by returning the inverse
condition and requesting that the caller invert the result of the condition.
The caller of LegalizeSetCCCondCode must handle the inverted CC, and they do
so as follows:
SETCC, BR_CC:
Invert the result of the SETCC with SelectionDAG::getNOT()
SELECT_CC:
Swap the true/false operands.
This is necessary for MSA which lacks an integer SETNE instruction.
Reviewers: resistor
CC: llvm-commits
Differential Revision: http://llvm-reviews.chandlerc.com/D2229
llvm-svn: 195355
It broke, at least, i686 target. It is reproducible with "llc -mtriple=i686-unknown".
FYI, it didn't appear to add either "-O0" or "-fast-isel".
llvm-svn: 195339
We now only allow breaking source order if the exit block frequency is
significantly higher than the other exit block. The actual bias is
currently under a flag so the best cut-off can be found; the flag
defaults to the old behavior. The idea is to get some benchmark coverage
over different values for the flag and pick the best one.
When we require the new frequency to be at least 20% higher than the old
frequency I see a 5% speedup on zlib's deflate when compressing a random
file on x86_64/westmere. Hal reported a small speedup on Fhourstones on
a BG/Q and no regressions in the test suite.
The test case is the full long_match function from zlib's deflate. I was
reluctant to add it for previous tweaks to branch probabilities because
it's large and potentially fragile, but changed my mind since it's an
important use case and more likely to break with all the current work
going into the PGO infrastructure.
Differential Revision: http://llvm-reviews.chandlerc.com/D2202
llvm-svn: 195265
While not strictly necessary (the class has an invariant that
"setDebugInfoOffset" is called before "getDebugInfoOffset" - anyone
client that actually gets the default zero offset is buggy/broken) this
is consistent with the code as originally written and the removal of the
initialization was an accident in r195166.
Suggested by Manman Ren.
llvm-svn: 195263
Emit DW_TAG_type_units into the debug_info section using compile unit
headers. This is bogus/unusable by debuggers, but testable and provides
more isolated review.
Subsequent patches will include support for type unit headers and
emission into the debug_types section, as well as comdat grouping the
types based on their hash. Also the CompileUnit type will be renamed
'Unit' and relevant portions pulled out into respective CompileUnit and
TypeUnit types.
llvm-svn: 195166
This is the first step to fix pr17918.
It extends the .section directive a bit, inspired by what the ELF one looks
like. The problem with using linkonce is that given
.section foo
.linkonce....
.section foo
.linkonce
we would already have switched sections when getting to .linkonce. The cleanest
solution seems to be to add the comdat information in the .section itself.
llvm-svn: 195148
This patch removes most of the trivial cases of weak vtables by pinning them to
a single object file. The memory leaks in this version have been fixed. Thanks
Alexey for pointing them out.
Differential Revision: http://llvm-reviews.chandlerc.com/D2068
Reviewed by Andy
llvm-svn: 195064
This change is incorrect. If you delete virtual destructor of both a base class
and a subclass, then the following code:
Base *foo = new Child();
delete foo;
will not cause the destructor for members of Child class. As a result, I observe
plently of memory leaks. Notable examples I investigated are:
ObjectBuffer and ObjectBufferStream, AttributeImpl and StringSAttributeImpl.
llvm-svn: 194997
This reverts commit f1d9fe9d04ce93f6d5dcebbd2cb6a07414d7a029.
This was causing PR17964. We need to use thread data before regular data.
llvm-svn: 194960
Implementing this on bigendian platforms could get strange. I added a
target hook, getStackSlotRange, per Jakob's recommendation to make
this as explicit as possible.
llvm-svn: 194942
and update test cases accordingly.
This doesn't affect the output dumped using llvm-dwarfdump, but
readelf does now dump the debug_loc section.
llvm-svn: 194898
This patch removes most of the trivial cases of weak vtables by pinning them to
a single object file.
Differential Revision: http://llvm-reviews.chandlerc.com/D2068
Reviewed by Andy
llvm-svn: 194865
The error reported the number of explicit operands,
but that isn't what is checked. In my case, this
resulted in the confusing errors
"Too few operands." followed shortly by
"8 operands expected, but 8 given."
llvm-svn: 194862
Stop folding constant adds into GEP when the type size doesn't match.
Otherwise, the adds' operands are effectively being promoted, changing the
conditions of an overflow. Results are different when:
sext(a) + sext(b) != sext(a + b)
Problem originally found on x86-64, but also fixed issues with ARM and PPC,
which used similar code.
<rdar://problem/15292280>
Patch by Duncan Exon Smith!
llvm-svn: 194840
Summary:
When getConstant() is called for an expanded vector type, it is split into
multiple scalar constants which are then combined using appropriate build_vector
and bitcast operations.
In addition to the usual big/little endian differences, the case where the
element-order of the vector does not have the same endianness as the elements
themselves is also accounted for. For example, for v4i32 on big-endian MIPS,
the byte-order of the vector is <3210,7654,BA98,FEDC>. For little-endian, it is
<0123,4567,89AB,CDEF>.
Handling this case turns out to be a nop since getConstant() returns a splatted
vector (so reversing the element order doesn't change the value)
This fixes a number of cases in MIPS MSA where calling getConstant() during
operation legalization introduces illegal types (e.g. to legalize v2i64 UNDEF
into a v2i64 BUILD_VECTOR of illegal i64 zeros). It should also handle bigger
differences between illegal and legal types such as legalizing v2i64 into v8i16.
lowerMSASplatImm() in the MIPS backend no longer needs to avoid calling
getConstant() so this function has been updated in the same patch.
For the sake of transparency, the steps I've taken since the review are:
* Added 'virtual' to isVectorEltOrderLittleEndian() as requested. This revealed
that the MIPS tests were falsely passing because a polymorphic function was
not actually polymorphic in the reviewed patch.
* Fixed the tests that were now failing. This involved deleting the code to
handle the MIPS MSA element-order (which was previously doing an byte-order
swap instead of an element-order swap). This left
isVectorEltOrderLittleEndian() unused and it was deleted.
* Fixed build failures caused by rebasing beyond r194467-r194472. These build
failures involved the bset, bneg, and bclr instructions added in these commits
using lowerMSASplatImm() in a way that was no longer valid after this patch.
Some of these were fixed by calling SelectionDAG::getConstant() instead,
others were fixed by a new function getBuildVectorSplat() that provided the
removed functionality of lowerMSASplatImm() in a more sensible way.
Reviewers: bkramer
Reviewed By: bkramer
CC: llvm-commits
Differential Revision: http://llvm-reviews.chandlerc.com/D1973
llvm-svn: 194811
This is to avoid this transformation in some cases:
fold (conv (load x)) -> (load (conv*)x)
On architectures that don't natively support some vector
loads efficiently casting the load to a smaller vector of
larger types and loading is more efficient.
Patch by Micah Villmow.
llvm-svn: 194783
This comes into play with patchpoint, which can fold multiple
operands. Since the patchpoint is already treated as a call, the
machine mem operands won't affect anything, and there's nothing to
test. But we still want to do the right thing here to be sure that our
MIs obey the rules.
llvm-svn: 194750
In ELF and COFF an alias is just another offset in a section. There is no way
to represent an alias to something in another file.
In MachO, the spec has the N_INDR type which should allow for exactly that, but
is not currently implemented. Given that it is specified but not implemented,
we error in codegen to avoid miscompiling but don't reject aliases to
declarations in the verifier to leave the option open of implementing it.
In the past we have used alias to declarations as a way of implementing
weakref, which is why it exists in some old tests which this patch updates.
llvm-svn: 194705
If a null call target is provided, don't emit a dummy call. This
allows the runtime to reserve as little nop space as it needs without
the requirement of emitting a call.
llvm-svn: 194676
This patch reapplies r193676 with an additional fix for the Hexagon backend. The
SystemZ backend has already been fixed by r194148.
The Type Legalizer recognizes that VSELECT needs to be split, because the type
is to wide for the given target. The same does not always apply to SETCC,
because less space is required to encode the result of a comparison. As a result
VSELECT is split and SETCC is unrolled into scalar comparisons.
This commit fixes the issue by checking for VSELECT-SETCC patterns in the DAG
Combiner. If a matching pattern is found, then the result mask of SETCC is
promoted to the expected vector mask type for the given target. Now the type
legalizer will split both VSELECT and SETCC.
This allows the following X86 DAG Combine code to sucessfully detect the MIN/MAX
pattern. This fixes PR16695, PR17002, and <rdar://problem/14594431>.
Reviewed by Nadav
llvm-svn: 194542
Based on discussions with Lang Hames and Jakob Stoklund Olesen at the hacker's lab, and in the light of upcoming work on the PBQP register allocator, it was though that CalcSpillWeights does not need to be a pass. This change will enable to customize / tune the spill weight computation depending on the allocator.
Update the documentation style while there.
No functionnal change.
llvm-svn: 194356
This patch moves the jump address materialization inside the noop slide. This
enables patching of the materialization itself or its complete removal. This
patch also adds the ability to define scratch registers that can be used safely
by the code called from the patchpoint intrinsic. At least one scratch register
is required, because that one is used for the materialization of the jump
address. This patch depends on D2009.
Differential Revision: http://llvm-reviews.chandlerc.com/D2074
Reviewed by Andy
llvm-svn: 194306
The new graph structure replaces the node and edge linked lists with vectors.
Free lists (well, free vectors) are used for fast insertion/deletion.
The ultimate aim is to make PBQP graphs cheap to clone. The motivation is that
the PBQP solver destructively consumes input graphs while computing a solution,
forcing the graph to be fully reconstructed for each round of PBQP. This
imposes a high cost on large functions, which often require several rounds of
solving/spilling to find a final register allocation. If we can cheaply clone
the PBQP graph and incrementally update it between rounds then hopefully we can
reduce this cost. Further, once we begin pooling matrix/vector values (future
work), we can cache some PBQP solver metadata and share it between cloned
graphs, allowing the PBQP solver to re-use some of the computation done in
earlier rounds.
For now this is just a data structure update. The allocator and solver still
use the graph the same way as before, fully reconstructing it between each
round. I expect no material change from this update, although it may change
the iteration order of the nodes, causing ties in the solver to break in
different directions, and this could perturb the generated allocations
(hopefully in a completely benign way).
Thanks very much to Arnaud Allard de Grandmaison for encouraging me to get back
to work on this, and for a lot of discussion and many useful PBQP test cases.
llvm-svn: 194300
The idea of the AnyReg Calling Convention is to provide the call arguments in
registers, but not to force them to be placed in a paticular order into a
specified set of registers. Instead it is up tp the register allocator to assign
any register as it sees fit. The same applies to the return value (if
applicable).
Differential Revision: http://llvm-reviews.chandlerc.com/D2009
Reviewed by Andy
llvm-svn: 194293
Based on discussions with Lang Hames and Jakob Stoklund Olesen at the hacker's lab, and in the light of upcoming work on the PBQP register allocator, it was though that CalcSpillWeights does not need to be a pass. This change will enable to customize / tune the spill weight computation depending on the allocator.
Update the documentation style while there.
No functionnal change.
llvm-svn: 194269
MorphNodeTo is not safe to call during DAG building. It eagerly
deletes dependent DAG nodes which invalidates the NodeMap. We could
expose a safe interface for morphing nodes, but I don't think it's
worth it. Just create a new MachineNode and replaceAllUsesWith.
My understaning of the SD design has been that we want to support
early target opcode selection. That isn't very well supported, but
generally works. It seems reasonable to rely on this feature even if
it isn't widely used.
llvm-svn: 194102
linkonce_odr_auto_hide was in incomplete attempt to implement a way
for the linker to hide symbols that are known to be available in every
TU and whose addresses are not relevant for a particular DSO.
It was redundant in that it all its uses are equivalent to
linkonce_odr+unnamed_addr. Unlike those, it has never been connected
to clang or llvm's optimizers, so it was effectively dead.
Given that nothing produces it, this patch just nukes it
(other than the llvm-c enum value).
llvm-svn: 193865
We add a map in DwarfDebug to map MDNodes that are shareable across CUs to the
corresponding DIEs: MDTypeNodeToDieMap. These DIEs can be shared across CUs,
that is why we keep the maps in DwarfDebug instead of CompileUnit.
We make the assumption that if a DIE is not added to an owner yet, we assume
it belongs to the current CU. Since DIEs for the type system are added to
their owners immediately after creation, and other DIEs belong to the current
CU, the assumption should be true.
A testing case is added to show that we only create a single DIE for a type
MDNode and we use ref_addr to refer to the type DIE.
We also add a testing case to show ref_addr relocations for non-darwin
platforms.
llvm-svn: 193779
When an extend more than doubles the size of the elements (e.g., a zext
from v16i8 to v16i32), the normal legalization method of splitting the
vectors will run into problems as by the time the destination vector is
legal, the source vector is illegal. The end result is the operation
often becoming scalarized, with the typical horrible performance. For
example, on x86_64, the simple input of:
define void @bar(<16 x i8> %a, <16 x i32>* %p) nounwind {
%tmp = zext <16 x i8> %a to <16 x i32>
store <16 x i32> %tmp, <16 x i32>*%p
ret void
}
Generates:
.section __TEXT,__text,regular,pure_instructions
.section __TEXT,__const
.align 5
LCPI0_0:
.long 255 ## 0xff
.long 255 ## 0xff
.long 255 ## 0xff
.long 255 ## 0xff
.long 255 ## 0xff
.long 255 ## 0xff
.long 255 ## 0xff
.long 255 ## 0xff
.section __TEXT,__text,regular,pure_instructions
.globl _bar
.align 4, 0x90
_bar:
vpunpckhbw %xmm0, %xmm0, %xmm1
vpunpckhwd %xmm0, %xmm1, %xmm2
vpmovzxwd %xmm1, %xmm1
vinsertf128 $1, %xmm2, %ymm1, %ymm1
vmovaps LCPI0_0(%rip), %ymm2
vandps %ymm2, %ymm1, %ymm1
vpmovzxbw %xmm0, %xmm3
vpunpckhwd %xmm0, %xmm3, %xmm3
vpmovzxbd %xmm0, %xmm0
vinsertf128 $1, %xmm3, %ymm0, %ymm0
vandps %ymm2, %ymm0, %ymm0
vmovaps %ymm0, (%rdi)
vmovaps %ymm1, 32(%rdi)
vzeroupper
ret
So instead we can check if there are legal types that enable us to split
more cleverly when the input vector is already legal such that we don't
turn it into an illegal type. If the extend is such that it's more than
doubling the size of the input we check if
- the number of vector elements is even,
- the source type is legal,
- the type of a split source is illegal,
- the type of an extended (by doubling element size) source is legal, and
- the type of that extended source when split is legal.
If the conditions are met, instead of just splitting both the
destination and the source types, we create an extend that only goes up
one "step" (doubling the element width), and the continue legalizing the
rest of the operation normally. The result is that this operates as a
new, more effecient, termination condition for the loop of "split the
operation until the destination type is legal."
With this change, the above example now compiles to:
_bar:
vpxor %xmm1, %xmm1, %xmm1
vpunpcklbw %xmm1, %xmm0, %xmm2
vpunpckhwd %xmm1, %xmm2, %xmm3
vpunpcklwd %xmm1, %xmm2, %xmm2
vinsertf128 $1, %xmm3, %ymm2, %ymm2
vpunpckhbw %xmm1, %xmm0, %xmm0
vpunpckhwd %xmm1, %xmm0, %xmm3
vpunpcklwd %xmm1, %xmm0, %xmm0
vinsertf128 $1, %xmm3, %ymm0, %ymm0
vmovaps %ymm0, 32(%rdi)
vmovaps %ymm2, (%rdi)
vzeroupper
ret
This generalizes a custom lowering that was added a while back to the
ARM backend. That lowering is no longer necessary, and is removed. The
testcases for it, however, provide excellent ARM tests for this change
and so remain.
rdar://14735100
llvm-svn: 193727
With this patch llvm produces a weak_def_can_be_hidden for linkonce_odr
if they are also unnamed_addr or don't have their address taken.
There is not a lot of documentation about .weak_def_can_be_hidden, but
from the old discussion about linkonce_odr_auto_hide and the name of
the directive this looks correct: these symbols can be hidden.
Testing this with the ld64 in Xcode 5 linking clang reduces the number of
exported symbols from 21053 to 19049.
llvm-svn: 193718
The Type Legalizer recognizes that VSELECT needs to be split, because the type
is to wide for the given target. The same does not always apply to SETCC,
because less space is required to encode the result of a comparison. As a result
VSELECT is split and SETCC is unrolled into scalar comparisons.
This commit fixes the issue by checking for VSELECT-SETCC patterns in the DAG
Combiner. If a matching pattern is found, then the result mask of SETCC is
promoted to the expected vector mask type for the given target. This mask has
usually the same size as the VSELECT return type (except for Intel KNL). Now the
type legalizer will split both VSELECT and SETCC.
This allows the following X86 DAG Combine code to sucessfully detect the MIN/MAX
pattern. This fixes PR16695, PR17002, and <rdar://problem/14594431>.
Reviewed by Nadav
llvm-svn: 193676
Use EmitLabelOffsetDifference for handling on darwin platform when
non-darwin platforms use EmitLabelPlusOffset.
Also fix a bug in EmitLabelOffsetDifference where the size is hard-coded
to 4 even though Size is passed in as an argument.
llvm-svn: 193660
To support ref_addr, we calculate the section offset of a DIE (i.e. offset
of a DIE from beginning of the debug info section). The Offset field in DIE
is currently CU-relative. To calculate the section offset, we add a
DebugInfoOffset field in CompileUnit to store the offset of a CU from beginning
of the debug info section. We set the value in DwarfUnits::computeSizeAndOffset
for each CompileUnit.
A helper function DIE::getCompileUnit is added to return the CU DIE that
the input DIE belongs to. We also add a map CUDieMap in DwarfDebug to help
finding the CU for a given CU DIE.
For a cross-referenced DIE, we first find the CU DIE it belongs to with
getCompileUnit, then we use CUDieMap to get the corresponding CU for the CU DIE.
Adding the section offset of the CU with the CU-relative offset of a DIE gives
us the seciton offset of the DIE.
We correctly emit ref_addr with relocation using EmitLabelPlusOffset when
doesDwarfUseRelocationsAcrossSections is true.
This commit handles the emission of DW_FORM_ref_addr when we have an attribute
with FORM_ref_addr. A follow-on patch will start using ref_addr when adding a
DIEEntry. This commit will be tested and verified in the follow-on patch.
Reviewed off-list by Eric, Thanks.
llvm-svn: 193658
after the DIE creation, we construct the context first.
Ensure that we create the context before we create a type so that we can add
the newly created type to the parent. Remove last use of addToContextOwner
now that it's not needed.
We use createAndAddDIE to wrap around "new DIE(". Now all shareable DIEs
should be added to their parents right after the creation.
Reviewed off-list by Eric, Thanks.
llvm-svn: 193657
This modifies the pass to classify every SSP-triggering AllocaInst according to
an SSPLayoutKind (LargeArray, SmallArray, AddrOf). This analysis is collected
by the pass and made available for use, but no other pass uses it yet.
The next patch will make use of this analysis in PEI and StackSlot
passes. The end goal is to support ssp-strong stack layout rules.
WIP.
Differential Revision: http://llvm-reviews.chandlerc.com/D1789
llvm-svn: 193653
after the DIE creation, we construct the context first.
This touches creation of namespaces and global variables. The purpose is to
handle all DIE creations similarly: constructs the context first, then creates
the DIE and immediately adds the DIE to its parent.
We use createAndAddDIE to wrap around "new DIE(".
llvm-svn: 193589
More patches will be submitted to convert "new DIE(" to use createAddAndDIE in
DwarfCompileUnit.cpp. This will simplify implementation of addDIEEntry where
we have to decide between ref4 and ref_addr, because DIEs that can be shared
across CU will be added to a CU already.
Reviewed off-list by Eric.
llvm-svn: 193567
It wraps around "new DIE(" and handles the bookkeeping part of the newly-created
DIE. It adds the DIE to its parent, and calls insertDIE if necessary. It makes
sure that bookkeeping is done at the earliest time and we should not see
parentless DIEs if all constructions of DIEs go through this helper function.
Later on, we can use an allocator for DIE allocation, and will only need to
change createAndAddDIE instead of modifying all the "new DIE(".
Reviewed off-list by Eric.
llvm-svn: 193566
Making useAA() default to true for SystemZ showed that the combiner alias
analysis wasn't handling volatile accesses. This hit many of the SystemZ
tests, but I arbitrarily picked one for the purpose of this patch.
llvm-svn: 193518
Most SelectionDAG code drops the TBAA info when creating a new form of a
load and store (e.g. during legalization, or when converting a plain
load to an extending one). This patch tries to catch all cases where
the TBAA information can legitimately be carried over.
The patch adds alternative forms of getLoad() and getExtLoad() that take
a MachineMemOperand instead of individual fields. (The corresponding
getTruncStore() already exists.) The idea is to use the MachineMemOperand
forms when all fields are carried over (size, pointer info, isVolatile,
isNonTemporal, alignment and TBAA info). If some adjustment is being
made, e.g. to narrow the load, then we still pass the individual fields
but also pass the TBAA info.
llvm-svn: 193517
ARM processors without ldrex/strex need to be able to make libcalls for all
atomic operations, including the newer min/max versions.
The alternative would probably be expanding these operations in terms of
cmpxchg (as x86 does always), but in the configurations where this matters
code-size tends to be paramount so the libcall is more desirable.
llvm-svn: 193398
This optimization is not SSE specific so I am moving it to DAGco.
The new scalar_to_vector dag node exposed a missing pattern in the AArch64 target that I needed to add.
llvm-svn: 193393
Also improve the implementation of EmitRawText(Twine) so it doesn't
bother using the SmallString buffer if the Twine is a simple StringRef
anyway.
llvm-svn: 193378
Since we never insert DIE for DITemplateTypeParameter to a map, there is no need
to call getDIE in getOrCreateTemplateTypeParameterDIE. It is also renamed to
constructTemplateTypeParameterDIE to match with other construct functions
in CompileUnit.
Same applies to getOrCreateTemplateValueParameterDIE.
llvm-svn: 193287
Remove the unneeded return values from createMemberDIE, constructEnumTypeDIE,
getOrCreateTemplateTypeParameterDIE, and getOrCreateTemplateValueParameterDIE.
llvm-svn: 193285
For some targets, it is useful to be able to look at the original
type of an argument without having to dig through the original IR.
This also fixes a bug in SelectionDAGBuilder where InputArg.PartOffset
was not taking into account the offset of structure elements.
Patch by: Justin Holewinski
Tom Stellard:
- Changed the type of ArgVT to EVT, so it can store non-simple types
like v3i32.
llvm-svn: 193214
Remove unnecessary creation of LexicalScope in collectDeadVariables.
The created LexicialScope was only used to get isAbstractScope, which
should be false from the creation:
"new LexicalScope(NULL, DIDescriptor(SP), NULL, false);".
We can also remove a DenseMap that holds the created LexicalScopes.
llvm-svn: 193196
Since (as of r190716) Clang no longer emits debug info for C++ friend
declarations (and it seems GCC never has/does, which was the motivation
for the Clang change), there's no actual reachable case for implementing
the part of DWARF 4, Section 7.27 part 5 that pertains to friends.
Leave an assert here so that if/when we do have a client producing
friends and using type units, we can fill in the gap and add appropriate
(unit and feature) tests.
llvm-svn: 193193
Includes a test case/FIXME demonstrating a bug/limitation in pointer to
member hashing. To be honest I'm not sure why we don't just always use
summary hashing for referenced types... but perhaps I'm missing
something.
llvm-svn: 193175
VTList has a long life cycle through the module and getVTList is frequently called. In current getVTList, sequential search over a std::vector is used, this is inefficient in big module.
This patch use FoldingSet to implement hashing mechanism when searching.
Reviewer: Nadav Rotem
Test : Pass unit tests & LNT test suite
llvm-svn: 193150
This uses a map, keeping the type DIE numbering separate from the DIEs
themselves - alternatively we could do things the way GCC does if we
want to add an integer to the DIE type to record the numbering there.
llvm-svn: 193105
This allows various variables to be more self-documenting and easier to
debug by being of specific types without overlapping enum values.
Precommit review by Eric Christopher.
llvm-svn: 193091
Found while adding type safety to the various DWARF enumerations (form,
attribute, tag, etc) that caused Clang to warn on an incompletely
covered switch. Converting the comment to a default/unreachable
uncovered this case of an unsupported form encoding. Seems we were
skipping fission strings entirely.
llvm-svn: 193089
This ensures that the prefix data is treated as part of the function for
the purpose of debug info. This provides a better debugging experience,
among other things by allowing a debug info client to correctly look up
a function in debug info given a function pointer.
llvm-svn: 193042
With this commit, all DIEs created in CompileUnit will be added to parents
inside the same function. Also make getOrCreateTemplateType|Value functions
private.
No functionality change.
llvm-svn: 193002
PR17168 describes a test case that fails when compiling for debug with
fast-isel. Investigation showed that the test was failing because a DBG_VALUE
machine instruction was placed prior to a PHI.
For this problem to occur requires the following:
* Compile for debug
* Compile with fast-isel
* In a block B, fast-isel must partially succeed before punting to DAG-isel
* B must start with a PHI
* The first unhandled node in the DAG must not generate a machine instruction
* A debug value with an order less than that of that first node exists
When all of these circumstances apply, the existing test that an instruction
was not inserted won't fire. Currently it tests whether the block is empty,
or whether the last instruction generated is a phi. When fast-isel has
partially succeeded, the last instruction generated will not be a phi.
Instead, we need to check whether the current insert position is immediately
following a phi. This patch adds that check, and adds the test case from the
PR as a regression test.
llvm-svn: 192976
There are targets that support i128 sized scalars but cannot emit
instructions that modify them directly. The proper thing to do is to
emit a libcall.
This fixes PR17481.
llvm-svn: 192957
When canonicalizing dags according to the rule
(shl (zext (shr X, c1) ), c1) ==> (zext (shl (shr X, c1), c1))
remember to add the new shl dag to the DAGCombiner worklist of nodes.
If we don't explicitly add it to the worklist of nodes to visit, we
may not trigger later on the rule that folds the shift left + logical
shift right into a AND instruction with bitmask.
llvm-svn: 192883
like C++ should be the fully qualified names for the type.
Add a routine that does a language specific context walk to build
up the qualified name and use it when we add types/names to the
tables. Expand the gnu pubnames testcase as it's the most complex
to make sure that qualified types are also being added.
llvm-svn: 192865
This happens e.g. with <2 x i64> -1 on x86_32. It cannot be generated directly
because i64 is illegal. It would be nice if getNOT would handle this
transparently, but I don't see a way to generate a legal constant there right
now. Fixes PR17487.
llvm-svn: 192795
This is really an extension of the current (shl (shr ...)) -> shl optimization.
The main difference is that certain upper bits must also not be demanded.
The motivating examples are the first two in the testcase, which occur
in llvmpipe output.
llvm-svn: 192783
1) Make sure we emit static member variables by checking
at the end of createGlobalVariableDIE rather than piecemeal
in the function.
(As a note, createGlobalVariableDIE needs rewriting.)
2) Make sure we use the definition rather than declaration DIE
for two things: a) determining linkage for gnu pubnames, and b)
as the address of the DIE for global variables.
(As a note, createGlobalVariableDIE really needs rewriting.)
Adjust the testcase to make sure we're checking the correct DIEs.
llvm-svn: 192761
twice and just look up the value. Fix the one case where
we were trying to create a subprogram DIE and we should already
have had one. Reflow formatting in collectDeadVariables while fixing.
llvm-svn: 192749
rdar:15221834 False AVX register dependencies cause 5x slowdown on
flops-5/6 and significant slowdown on several others.
This was blocking the switch to MI-Sched.
llvm-svn: 192669
This pass is needed to break false dependencies. Without it, unlucky
register assignment can result in wild (5x) swings in
performance. This pass was trying to handle AVX but not getting it
right. AVX doesn't have partial register defs, it has unused register
reads in which the high bits of a source operand are copied into the
unused bits of the dest.
Fixing this requires conservative liveness analysis. This is awkard
because the pass already has its own pseudo-liveness. However, proper
liveness is expensive, and we would like to use a generic utility to
compute it. The fix only invokes liveness on-demand. It is rare to
detect a case that needs undef-read dependence breaking, but when it
happens, it can be needed many times within a very large block.
I think the existing heuristic which uses a register window of 16 is
too conservative for loop-carried false dependencies. If the loop is a
reduction. The out-of-order engine may be able to execute several loop
iterations in parallel. However, I'll leave this tuning exercise for
next time.
llvm-svn: 192635
Clobbering is exclusive not inclusive on register units.
For liveness, we need to consider all the preserved registers.
e.g. A regmask that clobbers YMM0 may preserve XMM0.
Units are only clobbered when all super-registers are clobbered.
llvm-svn: 192623
Some clients may add block live ins and may track liveness over a
large scope. This guarantees an efficient implementation in all cases
with no memory allocation/deallocation, independent of the number of
target registers. It could be slightly less convenient but is fine in
the expected case.
llvm-svn: 192622
Clean up creation of static member DIEs. We can create static member DIEs from
two places, so we call getOrCreateStaticMemberDIE from the two places.
getOrCreateStaticMemberDIE will get or create the context DIE first, then it
will check if the DIE already exists, if not, we create the static member DIE
and add it to the context.
Creation of static member DIEs are handled in a similar way as subprogram DIEs.
llvm-svn: 192618
Per original comment, the intention of this loop
is to go ahead and break the critical edge
(in order to sink this instruction) if there's
reason to believe doing so might "unblock" the
sinking of additional instructions that define
registers used by this one. The idea is that if
we have a few instructions to sink "together"
breaking the edge might be worthwhile.
This commit makes a few small changes
to help better realize this goal:
First, modify the loop to ignore registers
defined by this instruction. We don't
sink definitions of physical registers,
and sinking an SSA definition isn't
going to unblock an upstream instruction.
Second, ignore uses of physical registers.
Instructions that define physical registers are
rejected for sinking, and so moving this one
won't enable moving any defining instructions.
As an added bonus, while virtual register
use-def chains are generally small due
to SSA goodness, iteration over the uses
and definitions (used by hasOneNonDBGUse)
for physical registers like EFLAGS
can be rather expensive in practice.
(This is the original reason for looking at this)
Finally, to keep things simple continue
to only consider this trick for registers that
have a single use (via hasOneNonDBGUse),
but to avoid spuriously breaking critical edges
only do so if the definition resides
in the same MBB and therefore this one directly
blocks it from being sunk as well.
If sinking them together is meant to be,
let the iterative nature of this pass
sink the definition into this block first.
Update tests to accomodate this change,
add new testcase where sinking avoids pipeline stalls.
llvm-svn: 192608
When if converting something like:
true:
... = R0<kill>
false:
... = R0<kill>
then the instructions of the true block must not have a <kill> flag
anymore, as the instruction of the false block follow and do still read
the R0 value.
Specifically this patch determines the set of register live-in in the
false block (possibly after simulating the liveness changes of the
duplicated instructions). Each of these live-in registers mustn't be
killed.
llvm-svn: 192482
This should fix the buildbots.
Original commit message:
[DAGCombiner] Slice a big load in two loads when the element are next to each
other in memory and the target has paired load and performs post-isel loads
combining.
E.g., this optimization will transform something like this:
a = load i64* addr
b = trunc i64 a to i32
c = lshr i64 a, 32
d = trunc i64 c to i32
into:
b = load i32* addr1
d = load i32* addr2
Where addr1 = addr2 +/- sizeof(i32), if the target supports paired load and
performs post-isel loads combining.
One should overload TargetLowering::hasPairedLoad to provide this information.
The default is false.
<rdar://problem/14477220>
llvm-svn: 192476
other in memory and the target has paired load and performs post-isel loads
combining.
E.g., this optimization will transform something like this:
a = load i64* addr
b = trunc i64 a to i32
c = lshr i64 a, 32
d = trunc i64 c to i32
into:
b = load i32* addr1
d = load i32* addr2
Where addr1 = addr2 +/- sizeof(i32), if the target supports paired load and
performs post-isel loads combining.
One should overload TargetLowering::hasPairedLoad to provide this information.
The default is false.
<rdar://problem/14477220>
llvm-svn: 192471
For NVPTX, this fixes a crash where the emitImplicitDef implementation was expecting physical registers,
while NVPTX uses virtual registers (with a couple of exceptions). Now, the implicit def comment will be
emitted as a true PTX register name. Other targets can use this to customize the output of implicit def
comments.
Fixes PR17519
llvm-svn: 192444
LiveRange just manages a list of segments and a list of value numbers
now as LiveInterval did previously, but without having details like spill
weight or a fixed register number.
LiveInterval is now a subclass of LiveRange and simply adds the spill weight
and the register number.
llvm-svn: 192393
The Segment struct contains a single interval; multiple instances of this struct
are used to construct a live range, but the struct is not a live range by
itself.
llvm-svn: 192392
This patch fixes an old FIXME by creating a MCTargetStreamer interface
and moving the target specific functions for ARM, Mips and PPC to it.
The ARM streamer is still declared in a common place because it is
used from lib/CodeGen/ARMException.cpp, but the Mips and PPC are
completely hidden in the corresponding Target directories.
I will send an email to llvmdev with instructions on how to use this.
llvm-svn: 192181
The most likely case where this error happens is when the user specifies
too many register operands. Don't make it look like an internal LLVM bug
when we can see that the error is coming from an inline asm instruction.
For other instructions we keep the "ran out of registers" error.
llvm-svn: 192041
When MC was first added, targets could use hasRawTextSupport to keep features
working before they were added to the MC interface.
The design goal of MC is to provide an uniform api for printing assembly and
object files. Short of relaxations and other corner cases, a object file is
just another representation of the assembly.
It was never the intention that targets would keep doing things like
if (hasRawTextSupport())
Set flags in one way.
else
Set flags in another way.
When they do that they create two code paths and the object file is no longer
just another representation of the assembly. This also then requires testing
with llc -filetype=obj, which is extremelly brittle.
This patch removes some of these hacks by replacing them with smaller ones.
The ARM flag setting is trivial, so I just moved it to the constructor. For
Mips, the patch adds two temporary hack directives that allow the assembly
to represent the same things as the object file was already able to.
The hope is that the mips developers will replace the hack directives with
the same ones that gas uses and drop the -print-hack-directives flag.
I will also try to implement a target streamer interface, so that we can
move this out of the common code.
In summary, for any new work, two rules of the thumb are
* Don't use "llc -filetype=obj" in tests.
* Don't add calls to hasRawTextSupport.
llvm-svn: 192035
is updated to use DITypeRef.
Move isUnsignedDIType and getOriginalTypeSize from DebugInfo.h to be static
helper functions in DwarfCompileUnit. We already have a static helper function
"isTypeSigned" in DwarfCompileUnit, and a pointer to DwarfDebug is added to
resolve the derived-from field. All three functions need to go across link
for derived-from fields, so we need to get hold of a type identifier map.
A pointer to DwarfDebug is also added to DbgVariable in order to resolve the
derived-from field.
Debug info verifier is updated to check a derived-from field is a TypeRef.
Verifier will not go across link for derived-from fields, in debug info finder,
we go across the link to add derived-from fields to types.
Function getDICompositeType is only used by dragonegg and since dragonegg does
not generate identifier for types, we use an empty map to resolve the
derived-from field.
When printing a derived-from field, we use DITypeRef::getName to either return
the type identifier or getName of the DIType.
A paired commit at clang is required due to changes to DIBuilder.
llvm-svn: 192018
DAGCombiner::visitFP_EXTEND will apply the following transformation:
fold (fpext (load x)) -> (fpext (fptrunc (extload x)))
but the implementation does not handle indexed loads (pre/post inc.), but did
not specifically ignore them either (unlike for extending loads, which it
already ignored), causing an assert when the transformation was applied to an
indexed load. This is the minimal fix for correctness (causing the
transformation to be skipped for indexed loads).
Unfortunately, I don't have an in-tree test case.
llvm-svn: 191989
In the case (shown in the attached test) where a member function
definition was emitted into debug info the following could occur:
1) build the debug info for the member function definition
2) in (1), build the debug info for the member function declaration
3) construct and add the member function declaration DIE
4) add it to its context
5) build its context (the type it is a member of)
6) construct the members and add them to the type
7) except don't add member functions because "getOrCreateSubprogram"
adds the function to its parent anyway
8) except we're only partway through building this subprogram
declaration so it hasn't been added yet - but we returned the partially
constructed DIE (since it's already in the MDNode->DIE mapping to avoid
infinitely recursing trying to create the member function DIE)
9) once the type is constructed, add the member function to it
10) now the members are out of order (the member function being defined
is listed as the last member, even though it was declared as the first)
To avoid this, construct the context of the subprogram DIE before we
query to see if it exists. That way we never end up creating it before
creating its context and ending up in this situation.
Alternatively, the type construction that visits/builds all the members
could call something like getOrCreateSubprogram, but that doesn't ever
do the "add to context" step. Then the type building code would always
be responsible for adding members (and the subprogram "addToContextDIE"
would no-op because the context building would have added the subprogram
declaration to the type/context DIE already).
(the test cases updated were overly-sensitive to offsets or abbreviation
numbers. We don't have a nice way to make these tests more robust as yet
- multiline FileCheck matches would be required)
llvm-svn: 191939
Changed the dwarf aranges code to not use getLabelEndName, as it turns out it's not reliable to call that given user-defined section names. Section names can have characters in that aren't representable as symbol names.
The dwarf-aranges test case has been updated to include a special character, to check this.
This fixes pr17416.
llvm-svn: 191932
DIE::addChild had a shortcircuit that silently no-op'd when a child was
readded to the same parent. This hid some quirky/redundant code in
DwarfDebug/CompileUnit. By removing that functionality and replacing it
with an assert I was able to find and cleanup those cases, mostly
centering around adding members to types in various circumstances.
1) The original oddity I noticed while working on type units (which
actually was helping me in the short term, by accident) was the
addToContextOwner call in constructTypeDIE. This call was completely
bogus (why was it only done for non-virtual types? what relevance does
that have at all) and redundant with the more uniform addToContextOwner
made in getOrCreateTypeDIE.
2) If a member function definition was visited (createSubprogramDIE), it
would attempt to build the member function declaration. The declaration
DIE would then be added to its context, but in building the context (the
type for which this function is a member) the members of the type would
be added to the type automatically, so by the time the context was
constructed, the member function was already associated with it.
3) The same as (2) but without the member function being constructed
first. Whenever a type was constructed, the members would be created and
member functions would be created by getOrCreateSubprogramDIE - this
would lead to the subprogram being added to the (incomplete) type
already, then the general member-construction code would add it again.
llvm-svn: 191928
r191052 added emitting .debug_aranges to Clang, but this
functionality is broken: it uses all MC labels added in DWARF Asm
printer, including the labels for build relocations between
different DWARF sections, like .Lsection_line or .Ldebug_loc0.
As a result, if any DIE .debug_info would contain "DW_AT_location=0x123"
attribute, .debug_aranges would also contain a range starting from 0x123,
breaking tools that rely on this section.
This patch fixes this by using only MC labels that corresponds to the
addresses in the user program.
llvm-svn: 191884
infrastructure.
This was essentially work toward PGO based on a design that had several
flaws, partially dating from a time when LLVM had a different
architecture, and with an effort to modernize it abandoned without being
completed. Since then, it has bitrotted for several years further. The
result is nearly unusable, and isn't helping any of the modern PGO
efforts. Instead, it is getting in the way, adding confusion about PGO
in LLVM and distracting everyone with maintenance on essentially dead
code. Removing it paves the way for modern efforts around PGO.
Among other effects, this removes the last of the runtime libraries from
LLVM. Those are being developed in the separate 'compiler-rt' project
now, with somewhat different licensing specifically more approriate for
runtimes.
llvm-svn: 191835
is updated to use DITypeRef.
Move isUnsignedDIType and getOriginalTypeSize from DebugInfo.h to be static
helper functions in DwarfCompileUnit. We already have a static helper function
"isTypeSigned" in DwarfCompileUnit, and a pointer to DwarfDebug is added to
resolve the derived-from field. All three functions need to go across link
for derived-from fields, so we need to get hold of a type identifier map.
A pointer to DwarfDebug is also added to DbgVariable in order to resolve the
derived-from field.
Debug info verifier is updated to check a derived-from field is a TypeRef.
Verifier will not go across link for derived-from fields, in debug info finder,
we go across the link to add derived-from fields to types.
Function getDICompositeType is only used by dragonegg and since dragonegg does
not generate identifier for types, we use an empty map to resolve the
derived-from field.
When printing a derived-from field, we use DITypeRef::getName to either return
the type identifier or getName of the DIType.
A paired commit at clang is required due to changes to DIBuilder.
llvm-svn: 191800
Rafael Espindola