Summary:
Tested with a unit test because we don't appear to have any transforms
that use this other than ASan, I think.
Fixes PR17935.
Reviewers: nicholas
CC: llvm-commits
Differential Revision: http://llvm-reviews.chandlerc.com/D3194
llvm-svn: 204866
Functions may in an instrumented binary but not in the original source
when they're inserted by the compiler or the runtime. These functions
aren't meaningful to the user, so teach llvm-cov to skip over them
instead of crashing.
llvm-svn: 204863
vector list parameter that is using all lanes "{d0[], d2[]}" but can
match and instruction with a ”{d0, d2}" parameter.
I’m finishing up a fix for proper checking of the unsupported
alignments on vld/vst instructions and ran into this. Thus I don’t
have a test case at this time. And adding all code that will
demonstrate the bug would obscure the very simple one line fix.
So if you would indulge me on not having a test case at this
time I’ll instead offer up a detailed explanation of what is
going on in this commit message.
This instruction:
vld2.8 {d0[], d2[]}, [r4:64]
is not legal as the alignment can only be 16 when the size is 8.
Per this documentation:
A8.8.325 VLD2 (single 2-element structure to all lanes)
<align> The alignment. It can be one of:
16 2-byte alignment, available only if <size> is 8, encoded as a = 1.
32 4-byte alignment, available only if <size> is 16, encoded as a = 1.
64 8-byte alignment, available only if <size> is 32, encoded as a = 1.
omitted Standard alignment, see Unaligned data access on page A3-108.
So when code is added to the llvm integrated assembler to not match
that instruction because of the alignment it then goes on to try to match
other instructions and comes across this:
vld2.8 {d0, d2}, [r4:64]
and and matches it. This is because of the method
ARMOperand::isVecListDPairSpaced() is missing the check of the Kind.
In this case the Kind is k_VectorListAllLanes . While the name of the method
may suggest that this is OK it really should check that the Kind is
k_VectorList.
As the method ARMOperand::isDoubleSpacedVectorAllLanes() is what was
used to match {d0[], d2[]} and correctly checks the Kind:
bool isDoubleSpacedVectorAllLanes() const {
return Kind == k_VectorListAllLanes && VectorList.isDoubleSpaced;
}
where the original ARMOperand::isVecListDPairSpaced() does not check
the Kind:
bool isVecListDPairSpaced() const {
if (isSingleSpacedVectorList()) return false;
return (ARMMCRegisterClasses[ARM::DPairSpcRegClassID]
.contains(VectorList.RegNum));
}
Jim Grosbach has reviewed the change and said: Yep, that sounds right. …
And by "right" I mean, "wow, that's a nasty latent bug I'm really, really
glad to see fixed." :)
rdar://16436683
llvm-svn: 204861
This commit consist of two parts.
The first part fix the PR15967. The wrong conclusion was made when the MaxLookup
limit was reached. The fix introduce a out parameter (MaxLookupReached) to
DecomposeGEPExpression that the function aliasGEP can act upon.
The second part is introducing the constant MaxLookupSearchDepth to make sure
that DecomposeGEPExpression and GetUnderlyingObject use the same search depth.
This is a small cleanup to clarify the original algorithm.
Patch by Karl-Johan Karlsson!
llvm-svn: 204859
In CallInst, op_end() points at the callee, which we don't want to iterate over
when just iterating over arguments. Now take this into account when returning
a iterator_range from arg_operands. Similar reasoning for InvokeInst.
Also adds a unit test to verify this actually works as expected.
llvm-svn: 204851
The edge data structure (EdgeEntry) now holds the indices of its entries in the
adjacency lists of the nodes it connects. This trades a little ugliness for
faster insertion/removal, which is now O(1) with a cheap constant factor. All
of this is implementation detail within the PBQP graph, the external API remains
unchanged.
Individual register allocations are likely to change, since the adjacency lists
will now be ordered differently (or rather, will now be unordered). This
shouldn't affect the average quality of allocations however.
llvm-svn: 204841
These patterns are dead (because v4f32 stores are currently promoted to v4i32
and stored using Altivec instructions), and also are likely not correct
(because they'd store the vector elements in the opposite order from that
assumed by the rest of the Altivec code).
llvm-svn: 204839
These instructions have access to the complete VSX register file. In addition,
they "swap" the order of the elements so that element 0 (the scalar part) comes
first in memory and element 1 follows at a higher address.
llvm-svn: 204838
This patch is in similar vein to what done earlier to Module::globals/aliases
etc. It allows to iterate over function arguments like this:
for (Argument Arg : F.args()) {
...
}
llvm-svn: 204835
In some cases it is possible for CGP to attempt to reuse a base address from
another basic block. In those cases we have to be sure that all the address
math was either done at the same bit width, or that none of it overflowed
before it was extended.
Patch by Louis Gerbarg <lgg@apple.com>
rdar://16307442
llvm-svn: 204833
> For functions where esi is used as base pointer, we would previously fall ba
> from lowering memcpy with "rep movs" because that clobbers esi.
>
> With this patch, we just store esi in another physical register, and restore
> it afterwards. This adds a little bit of register preassure, but the more
> efficient memcpy should be worth it.
>
> Differential Revision: http://llvm-reviews.chandlerc.com/D2968
This didn't work. I was ending up with code like this:
lea edi,[esi+38h]
mov ecx,0Fh
mov edx,esi
mov esi,ebx
rep movs dword ptr es:[edi],dword ptr [esi]
lea ecx,[esi+74h] <-- Ooops, we're now using esi before restoring it from edx.
add ebx,3Ch
mov esi,edx
I guess if we want to do this we need stronger glue or something, or doing the expansion
much later.
llvm-svn: 204829
v2i64 needs to be a legal VSX type because it is the SetCC result type from
v2f64 comparisons. We need to expand all non-arithmetic v2i64 operations.
This fixes the lowering for v2f64 VSELECT.
llvm-svn: 204828
This enables TableGen to generate an additional two operand matcher
for our ArithLogicR class of instructions (constituted by 3 register operands).
E.g.: and $1, $2 <=> and $1, $1, $2
llvm-svn: 204826
parseDirectiveWord is a generic function that parses an expression which
means there's no need for it to have such an specific name. Renaming it to
parseDataDirective so that it can also be used to handle .dword directives[1].
[1]To be added in a follow up commit.
No functional changes.
llvm-svn: 204818
The '.set mips64' directive enables the feature Mips:FeatureMips64
from assembly. Note that it doesn't modify the ELF header as opposed
to the use of -mips64 from the command-line. The reason for this
is that we want to be as compatible as possible with existing assemblers
like GAS.
llvm-svn: 204817
The '.set mips64r2' directive enables the feature Mips:FeatureMips64r2
from assembly. Note that it doesn't modify the ELF header as opposed
to the use of -mips64r2 from the command-line. The reason for this
is that we want to be as compatible as possible with existing assemblers
like GAS.
llvm-svn: 204815
We've already got versions without the barriers, so this just adds IR-level
support for generating the new v8 ones.
rdar://problem/16227836
llvm-svn: 204813
Given that we support multiple directives that enable a particular feature
(e.g. '.set mips16'), it's best to hoist that code into a new function
so that we don't repeat the same pattern w.r.t parsing and handling error cases.
No functional changes.
llvm-svn: 204811
After some discussion on IRC, emitting a call to the library function seems
like a better default, since it will move from a compiler internal error to
a linker error, that the user can work around until LLVM is fixed.
I'm also adding a note on the responsibility of the user to confirm that
the cache was cleared on platforms where nothing is done.
llvm-svn: 204806
The directive '.option pic2' enables PIC from assembly source.
At the moment none of the macros/directives check the PIC bit
but that's going to be fixed relatively soon. For example, the
expansion of macros like 'la' depend on the relocation model.
llvm-svn: 204803
Implementing the LLVM part of the call to __builtin___clear_cache
which translates into an intrinsic @llvm.clear_cache and is lowered
by each target, either to a call to __clear_cache or nothing at all
incase the caches are unified.
Updating LangRef and adding some tests for the implemented architectures.
Other archs will have to implement the method in case this builtin
has to be compiled for it, since the default behaviour is to bail
unimplemented.
A Clang patch is required for the builtin to be lowered into the
llvm intrinsic. This will be done next.
llvm-svn: 204802
With VSX there is a real vector select instruction, and so we should use it.
Note that VSELECT will still scalarize for v2f64 because the corresponding
SetCC result type (v2i64) is not currently a legal type.
llvm-svn: 204801
These are aliases of t4-t7 and are provided for compatibility with both the
original ABI documentation (using t4-t7) and GNU As (using t0-t3)
llvm-svn: 204797
Summary: Added test cases for O32 and N32 on MIPS64.
Reviewers: matheusalmeida
Reviewed By: matheusalmeida
Differential Revision: http://llvm-reviews.chandlerc.com/D3175
llvm-svn: 204796
This reveals a small mistake in mips-register-names.s ($sp is tested twice and
$s8 is not tested) which will be fixed in a follow-up commit.
llvm-svn: 204792
This reverts commit r204781.
I will follow up to with msan folks to see what is what they
were trying to do with aliases to weak aliases.
llvm-svn: 204784
Aliases are just another name for a position in a file. As such, the
regular symbol resolutions are not applied. For example, given
define void @my_func() {
ret void
}
@my_alias = alias weak void ()* @my_func
@my_alias2 = alias void ()* @my_alias
We produce without this patch:
.weak my_alias
my_alias = my_func
.globl my_alias2
my_alias2 = my_alias
That is, in the resulting ELF file my_alias, my_func and my_alias are
just 3 names pointing to offset 0 of .text. That is *not* the
semantics of IR linking. For example, linking in a
@my_alias = alias void ()* @other_func
would require the strong my_alias to override the weak one and
my_alias2 would end up pointing to other_func.
There is no way to represent that with aliases being just another
name, so the best solution seems to be to just disallow it, converting
a miscompile into an error.
llvm-svn: 204781
Adds the different broadcast instructions to the ReplaceableInstrsAVX2 table.
That way the ExeDepsFix pass can take better decisions when AVX2 broadcasts are
across domain (int <-> float).
In particular, prior to this patch we were generating:
vpbroadcastd LCPI1_0(%rip), %ymm2
vpand %ymm2, %ymm0, %ymm0
vmaxps %ymm1, %ymm0, %ymm0 ## <- domain change penalty
Now, we generate the following nice sequence where everything is in the float
domain:
vbroadcastss LCPI1_0(%rip), %ymm2
vandps %ymm2, %ymm0, %ymm0
vmaxps %ymm1, %ymm0, %ymm0
<rdar://problem/16354675>
llvm-svn: 204770
We need .symtab_shndxr if and only if a symbol references a section with an
index >= 0xff00.
The old code was trying to figure out if the section was needed ahead of time,
making it a fairly dependent on the code actually writing the table. It was
also somewhat conservative and would create the section in cases where it was
not needed.
If I remember correctly, the old structure was there so that the sections were
created in the same order gas creates them. That was valuable when MC's support
for ELF was new and we tested with elf-dump.py.
This patch refactors the symbol table creation to another class and makes it
obvious that .symtab_shndxr is really only created when we are about to output
a reference to a section index >= 0xff00.
While here, also improve the tests to use macros. One file is one section
short of needing .symtab_shndxr, the second one has just the right number.
llvm-svn: 204769
The VSX instruction set has two types of FMA instructions: A-type (where the
addend is taken from the output register) and M-type (where one of the product
operands is taken from the output register). This adds a small pass that runs
just after MI scheduling (and, thus, just before register allocation) that
mutates A-type instructions (that are created during isel) into M-type
instructions when:
1. This will eliminate an otherwise-necessary copy of the addend
2. One of the product operands is killed by the instruction
The "right" moment to make this decision is in between scheduling and register
allocation, because only there do we know whether or not one of the product
operands is killed by any particular instruction. Unfortunately, this also
makes the implementation somewhat complicated, because the MIs are not in SSA
form and we need to preserve the LiveIntervals analysis.
As a simple example, if we have:
%vreg5<def> = COPY %vreg9; VSLRC:%vreg5,%vreg9
%vreg5<def,tied1> = XSMADDADP %vreg5<tied0>, %vreg17, %vreg16,
%RM<imp-use>; VSLRC:%vreg5,%vreg17,%vreg16
...
%vreg9<def,tied1> = XSMADDADP %vreg9<tied0>, %vreg17, %vreg19,
%RM<imp-use>; VSLRC:%vreg9,%vreg17,%vreg19
...
We can eliminate the copy by changing from the A-type to the
M-type instruction. This means:
%vreg5<def,tied1> = XSMADDADP %vreg5<tied0>, %vreg17, %vreg16,
%RM<imp-use>; VSLRC:%vreg5,%vreg17,%vreg16
is replaced by:
%vreg16<def,tied1> = XSMADDMDP %vreg16<tied0>, %vreg18, %vreg9,
%RM<imp-use>; VSLRC:%vreg16,%vreg18,%vreg9
and we remove: %vreg5<def> = COPY %vreg9; VSLRC:%vreg5,%vreg9
llvm-svn: 204768
When cross-compiling LLVM itself the configure/make scripts get confused when
creating the needed build host tools. For example, building and configuring
like:
CC_FOR_BUILD='i686-pc-linux-gnu-gcc' CXX_FOR_BUILD='i686-pc-linux-gnu-g++'
CXX='i686-mingw32-g++' CC='i686-mingw32-gcc' LD='i686-mingw32-ld' /scratch
/meadori/llvm-trunk/src/trunk/configure --host=i686-mingw32
CC_FOR_BUILD='i686-pc-linux-gnu-gcc' CXX_FOR_BUILD='i686-pc-linux-gnu-g++'
CXX='i686-mingw32-g++' CC='i686-mingw32-gcc' LD='i686-mingw32-ld' make
causes the following build break:
checking whether the C compiler works... configure: error: cannot run C
compiled programs.
If you meant to cross compile, use `--host'.
See `config.log' for more details.
The 'config.log' shows that i686-mingw32-gcc is being used to create
executables for the build host.
This patch fixes the problem by propogating the names of the build host
tools via BUILD_* when configuring/making BuildTools.
Original patch by Ekaterina Sanina.
llvm-svn: 204760
Although the first two operands are the ones that can be swapped, the tied
input operand is listed before them, so we need to adjust for that.
I have a test case for this, but it goes along with an upcoming commit (so it
will come soon).
llvm-svn: 204748
TableGen will create a lookup table for the A-type FMA instructions providing
their corresponding M-form opcodes. This will be used by upcoming commits.
llvm-svn: 204746
Remove handling of select_cc, since it makes no sense to be there. This
now does nothing, but I'll be adding some handling of other target nodes
soon.
llvm-svn: 204743
Implement Pass::releaseMemory() in BlockFrequencyInfo and
MachineBlockFrequencyInfo. Just delete the private implementation when
not in use. Switch to a std::unique_ptr to make the logic more clear.
<rdar://problem/14292693>
llvm-svn: 204741
If getElementPtr uses a constant as base pointer, then make the constant opaque.
This prevents constant folding it with the offset. The offset can usually be
encoded in the load/store instruction itself and the base address doesn't have
to be rematerialized several times.
llvm-svn: 204739
The cost for the first four stackmap operands was always TCC_Free.
This is only true for the first two operands. All other operands
are TCC_Free if they are within 64bit.
llvm-svn: 204738
Usually opaque constants shouldn't be folded, unless they are simple unary
operations that don't create new constants. Although this shouldn't drop the
opaque constant flag. This commit fixes this.
Related to <rdar://problem/14774662>
llvm-svn: 204737
This used to resort to splitting the 256-bit operation into two 128-bit
shuffles and then recombining the results.
Fixes <rdar://problem/16167303>
llvm-svn: 204735
I found three implementations of this. This splits it out into a new function
and uses it from the three places.
My plan is to add a fourth use when lowering a vector_shuffle:v16i16.
Compared the assembly output of test/CodeGen/X86 before and after.
The only change is due to how the first PSHUFB was generated in
LowerVECTOR_SHUFFLEv8i16. If the shuffle mask specified undef (i.e. -1), the
old implementation would write -1 * 2 and -1 * 2 + 1 (254 and 255) in the
control mask. Now we write 0x80. These are of course interchangeable since
bit 7 decides if a constant zero is written in the result byte. The other
instances of this code use 0x80 consistently.
Related to <rdar://problem/16167303>
llvm-svn: 204734
Summary:
Previously the code didn't check if the before and after types for the
store were pointers to different address spaces. This resulted in
instcombine using a bitcast to convert between pointers to different
address spaces, causing an assertion due to the invalid cast.
It is not be appropriate to use addrspacecast this case because it is
not guaranteed to be a no-op cast. Instead bail out and do not do the
transformation.
CC: llvm-commits
Differential Revision: http://llvm-reviews.chandlerc.com/D3117
llvm-svn: 204733
If we have a loop of the form
for (unsigned n = 0; n != (k & -32); n += 32) {}
then we know that n is always divisible by 32 and the loop must
terminate. Even if we have a condition where the loop counter will
overflow it'll always hold this invariant.
PR19183. Our loop vectorizer creates this pattern and it's also
occasionally formed by loop counters derived from pointers.
llvm-svn: 204728
If GT/UGT or LT/ULT were set to expand, a comparison
with a constant would replace it with the illegal
cond code.
There are several more places later in this function that
will have the same basic problem.
Theoretically R600 should hit this problem for a test,
but for some reason it doesn't.
llvm-svn: 204727
It is impossible to create a hard link to a non existing file, so create a
dummy file, create the link an delete the dummy file.
On windows one cannot remove the current directory, so chdir first.
llvm-svn: 204719
Summary:
Remove the XFAIL added in my previous commit and correct the test such that
it correctly tests the expansion of the assembler temporary.
Also added a test to check that $at is always $1 when written by the
user.
Corrected the new assembler temporary warnings so that they are emitted for
numeric registers too.
Differential Revision: http://llvm-reviews.chandlerc.com/D3169
llvm-svn: 204711
Summary:
The assembler temporary is normally $at ($1) but can be reassigned using
'.set at=$reg'. Regardless of which register is nominated as the assembler
temporary, $at remains $1 when written by the user.
Adds warnings under the following conditions:
* The register nominated as the assembler temporary is used by the user.
* '.set noat' is in effect and $at is used by the user.
Both of these only work for named registers. I have a follow up commit that makes it work for numeric registers as well.
XFAIL set-at-directive.s since it incorrectly tests that $at is redefined by
'.set at=$reg'. Testcases will follow in a separate commit.
Patch by David Chisnall
His work was sponsored by: DARPA, AFRL
Differential Revision: http://llvm-reviews.chandlerc.com/D3167
llvm-svn: 204710
This is a pretty straight forward translation for COFF, we just need to
stick the data in a COMDAT section marked as
IMAGE_COMDAT_SELECT_NODUPLICATES.
N.B. We must be careful to avoid sticking entities with private linkage
in COMDAT groups. COFF is pretty hostile to the renaming of entities so
we must be careful to disallow GlobalVariables with unstable names.
llvm-svn: 204703
Extracts coming from phis were being hoisted, while all others were
sunk to their uses. This was inconsistent and didn't seem to serve a
purpose. Changing all extracts to be sunk to uses is a prerequisite
for adding block frequency to the SLP vectorizer's cost model.
I benchmarked the change in isolation (without block frequency). I
only saw noise on x86 and some potentially significant improvements on
ARM. No major regressions is good enough for me.
llvm-svn: 204699
Implement debug_loc.dwo, as well as llvm-dwarfdump support for dumping
this section.
Outlined in the DWARF5 spec and http://gcc.gnu.org/wiki/DebugFission the
debug_loc.dwo section has more variation than the standard debug_loc,
allowing 3 different forms of entry (plus the end of list entry). GCC
seems to, and Clang certainly, only use one form, so I've just
implemented dumping support for that for now.
It wasn't immediately obvious that there was a good refactoring to share
the implementation of dumping support between debug_loc and
debug_loc.dwo, so they're separate for now - ideas welcome or I may come
back to it at some point.
As per a comment in the code, we could choose different forms that may
reduce the number of debug_addr entries we emit, but that will require
further study.
llvm-svn: 204697
This seems excessive - switching section isn't expensive (or if it is
we're already being wasteful, since we emitted the debug_loc section
symbol earlier anyway) and otherwise there's no work that happens in
this function when the list is empty.
llvm-svn: 204696
This adds a function to Endian.h that reads from and updates a pointer
into a buffer with endian specific data. This is more convenient for
stream-like reading of data than endian::read.
llvm-svn: 204693
When register allocator's stage is RS_Spill, we choose spill over using the CSR
for the first time, if the spill cost is lower than CSRCost.
When register allocator's stage is < RS_Split, we choose pre-splitting over
using the CSR for the first time, if the cost of splitting is lower than
CSRCost.
CSRCost is set with command-line option "regalloc-csr-first-time-cost". The
default value is 0 to generate the same codes as before this commit.
With a value of 15 (1 << 14 is the entry frequency), I measured performance
gain of 3% on 253.perlbmk and 1.7% on 197.parser, with instrumented PGO,
on an arm device.
rdar://16162005
llvm-svn: 204690
Factor out two functions calculateRegionSplitCost and doRegionSplit
from tryRegionSplit. These two functions will be used in coming patches.
rdar://16162005
llvm-svn: 204684
No functional change intended.
Merging up-front rather than delaying this task until later. This just
seems simpler and more efficient (avoiding growing the debug loc list
only to have to skip over those post-merged entries, etc).
llvm-svn: 204679
This is used to avoid relocations in the dwo file by allowing
DW_AT_ranges specified in debug_info.dwo to be relative to this base
address. (r204667 implements the base-relative DW_AT_ranges side of
this)
llvm-svn: 204672
This removes the debug_ranges relocations from debug_info.dwo (but
doesn't implement the DW_AT_GNU_ranges_base which is also necessary for
correct functioning)
llvm-svn: 204668
Try to match scalar and first like the other instructions.
Expand 64-bit ands to a pair of 32-bit ands since that is not
available on the VALU.
llvm-svn: 204660
never returns, which is true by design.
Initially assumed that the reason is llvm_unreachable being dependent on NDEBUG.
However, even if llvm_unreachable is replaced by __assume(false), VC still warns in
Release modes but not in Debug modes...
The real reason turned out to be optimization flags.
With /Od in Debug modes the warning is not issued whereas with /O1 it is.
I could not find any documentation to this effect, but it is reproducable:
Try compiling http://msdn.microsoft.com/en-us/library/khwfyc5d(v=vs.90).aspx
with /O1 and then with /Od.
llvm-svn: 204659
As a first step towards real little-endian code generation, this patch
changes the PowerPC MC layer to actually generate little-endian object
files. This involves passing the little-endian flag through the various
layers, including down to createELFObjectWriter so we actually get basic
little-endian ELF objects, emitting instructions in little-endian order,
and handling fixups and relocations as appropriate for little-endian.
The bulk of the patch is to update most test cases in test/MC/PowerPC
to verify both big- and little-endian encodings. (The only test cases
*not* updated are those that create actual big-endian ABI code, like
the TLS tests.)
Note that while the object files are now little-endian, the generated
code itself is not yet updated, in particular, it still does not adhere
to the ELFv2 ABI.
llvm-svn: 204634
Those patterns are used when the load cannot be folded into the related broadcast
during the select phase.
This happens when the load gets additional uses that were not anticipated during
the previous lowering phases (constant vector to constant load, then constant
load reused) or when selection DAG is not able to prove that folding the load
will not create a cycle in the DAG.
<rdar://problem/16074331>
llvm-svn: 204631
Previously we would print an error message on machines where the only VS
version we find is 2013, even though we successfully install the integration
files for it.
Also, we shouldn't have two END labels.
llvm-svn: 204629
This can be observed with the old testcase of CodeGen/X86/pr12312.ll:
47c47
< vorps %ymm0, %ymm1, %ymm0
---
> vorps %ymm1, %ymm0, %ymm0
97c97
< vorps %ymm1, %ymm0, %ymm0
---
> vorps %ymm0, %ymm1, %ymm0
The vector VecIns is populated with all the values from VecInMap. This is done
while iterating VecInMap. VecInMap uses a hash of pointer values so the
resulting order can vary depending on the memory layout.
The fix is to populate the vector VecIns earlier as VecInMap is populated.
This is done in DAG traversal order.
Fixes <rdar://problem/16398806>
llvm-svn: 204623
[PPC64LE] ELFv2 ABI updates for the .opd section
The PPC64 Little Endian (PPC64LE) target supports the ELFv2 ABI, and as
such, does not have a ".opd" section. This is keyed off a _CALL_ELF=2
macro check.
The CALL_ELF check is not clearly documented at this time. The basis
for usage in this patch is from the gcc thread here:
http://gcc.gnu.org/ml/gcc-patches/2013-11/msg01144.html
> Adding comment from Uli:
Looks good to me. I think the old-style JIT doesn't really work
anyway for 64-bit, but at least with this patch LLVM will compile
and link again on a ppc64le host ...
llvm-svn: 204614
Summary:
These expressions already worked but weren't tested.
Patch by Robert N. M. Watson and David Chisnall (it was originally two patches)
Their work was sponsored by: DARPA, AFRL
Differential Revision: http://llvm-reviews.chandlerc.com/D3156
llvm-svn: 204612
I'm under the impression that we used to infer the isCommutable flag from the
instruction-associated pattern. Regardless, we don't seem to do this (at least
by default) any more. I've gone through all of our instruction definitions, and
marked as commutative all of those that should be trivial to commute (by
exchanging the first two operands). There has been special code for the RL*
instructions, and that's not changed.
Before this change, we had the following commutative instructions:
RLDIMI
RLDIMIo
RLWIMI
RLWIMI8
RLWIMI8o
RLWIMIo
XSADDDP
XSMULDP
XVADDDP
XVADDSP
XVMULDP
XVMULSP
After:
ADD4
ADD4o
ADD8
ADD8o
ADDC
ADDC8
ADDC8o
ADDCo
ADDE
ADDE8
ADDE8o
ADDEo
AND
AND8
AND8o
ANDo
CRAND
CREQV
CRNAND
CRNOR
CROR
CRXOR
EQV
EQV8
EQV8o
EQVo
FADD
FADDS
FADDSo
FADDo
FMADD
FMADDS
FMADDSo
FMADDo
FMSUB
FMSUBS
FMSUBSo
FMSUBo
FMUL
FMULS
FMULSo
FMULo
FNMADD
FNMADDS
FNMADDSo
FNMADDo
FNMSUB
FNMSUBS
FNMSUBSo
FNMSUBo
MULHD
MULHDU
MULHDUo
MULHDo
MULHW
MULHWU
MULHWUo
MULHWo
MULLD
MULLDo
MULLW
MULLWo
NAND
NAND8
NAND8o
NANDo
NOR
NOR8
NOR8o
NORo
OR
OR8
OR8o
ORo
RLDIMI
RLDIMIo
RLWIMI
RLWIMI8
RLWIMI8o
RLWIMIo
VADDCUW
VADDFP
VADDSBS
VADDSHS
VADDSWS
VADDUBM
VADDUBS
VADDUHM
VADDUHS
VADDUWM
VADDUWS
VAND
VAVGSB
VAVGSH
VAVGSW
VAVGUB
VAVGUH
VAVGUW
VMADDFP
VMAXFP
VMAXSB
VMAXSH
VMAXSW
VMAXUB
VMAXUH
VMAXUW
VMHADDSHS
VMHRADDSHS
VMINFP
VMINSB
VMINSH
VMINSW
VMINUB
VMINUH
VMINUW
VMLADDUHM
VMULESB
VMULESH
VMULEUB
VMULEUH
VMULOSB
VMULOSH
VMULOUB
VMULOUH
VNMSUBFP
VOR
VXOR
XOR
XOR8
XOR8o
XORo
XSADDDP
XSMADDADP
XSMAXDP
XSMINDP
XSMSUBADP
XSMULDP
XSNMADDADP
XSNMSUBADP
XVADDDP
XVADDSP
XVMADDADP
XVMADDASP
XVMAXDP
XVMAXSP
XVMINDP
XVMINSP
XVMSUBADP
XVMSUBASP
XVMULDP
XVMULSP
XVNMADDADP
XVNMADDASP
XVNMSUBADP
XVNMSUBASP
XXLAND
XXLNOR
XXLOR
XXLXOR
This is a by-inspection change, and I'm not sure how to write a reliable test
case. I would like advice on this, however.
llvm-svn: 204609
Summary:
- If only two registers are passed to a three-register operation, then the
first argument is both source and destination register.
- If a non-register is passed as the last argument, generate the immediate
version of the instruction.
Also mark DADD commutative and add scheduling information (to the generic
scheduler), and implement DSUB.
Patch by David Chisnall
His work was sponsored by: DARPA, AFRL
CC: theraven
Differential Revision: http://llvm-reviews.chandlerc.com/D3148
llvm-svn: 204605
I've done some experimentation with this, and it looks like using the
lower-latency (but lower throughput) copy instruction is essentially always the
right thing to do.
My assumption is that, in order to be relatively sure that the higher-latency
copy will increase throughput, we'd want to have it unlikely to be in-flight
with its use. On the P7, the global completion table (GCT) can hold a maximum
of 120 instructions, shared among all active threads (up to 4), giving 30
instructions per thread. So specifically, I'd require at least that many
instructions between the copy and the use before the high-latency variant is
used.
Trying this, however, over the entire test suite resulted in zero cases where
the high-latency form would be preferable. This may be a consequence of the
fact that the scheduler views copies as free, and so they tend to end up close
to their uses. For this experiment I created a function:
unsigned chooseVSXCopy(MachineBasicBlock &MBB,
MachineBasicBlock::iterator I,
unsigned DestReg, unsigned SrcReg,
unsigned StartDist = 1,
unsigned Depth = 3) const;
with an implementation like:
if (!Depth)
return PPC::XXLOR;
const unsigned MaxDist = 30;
unsigned Dist = StartDist;
for (auto J = I, JE = MBB.end(); J != JE && Dist <= MaxDist; ++J) {
if (J->isTransient() && !J->isCopy())
continue;
if (J->isCall() || J->isReturn() || J->readsRegister(DestReg, TRI))
return PPC::XXLOR;
++Dist;
}
// We've exceeded the required distance for the high-latency form, use it.
if (Dist > MaxDist)
return PPC::XVCPSGNDP;
// If this is only an exit block, use the low-latency form.
if (MBB.succ_empty())
return PPC::XXLOR;
// We've reached the end of the block, check the successor blocks (up to some
// depth), and use the high-latency form if that is okay with all successors.
for (auto J = MBB.succ_begin(), JE = MBB.succ_end(); J != JE; ++J) {
if (chooseVSXCopy(**J, (*J)->begin(), DestReg, SrcReg,
Dist, --Depth) == PPC::XXLOR)
return PPC::XXLOR;
}
// All of our successor blocks seem okay with the high-latency variant, so
// we'll use it.
return PPC::XVCPSGNDP;
and then changed the copy opcode selection from:
Opc = PPC::XXLOR;
to:
Opc = chooseVSXCopy(MBB, std::next(I), DestReg, SrcReg);
In conclusion, I'm removing the FIXME from the comment, because I believe that
there is, at least absent other examples, nothing to fix.
llvm-svn: 204591
This is a pretty straight forward translation for COFF, we just need to
stick the function in a COMDAT section marked as
IMAGE_COMDAT_SELECT_NODUPLICATES.
llvm-svn: 204565
Hal Finkel