usage of the shuffle bitmask. Both work in 128-bit lanes without
crossing, but in the former the mask of the high part is the same
used by the low part while in the later both lanes have independent
masks. Handle this properly and and add support for vpermilpd.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@136200 91177308-0d34-0410-b5e6-96231b3b80d8
When splitting global live ranges, it is now possible to split for
multiple destination intervals at once. Previously, we only had the main
and stack intervals.
Each edge bundle is assigned to a split candidate, and splitAroundRegion
will insert copies between the candidate intervals and the stack
interval as needed.
The multi-way splitting is used to split around compact regions when
enabled with -compact-regions. The best candidate register still gets
all the bundles it wants, but everything outside the main interval is
first split around compact regions before we create single-block
intervals.
Compact region splitting still causes some regressions, so it is not
enabled by default.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@136186 91177308-0d34-0410-b5e6-96231b3b80d8
These copies would coalesce easily, but the resulting value would be
defined by a deleted instruction. Now we also remove the undefined value
number from the destination register.
This fixes PR10503.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@136174 91177308-0d34-0410-b5e6-96231b3b80d8
On x86 we can't encode an immediate LHS of a sub directly. If the RHS comes from a XOR with a constant we can
fold the negation into the xor and add one to the immediate of the sub. Then we can turn the sub into an add,
which can be commuted and encoded efficiently.
This code is generated for __builtin_clz and friends.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@136167 91177308-0d34-0410-b5e6-96231b3b80d8
different from the previous 128-bit because they work in lanes.
Update a few comments and add testcases
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@136157 91177308-0d34-0410-b5e6-96231b3b80d8
Allow the rot_imm operand to be optional. This sets the stage for refactoring
away the "rr" versions from the multiclasses and replacing them with Pat<>s.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@136154 91177308-0d34-0410-b5e6-96231b3b80d8
Start of cleaning this up a bit. First step is to remove the encoder hook by
storing the operand as the bits it'll actually encode to so it can just be
directly used. Map it to the assembly source values 8/16/24 when we print it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@136152 91177308-0d34-0410-b5e6-96231b3b80d8
exit. Added an interfaces for querying either the loop's exact/max
backedge taken count or a specific loop exit's not-taken count.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@136100 91177308-0d34-0410-b5e6-96231b3b80d8
No intendeded functional change. Just cleaning up a bit to make things more
self-consistent in layout and style.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@136095 91177308-0d34-0410-b5e6-96231b3b80d8
When dead code elimination deletes a PHI value, the virtual register may
split into multiple connected components. In that case, revert each
component to the RS_Assign stage.
The new components are guaranteed to be smaller (the original value
numbers are distributed among the components), so this will always be
making progress. The components are now allowed to evict other live
ranges or be split again.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@136034 91177308-0d34-0410-b5e6-96231b3b80d8
The first problem to fix is to stop creating synthetic *Table_gen
targets next to all of the LLVM libraries. These had no real effect as
CMake specifies that add_custom_command(OUTPUT ...) directives (what the
'tablegen(...)' stuff expands to) are implicitly added as dependencies
to all the rules in that CMakeLists.txt.
These synthetic rules started to cause problems as we started more and
more heavily using tablegen files from *subdirectories* of the one where
they were generated. Within those directories, the set of tablegen
outputs was still available and so these synthetic rules added them as
dependencies of those subdirectories. However, they were no longer
properly associated with the custom command to generate them. Most of
the time this "just worked" because something would get to the parent
directory first, and run tablegen there. Once run, the files existed and
the build proceeded happily. However, as more and more subdirectories
have started using this, the probability of this failing to happen has
increased. Recently with the MC refactorings, it became quite common for
me when touching a large enough number of targets.
To add insult to injury, several of the backends *tried* to fix this by
adding explicit dependencies back to the parent directory's tablegen
rules, but those dependencies didn't work as expected -- they weren't
forming a linear chain, they were adding another thread in the race.
This patch removes these synthetic rules completely, and adds a much
simpler function to declare explicitly that a collection of tablegen'ed
files are referenced by other libraries. From that, we can add explicit
dependencies from the smaller libraries (such as every architectures
Desc library) on this and correctly form a linear sequence. All of the
backends are updated to use it, sometimes replacing the existing attempt
at adding a dependency, sometimes adding a previously missing dependency
edge.
Please let me know if this causes any problems, but it fixes a rather
persistent and problematic source of build flakiness on our end.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@136023 91177308-0d34-0410-b5e6-96231b3b80d8
This is just a LangRef entry and reading/writing/memory representation; optimizer+codegen support coming soon.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@136009 91177308-0d34-0410-b5e6-96231b3b80d8
shuffle before inserting on a 256-bit vector.
- Add AVX versions of movd/movq instructions
- Introduce a few COPY patterns to match insert_subvector instructions.
This turns a trivial insert_subvector instruction into a register copy,
coalescing the xmm into a ymm and avoid emiting on more instruction.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@136002 91177308-0d34-0410-b5e6-96231b3b80d8
Fix the Rn register encoding for both SSAT and USAT. Update the parsing of the
shift operand to correctly handle the allowed shift types and immediate ranges
and issue meaningful diagnostics when an illegal value or shift type is
specified. Add aliases to parse an ommitted shift operand (default value of
'lsl #0').
Add tests for diagnostics and proper encoding.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@135990 91177308-0d34-0410-b5e6-96231b3b80d8
Addresses PR10466, although the crash from that PR only triggers in cases where DAGCombine misses optimizing a shuffle.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@135980 91177308-0d34-0410-b5e6-96231b3b80d8
assert-path code, as previously we would have fallen off the end of the
function, but please review and let me know if this should go somewhere
else.
This fixes a Clang warning:
lib/MC/MCMachOStreamer.cpp:201:11: error: enumeration value 'MCSA_IndirectSymbol' not handled in switch [-Werror,-Wswitch-enum]
switch (Attribute) {
^
1 error generated.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@135976 91177308-0d34-0410-b5e6-96231b3b80d8
The shift immediate encoding, printing, etc. is handled directly by the
enclosing operand definition, so it should be a vanilla immediate, not a
nested complex operand (shift_imm).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@135968 91177308-0d34-0410-b5e6-96231b3b80d8
Remove some inititalizers that are the same as the default, move defs next to
their (singular) uses and generally simplify some formatting of asm operand
definitions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@135946 91177308-0d34-0410-b5e6-96231b3b80d8
unwind encoding for that function. This simply crawls through the prolog looking
for machine instrs marked as "frame setup". It can calculate from these what the
compact unwind should look like.
This is currently disabled because of needed linker support. But initial tests
look good.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@135922 91177308-0d34-0410-b5e6-96231b3b80d8
The .local, .hidden, .internal, and .protected are not legal for all supported
file formats (in particular, they're invalid for MachO). Move the parsing for
them into the ELF assembly parser since that's the format they're for.
Similarly, .weak is used by COFF and ELF, but not MachO, so move the parsing
to the COFF and ELF asm parsers. Previously, using any of these directives
on Darwin would result in an assertion failure in the parser; now we get
a diagnostic as we should.
rdar://9827089
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@135921 91177308-0d34-0410-b5e6-96231b3b80d8
This mechanism already exists, but the RS_Split2 stage makes it clearer.
When live range splitting creates ranges that may not be making
progress, they are marked RS_Split2 instead of RS_New. These ranges may
be split again, but only in a way that can be proven to make progress.
For local ranges, that means they must be split into ranges used by
strictly fewer instructions.
For global ranges, region splitting is bypassed and the RS_Split2
ranges go straight to per-block splitting.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@135912 91177308-0d34-0410-b5e6-96231b3b80d8
