RoundTripYAML pass is removed from the regular execution pass in r204296,
so the safeguard to protect it from OOM error is no longer needed, because
we are sure that the pass is only used for tests, and test files are all
small.
We also want to see RoundTripYAML pass to fail in tests if it fails,
rather than silently skipping failing tests.
llvm-svn: 204772
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
Conceptually one of these loops is just a while-loop, but the actual code-gen
is more complicated. We don't instrument all the different control flow edges
to get accurate counts for each conditional branch, nor do I think it makes
sense to do so. Instead, make the simplifying assumption that the loop behaves
like a while-loop. Use the same branch weights for the first check for an
empty collection as would be used for the back-edge of a while loop, and use
that same weighting for the innermost loop, ignoring the possibility that there
may be some extra code to go fetch more elements.
llvm-svn: 204767
if they didn't change, just like it does for
registers. This makes life easier for kernel
debugging and any other situation where values
are read-only.
<rdar://problem/16367795>
llvm-svn: 204764
For small structs, the frame format now prints them as one-liners
This follows the same definition that frame variable does for deciding what a "small struct" is, and as such should be fairly consistent with the variable display in general
llvm-svn: 204762
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
Add a GetFoundationVersion() to AppleObjCRuntime
This API is used to return and cache the major version of Foundation.framework, which is potentially a useful piece of data to key off of to enable or disable certain ObjC related behaviors (especially in data formatters)
llvm-svn: 204756
If a response file is given via command line, the final command line
arguments will not appear in the log because the actual arguments are
in the given file.
This patch is to show the final command line if /verbose is specified
to help users.
llvm-svn: 204754
This change makes significant improvements in the performance of
calculating a UUID within ObjectFileELF, and handles both running
processes and core files correctly. This does lazy evaluation of
UUID generation and caches the result when calculated.
Change by Piotr Rak.
llvm-svn: 204749
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
Also added 'import sys' on some tests that are using non-standard
unittest2.skipUnless blocks with code that is intended to do things
that we have more specializes @* attributes for. These skip
conditions were failing to execute due to missing import, causing
darwin-only tests to run on Linux regardless. Will file a bug for
that separately.
llvm-svn: 204747
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
When there was no process, the expression options were set to not ignore breakpoints. This causes debug info to be generated and causes errors when evaluating simple expressions.
llvm-svn: 204745
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
In gcc using -Ofast forces linking of crtfastmath.o.
In the current clang crtfastmath.o is only linked when -ffast-math/-funsafe-math-optimizations passed. It can lead to performance issues, when using only -Ofast without explicit -ffast-math (I faced with it).
My patch fixes inconsistency with gcc behaviour and also introduces few tests on it.
Patch by Zinovy Nis!
Differential Revision: http://llvm-reviews.chandlerc.com/D3114
llvm-svn: 204742
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
A refinement of r198953 to handle cases where an object is accessed both through
a property getter and through direct ivar access. An object accessed through a
property should always be treated as +0, i.e. not owned by the caller. However,
an object accessed through an ivar may be at +0 or at +1, depending on whether
the ivar is a strong reference. Outside of ARC, we don't have that information,
so we just don't track objects accessed through ivars.
With this change, accessing an ivar directly will deliberately override the +0
provided by a getter, but only if the +0 hasn't participated in other retain
counting yet. That isn't perfect, but it's already unusual for people to be
mixing property access with direct ivar access. (The primary use case for this
is in setters, init methods, and -dealloc.)
Thanks to Ted for spotting a few mistakes in private review.
<rdar://problem/16333368>
llvm-svn: 204730
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
We explicitly specifying all filenames instead of assuming some naming
convention used by clang and opt.
Contributed-by: Johannes Doerfert <doerfert@cs.uni-saarland.de>
llvm-svn: 204726
"Running multithreaded with n threads (from LLDB_TEST_THREADS)" is
incorrect in the common case, because the thread count is now obtained
from a --threads option, or a default of multiprocessing.cpu_count().
I left the messages in for now as there are some intermittent issues
that happen while running tests multithreaded and it may help in
triaging those. We may wish to just remove them later on.
llvm-svn: 204725
Michael J. Spencer