For some reason these were not included in the list of Neon intrinsics in
ARM's documentation, so they didn't make it into Clang either.
llvm-svn: 120815
declarations.
The motivation for this patch is that linkage/visibility computations
are linear in the number of redeclarations of an entity, and we've run
into a case where a single translation unit has > 6500 redeclarations
of the same (unused!) external variable. Since each redeclaration
involves a linkage check, the resulting quadratic behavior makes Clang
slow to a crawl. With this change, a simple test with 512
redeclarations of a variable syntax-checks ~20x faster than
before.
That said, I hate this change, and will probably end up reverting it
in a few hours. Reasons to hate it:
- It makes NamedDecl larger, since we don't have enough free bits in
Decl to squeeze in the extra information about caching.
- There are way too many places where we need to invalidate this
cache, because the visibility of a declaration can change due to
redeclarations (!). Despite self-hosting and passing the testsuite,
I have no confidence that I've found all of places where this cache
needs to be invalidated.
llvm-svn: 120808
This is currently the same as a lowercase version of the record name, but
it will allow us to have multiple records with the same name, which is
needed for intrinsics (e.g., vmul and vmull) that are implemented
differently depending on the type.
llvm-svn: 120807
a node in the trimmed graph might not always
correctly map back to the original error node.
This could cause a crash in some cases when
flagging memory leaks.
llvm-svn: 120795
about deprecated Objective-C pointer conversions. Plus, make sure to
actually set an appropriate AssignmentAction when performing an
implicit conversion from an InitializationSequence. Fixes regressions
in the GCC DejaGNU testsuite.
llvm-svn: 120744
TemplateArgumentLocInfo. Unfortunately, this means that we lose some
internal consistency checking when building a debug Clang. However,
having data structures change size/layout depending on NDEBUG causes
pain for clients of the Clang API.
llvm-svn: 120706
This does not work so well with the -fno-lax-vector-conversions option for
testing the arm_neon.h header but that is a really useful test, so I split
this out to a separate Sema test to check for the warning.
llvm-svn: 120694
when an initializer is variable (I handled the constant case in a previous
patch). This has three pieces:
1. Enhance AggValueSlot to have a 'isZeroed' bit to tell CGExprAgg that
the memory being stored into has previously been memset to zero.
2. Teach CGExprAgg to not emit stores of zero to isZeroed memory.
3. Teach CodeGenFunction::EmitAggExpr to scan initializers to determine
whether they are profitable to emit a memset + inividual stores vs
stores for everything.
The heuristic used is that a global has to be more than 16 bytes and
has to be 3/4 zero to be candidate for this xform. The two testcases
are illustrative of the scenarios this catches. We now codegen test9 into:
call void @llvm.memset.p0i8.i64(i8* %0, i8 0, i64 400, i32 4, i1 false)
%.array = getelementptr inbounds [100 x i32]* %Arr, i32 0, i32 0
%tmp = load i32* %X.addr, align 4
store i32 %tmp, i32* %.array
and test10 into:
call void @llvm.memset.p0i8.i64(i8* %0, i8 0, i64 392, i32 8, i1 false)
%tmp = getelementptr inbounds %struct.b* %S, i32 0, i32 0
%tmp1 = getelementptr inbounds %struct.a* %tmp, i32 0, i32 0
%tmp2 = load i32* %X.addr, align 4
store i32 %tmp2, i32* %tmp1, align 4
%tmp5 = getelementptr inbounds %struct.b* %S, i32 0, i32 3
%tmp10 = getelementptr inbounds %struct.a* %tmp5, i32 0, i32 4
%tmp11 = load i32* %X.addr, align 4
store i32 %tmp11, i32* %tmp10, align 4
Previously we produced 99 stores of zero for test9 and also tons for test10.
This xforms should substantially speed up -O0 builds when it kicks in as well
as reducing code size and optimizer heartburn on insane cases. This resolves
PR279.
llvm-svn: 120692
a global is larger than 32 bytes and has fewer than 6 non-zero values in the
initializer. Previously we'd turn something like this:
char test8(int X) {
char str[10000] = "abc";
into a 10K global variable which we then memcpy'd from. Now we generate:
%str = alloca [10000 x i8], align 16
%tmp = getelementptr inbounds [10000 x i8]* %str, i64 0, i64 0
call void @llvm.memset.p0i8.i64(i8* %tmp, i8 0, i64 10000, i32 16, i1 false)
store i8 97, i8* %tmp, align 16
%0 = getelementptr [10000 x i8]* %str, i64 0, i64 1
store i8 98, i8* %0, align 1
%1 = getelementptr [10000 x i8]* %str, i64 0, i64 2
store i8 99, i8* %1, align 2
Which is much smaller in space and also likely faster.
This is part of PR279
llvm-svn: 120645
Check for compatible gcc, Altivec and Neon vectors before handling the
lax-vector-conversions case. Otherwise there is no way to avoid the
warnings from -Wvector-conversions.
llvm-svn: 120633
Fariborz Jahanian