Summary:
Windows does not allow globals to be initialised to point to globals in
another DLL. Exported globals may be referenced only from code. Work
around this by creating an initialiser that runs in early library
initialisation and sets the isa pointer.
Reviewers: rjmccall
Reviewed By: rjmccall
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D50436
llvm-svn: 339317
gcc defines an intrinsic called __builtin_clrsb which counts the number of extra sign bits on a number. This is equivalent to counting the number of leading zeros on a positive number or the number of leading ones on a negative number and subtracting one from the result. Since we can't count leading ones we need to invert negative numbers to count zeros.
This patch will cause the builtin to be expanded inline while gcc uses a call to a function like clrsbdi2 that is implemented in libgcc. But this is similar to what we already do for popcnt. And I don't think compiler-rt supports clrsbdi2.
Differential Revision: https://reviews.llvm.org/D50168
llvm-svn: 339282
r330571 added a new FrontendTimesIsEnabled variable and replaced many usages of llvm::TimePassesIsEnabled. Including the place that set llvm::TimePassesIsEnabled for -ftime-report. The effect of this is that -ftime-report now only contains the timers specifically referenced in CodeGenAction.cpp and none of the timers in the backend.
This commit adds back the assignment, but otherwise leaves everything else unchanged.
llvm-svn: 339281
As suggested by @theraven on PR38210, this patch fixes the gcc -Woverloaded-virtual warnings by renaming the extra CGObjCGNU::GetSelector method to CGObjCGNU::GetTypedSelector
Differential Revision: https://reviews.llvm.org/D50448
llvm-svn: 339264
declare target.
According to OpenMP 5.0, variables captured in lambdas in declare target
regions must be considered as implicitly declare target.
llvm-svn: 339152
Always emit alloca in entry block for enqueue_kernel builtin.
Ensures the statically sized alloca is not converted to DYNAMIC_STACKALLOC
later because it is not in the entry block.
llvm-svn: 339150
These were intended to allow non-fragile and fragile ABI code to be
mixed, as long as the fragile classes were higher up the hierarchy than
the non-fragile ones. Unfortunately:
- No one actually wants to do this.
- Recent versions of Linux's run-time linker break it.
llvm-svn: 339128
Generate DILabel metadata and call llvm.dbg.label after label
statement to associate the metadata with the label.
After fixing PR37395.
Differential Revision: https://reviews.llvm.org/D45045
llvm-svn: 338989
After refactoring DbgInfoIntrinsic class hierarchy, we use
DbgVariableIntrinsic as the base class of variable debug info.
In resolveTopLevelMetadata() in CGVTables.cpp, we only care about
dbg.value, so we try to cast the instructions to DbgVariableIntrinsic
before resolving variables.
Differential Revision: https://reviews.llvm.org/D50226
llvm-svn: 338985
When a non-extended temporary object is created in a conditional branch, the
lifetime of that temporary ends outside the conditional (at the end of the
full-expression). If we're inserting lifetime markers, this means we could end
up generating
if (some_cond) {
lifetime.start(&tmp);
Tmp::Tmp(&tmp);
}
// ...
if (some_cond) {
lifetime.end(&tmp);
}
... for a full-expression containing a subexpression of the form `some_cond ?
Tmp().x : 0`. This patch moves the lifetime start for such a temporary out of
the conditional branch so that we don't need to generate an additional basic
block to hold the lifetime end marker.
This is disabled if we want precise lifetime markers (for asan's
stack-use-after-scope checks) or of the temporary has a non-trivial destructor
(in which case we'd generate an extra basic block anyway to hold the destructor
call).
Differential Revision: https://reviews.llvm.org/D50286
llvm-svn: 338945
Encoding offload target triples onto comdat group key for offload initialization
code guarantees that it will be executed once per each unique combination of
offload targets.
Differential Revision: https://reviews.llvm.org/D50218
llvm-svn: 338916
Found by KlockWorks, this variable is properly protected, however
the conditions in the test that initializes it and the one that uses
it could diverge, it seems to me that this is a 'free' init that will
prevent issues if one of the conditions is ever modified without the other.
llvm-svn: 338909
Ensures the statically sized alloca is not converted to DYNAMIC_STACKALLOC
later because it is not in the entry block.
Differential Revision: https://reviews.llvm.org/D50104
llvm-svn: 338899
Summary:
Emit !llvm.mem.parallel_loop_access metadata for memory accesses even if the parallel loop is not the top on the loop stack.
Fixes llvm.org/PR37558.
Reviewers: ABataev, hfinkel, amusman, tyler.nowicki
Reviewed By: hfinkel
Subscribers: Meinersbur, hfinkel, cfe-commits
Differential Revision: https://reviews.llvm.org/D48808
llvm-svn: 338810
The way address space declarations for builtins currently work
is nearly useless. The code assumes the address spaces used for
builtins is a confusingly named "target address space" from user
code using __attribute__((address_space(N))) that matches
the builtin declaration. There's no way to use this to declare
a builtin that returns a language specific address space.
The terminology used is highly cofusing since it has nothing
to do with the the address space selected by the target to use
for a language address space.
This feature is essentially unused as-is. AMDGPU and NVPTX
are the only in-tree targets attempting to use this. The AMDGPU
builtins certainly do not behave as intended (i.e. all of the
builtins returning pointers can never compile because the numbered
address space never matches the expected named address space).
The NVPTX builtins are missing tests for some, and the others
seem to rely on an implicit addrspacecast.
Change the used address space for builtins based on a target
hook to allow using a language address space for a builtin.
This allows the same builtin declaration to be used for multiple
languages with similarly purposed address spaces (e.g. the same
AMDGPU builtin can be used in OpenCL and CUDA even though the
constant address spaces are arbitarily different).
This breaks the possibility of using arbitrary numbered
address spaces alongside the named address spaces for builtins.
If this is an issue we probably need to introduce another builtin
declaration character to distinguish language address spaces from
so-called "target address spaces".
llvm-svn: 338707
This adds support for the unroll_and_jam pragma, to go with the recently
added unroll and jam pass. The name of the pragma is the same as is used
in the Intel compiler, and most of the code works the same as for unroll.
#pragma clang loop unroll_and_jam has been separated into a different
patch. This part adds #pragma unroll_and_jam with an optional count, and
#pragma no_unroll_and_jam to disable the transform.
Differential Revision: https://reviews.llvm.org/D47267
llvm-svn: 338566
offload targets.
Changed the linkage of omp_offloading.img_start.<triple> and omp_offloading.img_end.<triple> symbols from external to external weak to allow dropping of some targets during linking.
llvm-svn: 338413
No need to change the linkage, we can avoid the problem using special variable. That points to the original variable and, thus, prevent some of the optimizations that might break the compilation.
llvm-svn: 338399
OpenCL block literal structs have different fields which are now correctly
identified in the debug info.
Differential Revision: https://reviews.llvm.org/D49930
llvm-svn: 338299
Summary:
C and C++ are interesting languages. They are statically typed, but weakly.
The implicit conversions are allowed. This is nice, allows to write code
while balancing between getting drowned in everything being convertible,
and nothing being convertible. As usual, this comes with a price:
```
unsigned char store = 0;
bool consume(unsigned int val);
void test(unsigned long val) {
if (consume(val)) {
// the 'val' is `unsigned long`, but `consume()` takes `unsigned int`.
// If their bit widths are different on this platform, the implicit
// truncation happens. And if that `unsigned long` had a value bigger
// than UINT_MAX, then you may or may not have a bug.
// Similarly, integer addition happens on `int`s, so `store` will
// be promoted to an `int`, the sum calculated (0+768=768),
// and the result demoted to `unsigned char`, and stored to `store`.
// In this case, the `store` will still be 0. Again, not always intended.
store = store + 768; // before addition, 'store' was promoted to int.
}
// But yes, sometimes this is intentional.
// You can either make the conversion explicit
(void)consume((unsigned int)val);
// or mask the value so no bits will be *implicitly* lost.
(void)consume((~((unsigned int)0)) & val);
}
```
Yes, there is a `-Wconversion`` diagnostic group, but first, it is kinda
noisy, since it warns on everything (unlike sanitizers, warning on an
actual issues), and second, there are cases where it does **not** warn.
So a Sanitizer is needed. I don't have any motivational numbers, but i know
i had this kind of problem 10-20 times, and it was never easy to track down.
The logic to detect whether an truncation has happened is pretty simple
if you think about it - https://godbolt.org/g/NEzXbb - basically, just
extend (using the new, not original!, signedness) the 'truncated' value
back to it's original width, and equality-compare it with the original value.
The most non-trivial thing here is the logic to detect whether this
`ImplicitCastExpr` AST node is **actually** an implicit conversion, //or//
part of an explicit cast. Because the explicit casts are modeled as an outer
`ExplicitCastExpr` with some `ImplicitCastExpr`'s as **direct** children.
https://godbolt.org/g/eE1GkJ
Nowadays, we can just use the new `part_of_explicit_cast` flag, which is set
on all the implicitly-added `ImplicitCastExpr`'s of an `ExplicitCastExpr`.
So if that flag is **not** set, then it is an actual implicit conversion.
As you may have noted, this isn't just named `-fsanitize=implicit-integer-truncation`.
There are potentially some more implicit conversions to be warned about.
Namely, implicit conversions that result in sign change; implicit conversion
between different floating point types, or between fp and an integer,
when again, that conversion is lossy.
One thing i know isn't handled is bitfields.
This is a clang part.
The compiler-rt part is D48959.
Fixes [[ https://bugs.llvm.org/show_bug.cgi?id=21530 | PR21530 ]], [[ https://bugs.llvm.org/show_bug.cgi?id=37552 | PR37552 ]], [[ https://bugs.llvm.org/show_bug.cgi?id=35409 | PR35409 ]].
Partially fixes [[ https://bugs.llvm.org/show_bug.cgi?id=9821 | PR9821 ]].
Fixes https://github.com/google/sanitizers/issues/940. (other than sign-changing implicit conversions)
Reviewers: rjmccall, rsmith, samsonov, pcc, vsk, eugenis, efriedma, kcc, erichkeane
Reviewed By: rsmith, vsk, erichkeane
Subscribers: erichkeane, klimek, #sanitizers, aaron.ballman, RKSimon, dtzWill, filcab, danielaustin, ygribov, dvyukov, milianw, mclow.lists, cfe-commits, regehr
Tags: #sanitizers
Differential Revision: https://reviews.llvm.org/D48958
llvm-svn: 338288
The "Procedure Call Procedure Call Standard for the ARM® Architecture"
(https://static.docs.arm.com/ihi0042/f/IHI0042F_aapcs.pdf), specifies that
composite types are passed according to their "natural alignment", i.e. the
alignment before alignment adjustment on the entire composite is applied.
The same applies for AArch64 ABI.
Clang, however, used the adjusted alignment.
GCC already implements the ABI correctly. With this patch Clang becomes
compatible with GCC and passes such arguments in accordance with AAPCS.
Differential Revision: https://reviews.llvm.org/D46013
llvm-svn: 338279
This commit increases the number of sections and overall output size of
.o files by 10% and sometimes a bit more. This alone is challenging for
some users, but it also appears to trigger an as-yet unexplained
behavior in the Gold linker where the memory usage increases
considerably more than 10% (we think).
The increase is also frustrating because in many (if not all) cases we
end up with almost all of the growth coming from the ELF overhead of
-ffunction-sections and such, not from actual extra code being emitted.
Richard Smith and Eric Christopher are both going to investigate this
and try to get to the bottom of what is triggering this and whether the
kinds of increases here are sustainable or what options we might have to
minimize the impact they have. However, this is currently breaking
a pretty large number of our users' builds so reverting it while we sort
out how to make progress here. I've seen a longer and more detailed
update to the commit thread.
llvm-svn: 338209
With this change compiler generates alignment checks for wider range
of types. Previously such checks were generated only for the record types
with non-trivial default constructor. So the types like:
struct alignas(32) S2 { int x; };
typedef __attribute__((ext_vector_type(2), aligned(32))) float float32x2_t;
did not get checks when allocated by 'new' expression.
This change also optimizes the checks generated for the arrays created
in 'new' expressions. Previously the check was generated for each
invocation of type constructor. Now the check is generated only once
for entire array.
Differential Revision: https://reviews.llvm.org/D49589
llvm-svn: 338199
CUDA 8.0 E.3.9.4 says: Within the body of a __device__ or __global__
function, only __shared__ variables or variables without any device
memory qualifiers may be declared with static storage class.
It is unclear how a function-scope non-const static variable
without device memory qualifier is implemented, therefore only static
const variable without device memory qualifier is allowed, which
can be emitted as a global variable in constant address space.
Currently clang only allows function-scope static variable with
__shared__ qualifier.
This patch also allows function-scope static const variable without
device memory qualifier and emits it as a global variable in constant
address space.
Differential Revision: https://reviews.llvm.org/D49931
llvm-svn: 338188
Summary: Microsoft's C++ object model for ARM64 is the same as that for X86_64.
For example, small structs with non-trivial copy constructors or virtual
function tables are passed indirectly. Currently, they are passed in registers
when compiled with clang.
Reviewers: rnk, mstorsjo, TomTan, haripul, javed.absar
Reviewed By: rnk, mstorsjo
Subscribers: kristof.beyls, chrib, llvm-commits, cfe-commits
Differential Revision: https://reviews.llvm.org/D49770
llvm-svn: 338076
Summary:
Clang supports the GNU style ``__attribute__((interrupt))`` attribute on RISCV targets.
Permissible values for this parameter are user, supervisor, and machine.
If there is no parameter, then it defaults to machine.
Reference: https://gcc.gnu.org/onlinedocs/gcc/RISC-V-Function-Attributes.html
Based on initial patch by Zhaoshi Zheng.
Reviewers: asb, aaron.ballman
Reviewed By: asb, aaron.ballman
Subscribers: rkruppe, the_o, aaron.ballman, MartinMosbeck, brucehoult, rbar, johnrusso, simoncook, sabuasal, niosHD, kito-cheng, shiva0217, zzheng, edward-jones, mgrang, rogfer01, cfe-commits
Differential Revision: https://reviews.llvm.org/D48412
llvm-svn: 338045
When an exception is thrown in a block copy helper function, captured
objects that have previously been copied should be destructed or
released. Similarly, captured objects that are yet to be released should
be released when an exception is thrown in a dispose helper function.
rdar://problem/42410255
Differential Revision: https://reviews.llvm.org/D49718
llvm-svn: 338041
The first argument for the parallel outlined functions, called as
serialized parallel regions, should be a pointer to the global thread id
that always is 0.
llvm-svn: 337957
Summary:
Right now automatic variables are either initialized with bzero followed by a few stores, or memcpy'd from a synthesized global. We end up encountering a fair amount of code where memcpy of non-zero byte patterns would be better than memcpy from a global because it touches less memory and generates a smaller binary. The optimizer could reason about this, but it's not really worth it when clang already knows.
This code could definitely be more clever but I'm not sure it's worth it. In particular we could track a histogram of bytes seen and figure out (as we do with bzero) if a memset could be followed by a handful of stores. Similarly, we could tune the heuristics for GlobalSize, but using the same as for bzero seems conservatively OK for now.
<rdar://problem/42563091>
Reviewers: dexonsmith
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D49771
llvm-svn: 337887
Generate DILabel metadata and call llvm.dbg.label after label
statement to associate the metadata with the label.
After fixing PR37395.
Differential Revision: https://reviews.llvm.org/D45045
Patch by Hsiangkai Wang.
llvm-svn: 337800
This patch adds support for vrndi_f32() and vrndiq_f32()
intrinsics in AArch32 mode and for vrndns_f32() intrinsic in
AArch64 mode.
Differential Revision: https://reviews.llvm.org/D48829
llvm-svn: 337690
The optimization looks for opportunities to emit bzero, not memset. Rename the functions accordingly (and clang-format the diff) because I want to add a fallback optimization which actually tries to generate memset. bzero is still better and it would confuse the code to merge both.
llvm-svn: 337636
HIP generates one fat binary for all devices after linking. However, for each compilation
unit a ctor function is emitted which register the same fat binary. Measures need to be
taken to make sure the fat binary is only registered once.
Currently each ctor function calls __hipRegisterFatBinary and stores the returned value
to __hip_gpubin_handle. This patch changes the linkage of __hip_gpubin_handle to be linkonce
so that they are shared between LLVM modules. Then this patch adds check of value of
__hip_gpubin_handle to make sure __hipRegisterFatBinary is only called once. The code
is equivalent to
void *_gpubin_handle;
void ctor() {
if (__hip_gpubin_handle == 0) {
__hip_gpubin_handle = __hipRegisterFatBinary(...);
}
// register kernels and variables.
}
The patch also does similar change to dtors so that __hipUnregisterFatBinary
is called once.
Differential Revision: https://reviews.llvm.org/D49083
llvm-svn: 337631
MSVC doesn't, so neither should we.
Fixes PR38004, which is a crash that happens when we try to emit debug
info for a still-dependent partial variable template specialization.
As a follow-up, we should review what we're doing for function and class
member templates. It looks like we don't filter those out, but I can't
seem to get clang to emit any.
llvm-svn: 337616
no-ops.
A non-escaping block on the stack will never be called after its
lifetime ends, so it doesn't have to be copied to the heap. To prevent
a non-escaping block from being copied to the heap, this patch sets
field 'isa' of the block object to NSConcreteGlobalBlock and sets the
BLOCK_IS_GLOBAL bit of field 'flags', which causes the runtime to treat
the block as if it were a global block (calling _Block_copy on the block
just returns the original block and calling _Block_release is a no-op).
Also, a new flag bit 'BLOCK_IS_NOESCAPE' is added, which allows the
runtime or tools to distinguish between true global blocks and
non-escaping blocks.
rdar://problem/39352313
Differential Revision: https://reviews.llvm.org/D49303
llvm-svn: 337580
As documented here: https://software.intel.com/en-us/node/682969 and
https://software.intel.com/en-us/node/523346. cpu_dispatch multiversioning
is an ICC feature that provides for function multiversioning.
This feature is implemented with two attributes: First, cpu_specific,
which specifies the individual function versions. Second, cpu_dispatch,
which specifies the location of the resolver function and the list of
resolvable functions.
This is valuable since it provides a mechanism where the resolver's TU
can be specified in one location, and the individual implementions
each in their own translation units.
The goal of this patch is to be source-compatible with ICC, so this
implementation diverges from the ICC implementation in a few ways:
1- Linux x86/64 only: This implementation uses ifuncs in order to
properly dispatch functions. This is is a valuable performance benefit
over the ICC implementation. A future patch will be provided to enable
this feature on Windows, but it will obviously more closely fit ICC's
implementation.
2- CPU Identification functions: ICC uses a set of custom functions to identify
the feature list of the host processor. This patch uses the cpu_supports
functionality in order to better align with 'target' multiversioning.
1- cpu_dispatch function def/decl: ICC's cpu_dispatch requires that the function
marked cpu_dispatch be an empty definition. This patch supports that as well,
however declarations are also permitted, since the linker will solve the
issue of multiple emissions.
Differential Revision: https://reviews.llvm.org/D47474
llvm-svn: 337552
constant, don't convert the rest into a packed struct.
If an array constant has a large non-zero portion and a large zero
portion, we want to emit the first part as an array and the rest as a
zeroinitializer if possible. This fixes a memory usage regression from
r333141 when compiling PHP.
llvm-svn: 337498
device IDs are now 64-bit integers (as opposed to 32-bit)
map flags are 64-bit long (used to be 32-bit)
mappings for partially mapped structs are now calculated at compile time and members of partially mapped structs are flagged using the MEMBER_OF field
Support for is_device_ptr on struct members was dropped - this functionality is not supported by the OpenMP standard and its implementation is technically infeasible (however, use_device_ptr on struct members works as a non-standard extension of the compiler)
llvm-svn: 337468
The previous version of this patch (r332839) was reverted because it was
causing "definition with same mangled name as another definition" errors
in some module builds. This was caused by an unrelated bug in module
importing which it exposed. The importing problem was fixed in r336240,
so this recommits the original patch (r332839).
Differential Revision: https://reviews.llvm.org/D46685
llvm-svn: 337456
The codegen for this builtin was initially implemented to match GCC.
However, due to interest from users GCC changed behaviour to account for the
big endian bias of the instruction and correct it. This patch brings the
handling inline with GCC.
Fixes https://bugs.llvm.org/show_bug.cgi?id=38192
Differential Revision: https://reviews.llvm.org/D49424
llvm-svn: 337449
Summary:
Support for this option is needed for building Linux kernel.
This is a very frequently requested feature by kernel developers.
More details : https://lkml.org/lkml/2018/4/4/601
GCC option description for -fdelete-null-pointer-checks:
This Assume that programs cannot safely dereference null pointers,
and that no code or data element resides at address zero.
-fno-delete-null-pointer-checks is the inverse of this implying that
null pointer dereferencing is not undefined.
This feature is implemented in as the function attribute
"null-pointer-is-valid"="true".
This CL only adds the attribute on the function.
It also strips "nonnull" attributes from function arguments but
keeps the related warnings unchanged.
Corresponding LLVM change rL336613 already updated the
optimizations to not treat null pointer dereferencing
as undefined if the attribute is present.
Reviewers: t.p.northover, efriedma, jyknight, chandlerc, rnk, srhines, void, george.burgess.iv
Reviewed By: jyknight
Subscribers: drinkcat, xbolva00, cfe-commits
Differential Revision: https://reviews.llvm.org/D47894
llvm-svn: 337433
This patch uses CodeSegAttr to represent __declspec(code_seg) rather than
building on the existing support for #pragma code_seg.
The code_seg declspec is applied on functions and classes. This attribute
enables the placement of code into separate named segments, including compiler-
generated codes and template instantiations.
For more information, please see the following:
https://msdn.microsoft.com/en-us/library/dn636922.aspx
This patch fixes the regression for the support for attribute ((section).
746b78de78
Patch by Soumi Manna (Manna)
Differential Revision: https://reviews.llvm.org/D48841
llvm-svn: 337420
which was reverted in r337336.
The problem that required a revert was fixed in r337338.
Also added a missing "REQUIRES: x86-registered-target" to one of
the tests.
Original commit message:
> Teach Clang to emit address-significance tables.
>
> By default, we emit an address-significance table on all ELF
> targets when the integrated assembler is enabled. The emission of an
> address-significance table can be controlled with the -faddrsig and
> -fno-addrsig flags.
>
> Differential Revision: https://reviews.llvm.org/D48155
llvm-svn: 337339
Causing multiple failures on sanitizer bots due to TLS symbol errors,
e.g.
/usr/bin/ld: __msan_origin_tls: TLS definition in /home/buildbots/ppc64be-clang-test/clang-ppc64be/stage1/lib/clang/7.0.0/lib/linux/libclang_rt.msan-powerpc64.a(msan.cc.o) section .tbss.__msan_origin_tls mismatches non-TLS reference in /tmp/lit_tmp_0a71tA/mallinfo-3ca75e.o
llvm-svn: 337336
By default, we emit an address-significance table on all ELF
targets when the integrated assembler is enabled. The emission of an
address-significance table can be controlled with the -faddrsig and
-fno-addrsig flags.
Differential Revision: https://reviews.llvm.org/D48155
llvm-svn: 337333
If the declare target link entries are created but not used, the
compiler will produce an error message. Patch improves handling of such
situations + improves checks for possibly lost declare target variables.
llvm-svn: 337207
Summary: Automatic variable initialization was generating default-aligned stores (which are deprecated) instead of using the known alignment from the alloca. Further, they didn't specify inbounds.
Subscribers: dexonsmith, cfe-commits
Differential Revision: https://reviews.llvm.org/D49209
llvm-svn: 337041
Summary: In the SPMD case, we need to initialize the data sharing and globalization infrastructure. This covers the case when an SPMD region calls a function in a different compilation unit.
Reviewers: ABataev, carlo.bertolli, caomhin
Reviewed By: ABataev
Subscribers: Hahnfeld, jholewinski, guansong, cfe-commits
Differential Revision: https://reviews.llvm.org/D49188
llvm-svn: 337015
Code in `CodeGenModule::SetFunctionAttributes()` could set an empty
attribute `implicit-section-name` on a function that is affected by
`#pragma clang text="section"`. This is incorrect because the attribute
should contain a valid section name. If the function additionally also
used `__attribute__((section("section")))` then this could result in
emitting the function in a section with an empty name.
The patch fixes the issue by removing the problematic code that sets
empty `implicit-section-name` from
`CodeGenModule::SetFunctionAttributes()` because it is sufficient to set
this attribute only from a similar code in `setNonAliasAttributes()`
when the function is emitted.
Differential Revision: https://reviews.llvm.org/D48916
llvm-svn: 336842
The member init list for the sole constructor for CodeGenFunction
has gotten out of hand, so this patch moves the non-parameter-dependent
initializations into the member value inits.
Note: This is what was intended to be committed in r336726
llvm-svn: 336729
The member init list for the sole constructor for CodeGenFunction
has gotten out of hand, so this patch moves the non-parameter-dependent
initializations into the member value inits.
llvm-svn: 336726
Summary:
Make sure that loop metadata only is put on the backedge
when expanding a do-while loop.
Previously we added the loop metadata also on the branch
in the pre-header. That could confuse optimization passes
and result in the loop metadata being associated with the
wrong loop.
Fixes https://bugs.llvm.org/show_bug.cgi?id=38011
Committing on behalf of deepak2427 (Deepak Panickal)
Reviewers: #clang, ABataev, hfinkel, aaron.ballman, bjope
Reviewed By: bjope
Subscribers: bjope, rsmith, shenhan, zzheng, xbolva00, lebedev.ri, cfe-commits
Tags: #clang
Differential Revision: https://reviews.llvm.org/D48721
llvm-svn: 336717
This will convert the i8 mask argument to <8 x i1> and extract an i1 and then emit a select instruction. This replaces the '(__U & 1)" and ternary operator used in some of intrinsics. The old sequence was lowered to a scalar and and compare. The new sequence uses an i1 vector that will interoperate better with other mask intrinsics.
This removes the need to handle div_ss/sd specially in CGBuiltin.cpp. A follow up patch will add the GCCBuiltin name back in llvm and remove the custom handling.
I made some adjustments to legacy move_ss/sd intrinsics which we reused here to do a simpler extract and insert instead of 2 extracts and two inserts or a shuffle.
llvm-svn: 336622
This is part of an ongoing attempt at making 512 bit vectors illegal in the X86 backend type legalizer due to CPU frequency penalties associated with wide vectors on Skylake Server CPUs. We want the loop vectorizer to be able to emit IR containing wide vectors as intermediate operations in vectorized code and allow these wide vectors to be legalized to 256 bits by the X86 backend even though we are targetting a CPU that supports 512 bit vectors. This is similar to what happens with an AVX2 CPU, the vectorizer can emit wide vectors and the backend will split them. We want this splitting behavior, but still be able to use new Skylake instructions that work on 256-bit vectors and support things like masking and gather/scatter.
Of course if the user uses explicit vector code in their source code we need to not split those operations. Especially if they have used any of the 512-bit vector intrinsics from immintrin.h. And we need to make it so that merely using the intrinsics produces the expected code in order to be backwards compatible.
To support this goal, this patch adds a new IR function attribute "min-legal-vector-width" that can indicate the need for a minimum vector width to be legal in the backend. We need to ensure this attribute is set to the largest vector width needed by any intrinsics from immintrin.h that the function uses. The inliner will be reponsible for merging this attribute when a function is inlined. We may also need a way to limit inlining in the future as well, but we can discuss that in the future.
To make things more complicated, there are two different ways intrinsics are implemented in immintrin.h. Either as an always_inline function containing calls to builtins(can be target specific or target independent) or vector extension code. Or as a macro wrapper around a taget specific builtin. I believe I've removed all cases where the macro was around a target independent builtin.
To support the always_inline function case this patch adds attribute((min_vector_width(128))) that can be used to tag these functions with their vector width. All x86 intrinsic functions that operate on vectors have been tagged with this attribute.
To support the macro case, all x86 specific builtins have also been tagged with the vector width that they require. Use of any builtin with this property will implicitly increase the min_vector_width of the function that calls it. I've done this as a new property in the attribute string for the builtin rather than basing it on the type string so that we can opt into it on a per builtin basis and avoid any impact to target independent builtins.
There will be future work to support vectors passed as function arguments and supporting inline assembly. And whatever else we can find that isn't covered by this patch.
Special thanks to Chandler who suggested this direction and reviewed a preview version of this patch. And thanks to Eric Christopher who has had many conversations with me about this issue.
Differential Revision: https://reviews.llvm.org/D48617
llvm-svn: 336583
In generic data-sharing mode we are allowed to not globalize local
variables that escape their declaration context iff they are declared
inside of the parallel region. We can do this because L2 parallel
regions are executed sequentially and, thus, we do not need to put
shared local variables in the global memory.
llvm-svn: 336567
This case occurs in the intrinsic headers so we should avoid emitting the mask in those cases.
Factor the code into a helper function to make this easy.
llvm-svn: 336472
Shufflevector is easier to generate and matches what the backend pattern matches without relying on constant selects being turned into shuffles.
While I was there I also made the IR regular expressions a little stricter to ensure operand order on the shuffle.
llvm-svn: 336388
This patch removes on optimization used with the TRUE/FALSE
predicates, as was suggested in https://reviews.llvm.org/D45616
for r335339.
The optimization was buggy, since r335339 used it also
for *_mask builtins, without actually applying the mask -- the
mask argument was just ignored.
Reviewers: craig.topper, uriel.k, RKSimon, andrew.w.kaylor, spatel, scanon, efriedma
Reviewed By: spatel
Differential Revision: https://reviews.llvm.org/D48715
llvm-svn: 336355
Update clang to treat fp128 as a valid base type for homogeneous aggregate
passing and returning.
Differential Revision: https://reviews.llvm.org/D48044
llvm-svn: 336308
Summary:
Emmiting new intrinsic that strips invariant.groups to make
devirtulization sound, as described in RFC: Devirtualization v2.
Reviewers: rjmccall, rsmith, amharc, kuhar
Subscribers: llvm-commits, cfe-commits
Differential Revision: https://reviews.llvm.org/D47299
Co-authored-by: Krzysztof Pszeniczny <krzysztof.pszeniczny@gmail.com>
llvm-svn: 336137
This matches the way NVCC does it. Doing module cleanup at global
destructor phase used to work, but is, apparently, too late for
the CUDA runtime in CUDA-9.2, which ends up crashing with double-free.
Differential Revision: https://reviews.llvm.org/D48613
llvm-svn: 335763
Stephen Kelly