and is unrelated to the NEON intrinsics in arm_neon.h. On little
endian machines it works fine, however on big endian machines it
exhibits surprising behaviour:
uint32x2_t x = {42, 64};
return vget_lane_u32(x, 0); // Will return 64.
Because of this, explicitly call out that it is unsupported on big
endian machines.
This patch will emit the following warning in big-endian mode:
test.c:3:15: warning: vector initializers are a GNU extension and are not compatible with NEON intrinsics [-Wgnu]
int32x4_t x = {0, 1, 2, 3};
^
test.c:3:15: note: consider using vld1q_s32() to initialize a vector from memory, or vcombine_s32(vcreate_s32(), vcreate_s32()) to initialize from integer constants
1 warning generated.
llvm-svn: 211362
CL permits static redeclarations to follow extern declarations. The
storage specifier on the latter declaration has no effect.
This fixes PR20034.
Differential Revision: http://reviews.llvm.org/D4149
llvm-svn: 211238
There comes a time in the life of any amateur code generator when dumb string
concatenation just won't cut it any more. For NeonEmitter.cpp, that time has
come.
There were a bunch of magic type codes which meant different things depending on
the context. There were a bunch of special cases that really had no reason to be
there but the whole thing was so creaky that removing them would cause something
weird to fall over. There was a 1000 line switch statement for code generation
involving string concatenation, which actually did lexical scoping to an extent
(!!) with a bunch of semi-repeated cases.
I tried to refactor this three times in three different ways without
success. The only way forward was to rewrite the entire thing. Luckily the
testing coverage on this stuff is absolutely massive, both with regression tests
and the "emperor" random test case generator.
The main change is that previously, in arm_neon.td a bunch of "Operation"s were
defined with special names. NeonEmitter.cpp knew about these Operations and
would emit code based on a huge switch. Actually this doesn't make much sense -
the type information was held as strings, so type checking was impossible. Also
TableGen's DAG type actually suits this sort of code generation very well
(surprising that...)
So now every operation is defined in terms of TableGen DAGs. There are a bunch
of operators to use, including "op" (a generic unary or binary operator), "call"
(to call other intrinsics) and "shuffle" (take a guess...). One of the main
advantages of this apart from making it more obvious what is going on, is that
we have proper type inference. This has two obvious advantages:
1) TableGen can error on bad intrinsic definitions easier, instead of just
generating wrong code.
2) Calls to other intrinsics are typechecked too. So
we no longer need to work out whether the thing we call needs to be the Q-lane
version or the D-lane version - TableGen knows that itself!
Here's an example: before:
case OpAbdl: {
std::string abd = MangleName("vabd", typestr, ClassS) + "(__a, __b)";
if (typestr[0] != 'U') {
// vabd results are always unsigned and must be zero-extended.
std::string utype = "U" + typestr.str();
s += "(" + TypeString(proto[0], typestr) + ")";
abd = "(" + TypeString('d', utype) + ")" + abd;
s += Extend(utype, abd) + ";";
} else {
s += Extend(typestr, abd) + ";";
}
break;
}
after:
def OP_ABDL : Op<(cast "R", (call "vmovl", (cast $p0, "U",
(call "vabd", $p0, $p1))))>;
As an example of what happens if you do something wrong now, here's what happens
if you make $p0 unsigned before the call to "vabd" - that is, $p0 -> (cast "U",
$p0):
arm_neon.td:574:1: error: No compatible intrinsic found - looking up intrinsic 'vabd(uint8x8_t, int8x8_t)'
Available overloads:
- float64x2_t vabdq_v(float64x2_t, float64x2_t)
- float64x1_t vabd_v(float64x1_t, float64x1_t)
- float64_t vabdd_f64(float64_t, float64_t)
- float32_t vabds_f32(float32_t, float32_t)
... snip ...
This makes it seriously easy to work out what you've done wrong in fairly nasty
intrinsics.
As part of this I've massively beefed up the documentation in arm_neon.td too.
Things still to do / on the radar:
- Testcase generation. This was implemented in the previous version and not in
the new one, because
- Autogenerated tests are not being run. The testcase in test/ differs from
the autogenerated version.
- There were a whole slew of special cases in the testcase generation that just
felt (and looked) like hacks.
If someone really feels strongly about this, I can try and reimplement it too.
- Big endian. That's coming soon and should be a very small diff on top of this one.
llvm-svn: 211101
This patch adds support for pointer types in global named registers variables.
It'll be lowered as a pair of read/write_register and inttoptr/ptrtoint calls.
Also adds some early checks on types on SemaDecl to avoid the assert.
Tests changed accordingly. (PR19837)
llvm-svn: 210274
to the normal non-placement ::operator new and ::operator delete, but allow
optimizations like new-expressions and delete-expressions do.
llvm-svn: 210137
This allows us to be more careful when dealing with enums whose fixed
underlying type requires special handling in a format string, like
NSInteger.
A refinement of r163266 from a year and a half ago, which added the
special handling for NSInteger and friends in the first place.
<rdar://problem/16616623>
llvm-svn: 209966
A few (mostly CodeGen) parts of Clang were tightly coupled to the
AArch64 backend. Now that it's gone, they will not even compile.
I've also deduplicated RUN lines in many of the AArch64 tests. This
might improve "make check-all" time noticably: some of those NEON
tests were monsters.
llvm-svn: 209578
This patch implements global named registers in Clang, lowering to the just
created intrinsics in LLVM (@llvm.read/write_register). A new type of LValue
had to be created (Register), which just adds support to carry the metadata
node containing the name of the register. Two new methods to emit loads and
stores interoperate with another to emit the named metadata node.
No guarantees are being made and only non-allocatable global variable named
registers are being supported. Local named register support is unchanged.
llvm-svn: 209149
Now that llvm cannot represent alias cycles, we have to diagnose erros just
before trying to close the cycle. This degrades the errors a bit. The real
solution is what it was before: if we want to provide good errors for these
cases, we have to be able to find a clang level decl given a mangled name
and produce the error from Sema.
llvm-svn: 209008
We were emitting dynamic initializers for __thread variables if there
was no explicit initializer, as in this test case:
struct S { S(); };
__thread S s;
llvm-svn: 207675
glibc expects that stddef.h only defines a single thing if either of these
defines is set. For example, before this change, a C file containing
#include <stdlib.h>
int ptrdiff_t = 0;
would compile with gcc but not with clang. Now it compiles with clang too.
This also fixes PR12997, where older versions of the Linux headers would define
NULL incorrectly, and glibc would define __need_NULL and expect stddef.h to
redefine NULL with the correct definition.
llvm-svn: 207606
We never aka vector types because our attributed syntax for it is less
comprehensible than the typedefs. This leaves the user in the dark when
the typedef isn't named that well.
Example:
v2s v; v4f w;
w = v;
The naming in this cases isn't even that bad, but the error we give is
useless without looking up the actual typedefs.
t.c:6:5: error: assigning to 'v4f' from incompatible type 'v2s'
Now:
t.c:6:5: error: assigning to 'v4f' (vector of 4 'float' values) from
incompatible type 'v2s' (vector of 2 'int' values)
We do this for all diagnostics that print a vector type.
llvm-svn: 207267
through to the output even if the input comment comes from an untrusted source
Attribute filtering is currently based on a blacklist, which right now includes
all event handler attributes (they contain JavaScipt code). It should be
switched to a whitelist, but going over all of the HTML5 spec requires a
significant amount of time.
llvm-svn: 206882
This is a partial revert of 183015.
By not recognizing things like _setjmp we lose (returns_twice) attribute on
them, which leads to incorrect code generation.
Fixes PR16138.
llvm-svn: 206362
When instantiating an array that has an alignment attribute on it, we
were looking through the array type and only considering the element
type for the resulting alignment. We need to make sure we take the
array's requirements into account too.
llvm-svn: 206317
This patch adds support for the msvc pragmas section, bss_seg, code_seg,
const_seg and data_seg as well as support for __declspec(allocate()).
Additionally it corrects semantics and adds diagnostics for
__attribute__((section())) and the interaction between the attribute
and the msvc pragmas and declspec. In general conflicts should now be
well diganosed within and among these features.
In supporting the pragmas new machinery for uniform lexing for
msvc pragmas was introduced. The new machinery always lexes the
entire pragma and stores it on an annotation token. The parser
is responsible for parsing the pragma when the handling the
annotation token.
There is a known outstanding bug in this implementation in C mode.
Because these attributes and pragmas apply _only_ to definitions, we
process them at the time we detect a definition. Due to tentative
definitions in C, we end up processing the definition late. This means
that in C mode, everything that ends up in a BSS section will end up in
the _last_ BSS section rather than the one that was live at the time of
tentative definition, even if that turns out to be the point of actual
definition. This issue is not known to impact anything as of yet
because we are not aware of a clear use or use case for #pragma bss_seg
but should be fixed at some point.
Differential Revision=http://reviews.llvm.org/D3065#inline-16241
llvm-svn: 205810
which warns on compound conditionals that always evaluate to the same value.
For instance, (x > 5 && x < 3) will always be false since no value for x can
satisfy both conditions.
This patch also changes the CFG to use these tautological values for better
branch analysis. The test for -Wunreachable-code shows how this change catches
additional dead code.
Patch by Anders Rönnholm.
llvm-svn: 205665
better. This warning will now trigger on the following conditionals:
bool b;
int i;
if (b > 1) {} // always false
if (0 <= (i > 5)) {} // always true
if (-1 > b) {} // always false
Patch by Per Viberg.
llvm-svn: 205608
For namespaces, this is consistent with mangling and GCC's debug info
behavior. For structs, GCC uses <anonymous struct> but we prefer
consistency between all anonymous entities but don't want to confuse
them with template arguments, etc, so we'll just go with parens in all
cases.
llvm-svn: 205398
While investigating some debug info issues, Eric and I came across a
particular template case where the location of a decl was quite
different from the range of the same decl. It might've been rather
helpful if the dumper had actually showed us this.
llvm-svn: 205396
This removes a diagnostic that is no longer required (the semantic engine now properly handles attribute syntax so __declspec and __attribute__ spellings no longer get mismatched). This caused several testcases to need updating for a slightly different wording.
llvm-svn: 205234
A redeclaration may not add dllimport or dllexport attributes. dllexport is
sticky and can be omitted on redeclarations while dllimport cannot.
llvm-svn: 205197
This adds Clang support for the ARM64 backend. There are definitely
still some rough edges, so please bring up any issues you see with
this patch.
As with the LLVM commit though, we think it'll be more useful for
merging with AArch64 from within the tree.
llvm-svn: 205100
Taking a hint from -Wparentheses, use an extra '()' as a sigil that
a dead condition is intentionally dead. For example:
if ((0)) { dead }
When this sigil is found, do not emit a dead code warning. When the
analysis sees:
if (0)
it suggests inserting '()' as a Fix-It.
llvm-svn: 205069
This produces valid IR now that llvm rejects aliases to weak aliases and warns
the user that the resolution is not changed if the weak alias is overridden.
llvm-svn: 204935
The main difference between __va_start and __builtin_va_start is that
the address of the va_list has already been taken, and the va_list is
always a char*.
__va_end and __va_arg are not needed.
llvm-svn: 204821
Amends r204300 to not try to test fixing a wchar_t* to "%ls", which we don't
do correctly anyway. In C mode, wchar_t is just a typedef for a normal
primitive integer type, not a distinct type like it is in C++. To make this
work correctly, we'll need to look for the wchar_t typedef, not just the
builtin type.
Should fix the buildbots.
llvm-svn: 204349
Since "half" is an OpenCL keyword and clang accepts __fp16 as an extension for
other languages, error messages and metadata (and hence debug info) should refer
to the half-precision floating point as "__fp16" instead of "half" when
compiling for non-OpenCL languages. This patch creates a new printing policy for
half in a similar manner to what is done for bool and wchar_t.
Differential Revision: http://llvm-reviews.chandlerc.com/D2952
llvm-svn: 204164
Also relax unreachable 'break' and 'return' to not check for being
preceded by a call to 'noreturn'. That turns out to not be so
interesting in practice.
llvm-svn: 204000
Recent work on -Wunreachable-code has focused on suppressing uninteresting
unreachable code that center around "configuration values", but
there are still some set of cases that are sometimes interesting
or uninteresting depending on the codebase. For example, a dead
"break" statement may not be interesting for a particular codebase,
potentially because it is auto-generated or simply because code
is written defensively.
To address these workflow differences, -Wunreachable-code is now
broken into several diagnostic groups:
-Wunreachable-code: intended to be a reasonable "default" for
most users.
and then other groups that turn on more aggressive checking:
-Wunreachable-code-break: warn about dead break statements
-Wunreachable-code-trivial-return: warn about dead return statements
that return "trivial" values (e.g., return 0). Other return
statements that return non-trivial values are still reported
under -Wunreachable-code (this is an area subject to more refinement).
-Wunreachable-code-aggressive: supergroup that enables all these
groups.
The goal is to eventually make -Wunreachable-code good enough to
either be in -Wall or on-by-default, thus finessing these warnings
into different groups helps achieve maximum signal for more users.
TODO: the tests need to be updated to reflect this extra control
via diagnostic flags.
llvm-svn: 203994
Someone could write:
if (0) {
__c11_atomic_load(ptr, memory_order_release);
}
or the equivalent, which is perfectly valid, so we shouldn't outright reject
invalid orderings on purely static grounds.
rdar://problem/16242991
llvm-svn: 203564
This is a conservative check, because it's valid for the expression to be
non-constant, and in cases like that we just don't know whether it's valid.
rdar://problem/16242991
llvm-svn: 203561
I had forgotten that the same reachability code is used by both -Wreturn-type
and -Wunreachable-code, so the heuristics applied to the latter were indirectly
impacting the former.
To address this, the reachability code is more refactored so that whiled
the logic at its core is shared, the intention of the clients are better
captured and segregated in helper APIs.
Fixes PR19074, and also some false positives reported offline to me
by Nick Lewycky.
llvm-svn: 203209
getTypeSize (which rounds up sizes) in order to issue diagnostics
when casting to mismatched vector sizes; instead of crashing in IRGen.
// rdar:// 16196902. Reviewed offline by John McCall.
llvm-svn: 203175
Sometimes do..while() is used to create a scope that can be left early.
In such cases, the unreachable 'while()' test is not usually interesting
unless it actually does something that is observable.
llvm-svn: 203051
Sometimes do..while() is used to create a scope that can be left early.
In such cases, the unreachable 'while()' test is not usually interesting
unless it actually does something that is observable.
llvm-svn: 203036
Previously we only pruned dead returns preceded by a call to a
'noreturn' function. After looking at the results of the LLVM codebase,
there are many others that should be pruned as well.
llvm-svn: 203029
Note that for backwards compatibility, an unnamed capability will default to being a "mutex." This allows the deprecated lockable attribute to continue to function.
llvm-svn: 203012
Some unreachable code is only "sometimes unreachable" because it
is guarded by a configuration value that is determined at compile
time and is always constant. Sometimes those represent real bugs,
but often they do not. This patch causes the reachability analysis
to cover such branches even if they are technically unreachable
in the CFG itself. There are some conservative heuristics at
play here to determine a "configuration value"; these are intended
to be refined over time.
llvm-svn: 202912
This is a heuristic. Many switch statements, although they look covered
over an enum, may actually handle at runtime more values than in the enum.
This is overly conservative, as there are some cases that clearly
can be ruled as being clearly unreachable, e.g. 'switch (42) { case 1: ... }'.
We can refine this later.
llvm-svn: 202436
For example:
unreachable();
break;
This code is idiomatic and defensive. The fact that 'break' is
unreachable here is not interesting. This occurs frequently
in LLVM/Clang itself.
llvm-svn: 202328
null comparison when the pointer is known to be non-null.
This catches the array to pointer decay, function to pointer decay and
address of variables. This does not catch address of function since this
has been previously used to silence a warning.
Pointer to bool conversion is under -Wbool-conversion.
Pointer to null comparison is under -Wtautological-pointer-compare, a sub-group
of -Wtautological-compare.
void foo() {
int arr[5];
int x;
// warn on these conditionals
if (foo);
if (arr);
if (&x);
if (foo == null);
if (arr == null);
if (&x == null);
if (&foo); // no warning
}
llvm-svn: 202216
The warnings fall into three groups.
1) Using an absolute value function of the wrong type, for instance, using the
int absolute value function when the argument is a floating point type.
2) Using the improper sized absolute value function, for instance, using abs
when the argument is a long long. llabs should be used instead.
From these two cases, an implicit conversion will occur which may cause
unexpected behavior. Where possible, suggest the proper absolute value
function to use, and which header to include if the function is not available.
3) Taking the absolute value of an unsigned value. In addition to this warning,
suggest to remove the function call. This usually indicates a logic error
since the programmer assumed negative values would have been possible.
llvm-svn: 202211
When calculating the preferred alignment of a type, consider if a alignment
attribute came from a typedef declaration. If one did, do not naturally align
the type.
Patch by Stephan Tolksdorf, with a little tweaking and an additional testcase by me.
llvm-svn: 202088
The language forbids defining enums in prototypes, so this check is normally
redundant, but if an enum is defined during template instantiation it should
not be added to the prototype scope.
While at it, clean up the code that deals with tag definitions in prototype
scope and expand the visibility warning to cover the case where an anonymous
enum is defined.
Differential Revision: http://llvm-reviews.chandlerc.com/D2742
llvm-svn: 201927
CGRecordLayoutBuilder was aging, complex, multi-pass, and shows signs of
existing before ASTRecordLayoutBuilder. It redundantly performed many
layout operations that are now performed by ASTRecordLayoutBuilder and
asserted that the results were the same. With the addition of support
for the MS-ABI, such as placement of vbptrs, vtordisps, different
bitfield layout and a variety of other features, CGRecordLayoutBuilder
was growing unwieldy in its redundancy.
This patch re-architects CGRecordLayoutBuilder to not perform any
redundant layout but rather, as directly as possible, lower an
ASTRecordLayout to an llvm::type. The new architecture is significantly
smaller and simpler than the CGRecordLayoutBuilder and contains fewer
ABI-specific code paths. It's also one pass.
The architecture of the new system is described in the comments. For the
most part, the new system simply takes all of the fields and bases from
an ASTRecordLayout, sorts them, inserts padding and dumps a record.
Bitfields, unions and primary virtual bases make this process a bit more
complicated. See the inline comments.
In addition, this patch updates a few lit tests due to the fact that the
new system computes more accurate llvm types than CGRecordLayoutBuilder.
Each change is commented individually in the review.
Differential Revision: http://llvm-reviews.chandlerc.com/D2795
llvm-svn: 201907
The following attributes have been (silently) deprecated, with their replacements listed:
lockable => capability
exclusive_locks_required => requires_capability
shared_locks_required => requires_shared_capability
locks_excluded => requires_capability
There are no functional changes intended.
llvm-svn: 201585
According to the GNU docs, zero-sized bitfields should not be affected by the
packed attribute.
Differential Revision: http://llvm-reviews.chandlerc.com/D2693
llvm-svn: 201288
It is actually useful to warn in such cases, thanks to Dmitri for pushing on this and making us see the light!
Related to rdar://15925483 and rdar://15922612. The latter radar is where the usefulness of the warning is most clear.
llvm-svn: 201165
Thanks to r199467, __attribute__((nonnull)) (without arguments) can apply
directly to parameters, instead of being applied to the whole function.
However, the old form of nonnull (with an argument index) could also apply
to the arguments of function and block pointers, and both of these can be
passed as parameters.
Now, if 'nonnull' with an argument is found on a parameter, /and/ the
parameter is a function or block pointer, it is handled the old way.
PR18795
llvm-svn: 201162
When a lax conversion featured a vector and a non-vector, we were
only requiring the non-vector to be a scalar type, but really it
needs to be a real type (i.e. integral or real floating); it is
not reasonable to allow a pointer, member pointer, or complex
type here.
r198474 required lax conversions to match in "data size", i.e.
element size * element count, forbidding matches that happen
only because a vector is rounded up to the nearest power of two
in size. Unfortunately, the erroneous logic was repeated in
several different places; unify them to use the new condition,
so that it triggers for arbitrary conversions and not just
those performed as part of binary operator checking.
rdar://15931426
llvm-svn: 200810
Due to statement expressions supported as GCC extension, it is possible
to put 'break' or 'continue' into a loop/switch statement but outside
its body, for example:
for ( ; ({ if (first) { first = 0; continue; } 0; }); )
This code is rejected by GCC if compiled in C mode but is accepted in C++
code. GCC bug 44715 tracks this discrepancy. Clang used code generation
that differs from GCC in both modes: only statement of the third
expression of 'for' behaves as if it was inside loop body.
This change makes code generation more close to GCC, considering 'break'
or 'continue' statement in condition and increment expressions of a
loop as it was inside the loop body. It also adds error for the cases
when 'break'/'continue' appear outside loop due to this syntax. If
code generation differ from GCC, warning is issued.
Differential Revision: http://llvm-reviews.chandlerc.com/D2518
llvm-svn: 199897
create an implicit declaration of it (because some type it depends on is
unavailable). This had the effect of causing us to not implicitly give it the
right attributes. It turns out that glibc's __sigsetjmp is declared before
sigjmp_buf is declared, and this resulted in us not implicitly giving it
__attribute__((returns_twice)), which in turn resulted in miscompiles in any C
code calling glibc's sigsetjmp.
(See also the vaguely-related sourceware.org/PR4662.)
llvm-svn: 199850
This involved making CheckReturnStackAddr into a static function, which
is now called by a top-level return value checking routine called
CheckReturnValExpr.
llvm-svn: 199790
This attribute is supported by GCC. More generally it should
probably be a type attribute, but this behavior matches 'nonnull'.
This patch does not include warning logic for checking if a null
value is returned from a function annotated with this attribute.
That will come in subsequent patches.
llvm-svn: 199626
This allows the following syntax:
void baz(__attribute__((nonnull)) const char *str);
instead of:
void baz(const char *str) __attribute__((nonnull(1)));
This also extends to Objective-C methods.
The checking logic in Sema is not as clean as I would like. Effectively
now we need to check both the FunctionDecl/ObjCMethodDecl and the parameters,
so the point of truth is spread in two places, but the logic isn't that
cumbersome.
Implements <rdar://problem/14691443>.
llvm-svn: 199467
This ports the last Sema tests over to use the frontend directly, and adds a
local lit substitution to disable inappropriate %clang usage under this
directory.
llvm-svn: 199348
-verify was simply ignored by the driver.
This commit fixes the RUN line and XFAILs the test, unblocking changes to ban
use of the driver in Sema tests and avoid problems like this.
llvm-svn: 199347
a subprocess invocation which is pretty significant on Windows. It also
likely saves a bunch of thrashing the host machine needlessly. Finally
it makes the tests much more predictable and less dependent on the host.
For example 'header_lookup1.c' was passing '-fno-ms-extensions' just to
thwart the host detection adding it into the compilation. By runnig CC1
directly we don't have to deal with such oddities.
llvm-svn: 199308
This makes the C++ ABI depend entirely on the target: MS ABI for -win32 triples,
Itanium otherwise. It's no longer possible to do weird combinations.
To be able to run a test with a specific ABI without constraining it to a
specific triple, new substitutions are added to lit: %itanium_abi_triple and
%ms_abi_triple can be used to get the current target triple adjusted to the
desired ABI. For example, if the test suite is running with the i686-pc-win32
target, %itanium_abi_triple will expand to i686-pc-mingw32.
Differential Revision: http://llvm-reviews.chandlerc.com/D2545
llvm-svn: 199250
In preparation for making the Win32 triple imply MS ABI mode,
make all tests pass in this mode, or make them use the Itanium
mode explicitly.
Differential Revision: http://llvm-reviews.chandlerc.com/D2401
llvm-svn: 199130
consideration the num-of-elements*width-of-element width.
Disallow casts when such width is not equal between the vector types otherwise
we may end up with an invalid LLVM bitcast.
rdar://15722308.
llvm-svn: 198474
Since this warning was generalized, it was also given a sensible warning group flag and the corresponding test was updated to reflect this.
llvm-svn: 198053
Fixes <rdar://problem/15584219> and <rdar://problem/12241361>.
This change looks large, but all it does is reuse and consolidate
the delayed diagnostic logic for deprecation warnings with unavailability
warnings. By doing so, it showed various inconsistencies between the
diagnostics, which were close, but not consistent. It also revealed
some missing "note:"'s in the deprecated diagnostics that were showing
up in the unavailable diagnostics, etc.
This change also changes the wording of the core deprecation diagnostics.
Instead of saying "function has been explicitly marked deprecated"
we now saw "'X' has been been explicitly marked deprecated". It
turns out providing a bit more context is useful, and often we
got the actual term wrong or it was not very precise
(e.g., "function" instead of "destructor"). By just saying the name
of the thing that is deprecated/deleted/unavailable we define
this issue away. This diagnostic can likely be further wordsmithed
to be shorter.
llvm-svn: 197627
An empty string for an ASM input constraint is invalid, and will crash
during clang CodeGen. Change TargetInfo::validateInputConstraint to
reject an empty string.
<rdar://problem/15552191>
llvm-svn: 197362
This patch was submitted to the list for review and didn't receive a LGTM.
(In fact one explicit objection and one query were raised.)
This reverts commit r197295.
llvm-svn: 197299
Previously, a line like
// expected-error-re {{foo}}
treats the entirety of foo as a regex. This is inconvenient when matching type
names containing regex characters. For example, to match
"void *(class test8::A::*)(void)" inside such a regex, one would have to type
"void \*\(class test8::A::\*\)\(void\)".
This patch changes the semantics of expected-error-re to only treat the parts
of the directive wrapped in double curly braces as regexes. This avoids the
escaping problem and leads to nicer patterns for those cases; see e.g. the
change to test/Sema/format-strings-scanf.c.
(The balanced search for closing }} of a directive also makes us handle the
full directive in test\SemaCXX\constexpr-printing.cpp:41 and :53.)
Differential Revision: http://llvm-reviews.chandlerc.com/D2388
llvm-svn: 197092
Changed from:
keyword '__is_empty' will be treated as an identifier for the remainder of the translation unit
To:
keyword '__is_empty' will be made available as an identifier for the remainder of the translation unit
This is a more accurate description of clang's keyword compatibility feature,
given that some of the keywords are turned into context-sensitive keywords
(e.g. REVERTIBLE_TYPE_TRAIT) rather than being fully disabled.
llvm-svn: 196776
Add -verify and update the test directives to match current expectations.
Also add a FIXME to an ObjC test that has expected-* directives but no -verify.
llvm-svn: 196737
Going by PR6913 it looks like this one can no longer reach CodeGen so remove
the redundant -emit-llvm case and treat it as an ordinary Sema test.
llvm-svn: 196736
Due to a missing -verify, 2007-10-01-BuildArrayRef.c was a no-op.
The message was changed 5 years ago so also update the test to reflect the new wording.
llvm-svn: 196729
We already support using "r" on 64-bit values (a GPRPair is
allocated), but Sema doesn't know this yet so issues a warning. This
should fix it.
llvm-svn: 196724
This commit changes -Wassign-enum to compare unqualified types. One could
think that this does not matter much, because who wants a value of enum type
that is const-qualified? But this breaks the intended pattern to silence this
warning with an explicit cast:
static const enum Foo z = (enum Foo) 42;
In this case, source type is 'enum Foo', and destination type is 'const enum
Foo', and if we compare qualified types, they don't match, so we used warn.
llvm-svn: 196548
the following pattern.
If 'case' expression refers to a static const variable of the correct enum
type, then we count this as a sufficient declaration of intent by the user,
so we silence the warning.
llvm-svn: 196546
clang converts keywords to identifiers for compatibility with various system
headers such as GNU libc.
Implement a -Wkeyword-compat extension warning to diagnose those cases. The
warning is on by default but will generally be ignored in system headers. It
can however be enabled globally to aid standards conformance testing.
This also changes the __uptr keyword avoidance from r195710 to no longer
special-case system headers, bringing it in line with other similar workarounds
in clang.
Implementation returns bool for symmetry with token annotation functions.
Some examples:
warning: keyword '__is_pod' will be treated as an identifier for the remainder of the translation unit [-Wkeyword-compat]
struct __is_pod
warning: keyword '__uptr' will be treated as an identifier here [-Wkeyword-compat]
union w *__uptr;
llvm-svn: 196212
lookup, if parsing failed, we did not restore the lexer state properly, and
eventually crashed. This change ensures that we always consume all the tokens
from the new token stream we started to parse the name from inline asm.
llvm-svn: 196182
GNU libc uses '__uptr' as a member name in C mode, conflicting with the
eponymous MSVC pointer modifier keyword.
Detect and mark the token as an identifier when these specific conditions are
met. __uptr will continue to work as a keyword for the remainder of the
translation unit.
Fixes PR17824.
llvm-svn: 195710
the GNU documentation: the attribute only appertains to the label if it is
followed by a semicolon. Based on a patch by Aaron Ballman!
llvm-svn: 194869
We already have builtins that are only available in GNU mode, so this
mirrors that.
Reviewers: rsmith
Differential Revision: http://llvm-reviews.chandlerc.com/D2128
llvm-svn: 194615
This patch fixes PR8264. Duplicate qualifiers already are diagnozed,
now the same diagnostics is issued for duplicate function specifiers.
Differential Revision: http://llvm-reviews.chandlerc.com/D2025
llvm-svn: 194559
The determination of which diagnostics would be issued for certain
anonymous unions started to get a little ridiculous. Clean this up by
inverting the condition-tree's logic from dialect -> issue to
issue -> diagnostic.
As part of this cleanup, move ext_c99_flexible_array_member from
DiagnosticParseKinds.td to DiagnosticSemaKinds.td because it's driven by
Sema, not Parse.
Also, the liberty was taken to edit ext_c99_flexible_array_member to
match other, similar, diagnostics.
llvm-svn: 193919
The AST was constructed so that this builtin returned the default BoolTy and
since I'd opted for custom SemaChecking, I should have set it properly at that
point.
This caused an assertion failure when the types didn't match up with what we
generated. This makes it return an IntTy, which is as good as anything.
llvm-svn: 193606
Specifically, this warns when a character literal is added (using '+') to a
variable with type 'char *' (or any other pointer to character type). Like
-Wstring-plus-int, there is a fix-it to change "foo + 'a'" to "&foo['a']"
iff the character literal is on the right side of the string.
Patch by Anders Rönnholm!
llvm-svn: 193418
which we don't think can't have one, only allow it in the tiny number of
attributes which opts into this weird parse rule.
I've manually checked that the handlers for all these attributes can in fact
cope with an identifier as the argument. This is still somewhat terrible; we
should move more fully towards picking the parsing rules based on the
attribute, and make the Parse -> Sema interface more type-safe.
llvm-svn: 193295
it. Also removes all of the microsoft C++ ABI related code from the
itanium layout builder.
Differential Revision: http://llvm-reviews.chandlerc.com/D2003
llvm-svn: 193290
This fixes pr17624.
A FIXME from Richard Smith:
It seems to me that the root cause is that a per-Decl 'used' flag doesn't
really make much sense in the way we use it now. I think we should either track
whether that particular declaration is used (with isUsed scanning the entire
redecl chain), or we should only have one flag for the entire redeclaration
chain (perhaps by always looking at the flag on either the most recent decl or
the canonical decl). Modeling it as "is this declaration or any previous
declaration used" is weird, and requires contortions like the loop at the end
of Sema::MarkFunctionReferenced.
llvm-svn: 193202
This fixes pr17639.
Before this patch clang would consider
void foo(void) __attribute((alias("__foo")));
a declaration. It now correctly handles it as a definition.
Initial patch by Alp Toker. I added support for variables.
llvm-svn: 193200
This is a fixed version of r193161. In order to handle
void foo() __attribute__((alias("bar")));
void bar() {}
void zed() __attribute__((alias("foo")));
it is not enough to delay aliases to the end of the TU, we have to do two
passes over them to find if they are defined or not.
This can be implemented by producing alias as we go and just doing the second
pass at the end. This has the advantage that other parts of clang that were
expecting alias to be processed in order don't have to be changed.
This patch also handles cyclic aliases.
llvm-svn: 193188
This patch wasn't reviewed, and isn't correctly preserving the behaviors
relied upon by QT. I don't have a direct example of fallout, but it
should go through the standard code review process. For example, it
should never have removed the QT test case that was added when fixing
those users.
llvm-svn: 193174
This reverts commit r193161.
It broke
void foo() __attribute__((alias("bar")));
void bar() {}
void zed() __attribute__((alias("foo")));
Looks like we have to fix pr17639 first :-(
llvm-svn: 193162
names. For example, with this patch we now reject
void f1(void) __attribute__((alias("g1")));
This patch is implemented in CodeGen. It is quiet a bit simpler and more
compatible with gcc than implementing it in Sema. The downside is that the
errors only fire during -emit-llvm.
llvm-svn: 193161
Due to statement expressions supported as GCC extension, it is possible
to put 'break' or 'continue' into a loop/switch statement but outside its
body, for example:
for ( ; ({ if (first) { first = 0; continue; } 0; }); )
Such usage must be diagnosed as an error, GCC rejects it. To recognize
this and similar patterns the flags BreakScope and ContinueScope are
temporarily turned off while parsing condition expression.
Differential Revision: http://llvm-reviews.chandlerc.com/D1762
llvm-svn: 193073
Follow-up from r192240.
This makes it an error to use callee-cleanup conventions on variadic
functions, except for __fastcall and __stdcall, which we ignore with
a warning for GCC and MSVC compatibility.
Differential Revision: http://llvm-reviews.chandlerc.com/D1870
llvm-svn: 192308
MSVC allows this and silently falls back to __cdecl for variadic functions.
This patch turns Clang's error into a warning in MS mode and adds a test
to make sure we generate correct code.
Differential Revision: http://llvm-reviews.chandlerc.com/D1861
llvm-svn: 192240
Fixes <rdar://problem/10679282>.
I'm not completely satisfied with this patch. Sprinkling "diagnostic ignored"
_Pragmas throughout this file is gross, but I couldn't suppress
it for the entire file.
llvm-svn: 192143
This attribute allows users to use a modified C or C++ function as an ARM
exception-handling function and, with care, to successfully return control to
user-space after the issue has been dealt with.
rdar://problem/14207019
llvm-svn: 191769
variable from being the function to being the enclosing namespace scope (in
C++) or the TU (in C). This allows us to fix a selection of related issues
where we would build incorrect redeclaration chains for such declarations, and
fail to notice type mismatches.
Such declarations are put into a new IdentifierNamespace, IDNS_LocalExtern,
which is only found when searching scopes, and not found when searching
DeclContexts. Such a declaration is only made visible in its DeclContext if
there are no non-LocalExtern declarations.
llvm-svn: 191064
This patch adds the following, more specific warning flags:
gnu-anonymous-struct
gnu-compound-literal-initializer
gnu-empty-struct
gnu-flexible-array-initializer
gnu-flexible-array-union-member
gnu-folding-constant
redeclared-class-member
gnu-redeclared-enum
gnu-union-cast
gnu-variable-sized-type-not-at-end
Patch by Peter Lewis.
llvm-svn: 190972
Fix for PR16752. Second commit.
PR16752: 'mode' attribute for unusual targets doesn't work properly
Description:
Troubles could be happened due to some assumptions in handleModeAttr function (see SemaDeclAttr.cpp).
For example, it assumes that 32 bit integer is 'int', while it could be 16 bit only.
Instead of asking target: 'which type do you want to use for int32_t ?' it just hardcodes general opinion. That doesn't looks pretty correct.
Please consider the next solution:
1. In Basic/TargetInfo add getIntTypeByWidth and getRealTypeByWidth methods. Methods asks target for proper type for given bit width.
2. Fix handleModeAttr according to new methods in TargetInfo.
Fixes:
1st Commit (Done): Add new methods for TargetInfo:
getRealTypeByWidth and getIntTypeByWidth
for ASTContext names are almost same(invokes new methods from TargetInfo):
getIntTypeForBitwidth and getRealTypeForBitwidth
2nd Commit (Current): Fix SemaDeclAttr, handleModeAttr function.
Also test/Sema/attr-mode.c was fixed. 'XC' mode test was disabled for PPC64 machines.
llvm-svn: 190926
LLVM supports applying conversion instructions to vectors of the same number of
elements (fptrunc, fptosi, etc.) but there had been no way for a Clang user to
cause such instructions to be generated when using builtin vector types.
C-style casting on vectors is already defined in terms of bitcasts, and so
cannot be used for these conversions as well (without leading to a very
confusing set of semantics). As a result, this adds a __builtin_convertvector
intrinsic (patterned after the OpenCL __builtin_astype intrinsic). This is
intended to aid the creation of vector intrinsic headers that create generic IR
instead of target-dependent intrinsics (in other words, this is a generic
_mm_cvtepi32_ps). As noted in the documentation, the action of
__builtin_convertvector is defined in terms of the action of a C-style cast on
each vector element.
llvm-svn: 190915
Summary:
Makes functions with implicit calling convention compatible with
function types with a matching explicit calling convention. This fixes
things like calls to qsort(), which has an explicit __cdecl attribute on
the comparator in Windows headers.
Clang will now infer the calling convention from the declarator. There
are two cases when the CC must be adjusted during redeclaration:
1. When defining a non-inline static method.
2. When redeclaring a function with an implicit or mismatched
convention.
Fixes PR13457, and allows clang to compile CommandLine.cpp for the
Microsoft C++ ABI.
Excellent test cases provided by Alexander Zinenko!
Reviewers: rsmith
Differential Revision: http://llvm-reviews.chandlerc.com/D1231
llvm-svn: 189412
Basically, isInMainFile considers line markers, and isWrittenInMainFile
doesn't. Distinguishing between the two is useful when dealing with
files which are preprocessed files or rewritten with -frewrite-includes
(so we don't, for example, print useless warnings).
llvm-svn: 188968
We generally don't warn about extensions involving keywords reserved
for the implementation, so we shouldn't warn here either: the
standard doesn't require it, and it doesn't provide useful information
to the user.
llvm-svn: 188840
This adds the following as subgroups of -Wgnu: -Wgnu-alignof-expression,
-Wgnu-case-range, -Wgnu-complex-integer, -Wgnu-conditional-omitted-operand,
-Wgnu-empty-initializer, -Wgnu-label-as-value, -Wgnu-local-label,
and -Wgnu-statement-expression,
Patch by Peter Lewis.
llvm-svn: 188839
When a local extern declaration redeclares some other entity, the type of that
entity is merged with the prior type if the prior declaration is visible (in C)
or is declared in the same scope (in C++).
- Make LookupRedeclarationWithLinkage actually work in C++, use it in the right
set of cases, and make it track whether it found a shadowed declaration.
- Track whether we found a declaration in the same scope (for C++) including
across serialization and template instantiation.
llvm-svn: 188307
We were exposing the extra alignment given to large arrays. The new behavior
matches gcc, which is a good thing since this is a gcc extension.
Thanks to Joerg Sonnenberger for noticing it.
While at it, centralize the method description in the .h file.
llvm-svn: 187999
James Molloy