hwasan: Improve precision of checks using short granule tags.
A short granule is a granule of size between 1 and `TG-1` bytes. The size
of a short granule is stored at the location in shadow memory where the
granule's tag is normally stored, while the granule's actual tag is stored
in the last byte of the granule. This means that in order to verify that a
pointer tag matches a memory tag, HWASAN must check for two possibilities:
* the pointer tag is equal to the memory tag in shadow memory, or
* the shadow memory tag is actually a short granule size, the value being loaded
is in bounds of the granule and the pointer tag is equal to the last byte of
the granule.
Pointer tags between 1 to `TG-1` are possible and are as likely as any other
tag. This means that these tags in memory have two interpretations: the full
tag interpretation (where the pointer tag is between 1 and `TG-1` and the
last byte of the granule is ordinary data) and the short tag interpretation
(where the pointer tag is stored in the granule).
When HWASAN detects an error near a memory tag between 1 and `TG-1`, it
will show both the memory tag and the last byte of the granule. Currently,
it is up to the user to disambiguate the two possibilities.
Because this functionality obsoletes the right aligned heap feature of
the HWASAN memory allocator (and because we can no longer easily test
it), the feature is removed.
Also update the documentation to cover both short granule tags and
outlined checks.
Fangrui Song [Tue, 9 Jul 2019 19:36:22 +0000 (19:36 +0000)]
[OpenMP] Simplify getFloatTypeSemantics
When the float point representations are the same on the host and on the target device,
(`&Target->getLongDoubleFormat() == &AuxTarget->getLongDoubleFormat()`),
we can just use `AuxTarget->getLongDoubleFormat()`.
Ignore trailing NullStmts in StmtExprs for GCC compatibility.
Ignore trailing NullStmts in compound expressions when determining the result type and value. This is to match the GCC behavior which ignores semicolons at the end of compound expressions.
[OPENMP]Fix the float point semantics handling on the device.
The device should use the same float point representation as the host.
Previous patch fixed the handling of the sizes of the float point types,
but did not fixed the fp semantics. This patch makes target device to
use the host fp semantics. this is required for the correct data
transfer between host and device and correct codegen.
The current useFloat128ManglingForLongDouble() bisection is not suitable
when we support -mlong-double-128 in clang (D64277). Replace
useFloat128ManglingForLongDouble() with getLongDoubleMangling() and
getFloat128Mangling() to allow 3 mangling schemes.
I also deleted the `getTriple().isOSBinFormatELF()` check (the Darwin
support has gone: https://reviews.llvm.org/D50988).
For x86, change the mangled code of __float128 from `U10__float128` to `g`. `U10__float128` was wrongly copied from PowerPC.
The test will be added to `test/CodeGen/x86-long-double.cpp` in D64277.
Simon Pilgrim [Tue, 9 Jul 2019 11:26:35 +0000 (11:26 +0000)]
Revert rL365355 : [Syntax] Introduce syntax trees
Summary:
A tooling-focused alternative to the AST. This commit focuses on the
memory-management strategy and the structure of the AST.
More to follow later:
- Operations to mutate the syntax trees and corresponding textual
replacements.
- Mapping between clang AST nodes and syntax tree nodes.
- More node types corresponding to the language constructs.
Reviewers: sammccall
Reviewed By: sammccall
Subscribers: llvm-commits, mgorny, cfe-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D61637
........
Fixes buildbots which were crashing on SyntaxTests.exe
[ASTImporter] Added visibility context check for EnumDecl.
Summary:
ASTImporter makes now difference between enums with same name in different translation
units if these are not visible outside.
("Scoped enums" are not handled yet.)
Yonghong Song [Tue, 9 Jul 2019 04:21:50 +0000 (04:21 +0000)]
[BPF] Preserve debuginfo array/union/struct type/access index
For background of BPF CO-RE project, please refer to
http://vger.kernel.org/bpfconf2019.html
In summary, BPF CO-RE intends to compile bpf programs
adjustable on struct/union layout change so the same
program can run on multiple kernels with adjustment
before loading based on native kernel structures.
In order to do this, we need keep track of GEP(getelementptr)
instruction base and result debuginfo types, so we
can adjust on the host based on kernel BTF info.
Capturing such information as an IR optimization is hard
as various optimization may have tweaked GEP and also
union is replaced by structure it is impossible to track
fieldindex for union member accesses.
Three intrinsic functions, preserve_{array,union,struct}_access_index,
are introducted.
addr = preserve_array_access_index(base, index, dimension)
addr = preserve_union_access_index(base, di_index)
addr = preserve_struct_access_index(base, gep_index, di_index)
here,
base: the base pointer for the array/union/struct access.
index: the last access index for array, the same for IR/DebugInfo layout.
dimension: the array dimension.
gep_index: the access index based on IR layout.
di_index: the access index based on user/debuginfo types.
If using these intrinsics blindly, i.e., transforming all GEPs
to these intrinsics and later on reducing them to GEPs, we have
seen up to 7% more instructions generated. To avoid such an overhead,
a clang builtin is proposed:
base = __builtin_preserve_access_index(base)
such that user wraps to-be-relocated GEPs in this builtin
and preserve_*_access_index intrinsics only apply to
those GEPs. Such a buyin will prevent performance degradation
if people do not use CO-RE, even for programs which use
bpf_probe_read().
For example, for the following example,
$ cat test.c
struct sk_buff {
int i;
int b1:1;
int b2:2;
union {
struct {
int o1;
int o2;
} o;
struct {
char flags;
char dev_id;
} dev;
int netid;
} u[10];
};
static int (*bpf_probe_read)(void *dst, int size, const void *unsafe_ptr)
= (void *) 4;
Note that @llvm.preserve.{struct,union}.access.index calls have metadata llvm.preserve.access.index
attached to instructions to provide struct/union debuginfo type information.
For &ctx->u[5].dev.dev_id,
. The "%6 = ..." represents struct member "u" with index 2 for IR layout and index 3 for DI layout.
. The "%7 = ..." represents array subscript "5".
. The "%8 = ..." represents union member "dev" with index 1 for DI layout.
. The "%10 = ..." represents struct member "dev_id" with index 1 for both IR and DI layout.
Basically, traversing the use-def chain recursively for the 3rd argument of bpf_probe_read() and
examining all preserve_*_access_index calls, the debuginfo struct/union/array access index
can be achieved.
The intrinsics also contain enough information to regenerate codes for IR layout.
For array and structure intrinsics, the proper GEP can be constructed.
For union intrinsics, replacing all uses of "addr" with "base" should be enough.
Signed-off-by: Yonghong Song <yhs@fb.com>
Differential Revision: https://reviews.llvm.org/D61809
Yonghong Song [Tue, 9 Jul 2019 04:04:21 +0000 (04:04 +0000)]
[BPF] Preserve debuginfo array/union/struct type/access index
For background of BPF CO-RE project, please refer to
http://vger.kernel.org/bpfconf2019.html
In summary, BPF CO-RE intends to compile bpf programs
adjustable on struct/union layout change so the same
program can run on multiple kernels with adjustment
before loading based on native kernel structures.
In order to do this, we need keep track of GEP(getelementptr)
instruction base and result debuginfo types, so we
can adjust on the host based on kernel BTF info.
Capturing such information as an IR optimization is hard
as various optimization may have tweaked GEP and also
union is replaced by structure it is impossible to track
fieldindex for union member accesses.
Three intrinsic functions, preserve_{array,union,struct}_access_index,
are introducted.
addr = preserve_array_access_index(base, index, dimension)
addr = preserve_union_access_index(base, di_index)
addr = preserve_struct_access_index(base, gep_index, di_index)
here,
base: the base pointer for the array/union/struct access.
index: the last access index for array, the same for IR/DebugInfo layout.
dimension: the array dimension.
gep_index: the access index based on IR layout.
di_index: the access index based on user/debuginfo types.
If using these intrinsics blindly, i.e., transforming all GEPs
to these intrinsics and later on reducing them to GEPs, we have
seen up to 7% more instructions generated. To avoid such an overhead,
a clang builtin is proposed:
base = __builtin_preserve_access_index(base)
such that user wraps to-be-relocated GEPs in this builtin
and preserve_*_access_index intrinsics only apply to
those GEPs. Such a buyin will prevent performance degradation
if people do not use CO-RE, even for programs which use
bpf_probe_read().
For example, for the following example,
$ cat test.c
struct sk_buff {
int i;
int b1:1;
int b2:2;
union {
struct {
int o1;
int o2;
} o;
struct {
char flags;
char dev_id;
} dev;
int netid;
} u[10];
};
static int (*bpf_probe_read)(void *dst, int size, const void *unsafe_ptr)
= (void *) 4;
Note that @llvm.preserve.{struct,union}.access.index calls have metadata llvm.preserve.access.index
attached to instructions to provide struct/union debuginfo type information.
For &ctx->u[5].dev.dev_id,
. The "%6 = ..." represents struct member "u" with index 2 for IR layout and index 3 for DI layout.
. The "%7 = ..." represents array subscript "5".
. The "%8 = ..." represents union member "dev" with index 1 for DI layout.
. The "%10 = ..." represents struct member "dev_id" with index 1 for both IR and DI layout.
Basically, traversing the use-def chain recursively for the 3rd argument of bpf_probe_read() and
examining all preserve_*_access_index calls, the debuginfo struct/union/array access index
can be achieved.
The intrinsics also contain enough information to regenerate codes for IR layout.
For array and structure intrinsics, the proper GEP can be constructed.
For union intrinsics, replacing all uses of "addr" with "base" should be enough.
Signed-off-by: Yonghong Song <yhs@fb.com>
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@365435 91177308-0d34-0410-b5e6-96231b3b80d8
Let unaliased Args track which Alias they were created from, and use that in Arg::getAsString() for diagnostics
With this, `clang-cl /source-charset:utf-16 test.cc` now prints `invalid
value 'utf-16' in '/source-charset:utf-16'` instead of `invalid value
'utf-16' in '-finput-charset=utf-16'` before, and several other clang-cl
flags produce much less confusing output as well.
Fixes PR29106.
Since an arg and its alias can have different arg types (joined vs not)
and different values (because of AliasArgs<>), I chose to give the Alias
its own Arg object. For convenience, I just store the alias directly in
the unaliased arg – there aren't many arg objects at runtime, so that
seems ok.
Finally, I changed Arg::getAsString() to use the alias's representation
if it's present – that function was already documented as being the
suitable function for diagnostics, and most callers already used it for
diagnostics.
Implementation-wise, Arg::accept() previously used to parse things as
the unaliased option. The core of that switch is now extracted into a
new function acceptInternal() which parses as the _aliased_ option, and
the previously-intermingled unaliasing is now done as an explicit step
afterwards.
(This also changes one place in lld that didn't use getAsString() for
diagnostics, so that that one place now also prints the flag as the user
wrote it, not as it looks after it went through unaliasing.)
Fangrui Song [Tue, 9 Jul 2019 00:27:43 +0000 (00:27 +0000)]
[X86][PPC] Support -mlong-double-64
-mlong-double-64 is supported on some ports of gcc (i386, x86_64, and ppc{32,64}).
On many other targets, there will be an error:
error: unrecognized command line option '-mlong-double-64'
This patch makes the driver option -mlong-double-64 available for x86
and ppc. The CC1 option -mlong-double-64 is available on all targets for
users to test on unsupported targets.
LongDoubleSize is added as a VALUE_LANGOPT so that the option can be
shared with -mlong-double-128 when we support it in clang.
Also, make powerpc*-linux-musl default to use 64-bit long double. It is
currently the only supported ABI on musl and is also how people
configure powerpc*-linux-musl-gcc.
[analyzer] exploded-graph-rewriter: Implement a single-path mode.
Instead of rewriting the whole graph, rewrite the leftmost path in the
graph. Useful for trimmed graphs that are still too large to display due
to multiple equivalent reports mixed into them.
Erik Pilkington [Mon, 8 Jul 2019 23:42:52 +0000 (23:42 +0000)]
[ObjC] Add a -Wtautological-compare warning for BOOL
On macOS, BOOL is a typedef for signed char, but it should never hold a value
that isn't 1 or 0. Any code that expects a different value in their BOOL should
be fixed.
David Blaikie [Mon, 8 Jul 2019 23:24:41 +0000 (23:24 +0000)]
[cxx2a] P0624R2 fix: only lambdas with no lambda-capture are default-constructible and assignable.
This is a fix for rG864949 which only disabled default construction and
assignment for lambdas with capture-defaults, where the C++2a draft
disables them for lambdas with any lambda-capture at all.
[Frontend] Explicitly include Bitstream/BitCodes.h and BitstreamWriter.h
This fixes a modules issue:
error: declaration of 'bitc' must be imported from module
'Clang_Serialization.ASTBitCodes' before it is required
Stream.EmitRecord(llvm::bitc::BLOCKINFO_CODE_SETBID, Record);
[Sema] Resolve placeholder types before type deduction to silence
spurious `-Warc-repeated-use-of-weak` warnings
The spurious -Warc-repeated-use-of-weak warnings are issued when an
initializer expression uses a weak ObjC pointer.
My first attempt to silence the warnings (r350917) caused clang to
reject code that is legal in C++17. The patch is based on the feedback I
received from Richard when the patch was reverted.
Summary:
A tooling-focused alternative to the AST. This commit focuses on the
memory-management strategy and the structure of the AST.
More to follow later:
- Operations to mutate the syntax trees and corresponding textual
replacements.
- Mapping between clang AST nodes and syntax tree nodes.
- More node types corresponding to the language constructs.
[ASTImporter] Fix import of lambda in function param
Summary:
The current import implementation fails to import the definition of a
lambda class if the lambda class is defined in a function param.
E.g., the lambda class below will be imported without any methods:
```
template <typename F>
void f(F L = [](){}) {}
```
Endre Fulop [Mon, 8 Jul 2019 12:37:10 +0000 (12:37 +0000)]
[analyzer] Add analyzer option to limit the number of imported TUs
Summary:
During CTU analysis of complex projects, the loaded AST-contents of
imported TUs can grow bigger than available system memory. This option
introduces a threshold on the number of TUs to be imported for a single
TU in order to prevent such cases.
Summary:
Change the vuqadd scalar instrinsics to have the second argument as unsigned values, not signed,
accordingly to https://developer.arm.com/architectures/instruction-sets/simd-isas/neon/intrinsics
So now the compiler correctly warns that a undefined negative float conversion is being done.
Summary:
Change the vsqadd scalar instrinsics to have the second argument as signed values, not unsigned,
accordingly to https://developer.arm.com/architectures/instruction-sets/simd-isas/neon/intrinsics
The existing unsigned argument can cause faulty code as negative float to unsigned conversion is
undefined, which llvm/clang optimizes away.
Richard Smith [Sat, 6 Jul 2019 21:05:52 +0000 (21:05 +0000)]
Treat the range of representable values of floating-point types as [-inf, +inf] not as [-max, +max].
Summary:
Prior to r329065, we used [-max, max] as the range of representable
values because LLVM's `fptrunc` did not guarantee defined behavior when
truncating from a larger floating-point type to a smaller one. Now that
has been fixed, we can make clang follow normal IEEE 754 semantics in this
regard and take the larger range [-inf, +inf] as the range of representable
values.
In practice, this affects two parts of the frontend:
* the constant evaluator no longer treats floating-point evaluations
that result in +-inf as being undefined (because they no longer leave
the range of representable values of the type)
* UBSan no longer treats conversions to floating-point type that are
outside the [-max, +max] range as being undefined
In passing, also remove the float-divide-by-zero sanitizer from
-fsanitize=undefined, on the basis that while it's undefined per C++
rules (and we disallow it in constant expressions for that reason), it
is defined by Clang / LLVM / IEEE 754.
Joel E. Denny [Sat, 6 Jul 2019 02:55:06 +0000 (02:55 +0000)]
[Rewrite] Extend to further accept CharSourceRange
Some Rewrite functions are already overloaded to accept
CharSourceRange, and this extends others in the same manner. I'm
calling these in code that's not ready to upstream, but I figure they
might be useful to others in the meantime.
Bitstream reader: Fix undefined behavior seen after rL364464
Summary:
After rL364464 the following tests started to fail when
running the clang-doc tests with an ubsan instrumented
build of clang-doc:
Clang Tools :: clang-doc/single-file-public.cpp
Extra Tools Unit Tests :: clang-doc/./ClangDocTests/BitcodeTest.emitEnumInfoBitcode
Extra Tools Unit Tests :: clang-doc/./ClangDocTests/BitcodeTest.emitMethodInfoBitcode
Extra Tools Unit Tests :: clang-doc/./ClangDocTests/BitcodeTest.emitRecordInfoBitcode
Extra Tools Unit Tests :: clang-doc/./ClangDocTests/SerializeTest.emitInfoWithCommentBitcode
We need to check that the read value is in range for being
casted to the llvm::bitc::FixedAbbrevIDs enum, before the
cast in ClangDocBitcodeReader::skipUntilRecordOrBlock.
SerializedDiagnosticReader::skipUntilRecordOrBlock was updated
in the same way.
[analyzer] Track terminator conditions on which a tracked expression depends
This patch is a major part of my GSoC project, aimed to improve the bug
reports of the analyzer.
TL;DR: Help the analyzer understand that some conditions are important,
and should be explained better. If an CFGBlock is a control dependency
of a block where an expression value is tracked, explain the condition
expression better by tracking it.
if (A) // let's explain why we believe A to be true
10 / x; // division by zero
This is an experimental feature, and can be enabled by the
off-by-default analyzer configuration "track-conditions".
In detail:
This idea was inspired by the program slicing algorithm. Essentially,
two things are used to produce a program slice (a subset of the program
relevant to a (statement, variable) pair): data and control
dependencies. The bug path (the linear path in the ExplodedGraph that leads
from the beginning of the analysis to the error node) enables to
analyzer to argue about data dependencies with relative ease.
Control dependencies are a different slice of the cake entirely.
Just because we reached a branch during symbolic execution, it
doesn't mean that that particular branch has any effect on whether the
bug would've occured. This means that we can't simply rely on the bug
path to gather control dependencies.
In previous patches, LLVM's IDFCalculator, which works on a control flow
graph rather than the ExplodedGraph was generalized to solve this issue.
We use this information to heuristically guess that the value of a tracked
expression depends greatly on it's control dependencies, and start
tracking them as well.
After plenty of evaluations this was seen as great idea, but still
lacking refinements (we should have different descriptions about a
conditions value), hence it's off-by-default.
[analyzer][IDF] Add a control dependency calculator + a new debug checker
I intend to improve the analyzer's bug reports by tracking condition
expressions.
01 bool b = messyComputation();
02 int i = 0;
03 if (b) // control dependency of the bug site, let's explain why we assume val
04 // to be true
05 10 / i; // warn: division by zero
I'll detail this heuristic in the followup patch, strictly related to this one
however:
* Create the new ControlDependencyCalculator class that uses llvm::IDFCalculator
to (lazily) calculate control dependencies for Clang's CFG.
* A new debug checker debug.DumpControlDependencies is added for lit tests
* Add unittests
Make joined instances of JoinedOrSeparate flags point to the unaliased args, like all other arg types do
This fixes an 8-year-old regression. r105763 made it so that aliases
always refer to the unaliased option – but it missed the "joined" branch
of JoinedOrSeparate flags. (r162231 then made the Args classes
non-virtual, and r169344 moved them from clang to llvm.)
Back then, there was no JoinedOrSeparate flag that was an alias, so it
wasn't observable. Now /U in CLCompatOptions is a JoinedOrSeparate alias
in clang, and warn_slash_u_filename incorrectly used the aliased arg id
(using the unaliased one isn't really a regression since that warning
checks if the undefined macro contains slash or backslash and only then
emits the warning – and no valid use will pass "-Ufoo/bar" or similar).
Also, lld has many JoinedOrSeparate aliases, and due to this bug it had
to explicitly call `getUnaliasedOption()` in a bunch of places, even
though that shouldn't be necessary by design. After this fix in Option,
these calls really don't have an effect any more, so remove them.
No intended behavior change.
(I accidentally fixed this bug while working on PR29106 but then
wondered why the warn_slash_u_filename broke. When I figured it out, I
thought it would make sense to land this in a separate commit.)
Mikael Holmen [Fri, 5 Jul 2019 06:12:24 +0000 (06:12 +0000)]
Silence gcc warning "control reaches end of non-void function" [NFCI]
Without this fix gcc (7.4) complains with
/data/repo/master/clang/lib/CodeGen/CGObjCMac.cpp: In member function 'std::__cxx11::string {anonymous}::CGObjCCommonMac::GetSectionName(llvm::StringRef, llvm::StringRef)':
/data/repo/master/clang/lib/CodeGen/CGObjCMac.cpp:4944:1: error: control reaches end of non-void function [-Werror=return-type]
}
^
All values in the ObjectFormatType enum are currently handled in the switch
but gcc complains anyway.
[NFC] Make some ObjectFormatType switches covering
Summary:
This patch removes the `default` case from some switches on
`llvm::Triple::ObjectFormatType`, and cases for the missing enumerators
are then added.
For `UnknownObjectFormat`, the action (`llvm_unreachable`) for the
`default` case is kept.
For the other unhandled cases, `report_fatal_error` is used instead.
Fangrui Song [Thu, 4 Jul 2019 04:44:42 +0000 (04:44 +0000)]
[PowerPC] Support constraint code "ww"
Summary:
"ww" and "ws" are both constraint codes for VSX vector registers that
hold scalar double data. "ww" is preferred for float while "ws" is
preferred for double.
However, even though the path of execution in B2 only depends on C's value,
CFGBlock::getCondition() would return the entire condition (A && B && C). For
B3, it would return A && B. I changed this the actual condition.
[analyzer][Dominator] Add post dominators to CFG + a new debug checker
Transform clang::DominatorTree to be able to also calculate post dominators.
* Tidy up the documentation
* Make it clang::DominatorTree template class (similarly to how
llvm::DominatorTreeBase works), rename it to clang::CFGDominatorTreeImpl
* Clang's dominator tree is now called clang::CFGDomTree
* Clang's brand new post dominator tree is called clang::CFGPostDomTree
* Add a lot of asserts to the dump() function
* Create a new checker to test the functionality
In Clang's CFG, we use nullpointers to represent unreachable nodes, for
example, in the included testfile, block B0 is unreachable from block
B1, resulting in a nullpointer dereference somewhere in
llvm::DominatorTreeBase<clang::CFGBlock, false>::recalculate.
This patch fixes this issue by specializing
llvm::DomTreeBuilder::SemiNCAInfo::ChildrenGetter::Get for
clang::CFG to not contain nullpointer successors.
[clang-tidy] Fix the YAML created for checks like modernize-pass-by-value
Currently this check generates the replacement with the newline in the end.
The proper way to export it to YAML is to have two \n\n instead of one.
Without this fix clients should reinterpret the replacement as
"#include <utility> " instead of "#include <utility>\n"
[clang][HeaderSearch] Shorten paths for includes in mainfile's directory
Summary:
Currently HeaderSearch only looks at SearchDir's passed into it, but in
addition to those paths headers can be relative to including file's directory.
[analyzer] exploded-graph-rewriter: Collapse very long statement pretty-prints.
When printing various statements that include braces (compound
statements, lambda expressions, statement-expressions, etc.),
replace the code between braces with '...'.
Erik Pilkington [Tue, 2 Jul 2019 18:28:13 +0000 (18:28 +0000)]
[C++2a] Add __builtin_bit_cast, used to implement std::bit_cast
This commit adds a new builtin, __builtin_bit_cast(T, v), which performs a
bit_cast from a value v to a type T. This expression can be evaluated at
compile time under specific circumstances.
The compile time evaluation currently doesn't support bit-fields, but I'm
planning on fixing this in a follow up (some of the logic for figuring this out
is in CodeGen). I'm also planning follow-ups for supporting some more esoteric
types that the constexpr evaluator supports, as well as extending
__builtin_memcpy constexpr evaluation to use the same infrastructure.
Sam McCall [Tue, 2 Jul 2019 15:53:14 +0000 (15:53 +0000)]
clang-format: Add new style option AlignConsecutiveMacros
This option behaves similarly to AlignConsecutiveDeclarations and
AlignConsecutiveAssignments, aligning the assignment of C/C++
preprocessor macros on consecutive lines.
I've worked in many projects (embedded, mostly) where header files full
of large, well-aligned "#define" blocks are a common pattern. We
normally avoid using clang-format on these files, since it ruins any
existing alignment in said blocks. This style option will align "simple"
PP macros (no parameters) and PP macros with parameter lists on
consecutive lines.
Related Bugzilla entry (thanks mcuddie):
https://llvm.org/bugs/show_bug.cgi?id=20637