Summary:
General idea is to utilize generic (mostly Generic_GCC) code
and get rid of Solaris-specific handling as much as possible.
In particular:
- scanLibDirForGCCTripleSolaris was removed, relying on generic
CollectLibDirsAndTriples
- findBiarchMultilibs is now properly utilized to switch between
m32 and m64 include & lib paths on Solaris
- C system include handling copied from Linux (bar multilib hacks)
Fixes PR24606.
Reviewers: dlj, rafael, jyknight, theraven, tstellar
Reviewed By: jyknight
Subscribers: aaron.ballman, mgorny, krytarowski, ro, joerg, cfe-commits
Differential Revision: https://reviews.llvm.org/D35755
llvm-svn: 323193
Pass and return _Float16 as if it were an int or float for ARM, but with the
top 16 bits unspecified, similarly like we already do for __fp16.
We will implement proper half-precision function argument lowering in the ARM
backend soon, but want to use this workaround in the mean time.
Differential Revision: https://reviews.llvm.org/D42318
llvm-svn: 323185
When a function taking transparent union is declared as taking one of
union members earlier in the translation unit, clang would hit an
"Invalid cast" assertion during EmitFunctionProlog. This case
corresponds to function f1 in test/CodeGen/transparent-union-redecl.c.
We decided to cast i32 to union because after merging function
declarations function parameter type becomes int,
CGFunctionInfo::ArgInfo type matches with ABIArgInfo type, so we decide
it is a trivial case. But these types should also be castable to
parameter declaration type which is not the case here.
Now the fix is in converting from ABIArgInfo type to VarDecl type and using
argument demotion when necessary.
Additional tests in Sema/transparent-union.c capture current behavior and make
sure there are no regressions.
rdar://problem/34949329
Reviewers: rjmccall, rafael
Reviewed By: rjmccall
Subscribers: aemerson, cfe-commits, kristof.beyls, ahatanak
Differential Revision: https://reviews.llvm.org/D41311
llvm-svn: 323156
Summary:
First, we need to explain the core of the vulnerability. Note that this
is a very incomplete description, please see the Project Zero blog post
for details:
https://googleprojectzero.blogspot.com/2018/01/reading-privileged-memory-with-side.html
The basis for branch target injection is to direct speculative execution
of the processor to some "gadget" of executable code by poisoning the
prediction of indirect branches with the address of that gadget. The
gadget in turn contains an operation that provides a side channel for
reading data. Most commonly, this will look like a load of secret data
followed by a branch on the loaded value and then a load of some
predictable cache line. The attacker then uses timing of the processors
cache to determine which direction the branch took *in the speculative
execution*, and in turn what one bit of the loaded value was. Due to the
nature of these timing side channels and the branch predictor on Intel
processors, this allows an attacker to leak data only accessible to
a privileged domain (like the kernel) back into an unprivileged domain.
The goal is simple: avoid generating code which contains an indirect
branch that could have its prediction poisoned by an attacker. In many
cases, the compiler can simply use directed conditional branches and
a small search tree. LLVM already has support for lowering switches in
this way and the first step of this patch is to disable jump-table
lowering of switches and introduce a pass to rewrite explicit indirectbr
sequences into a switch over integers.
However, there is no fully general alternative to indirect calls. We
introduce a new construct we call a "retpoline" to implement indirect
calls in a non-speculatable way. It can be thought of loosely as
a trampoline for indirect calls which uses the RET instruction on x86.
Further, we arrange for a specific call->ret sequence which ensures the
processor predicts the return to go to a controlled, known location. The
retpoline then "smashes" the return address pushed onto the stack by the
call with the desired target of the original indirect call. The result
is a predicted return to the next instruction after a call (which can be
used to trap speculative execution within an infinite loop) and an
actual indirect branch to an arbitrary address.
On 64-bit x86 ABIs, this is especially easily done in the compiler by
using a guaranteed scratch register to pass the target into this device.
For 32-bit ABIs there isn't a guaranteed scratch register and so several
different retpoline variants are introduced to use a scratch register if
one is available in the calling convention and to otherwise use direct
stack push/pop sequences to pass the target address.
This "retpoline" mitigation is fully described in the following blog
post: https://support.google.com/faqs/answer/7625886
We also support a target feature that disables emission of the retpoline
thunk by the compiler to allow for custom thunks if users want them.
These are particularly useful in environments like kernels that
routinely do hot-patching on boot and want to hot-patch their thunk to
different code sequences. They can write this custom thunk and use
`-mretpoline-external-thunk` *in addition* to `-mretpoline`. In this
case, on x86-64 thu thunk names must be:
```
__llvm_external_retpoline_r11
```
or on 32-bit:
```
__llvm_external_retpoline_eax
__llvm_external_retpoline_ecx
__llvm_external_retpoline_edx
__llvm_external_retpoline_push
```
And the target of the retpoline is passed in the named register, or in
the case of the `push` suffix on the top of the stack via a `pushl`
instruction.
There is one other important source of indirect branches in x86 ELF
binaries: the PLT. These patches also include support for LLD to
generate PLT entries that perform a retpoline-style indirection.
The only other indirect branches remaining that we are aware of are from
precompiled runtimes (such as crt0.o and similar). The ones we have
found are not really attackable, and so we have not focused on them
here, but eventually these runtimes should also be replicated for
retpoline-ed configurations for completeness.
For kernels or other freestanding or fully static executables, the
compiler switch `-mretpoline` is sufficient to fully mitigate this
particular attack. For dynamic executables, you must compile *all*
libraries with `-mretpoline` and additionally link the dynamic
executable and all shared libraries with LLD and pass `-z retpolineplt`
(or use similar functionality from some other linker). We strongly
recommend also using `-z now` as non-lazy binding allows the
retpoline-mitigated PLT to be substantially smaller.
When manually apply similar transformations to `-mretpoline` to the
Linux kernel we observed very small performance hits to applications
running typical workloads, and relatively minor hits (approximately 2%)
even for extremely syscall-heavy applications. This is largely due to
the small number of indirect branches that occur in performance
sensitive paths of the kernel.
When using these patches on statically linked applications, especially
C++ applications, you should expect to see a much more dramatic
performance hit. For microbenchmarks that are switch, indirect-, or
virtual-call heavy we have seen overheads ranging from 10% to 50%.
However, real-world workloads exhibit substantially lower performance
impact. Notably, techniques such as PGO and ThinLTO dramatically reduce
the impact of hot indirect calls (by speculatively promoting them to
direct calls) and allow optimized search trees to be used to lower
switches. If you need to deploy these techniques in C++ applications, we
*strongly* recommend that you ensure all hot call targets are statically
linked (avoiding PLT indirection) and use both PGO and ThinLTO. Well
tuned servers using all of these techniques saw 5% - 10% overhead from
the use of retpoline.
We will add detailed documentation covering these components in
subsequent patches, but wanted to make the core functionality available
as soon as possible. Happy for more code review, but we'd really like to
get these patches landed and backported ASAP for obvious reasons. We're
planning to backport this to both 6.0 and 5.0 release streams and get
a 5.0 release with just this cherry picked ASAP for distros and vendors.
This patch is the work of a number of people over the past month: Eric, Reid,
Rui, and myself. I'm mailing it out as a single commit due to the time
sensitive nature of landing this and the need to backport it. Huge thanks to
everyone who helped out here, and everyone at Intel who helped out in
discussions about how to craft this. Also, credit goes to Paul Turner (at
Google, but not an LLVM contributor) for much of the underlying retpoline
design.
Reviewers: echristo, rnk, ruiu, craig.topper, DavidKreitzer
Subscribers: sanjoy, emaste, mcrosier, mgorny, mehdi_amini, hiraditya, llvm-commits
Differential Revision: https://reviews.llvm.org/D41723
llvm-svn: 323155
Summary:
This patch removes IdentifierInfo from completion token after remembering
the identifier in the preprocessor.
Prior to this patch, completion token had the IdentifierInfo set to null when
completing at the start of identifier and to the II for completion prefix
when in the middle of identifier.
This patch unifies how code completion token is handled when it is insterted
before the identifier and in the middle of the identifier.
The actual IdentifierInfo can still be obtained from the Preprocessor.
Reviewers: bkramer, arphaman
Reviewed By: bkramer
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D42241
llvm-svn: 323133
Fix an assertion failure caused by a missing CheckName. The malloc checker
enables "basic" support in the CStringChecker, which causes some CString
bounds checks to be enabled. In this case, make sure that we have a
valid CheckName for the BugType.
llvm-svn: 323052
This fixes PR32732 by updating CurLexerKind to reflect available lexers.
We were hitting null pointer in Preprocessor::Lex because CurLexerKind
was CLK_Lexer but CurLexer was null. And we set it to null in
Preprocessor::HandleEndOfFile when exiting a file with code completion
point.
To reproduce the crash it is important for a comment to be inside a
class specifier. In this case in Parser::ParseClassSpecifier we improve
error recovery by pushing a semicolon token back into the preprocessor
and later on try to lex a token because we haven't reached the end of
file.
Also clang crashes only on code completion in included file, i.e. when
IncludeMacroStack is not empty. Though we reset CurLexer even if include
stack is empty. The difference is that during pushing back a semicolon
token, preprocessor calls EnterCachingLexMode which decides it is
already in caching mode because various lexers are null and
IncludeMacroStack is not empty. As the result, CurLexerKind remains
CLK_Lexer instead of updating to CLK_CachingLexer.
rdar://problem/34787685
Reviewers: akyrtzi, doug.gregor, arphaman
Reviewed By: arphaman
Subscribers: cfe-commits, kfunk, arphaman, nemanjai, kbarton
Differential Revision: https://reviews.llvm.org/D41688
llvm-svn: 323008
Using a BlockDecl in a default member initializer causes it to be attached to
CXXMethodDecl without its access specifier being set. This prevents a crash
where getAccess is called on this BlockDecl, since that method expects any
Decl in CXXRecord scope to have an access specifier.
llvm-svn: 322984
Summary:
Upstream LLVM is changing the the prototypes of the @llvm.memcpy/memmove/memset
intrinsics. This change updates the Clang tests for this change.
The @llvm.memcpy/memmove/memset intrinsics currently have an explicit argument
which is required to be a constant integer. It represents the alignment of the
dest (and source), and so must be the minimum of the actual alignment of the
two.
This change removes the alignment argument in favour of placing the alignment
attribute on the source and destination pointers of the memory intrinsic call.
For example, code which used to read:
call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 100, i32 4, i1 false)
will now read
call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %dest, i8* align 4 %src, i32 100, i1 false)
At this time the source and destination alignments must be the same (Step 1).
Step 2 of the change, to be landed shortly, will relax that contraint and allow
the source and destination to have different alignments.
llvm-svn: 322964
The standard says:
[expr.static.cast] p11: "If the prvalue of type “pointer to cv1 B” points to a B
that is actually a subobject of an object of type D, the resulting pointer points
to the enclosing object of type D. Otherwise, the behavior is undefined."
Therefore, the GEP must be inbounds.
This should solve the failure to optimize away a null check shown in PR35909:
https://bugs.llvm.org/show_bug.cgi?id=35909
Differential Revision: https://reviews.llvm.org/D42249
llvm-svn: 322950
This removes the following (already default-off) warnings from -Wextra:
-Wtautological-type-limit-compare,
-Wtautological-unsigned-zero-compare
-Wtautological-unsigned-enum-zero-compare
On the thread "[cfe-dev] -Wtautological-constant-compare issues", clang
code owners Richard Smith, John McCall, and Reid Kleckner as well as
libc++ code owner Marshall Clow stated that these new warnings are not
yet ready for prime time and shouldn't be part of -Wextra.
Furthermore, Vedant Kumar (Apple), Peter Hosek (Fuchsia), and me (Chromium)
expressed the same concerns (Vedant on that thread, Peter on
https://reviews.llvm.org/D39462, me on https://reviews.llvm.org/D41512).
So remove them from -Wextra, and remove TautologicalInRangeCompare from
TautologicalCompare too until they're usable with real-world code.
llvm-svn: 322901
Firstly, each offloading entry must have a unique name or the
linker will complain if there are multiple files with target
regions. Secondly, the compiler must not introduce padding so
mark the struct with a PackedAttr.
Differential Revision: https://reviews.llvm.org/D42168
llvm-svn: 322858
PreStmt<CXXNewExpr> was never called.
Additionally, under c++-allocator-inlining=true, PostStmt<CXXNewExpr> was
called twice when the allocator was inlined: once after evaluating the
new-expression itself, once after evaluating the allocator call which, for the
lack of better options, uses the new-expression as the call site.
This patch fixes both problems.
Differential Revision: https://reviews.llvm.org/D41934
rdar://problem/12180598
llvm-svn: 322797
Add PostAllocatorCall program point to represent the moment in the analysis
between the operator new() call and the constructor call. Pointer cast from
"void *" to the correct object pointer type has already happened by this point.
The new program point, unlike the previously used PostImplicitCall, contains a
reference to the new-expression, which allows adding path diagnostics over it.
Differential Revision: https://reviews.llvm.org/D41800
rdar://problem/12180598
llvm-svn: 322796
Pointer escape event notifies checkers that a pointer can no longer be reliably
tracked by the analyzer. For example, if a pointer is passed into a function
that has no body available, or written into a global, MallocChecker would
no longer report memory leaks for such pointer.
In case of operator new() under -analyzer-config c++-allocator-inlining=true,
MallocChecker would start tracking the pointer allocated by operator new()
only to immediately meet a pointer escape event notifying the checker that the
pointer has escaped into a constructor (assuming that the body of the
constructor is not available) and immediately stop tracking it. Even though
it is theoretically possible for such constructor to put "this" into
a global container that would later be freed, we prefer to preserve the old
behavior of MallocChecker, i.e. a memory leak warning, in order to
be able to find any memory leaks in C++ at all. In fact, c++-allocator-inlining
*reduces* the amount of false positives coming from this-pointers escaping in
constructors, because it'd be able to inline constructors in some cases.
With other checkers working similarly, we simply suppress the escape event for
this-value of the constructor, regardless of analyzer options.
Differential Revision: https://reviews.llvm.org/D41797
rdar://problem/12180598
llvm-svn: 322795
Implements finding appropriate source locations for intermediate diagnostic
pieces in path-sensitive bug reports that need to descend into an inlined
operator new() call that was called via new-expression. The diagnostics have
worked correctly when operator new() was called "directly".
Differential Revision: https://reviews.llvm.org/D41409
rdar://problem/12180598
llvm-svn: 322791
Fix the const qualifier so that the operator defined in the tests indeed does
override the default global nothrow version of new.
Differential Revision: https://reviews.llvm.org/D41408
llvm-svn: 322790
The callback runs after operator new() and before the construction and allows
the checker to access the casted return value of operator new() (in the
sense of r322780) which is not available in the PostCall callback for the
allocator call.
Update MallocChecker to use the new callback instead of PostStmt<CXXNewExpr>,
which gets called after the constructor.
Differential Revision: https://reviews.llvm.org/D41406
rdar://problem/12180598
llvm-svn: 322787
Make sure that with c++-allocator-inlining=true we have the return value of
conservatively evaluated operator new() in the correct memory space (heap).
This is a regression/omission that worked well in c++-allocator-inlining=false.
Heap regions are superior to regular symbolic regions because they have
stricter aliasing constraints: heap regions do not alias each other or global
variables.
Differential Revision: https://reviews.llvm.org/D41266
rdar://problem/12180598
llvm-svn: 322780
According to [basic.stc.dynamic.allocation], the return type of any C++
overloaded operator new() is "void *". However, type of the new-expression
"new T()" and the type of "this" during construction of "T" are both "T *".
Hence an implicit cast, which is not present in the AST, needs to be performed
before the construction. This patch adds such cast in the case when the
allocator was indeed inlined. For now, in the case where the allocator was *not*
inlined we still use the same symbolic value (which is a pure SymbolicRegion of
type "T *") because it is consistent with how we represent the casts and causes
less surprise in the checkers after switching to the new behavior.
The better approach would be to represent that value as a cast over a
SymbolicRegion of type "void *", however we have technical difficulties
conjuring such region without any actual expression of type "void *" present in
the AST.
Differential Revision: https://reviews.llvm.org/D41250
rdar://problem/12180598
llvm-svn: 322777
The -analyzer-config c++-allocator-inlining experimental option allows the
analyzer to reason about C++ operator new() similarly to how it reasons about
regular functions. In this mode, operator new() is correctly called before the
construction of an object, with the help of a special CFG element.
However, the subsequent construction of the object was still not performed into
the region of memory returned by operator new(). The patch fixes it.
Passing the value from operator new() to the constructor and then to the
new-expression itself was tricky because operator new() has no call site of its
own in the AST. The new expression itself is not a good call site because it
has an incorrect type (operator new() returns 'void *', while the new expression
is a pointer to the allocated object type). Additionally, lifetime of the new
expression in the environment makes it unsuitable for passing the value.
For that reason, an additional program state trait is introduced to keep track
of the return value.
Finally this patch relaxes restrictions on the memory region class that are
required for inlining the constructor. This change affects the old mode as well
(c++-allocator-inlining=false) and seems safe because these restrictions were
an overkill compared to the actual problems observed.
Differential Revision: https://reviews.llvm.org/D40560
rdar://problem/12180598
llvm-svn: 322774
When using -fno-integrated-as flag, the gnu assembler produces code
with some default march/mabi which later causes linker failure due
to incompatible mabi/march.
In this patch we explicitly propagate -mabi and -march flags to the
GNU assembler.
In this patch we explicitly propagate -mabi and -march flags to the GNU assembler.
Differential Revision: https://reviews.llvm.org/D41271
llvm-svn: 322769
Both are related to handling anonymous structures.
* clang didn't handle () around an anonymous struct variable.
* clang also crashed on syntax errors that could lead to other
syntactic constructs following the declaration of an
anonymous struct. While the code is invalid, that's not
a good reason to panic compiler.
Differential Revision: https://reviews.llvm.org/D41788
llvm-svn: 322742
When parsing C++ type construction expressions with list initialization,
forward the locations of the braces to Sema.
Without these locations, the code coverage pass crashes on the given test
case, because the pass relies on getLocEnd() returning a valid location.
Here is what this patch does in more detail:
- Forwards init-list brace locations to Sema (ParseExprCXX),
- Builds an InitializationKind with these locations (SemaExprCXX), and
- Uses these locations for constructor initialization (SemaInit).
The remaining changes fall out of introducing a new overload for
creating direct-list InitializationKinds.
Testing: check-clang, and a stage2 coverage-enabled build of clang with
asserts enabled.
Differential Revision: https://reviews.llvm.org/D41921
llvm-svn: 322729
HTML diagnostics can be an overwhelming blob of pages of code.
This patch adds a checkbox which filters this list down to only the
lines *relevant* to the counterexample by e.g. skipping branches which
analyzer has assumed to be infeasible at a time.
The resulting amount of output is much smaller, and often fits on one
screen, and also provides a much more readable diagnostics.
Differential Revision: https://reviews.llvm.org/D41378
llvm-svn: 322612
parallel for simd` directives.
Added codegen for `depend` clauses on `#pragma omp target teams
distribute parallel for simd` directives.
llvm-svn: 322587
Dan Gohman